MySQL Quietly Drops Support For Debian Linux [UPDATED]

volts writes "MySQL quietly deprecated support for most Linux distributions on October 16, when its 'MySQL Network' support plan was replaced by 'MySQL Enterprise.' MySQL now supports only two Linux distributions — Red Hat Enterprise Linux and SUSE Linux Enterprise Server. We learned of this when MySQL declined to sell us support for some new Debian-based servers. Our sales rep 'found out from engineering that the current Enterprise offering is no longer supported on Debian OS.' We were told that 'Generic Linux' in MySQL's list of supported platforms means 'generic versions of the implementations listed above'; not support for Linux in general." Update: 12/13 20:52 GMT by J : MySQL AB's Director of Architecture (and former Slash programmer) Brian Aker corrects an apparent miscommunication in a blog post: "we are just starting to roll out [Enterprise] binaries... We don't build binaries for Debian in part because the Debian community does a good job themselves... If you call MySQL and you have support we support you if you are running Debian (the same with Suse, RHEL, Fedora, Ubuntu and others)... someone in Sales was left with the wrong information"

Comments

Let's fork it!

[bogaboga]

Guys, it's time to fork MySQL. I am ready [and willing] to contribute. What do you think?

Re:Let's fork it!

[eln]

I suppose you could do that, but unless you're planning on offering Enterprise support for your offering on a wide variety of platforms, you're not really gaining anything. MySQL will presumably still run on Debian, at least for now, but without the ability to buy support for it on that platform, you're not going to get approval to put it on that platform in any sort of business-critical environment.

Now, if you wanted to start a new company that offered Enterprise support for MySQL on Debian, you might have something there. I don't know that you would make any money, but at least you'd be offering something that isn't currently offered.

Re:Let's fork it!

[Nasarius]

Why? Is there a problem with the code, or the license? You're free to start your own company and offer tech support and other services for MySQL, and there's always PostgreSQL. But if the MySQL coders are still doing good work, I see no reason to fork.

Re:Let's fork it!

[Alchemar]

The problem is support, not inoperability. The software still works, you just don't have anyone to call when it doesn't work the way you expected. Forking the project does not solve this problem. If a third party wanted to sell a customer support contract for it, they could do so without needing a fork. If MySQL started releasing later versions of the software without the source, then a fork would be needed to have a branch that could be supported by another company.

Re:Let's fork it!

[jmorris42]

> I don't know that you would make any money, but at least you'd be offering something that isn't currently offered.

I doubt it. And more important than my opinion, MySQL doubts it and has the sales figures to show it. Companies don't normally kill off profitable products and services, not even evil/stupid corporations.

Re:Let's fork it!

[Drasil]

Guys, it's time to fork MySQL.

...or switch to the excellent Postgres which is more open and a more complete SQL implementation than MySQL anyway.

Expect to see more things like this happening as the IT landscape undergoes it's coming changes.

Re:Let's fork it!

[suntac]

Mmm fork MySQL? Why? There is nothing wrong with the code. You could try to fork the support and start a company specialized in MySQL support on Debian.....

I think there is a market for this. The only thing you need is a couple of good people. You/we(the community) could also create a company GPL style. Create a pool of people willing to devote there time on solving MySQL Debian support problems. Create a ticket like system and assign questions to people in the pool.

This way you can quickly create a non-profit company with little to non investments. The biggest "problem" is that you have to attract people willing to become part of you expert pool.

While writing this, it might even be a good challenge to start this..... I will think some more about this. :-) Anyone in? ;-)

Regards,
Johan Louwers.

Re:Let's fork it!

[Anonymous Coward]

But if the MySQL coders are still doing good work

What do you mean by still? It makes you sound like you are trying to claim MySQL was not a fucking toy database for girlymen.

Re:Let's fork it!

[Orange Crush]

but without the ability to buy support for it on that platform, you're not going to get approval to put it on that platform in any sort of business-critical environment.

. . . without the ability to buy support from MySQL for it, that is. Third parties, system integrators, etc. will continue to support whatever their customers pay them for. So while this is a blow for Debian in big enterprise, let's face it, how many big enterprise environments were running straight Debian in the first place? Red Hat's king with SUSE buzzing around their ankles. This won't affect small to mid sized organizations with outside IT people.

Re:Let's fork it!

[Anonymous Coward]

...or switch to the excellent Postgres which is more open and a more complete SQL implementation than MySQL anyway.

Yes ! It even includes a vaccuum !

Re:Let's fork it!

[Dan Farina]

Postgres is getting faster, too. With SQLite and Postgres, I find the obvious uses for MySQL to be shrinking and becoming more dubious.

Re:Let's fork it!

[Nasarius]

Funny and true. Great reply.

Re:Let's fork it!

[Vlad_the_Inhaler]

The other explanation would be that Debian has done something to seriously piss the MySQL people off (my speculation).

I *know* that they went this way with the Seamonkey crew. Here is a reproduced Newsgroup response from a Seamonkey developer on the subject of Debian and Iceape (the previous thread entry is in italics and the developers response is bold:

The "SeaMonkey" trademark is held by MoFo, but AIUI, they allow the Council to grant people the right to use it.

Well, MoFo applied for the trademarks, but doesn't hold them yet, as they've not yet been granted. They applied for them representing us though, and they will leave management of the trademark in the hands of the SeaMonkey Council.

But AIUI, Debian has moved past caring about using MoFo's trademarks.
And AFAICT from this thread, the level of bitterness on the SeaMonkey side seems even higher than in the Mozilla community in general.

That may very much be true, as they pre-judged us of being the same as MoCo and not even listening to what we wanted to say. Us being legally backend by MoFo was enough for them to not even really discuss this topic, i.e. not even asking what the terms for using the SeaMonkey trademarks would be.

And their choice of name for the clone they are shipping is an insult in my ears anyways, but that's just my personal opinion.

BTW, I really think their inconsistent treatment of trademarks is enough reason for not understanding them anyways. Their own trademarks are protected with one of the strictest possible policies (no use except explicitely granted by Debian) and then they accuse other of being too strict - and it seems some of their responsible people have not yet understood that trademarks and copyright are two completely different things legally.

Anyways, for me, that discussion is over and Debian itself is dead meat in this regard for me personally (note that ubuntu even departs from Debian's path for MoCo trademarks already).

Elsewhere in the thread, IceApe was described by the same person as a 'Crappy Clone'.

Re:Let's fork it!

[smellsofbikes]

>Companies don't normally kill off profitable products and services, not even evil/stupid corporations.

I'd have to disagree with that. I've watched three large companies for whom I've worked -- all Fortune 500 companies -- kill off profitable products and services that were not as profitable as they wanted. The company I'm working for right now sold off three business units because they didn't have a profit margin above 30%. We're only keeping the parts of the company that can beat 30%: if you don't, you're out the door. There are probably a lot of fields where companies can't afford to throw away marginal profit, but there are plenty of fields where it's not worth chasing chump change when there's a 50% profit margin to be hunted down and seized.

If that's the case, it's quite possible another, smaller and more agile, company could live very comfortably on the profits from this discard.

Re:Let's fork it!

[newt0311]

atleast it has decent support for transactions, key constraints, and procedural languages.

Re:Let's fork it!

[fritsd]

I'm game for that :-)

Re:Let's fork it!

[IdleTime]

LOLOLOLOL Funniest comment today.

I suppose you are an expert in database design and programming? Wanna start writing a cost based optimizer? How about an obfuscation toolkit? Or even better, write a new writer process that can perform writes to multiple database files based on requests over a network connection?

ROFLMAO!

Re:Let's fork it!

[Da_Weasel]

I have a better idea...why don't you go fork yourself!

(I couldn't resist!)

Re:Let's fork it!

[ciggieposeur]

First, by support the article refers to technical support contracts, not whether or not the software will actually run on Debian. And MySQL has decided that they will provide technical support only for a very limited subset of the popular Linux distros. As far is this issue is concerned, Debian is in the same boat as a lot of other distros and was not singled out for special treatment.

Second, the Mozilla trademark issue was at its core unavoidable. Debian has to be able to say to its derivative distros that everything in "main" is really free, Mozilla had copyrighted images that were NOT free, so Debian couldn't use them and Mozilla responded by saying they had to rename the browser. So they did, and the Mozilla-branded browser remains in "non-free" due to the copyrighted images. Everyone accusing Debian of hypocrisy on the trademark issue because they have an official logo is (to be blunt) wrong. Debian has an official logo (that they hardly ever use) to provide legal recourse to stop anyone else claiming to be Debian. It is otherwise of no use in the project and does nothing to prevent derivative distros from doing their own thing when they want to.

Incidentally, the Mozilla trademark dispute has caused me to reinvestigate my use of ALL software from Mozilla. I'm finding that KDE software is far more user-friendly and powerful than the Mozilla software across a number of applications. KMail can be made (rather easily) to store mail in ~/Mail in mbox format, its mail filters execute much faster, I can right-click -> "Create Filter" -> "Filter on From" in seconds, and in dozens of other ways it kicks mozilla-mail's ass. Likewise KNode, Konqueror, and Kontact.

Re:Let's fork it!

[aristotle-dude]

Guys, it's time to fork MySQL. I am ready [and willing] to contribute. What do you think?

Fork you.

Re:Let's fork it!

[Anonymous Coward]

"The software still works, you just don't have anyone to call when it doesn't work the way you expected."

Sorry man, no candy.

That might be the case with a closed source product, but you must rewrite your assertion since MySQL is GPL:

"The software still works, you just don't call *MySQL AB* when it doesn't work the way you expected".

Re:Let's fork it!

[killjoe]

postgres recently came out with a new version and I don't think /. even covered it.

What gives?

Re:Let's fork it!

[killjoe]

Yes but it doesn't have what people REALLY want. Replication, clustering, failover, case insensitive where clauses.

If you want high availibility you have to cobble together slony and pgpool (which does not support multi master replication) neither of which is suitable for working over a WAN.

There is a reason why people choose MySql and that's because it delivers the features people really want first. Even the features are not 100% "correct" they are delivered "good enough" to get "real work" done.

Take case insesntive where clauses for example. For the last five years or so that I have been following the pg mailing lists there must have been hundreds of requests from people who want to switch over from mysql, ms-sql, oracle, informix, firebird etc for a case insensitive collation option. They just get ignored and told to change all their queries to use ILIKE or *~ or some other stupid non standard postgres only SQL. Oddly enough their primary excuse for not providing it is that it's not a SQL standard.

So if you using any kind of an ORM and you can not stomach asking your employees or web users to remember the exact capitalization of everything they have ever typed into your database then postgres is not an option.

Sorry.

Re:Let's fork it!

[IpalindromeI]

Yes, they did cover it.

Re:Let's fork it!

[Jamesday]

Who do I contact at PostgreSQL there to buy their support? Oh. They don't sell it. You seem to have missed the point of this: it's about the developers of the database selling support with service level agreements and contractual service obligations for what they produce. PostgreSQL doesn't do that at all, for any platform. MySQL does.

Like PostgreSQL, MySQL also has third parties like Red hat who sell the server and also sell support for what they build. Those partners can escalate issues to the MySQL support team if they encounter a problem they can't deal with. All with contractual service levels and warranties for both what they deliver to their customers and what MySQL delivers to them.

Mailing lists are great. They aren't the same as MySQL having someone who knows the server around to answer the phone 24/7 to deal with an emergency, who has the phone numbers of anyone in the company who they need to contact to get a problem resolved.

Re:Let's fork it!

[Generic Player]

"Yes but it doesn't have what people REALLY want. Replication, clustering, failover, case insensitive where clauses."

Yes, yes, yes, and yes. I realize you are just trolling, but I will still point out the facts so that you don't breed new mysqltards.

There are a few replication options, that's trivial. Clustering and failover are provided by pgcluster. Case insensitive where clauses are provided by ilike, as you already know.

"There is a reason why people choose MySql and that's because it delivers the features people really want first. Even the features are not 100% "correct" they are delivered "good enough" to get "real work" done."

Most people I know who use mysql do so because they are using some shitty software that requires it, because they didn't know there were other free DBs, or because they are doing some high performance task where data safety doesn't matter, so they are using the fast myisam tables.

"Take case insesntive where clauses for example. For the last five years or so that I have been following the pg mailing lists there must have been hundreds of requests from people who want to switch over from mysql, ms-sql, oracle, informix, firebird etc for a case insensitive collation option. They just get ignored and told to change all their queries to use ILIKE or *~ or some other stupid non standard postgres only SQL. Oddly enough their primary excuse for not providing it is that it's not a SQL standard."

No, only from idiots who think "a" and "A" are the same thing. If you want to do a case insensitive comparison, use ilike or use upper() or lower() (depending on the circumstances).

"So if you using any kind of an ORM and you can not stomach asking your employees or web users to remember the exact capitalization of everything they have ever typed into your database then postgres is not an option."

Really? That's why thousands of people use it and have no problem at all with case insensitivity right? Being case sensitive is normal behaviour. If you want to check an email then do lower(email), this is basic stuff people learn in their first ever foray into SQL.

Re:Let's fork it!

[killjoe]

"There are a few replication options, that's trivial."

Really. First of all I would hardly call slony trivial. It's a bitch to set up. Secondly it does not support multi master replication. Finally it's not asyncronous and should not be used WANs.

"Clustering and failover are provided by pgcluster."

Pgcluster does not work over a WAN. Pgcluster is not a part of postgres it's a separate product.

"Case insensitive where clauses are provided by ilike, as you already know."

Ilike is not an SQL standard. If you have hundreds of queries already written then you have to rewrite every single one of them to use ilike. Ilike is slower then where x='y'. I don't know of any ORM on the market that uses ilike this means you can not use hibernate, activerecord, etc.

"No, only from idiots who think "a" and "A" are the same thing."

Unfortunately for postgres that's most human beings on the planet. That's why virtually every single database on the planet gives you the option of a case insensitive collation except postgres. Oracle, DB/2, MSSQL, Mysql, firebird, informix, all have that option. Postgres does not.

"If you want to do a case insensitive comparison, use ilike or use upper() or lower() (depending on the circumstance"

In other words stop using an ORM and write all all your queries by hand. If you are not using an ORM and are porting an app from mysql then rewrite every single query that has a where clause to use upper and lower.

"Really? That's why thousands of people use it and have no problem at all with case insensitivity right? "

Yes thousands use it. Alas millions use oracle, mysql, mssql, and db/2. As you may have noticed postgres is not the most popular database. There is a reason for that.

"If you want to check an email then do lower(email)"

Hey I have another idea. I could just use mysql, firebird, oracle, mssql, informix, db/2 instead then I would not have to rewrite all my queries.

I do have a question for you though...

How come the postgres developers feel fine about completely ignoring the hundreds of requests they have had for this feature? Every other database has it so why not just write a new collation? Is it really that hard?

I'll answer the question for you. No it's not that hard. The developers are arrogant pricks who feel fine about non standard SQL like ILIKE and *~ when it suits them but throw a knipshin about standards when somebody asks for a case insensitive collation.

People have been asking for this for years now, just google their mailing list. They just don't give a shit. If you want to know why people are choosing mysql instead of postgres there is the answer. Mysql is more responsive to their users, postgres developers don't give a shit. They do what they want not what their users want.

Re:Let's fork it!

[Generic Player]

"Really. First of all I would hardly call slony trivial. It's a bitch to set up. Secondly it does not support multi master replication. Finally it's not asyncronous and should not be used WANs."

Hmm, I must just be so super awesome then since slony doesn't seem to be so hard to me. Replication is master->slave. That is the definition. Master-master is clustering, which is what pgcluster is for (shocking I know). And if you want async replication then use async replication (rserv). Why do you think a sync replication package not supporting async replication is a problem?

"Pgcluster does not work over a WAN. Pgcluster is not a part of postgres it's a separate product."

Oh of course, you are doing multi master clustering over a WAN using mysql. Just because mysql isn't up front about the limitations of the replication and clustering options, doesn't mean those limitations aren't there.

"Ilike is not an SQL standard. If you have hundreds of queries already written then you have to rewrite every single one of them to use ilike. Ilike is slower then where x='y'. I don't know of any ORM on the market that uses ilike this means you can not use hibernate, activerecord, etc."

Yeah, having to change your queries never happens with migrating between other DBs. Come on, at least try to make sense. If you cared about portability, you would be using upper() or lower() in the first place. Expecting postgresql to support the vendor specific hack you like instead of the vendor specific hack they already have is just dumb. I'm lucky enough not to have to touch java, but the folks at work had no problem with getting hibernate to co-operate for this. Activerecord may well be a problem given that it is a mysql mapping, not an SQL mapping, talk to DHH about his damage not postgresql devs. And why do you think ilike is slower than x='y' in a case insensitive DB? Here's a hint, that case insensitivity isn't put there by magic fairies, that's actually code that has to execute to do that for you. Hiding the upper() from you doesn't make it faster.

"Yes thousands use it. Alas millions use oracle, mysql, mssql, and db/2. As you may have noticed postgres is not the most popular database. There is a reason for that."

Its unlikely that millions of people actually care about what you are talking about. Thousands of postgresql users think you are wrong. It doesn't seem that many other DB users think you are right. In fact, google seems to support the idea that its mostly just you posting this same shit all over the web because you like to whine about not getting your way. I don't care that postgresql is not the most popular. It never will be and that's great. I can't think of any software I use that is the most popular other than apache, and I don't use it because of its popularity. Popular != good.

Re:Let's fork it!

[DeathElk]

Get forked!

Re:Let's fork it!

[killjoe]

"Oh of course, you are doing multi master clustering over a WAN using mysql."

I have done that for years. It's called ring replication look it up sometime. Now mysql also offers Mysql cluster which also works over a wan.

"Yeah, having to change your queries never happens with migrating between other DBs. Come on, at least try to make sense. "

If you are using an ORM then it should be a trivial exercize. Ooops no ORM uses ILIKE or *~ or UPPER() to do where clauses so if you are trying to migrate from mysql to postgres you better forget it. You have to get rid of your ORM and write all your queries by hand. So sorry, the developers of postgres don't really care about that.

"Expecting postgresql to support the vendor specific hack you like instead of the vendor specific hack they already have is just dumb."

But every vendor of databases supports case insensitive queries. Except postgres of course.

"And why do you think ilike is slower than x='y' in a case insensitive DB? "

Does ilike use indexes?

"It doesn't seem that many other DB users think you are right."

Then why do all of them give you the option of case insensitive collations?

" In fact, google seems to support the idea that its mostly just you posting this same shit all over the web because you like to whine about not getting your way."

Really. Over the last five years I am the only one asking for this feature? Google says that? Really?

"Popular != good."

Yes but unpopular != good either.

Here is a hint for you. I need feature X. Every database in the world has feature X except postgres. I guess I won't use postgres.

This topic was about why people use mysql when postgres is so much better. There is a reason for you. Postgres does not support feature X.

Re:Let's fork it!

[Generic Player]

"So sorry, the developers of postgres don't really care about that."

No need to be sorry, lots of people don't care about that, at all, no really involved.

"But every vendor of databases supports case insensitive queries. Except postgres of course."

Postgres does impliment it, just not the way you want it to.

"Does ilike use indexes?"

Off the top of my head, I couldn't tell you. If not then use the portable upper() or lower() like you should have been doing all along.

"Then why do all of them give you the option of case insensitive collations?"

Users of other databases give you that option?

"Over the last five years I am the only one asking for this feature? Google says that? Really?"

No, google doesn't say that. And neither did I.

"Here is a hint for you. I need feature X. Every database in the world has feature X except postgres. I guess I won't use postgres."

Good for you.

"This topic was about why people use mysql when postgres is so much better. There is a reason for you. Postgres does not support feature X."

And as I said, you don't matter. I am sure this is a problem for you. I am sure your refusal to do things correctly is for a good reason. That doesn't mean its a problem for everyone else, or even a significant number of people. "I can't switch to postgres easily because I relied on non-standard vendor specific features" isn't very compelling. You already chose your database before you made the mess, so clearly the mess is not the reason you didn't choose postgres, its just a reason you haven't migrated to it. A good reason not to migrate I might add. And not a good reason to put bad "features" that shouldn't be there into postgres. Its not GNU software, they won't just add any and every feature anyone ever requests.

Re:Let's fork it!

[jetxee]

Incidentally, the Mozilla trademark dispute has caused me to reinvestigate my use of ALL software from Mozilla.

So did I. I switched to Epiphany and find its bookmarking functionality much more flexible.

Re:Let's fork it!

[killjoe]

"No need to be sorry, lots of people don't care about that, at all, no really involved."

Then why does every other database support it?

"Postgres does impliment it, just not the way you want it to."

No postgres implements a workaround which is not an SQL standard and is not supported by any ORM.

"Off the top of my head, I couldn't tell you. If not then use the portable upper() or lower() like you should have been doing all along."

Because littering your code with SQL is a bad programming practice. Modern programming methods use an object relational layer. You may want to read about them sometime. Oh I will answer the question for you, ilike does not use indexes.

""I can't switch to postgres easily because I relied on non-standard vendor specific features" isn't very compelling. "

Except that I did no such thing, "where lastname = 'McNab'" is standard. "Where lastname *~ 'McNab'" is not. See the difference?

Re:Let's fork it!

[Generic Player]

"Then why does every other database support it?"

Vendor A impliments "feature" designed to lock in customers. Vendor B is forced to impliment it too. Postgresql isn't a company, there is no vendor, so they aren't under the presure to screw up their software.

"No postgres implements a workaround which is not an SQL standard and is not supported by any ORM."

Right, just not the workaround you like. If you want to be portable and SQL standard, then YOU MUST USE upper() or lower(). There is no other option. Just because you are relying on a vendor extension doesn't mean postgresql should impliment it for you. And I don't believe that any ORM sucks that hard, they ALL let you define your searches. If you want to search case insensitive, then do so.

"Because littering your code with SQL is a bad programming practice. Modern programming methods use an object relational layer. You may want to read about them sometime. Oh I will answer the question for you, ilike does not use indexes."

You don't need to litter your code with SQL. You already have to build a where clause. Any ORM will let you put upper() in there, or any other SQL. You can even configure them (hibernate at least) to specify what function upper() is if you really need to. And "modern programming methods" and ORM are miles apart. OO programming is very old, as is the bizzare fetish of trying to mash anything and everything into an OO layer for no reason.

"Except that I did no such thing, "where lastname = 'McNab'" is standard. "Where lastname *~ 'McNab'" is not. See the difference?"

You are ignoring your "be case insensitive" setting. "where lastname = 'McNab'" is standard. And as per the standard, will only return rows where lastname = 'McNab', not 'mcnab'. If you want case insensitivity, then use the also standard:
"where lower(lastname) = 'mcnab'". Duh.

Setting the database to return the wrong information to a standard query means its no longer standard. This should be pretty easy to understand.

Re:Let's fork it!

[suntac]

As promised in my previous post I would give the idea some more thoughts. Forking MySQL Debian support might not be such a bad idea. MySQL is giving up support on Debian and thereby leaving open the possibilities for this to the community.

The opensource community has proven in the past that they are quite capable of handling things like this. I might even be an idea for a small startup (profit or non-profit) I will post some more info about my idea around this on in this discussion and on my weblog http://johanlouwers.blogspot.com/

Regards,
Johan Louwers.

Re:Let's fork it!

[Jamesday]

Before starting that business, see http://developers.slashdot.org/comments.pl?sid=211 584&cid=17233206 and note that MySQL is increasing its support for Debian, not decreasing it.

MySQL is removing the public downloads for version 4.0, so if you do want a business idea, providing those might work.

Re:Let's fork it!

[killjoe]

"Vendor A impliments "feature" designed to lock in customers. Vendor B is forced to impliment it too. Postgresql isn't a company, there is no vendor, so they aren't under the presure to screw up their software."

Explain why all the open source databases support this feature too then?

"There is no other option. "

Yes there is. It's called collations. It's a standard. Look into it sometime. In fact the SQL standard says you should be able to set collations on a per column basis which MSSQL and firebird support but postgres does not.

"Any ORM will let you put upper() in there, or any other SQL. "

Really? So when I am using activerecord and I call Person.find_by_last_name('mcnab') it puts UPPER in there for me?

"OO programming is very old, as is the bizzare fetish of trying to mash anything and everything into an OO layer for no reason."

Aaaah you are one of those people are you.

"Setting the database to return the wrong information to a standard query means its no longer standard. This should be pretty easy to understand."

You know this conversation reminds of so many I had with postgres users. I remember when postgres didn't support alter table and dickwads like you were saying "there is no need for alter table just use this workaround". I remember when posgtres didn't support outer joins (I do believe it was the last database in the world to support outer joins) and dickwads like you were saying "there is no need for outer joins just your this query instead".

No difference now "there is no need for case insensitive collations just use ilike, *~ or upper() instead". You are wrong. Eventually postgres will implement case insensitive collations or datatypes thereby upholding their long held tradition of being the last database in the world to implement a feature.

Last to implement alter table, last to implement outer joins, last to implement replication, last to implement clustering, last to implement failover, last to implement case insensitive collations. Due in large part to dickwads like you.

Re:Let's fork it!

[Generic Player]

"Explain why all the open source databases support this feature too then?"

The ones with vendors and customers? How does being open source have any relevance?

"Yes there is. It's called collations. It's a standard. Look into it sometime. In fact the SQL standard says you should be able to set collations on a per column basis which MSSQL and firebird support but postgres does not."

Implimenting collations correctly is more involved than you are pretending. It will need to be flexible enough to support other datatypes besides text based ones. Again, all 2 people that care about this are free to use MSSQL. You are under the impression that because you are one of those 2 people, that there is a huge massive majority of people actively avoiding postgresql because of this. Its simply not the case. Easily THOUSANDS of times more people avoid postgresql because its not a PHB approved corporate supported database. Nobody cares about that "problem" either.

"Really? So when I am using activerecord and I call Person.find_by_last_name('mcnab') it puts UPPER in there for me?"

I sure as hell hope not. I know this is going to shock you, but even activerecord is flexible enough to let you define whatever searches you want, and sane enough not to produce incorrect results by default. If you will only ever search last name in a case insensitive manner, then define find_by_last_name to use upper() or lower(). One of the features people like in an ORM is the ability to let you make customizations like this in a single place instead of all over the place. You still have to tell it what to do, its not magic.

"Aaaah you are one of those people are you."

"those people"? You mean "those people" who use the right tools and techniques for the task? Yep. You mean "those people" who aren't stupid enough to think "littering your code with SQL" and using an ORM are the only possibilities? Yep. Do you normally make stuff up and then fall back on "those people" bullshit when called on it? Or did you mean to say "PHB approved industry standard practices" and "modern" was a typo? Or do you seriously think decade and a half old techniques are modern? Welcome to the 1990s, enjoy your stay.

"You know this conversation reminds of so many I had with postgres users. I remember when postgres didn't support alter table and dickwads like you were saying "there is no need for alter table just use this workaround". I remember when posgtres didn't support outer joins (I do believe it was the last database in the world to support outer joins) and dickwads like you were saying "there is no need for outer joins just your this query instead"."

No, dickwads like me were not using postgresql because it was a pain in the ass, exactly because of the lack of alter table and a few other annoying things that needed to be "worked around". Just because I don't think your fabricated problem is a real problem doesn't mean I think alter table is useless.

"Last to implement alter table, last to implement outer joins, last to implement replication, last to implement clustering, last to implement failover, last to implement case insensitive collations. Due in large part to dickwads like you."

No, due exclusively to the nature of open source projects. The developers impliment what the developers want. If you share their needs, lucky you, you can use their project. If you don't, then don't use it. Pretty simple huh? Rather amusing that you rag on postgresql for being "last to support" stuff though. If you wanted to use postgresql for real, this would not be an issue. You would either use upper() or lower() like you should, or you would install a case-insensitive data type. You just want to whine about postgresql. Feel free to do so, but its not going to suddenly make you important, and its not going to suddenly make postgresql developers do your bidding.

Re:Let's fork it!

[killjoe]

"Implimenting collations correctly is more involved than you are pretending."

Apparently it is very difficult for the postgres developers. No problem for oracle, firebird, mssql, db/2, mysql programmers though. Collations are so fucked up in postgres most languages don't sort properly. Go and search the postgres mailing list for all the complaints.

""those people"? You mean "those people" who use the right tools and techniques for the task? "

I mean those people who rant against OO programing and ORMs.

"If you wanted to use postgresql for real, this would not be an issue."

I will use postgres if and when it has the features it I need and want. So far whenever I have started a project postgres has not been good enough. It's still not good enough. So I use databases that have the features I want.

"You just want to whine about postgresql."

No I am trying to explain to the dickwad fanbois why people don't use postgres and choose mysql instead. Mysql has more features then postgres, features people want. Postgres has lots of features nobody really cares about like *~

Re:Let's fork it!

[Generic Player]

"I mean those people who rant against OO programing and ORMs."

No, I simply pointed out that there is nothing modern about it. I personally don't use ORMs because if I wanted to work in an OO style I would use an object database instead of a relational database. Just because I have common sense doesn't mean I "rant against" OO programming (which I use all the time) or ORMs (which have their place). Do you really want attention so badly that you make up random nonsense to try and keep arguments going just so you have someone to talk to?

"I will use postgres if and when it has the features it I need and want."

Like I said, if you wanted to use postgresql you would. You are inventing the most trivial problem, which has several trival solutions. Clearly you don't want to use postgres, so don't.

"No I am trying to explain to the dickwad fanbois why people don't use postgres and choose mysql instead. Mysql has more features then postgres, features people want. Postgres has lots of features nobody really cares about like *~"

You are a megalomaniac. The universe does not revolve around you. Just because you invent excuses not to use postgres, that doesn't mean all the people choosing mysql are doing it for your invented reason. And just because someone disagrees with your opinion, doesn't make them dickwads or "fanbois". You can bitch about this for as long as you want, on as many websites as you want, and get laughed at by as many people as you want, but that will never make your bizzare self absorbed fantasy become reality. Most mysql users don't even know that postgresql doesn't have this, I doubt its affected the choice for even .1% of mysql users.

Re:Let's fork it!

[killjoe]

"No, I simply pointed out that there is nothing modern about it. "

Well other then the fact virtually every program written today is written using objects there is nothing modern about it all. What a stupid thing to say. You really thought you were trying to say something profound with that?

"I personally don't use ORMs because if I wanted to work in an OO style I would use an object database instead of a relational database."

Ah you are one of those people too.

"Just because you invent excuses not to use postgres, that doesn't mean all the people choosing mysql are doing it for your invented reason."

No they are using for their own reasons. There are features they want the mysql has and postgres doesn't It's not just one feature, I pointed out several that postgres doesn't have, there are dozens more. There are lots of features people want and postgres doesn't offer them so they choose a different database. That's why postgres has such a tiny percentage of the database market. It lacks a lot of really important features people want.

"And just because someone disagrees with your opinion, doesn't make them dickwads or "fanbois"."

No it doesn't. It's how they disagree that makes them dickwads and fanbois. Remember when postgres didn't have alter table? Go back and read the posts about how dickwads (like you!) gave workarounds and pretended that the only people who wanted alter table were stupid and how they were a small minority etc. Those dickwads were making the exact same arguments you are. Those dickwads were wrong. Eventually even the stubborn postgres developers implemented alter table thereby making the dickwads the joke of history. Postgres will one day clean up their collation mess. They will one day implement a case insensitive collation thereby joining every other database in the world. That day you too will be a joke of history and this thread will one day be resurrected by some people looking for an example of how close minded dickwad-ish postgres users are.

"Most mysql users don't even know that postgresql doesn't have this, I doubt its affected the choice for even .1% of mysql users."

Oh I keep forgetting. Mysql users are stupid, they are unable to compare features of databases. Postgres users are smart that's why they don't think outer joins are such a great idea.

Re:Let's fork it!

[Anonymous Coward]

This is a wonderfully entertaining flamewar. I'm on the side of the Generic Player guy. killjoe, you've missed a fundamental point: you can have the source to postgres if you want it. If these features are important to you (and the legions of others you refer to), then you and they can implement the features yourselves.

Re:Let's fork it!

[killjoe]

"If these features are important to you (and the legions of others you refer to), then you and they can implement the features yourselves."

Adding asynchronous multi master replication is not a trivial task. I could understand your point if it a minor feature though. Even then it's easier to use something else then to hack postgres code. People who want those feature end up taking the path of least resistance which is to use mysql or firebird or oracle or whatever.

Oh well

[0racle]

Is it really a problem? If you worried about support wouldn't you be using a distro that also offers support contracts?

Re:Oh well

[morgan_greywolf]

Canonical, Ltd. offers support contracts for Ubuntu, but MySQL won't support anything but Red Hat or SuSE.

Re:Oh well

[mr_mischief]

Maybe Canonical should step up and offer MySQL support on Ubuntu.

Re:Oh well

[MECC]

Can you buy support from Progeny if you've got a Debian system? I looked at their site, and it looked as though they want to build a system for you, then support it.

Re:Oh well

[John Hasler]

There is plenty of support available for Debian.

Re:Oh well

[Anonymous Coward]

This is a problem. The shifting to Enterprise Only support is destroying free software implementations in business. I just bought a pair of dell servers, and they are set up so the only software I can install on them is Server 2003, RHENT, or SuSE Enterprise. The bios will only load dell signed disks. If this causes a "tiering" of the Free Software community then everyone looses.

Re:Oh well

[Da_Weasel]

You bought servers that would only allow you to boot signed disc? That was dumb!

Re:Oh well

[Doctor Memory]

And does (can?) anyone offer enterprise-level Debian? Canonical offers "enterprise-level" support (whatever that means, apparently guaranteed response times [note that "You don't generally see behavior like that in an enterprise-grade appliance" is, technically, a response]). But has anyone certified its performance on specific hardware? I see that HP offers Debian support "packs" (as well as installation assistance), but doesn't certify it on any of their servers.

Maybe I'm off track here, but I could certainly understand MySQL not wanting to offer an enterprise-level product for a platform that wasn't also enterprise-level.

Re:Oh well

[Nutria]

You bought servers that would only allow you to boot signed disc? That was dumb!

No, in some facilities it's smart. Why? It prevents unauthorized personnel from booting with a live CD they bring from home.

Re:Oh well

[ultranova]

Maybe I'm off track here, but I could certainly understand MySQL not wanting to offer an enterprise-level product for a platform that wasn't also enterprise-level.

Is MySQL "enterprise-level" nowadays ? Every time there's been a story about databases, people have told horror stories about MySQL quietly corrupting data in database.

And just what does "enterprise-level" mean, anyway ? Scales to infinity ? Reliable ? Costly ? Doesn't get the IT manager fired when the CEO find out he bought it ?-)

Re:Oh well

[Anonymous Coward]

So they rip the hard drives out and stick them in another box, and your "signed boot disk" is still useless.

Re:Oh well

[turbidostato]

"No, in some facilities it's smart"

No; it's never smart to limit yourself to whatever nuisances a third party might inflict to you.

It might be smart to have systems that won't allow but *your* signed boot CDs, but this move (if true, I really doubt it) can only be smart... if you are Dell Corp.

Re:Oh well

[rainman_bc]

No, in some facilities it's smart. Why? It prevents unauthorized personnel from booting with a live CD they bring from home.

Er - why does unauthorized personnel have access to the server anyway???

Re:Oh well

[xenocide2]

I'd imagine they wouldn't turn away money. Whether you two can agree to how much is another question altogether.

Re:Oh well

[Anonymous Coward]

Every time there's been a story about databases, people have told horror stories about MySQL quietly corrupting data in database.

MySQL is terrible about this. You can have the database shutdown gracefully but start it up again only to find whole columns of data completly corrupted. Not to mention how MySQL often breaks compatibility between minor security releases often causing data to be corrupted or completly destroyed because of it.

Re:Oh well

[Nutria]

Er - why does unauthorized personnel have access to the server anyway???

Not every computer is in a secure vault.

Those mother...

[Unoti]

Clearly we need to get some tough mother forkin programmers on this...

Re:Those mother...

[sbrown123]

Nah, just use PostgreSQL .... or sell your own MySQL Debian support :)

Bit misleading

[Anonymous Coward]

MySQL (the database) still works with Debian, but MySQL (the support company) no longer sells support for Debian.

Re:Bit misleading

[McDutchie]

MySQL (the database) still works with Debian, but MySQL (the support company) no longer sells support for Debian.

For medium and large companies (which are the only entities that would buy support to begin with), that difference is purely academic. If it isn't supported, it isn't worth running.

Re:Bit misleading

[Anonymous Coward]

And if there were a significant market for Debian in the enterprise, I doubt they'd stop supporting it. Most of the places I've worked have used Red Hat, because it has an enterprise reputation.

Re:Bit misleading

[xantho]

Yeah, it's great for name recognition that the two are named the same, but for getting information out like this, it's really confusing to have then named the same. You have to play stupid word games like saying, "The MySQL Project," and "The MySql Company," or something equally inane.

Re:Bit misleading

[dsci]

Most of the places I've worked have used Red Hat, because it has an enterprise reputation.
Point of clarification: places have RH because they offer support to their enterprise product. Debian's reputation for stability and such is pretty strong, but that only carries so far in the business setting. It's not reputation that drives RH over Deb to the enterprise...it's "I can pay YOU to fix it when it's broke." JMO.

Re:Bit misleading

[Matey-O]

I tend to avoid Redhat as they gave me the impression that they were the Microsoft of Linux. I loved Suse, then they got bought by Novell. I moved to Debian as they allowed me to install a bare system in 300 mb with text only and install JUST the parts I needed. Now I get to look for something else? (actually, it doesn't look like it as I haven't needed paid support for MySQL yet.)

Re:Bit misleading

[stefanlasiewski]

If it isn't supported, it isn't worth running.

If that's the case, would they be running Debian?

I would argue that most Medium businesses are fine with using unsupported technology, especially for non-mission-critical or non-user-facing services. Debian and MySQL might work perfectly fine in some situations. In my experience, support for many smaller commercial products is worthless 95% of the time.

Re:Bit misleading

[walt-sjc]

If you need a support contract for Linux, it's because you have A) incompetent system administrators or B) insufficient sys admin staff. Now that is not necessarily a BAD thing... "Incompetent" doesn't mean that you are an idiot, it just means you don't have the required level of skills needed to to what you are doing. Small businesses for example may not have the budget for a "Real" senior system administrator who has the skills / time to solve more difficult problems. But hey, everyone needs to start somewhere... You don't become a senior sys admin overnight.

This is a big "ditto" for MySQL. No point in buying a license unless you NEED that certain level of support, and dropping support for Debian won't mean a thing in the grand scheme of things. Most people that use Debian use Debian packages and not MySQL versions anyway. Debian has their own support channel, and who says that it's worse than what you get directly from MySQL?

Re:Bit misleading

[epiphani]

Well, I'm pissed off by this. We were looking at buying top-end mysql support shortly. Now we cant, because we run a home-brew version of Linux.

We're technically competent people, but we don't know MySQL inside and out. We wanted support so we could go to mysqlab and present them, the MySQL experts, with some of the problems we have and we could work WITH them to fix them. Now, instead of being able to go to the developers, and PAY them for their time, we're stuck on our own trying to figure things out. This is just dumb, they're throwing out a lot of revenue. I know several large companies that do not run Suse or Redhat, but make heavy use of MySQL. They're cutting off the group of large companies that maintain their own linux distributions in house. There are a surprising number of us.

Re:Bit misleading

[Omnifarious]

I've always wondered where the "Microsoft of Linux" thing came from. They don't seem to be at all like Microsoft to me, so the comparison makes no sense to me.

Re:Bit misleading

[jellomizer]

Get one system with Red Hat. Put MySql on it get the suport. If the problem doesn't work on red hat and your own. Then call them up and tinker with the one RedHat box until it works and do the same on your box. Supporing every Linux Distro is disasterious for a company. To many of them all with their own quarks it make offering support near impossible. By sticking to a few Distros they can quickly figure out if it is an OS Problem or a MySQL problem.

Re:Bit misleading

[Anonymous Coward]

Which also means that you can't depend on MySQL to WORK on nonsupported OSes - regardless of whether you want to pay for support or not. They won't put any development or testing effort into nonsupported systems.

I have one acronym for you all to consider: LAPP (Linux Apache POSTGRESQL PHP)

Re:Bit misleading

[AmigaBen]

If you need a support contract for Linux, it's because you have A) incompetent system administrators or B) insufficient sys admin staff.

If you make a statement like this, it's because you have A) no experience in a real company or B) you do but are still naive as to how your company actually works.

Re:Bit misleading

[Anonymous Coward]

DUDE!!

They sold out to the THE MAN!!
Anytime anybody has money it is automatically considered greed or stealing from everyone.

Where have you been the last 4 years of intense modding of rubbish stuff like that.

Re:Bit misleading

[jedidiah]

Having worked in "real companies", I would say that any environment that would tolerate MySQL would tolerate Debian for exactly the same reasons. No form of MySQL is considered "supported" when compared to things vendors as Oracle or IBM. Insisting on getting a "supported enterprise distribution" to run something like MySQL is simply assinine.

            Some of the bigger companies that are big enough to dwarf nations might be that idiotic. However, many smaller and equally robust enterprises aren't going to buy into that sort of stupidity.

Re:Bit misleading

[modir]

The article here on Slashdot is a little bit misleading. You still can get support from them. Them main part is this:
Will you support MySQL Binaries built by third-party vendors? No.
http://www.mysql.com/company/legal/supportpolicies /policies-04.html#q04

The person who wrote this article wanted to take the binaries provided by Debian. And this doesn't work. But if you take the binaries from MySQL you should still get support.

Re:Bit misleading

[Anonymous Coward]

Mod parent funny. He wrote "For medium and large companies (which are the only entities that would buy support to begin with), that difference is purely academic. If it isn't supported, it isn't worth running."

Uh, for large customers they're the most likely ones to have the resources and the needs to manage their own distro. Think Google and Munich as two who are known to do this.

Re:Bit misleading

[Moofie]

It's pretty simple. The biggest player in any given space is the Microsoft (or GM) of that space. Having drawn the parallel, you can carry over all the preconceptions about the other space, and stop thinking right there.

It's apparently very comforting.

Re:Bit misleading

[walt-sjc]

Please provide proof to backup that statement.

Re:Bit misleading

[acidrain69]

I have experience in a real company using MySQL on Debian, and I'd say that's pretty accurate. I have the benefit of building the system from scratch myself; this is not some purchased solution. I have to do the research when something goes awry. Incompetent is kind of a harsh generality, but when you break down the word, it makes sense. If something breaks and I don't know how to fix it, I am at fault/responsibility, and thus deficient in some sense. Luckily, most management that I deal with understand that nobody knows everything, especially in the tech world. And I have proven enough times that if I don't know something, I can research it. I have not needed support on Debian, and I find sometimes that the software we use that DOES have support contracts, the support contract isn't always worth the paper it is printed on.

Re:Bit misleading

[Mathetes]

Seems to me that the Microsoft-Novell deal proves that Microsoft is indeed the Microsoft of Linux. So sorry, that title has been taken!

Re:Bit misleading

[Anonymous Coward]

"I've always wondered where the "Microsoft of Linux" thing came from"

Their ingineering/technical efforts are strongly subjugated to their marketing division.
They prefer dumb sysadmins/users to technically knowledgeable ones as clients.
They prefer selling their products to PHBs better than to sysadmins.
These kind of practices in the IT world were first succesfully used by, and attained a master level by... Microsoft.

Now you know where it came from.

Re:Bit misleading

[Nutria]

I've always wondered where the "Microsoft of Linux" thing came from.

Google is your friend.

http://linux.slashdot.org/article.pl?sid=02/08/27/ 0145209

Re:Bit misleading

[doom]

Omnifarious (11933) wrote:

I've always wondered where the "Microsoft of Linux" thing came from. They don't seem to be at all like Microsoft to me, so the comparison makes no sense to me.

No, it doesn't make any sense, unless Microsoft has started releasing all the code they write under the GPL.

I gave up on Red Hat because (a) I got tired of being burned by weak technical decisions and an apparent absence of any QA (circa RH 5-8) -- and my brief look at FC3 didn't impress me much, either; (b) I wasn't impressed with them suddenly dropping support for their server distro, announcing "linux isn't ready for the desktop" (too bad no one told SUSE that, eh?).

The pods are just sticking with RedHat because they've heard about them in the funny pages -- I mean, business pages. With an IPO that big it's got to be good.

(And if your IT staff is good enough, it doesn't matter what distro you start with -- a good sysadmin can create a Debian work-alike with any distro.)

Re:Bit misleading

[sd.fhasldff]

To many of them all with their own quarks it make offering support near impossible.

Yeah, quarks will do that to ya!

There has to be an "entanglement" joke in there somewhere, but I'll be darned if I can find it.

Re:Bit misleading

[Nutria]

Their ingineering/technical efforts are strongly subjugated to their marketing division.
They prefer dumb sysadmins/users to technically knowledgeable ones as clients.
They prefer selling their products to PHBs better than to sysadmins.
These kind of practices in the IT world were first succesfully used by, and attained a master level by... Microsoft.

When were you born? 1985?

Big Blue was (is?) the Past Master at subjugating engineering to marketing and selling directly to PHBs. Field Circus always likes smart system programmers, though.

Re:Bit misleading

[Nutria]

I have experience in a real company using MySQL on Debian, and I'd say that's pretty accurate. I have the benefit of building the system from scratch myself; this is not some purchased solution. I have to do the research when something goes awry. Incompetent is kind of a harsh generality, but when you break down the word, it makes sense. If something breaks and I don't know how to fix it, I am at fault/responsibility

Either this app is not critical, or management is foolish for depending on one person for a critical app.

Re:Bit misleading

[Anonymous Coward]

It sounds like Slackware might be a useful distribution for you to consider. The package management is nowhere near as good as Debian's, but it has the other options you like.

-M

Re:Bit misleading

[shmlco]

"Either this app is not critical, or management is foolish for depending on one person for a critical app."

How many businesses can afford to have a fully qualified individual on staff for every single piece of "critical" hardware and software they use?

How about hiring a automobile mechanic full-time just in case the company car breaks? A telcom engineer just in case the boss breaks his phone? A HP and Dell and Apple guy just in case a server or desktop goes down?

Maybe in a ideal world this would happen, but since NO company has an infinite number of resources, most hire what they need (or can afford) and contract out the rest.

Re:Bit misleading

[gbjbaanb]

sure. Get a job. :-)

Seriously, businesses are about money, and all of them will pay for support because paying a full-time dev to fix things will work out more expensive than paying the support fees. If you have a part-time dev, there's the chance that he will not be good enough (ie as familiar with the code/system as he needs to be) to fix things.

Then there's the risk you have to cover that your devs are on holiday when something breaks, or the risk that he just cannot fix it. eg. the company I work for has Oracle support so when the customer site broke (100% cpu usage for far too long), a call to Oracle resulted in things to look for and things to do to prevent it happening again. In terms of the cost of that downtime, the support fee was insignificant.

Also support fees are a business expense so the taxman pays for most of it.

Re:Bit misleading

[Omnifarious]

Hmm, well, those have never been how I define Microsoft. I've defined Microsoft by their abuse of their monopoly power and predatory relationships with their 'partners'. A close second has been the shoddy nature of their products, since they don't have to actually make a decent product in order to sell it.

I don't think any of those things are true with RedHat.

Re:Bit misleading

[Anonymous Coward]

Red Hat have not sold out to anyone..

But Novell has. Novell sold out big time. Do I trust Novell? Not anymore.

Do I trust Red Hat? They have a very good history.

Re:Bit misleading

[Matey-O]

I switched from Redhat because they were the 'defacto' Linux for business. As such, they behave like the 800 lb. gorilla of the group. "Our product only works with redhat, and only if you patch and recompile the kernel. " = no thanks, that's a support nightmare where you end up 8 versions behind when you can't compile a new kernel. "We have a wonderful GUI for administration" = It works great...except when it doesn't. Then I get to muddle about the conf files. If I've gotta do that, I'll just start there. "We're marvellously responsive with security fixes...but only if you subscribe. And the Licensing is just about the same price as Microsoft. (This was a good 6-8 years ago. So I switched to SUSE. SUSE had GREAT QA. Yast worked as it should AND worked identically so in a text only environment. That let my IDS boxes spend more time looking at packets, rather than drawing X11. They had a free package management and patching system. Then Novell bought them and had to 'monetize' the purchase. Good bye security updates. I flirted with Gentoo a bit, but never felt completely confortable there. I played with BSD, but it became obvious that low level stuff is DIFFERENT than Linux. I could learn it, but I'd be starting from scratch for the stuff I was doing. Debian was _stable_. It wasn't on a constant upgrade cycle, has a great package management system, and security updates will always be free. Debian is a forking point for a BUNCH of distros...ever wonder why that is? Are you happy with my bigger assessment?

Re:Bit misleading

[Nutria]

How many businesses can afford to have a fully qualified individual on staff for every single piece of "critical" hardware and software they use?

That's why companies try to standardize on a single platform. Single out-lying systems are a liability if the only person who can support it gets another job or is hit by a bus.

Our company has been using OpenVMS for 13 years, and it works well. But we're phasing it out, even though HP is still releasing new versions, adding new features. Why? Hard to find qualified personnel. Only greybeards want to work on it, and they are either (a) retiring or (b) see the writing on the wall and have retooled their skillsets.

How about hiring a automobile mechanic full-time just in case the company car breaks? A telcom engineer just in case the boss breaks his phone? A HP and Dell and Apple guy just in case a server or desktop goes down?

Never heard of consultants per-incident charges and support contracts?

Re:Bit misleading

[walt-sjc]

From your "proof", it's clear that you don't know what you are talking about.

Oracle is not open source and therefore totally fails as an example in this case. This is about supporting open source products - not closed source.

First, you don't need a dedicated developer. A senior sysadmin has all the developer skills needed. Competent senior sysadmins can recompile and patch the linux kernel in their sleep, and a good many are skilled enough to even write device drivers. In fact, the best system administrators actually write a lot of code, and were former developers who wanted more of a challenge. They are generally compensated better than developers to reflect their level of knowledge (and the more non-traditional work hours.)

  A support contract from RedHat or MySQL isn't going to get you a temporary sysadmin (or developer) to work the holidays / weekends. If you think RedHat is going to pull in Alan Cox on Christmas eve to write a patch for you just because you have a support contract, you are out of your tree. A support contract is no guarantee that you will actually get your problem FIXED on a weekend / holiday.

If you run a 24x7 shop like I do, you need coverage at all times. This means a staff of at LEAST 4, or if you are REALLY cheap and overwork your people, you can get by with 2. When you have a problem, you are going to have one of those sysadmins working WITH the company (Sun, Oracle, RedHat, or MySQL) ANYWAY until the problem is solved (remember, you have to repeat your troubleshooting steps that you already did all over again for the first-level support person who won't be able to solve your problem anyway, and finally passes you off to second level. Been there, done that.)

Consider a largish infrastructure of HUNDREDS of servers - your licensing costs for RedHat (for example, as their quantity discounts are near zero) will cost you upwards of $500K / year (that was a quote we got.) For about 1/3rd of that, you can hire another top-notch sysadmin who can be the dedicated internal support person. Chances are that you won't NEED another dedicated person as you hire competent people that can solve the problems on their own anyway, and you save your company $500K - $500K that can be used for something else (thus destroying your tax savings argument.) Now with smaller companies (as I stated,) it may make sense because the cost of support is so low.

Clearly, you don't understand the issue as you haven't lived it. You seem to think that it costs more to internally support MySQL and Linux (RedHat) than a support contract costs because you don't have any experience to the contrary. I do.

As a final nail in your baseless "theory", do you really think Google would be better off buying support for MySQL and Redhat than rolling their own as they do? Not a chance.

Re:Bit misleading

[Omnifarious]

*nod* That analysis has some meat on it and makes sense. But I still don't think that means that RedHat is the Microsoft of Linux. :-)

I've used Ubuntu and I was extremely disappointed. Important security updates were delayed for the next release. For example, I made extensive use of ethereal at my last job and when it changed to wireshark along with a whole slew of security updates I waited in vain for it to show up in Ubuntu. It even showed up in fink. But it never made it into Ubuntu until the next release of the entire distribution.

And the same thing with Mercurial. And the packaging system was so arcane and ill-documented that creating my own Mercurial packages from my personal development branch wasn't something I ever figured out how to do.

I've considered switching to Suse because I really hate RedHat's attitude towards ReiserFS.

I run several headless server boxes in text-only mode under RedHat. And while I feel that I have way more software on them than I probably need, they aren't running X, and I've been able to work with them just fine without it.

But experiences and tastes vary. I'm glad that you've found something you like even if it isn't what I like. And the fact the I can write software you can use despite the fact we use wildly different distributions is the reason I can't bring myself to think of RedHat as being anything like Microsoft.

Re:Bit misleading

[TeraCo]

A support contract is no guarantee that you will actually get your problem FIXED on a weekend / holiday.

Maybe not with a redhat contract. With Microsoft, Oracle or Cisco, a support contract means I'll have engineers working around the clock with me (from all around the world in the case of the TAC) until it's fixed.

Re:Bit misleading

[Jamesday]

'As a final nail in your baseless "theory", do you really think Google would be better off buying support for MySQL and Redhat than rolling their own as they do? Not a chance.'

Ahem: http://www.mysql.com/customers/customer.php?id=75 . Customer = paying MySQL.

If they are a major player, they probably have a support contract with MySQL. That is not a substitute for excellent in-house people when you get big needs but it is a useful supplement to them. Google isn't spending the money just to get a tax write-off.

Wow... this is the beginning of the end

[jarich]

I've used MySql in a number of different companies. In many cases knowing that they could buy a support contract let me bring in MySql. I also really like Kubuntu these days.

I guess it time to dig in and learn another tool to replace it.

Re:Wow... this is the beginning of the end

[Professor_UNIX]

I guess it time to dig in and learn another tool to replace it.

Meh, I'd rather replace MySQL than my Debian distribution. If I was truly concerned about commercial support for my database then I'd buy a commercial database like Sybase or Oracle. People use MySQL because it's free, not necessarily because it's better, or even comparable, to commercial offerings.

Re:Wow... this is the beginning of the end

[jarich]

Meh, I'd rather replace MySQL than my Debian distribution.

That's what I meant. Sorry for the ambiguity. I love Kubuntu and wouldn't ditch it.

If I was truly concerned about commercial support for my database then I'd buy a commercial database like Sybase or Oracle.

I'm not concerned about the support. I've never needed it... but I've found it makes C level execs (especially at startups) feel better about a tool if they know there's someone out there they can throw money at and "get help".

People use MySQL because it's free, not necessarily because it's better, or even comparable, to commercial offerings.

Not me... I like MySql. :) It's fast, got enough features, lightweight... and the most important feature for a tool... I already know it.

Re:Wow... this is the beginning of the end

[Midnight Thunder]

We just need Debian enterprise or Kubuntu enterprise to have $$$ support. Then again given the amount of helpful people around not wanting a dime for the help they provide, the only people giving a damn are those people paying for an enterprise version of Linux.

Re:Wow... this is the beginning of the end

[dsci]

Then again given the amount of helpful people around not wanting a dime for the help they provide, the only people giving a damn are those people paying for an enterprise version of Linux.

Community support is a great thing, and hopefully all of us that USE F/OSS software give back to that in some way. But the business world, and many individuals, operate on the principle of "you get what you pay for." Most of the time this is a good guideline, but F/OSS is an exception. There are QUALITY products out there, and quality support, available for no upfront financial cost. But in the minds of many business types, if you pay nothing, it must be worth nothing.

(car analogy to follow)

Think about it this way; would you take a FREE car without ANY suspicion that there's something wrong with it? Perhaps, if you knew the seller and trusted him. You and I trust the seller (the OSS community) to provide good products and services, but the average PHB does not know this community - he cannot trust his enterprise with such an unknown.

Another way to put it is that you and I can see the VALUE, independent of price, of OSS, but many others don't. They associate the value with the price tag. Without PAID support, the support is worthless.

Re:Wow... this is the beginning of the end

[Locutus]

My first thought was how strange this was considering how Shuttleworth is trying to move Ubuntu(Debian based) to the server( Ubuntu 6.06 LTS-long term support ) amongst the other Debian distro's out there. I wonder why the had originally included Debian for MySQL Enterprise Support in the first place since the business case is still growing.

My second thought was to some old news that MySQL Inc was partnering with Microsoft and are they getting any 'incentives' to move away from anything but Novell Suse products. Debian being the first to go.

But, this is MySQL Enterprise Support we're talking about and you won't find as much enterprise support for Debian, or at the same scale, as you'd find for RedHat or Suse. It is a shame though and I hope their "partnership" with Microsoft is not an issue here. If it is, I suspect we'll next hear of another deal with Microsoft and the dropping of RedHat Enterprise support. Not necessarily in that order. IMO.

LoB

Re:Wow... this is the beginning of the end

[chris_mahan]

> But in the minds of many business types, if you pay nothing, it must be worth nothing.

If that is the case, why isn't every single computer on the planet an IBM?

Oh. Wait. Maybe there is such a thing as Do-It-Yourselfers who prefer a FOSS + "elbow grease" approach.

Let me see... Can you say GOOG?

In any case, I see projects moving off mySQL and onto Postgres, or at least adding postgres support. I also see projects adding or using sqlite as an alternative to mySQL.

Re:Wow... this is the beginning of the end

[Anonymous Coward]

>> But in the minds of many business types, if you pay nothing, it must be worth nothing.

> If that is the case, why isn't every single computer on the planet an IBM?

because "many" != "all"

Re:Wow... this is the beginning of the end

[JayAEU]

My first thought was how strange this was considering how Shuttleworth is trying to move Ubuntu(Debian based) to the server( Ubuntu 6.06 LTS-long term support ) amongst the other Debian distro's out there. I wonder why the had originally included Debian for MySQL Enterprise Support in the first place since the business case is still growing. While I appreciate what Mr. Shuttleworth is doing with Ubuntu, I think those not using RedHat or SuSE will go with the traditional choice, Debian proper.

I'd be hard pressed to come up with more stable alternatives when choosing a distribution for a Linux server. In any case, I think not offering support for any Debian based distributions is a bad idea, mostly because there's so many of them. This really is lots of revenue they'll be missing out on.

Re:Wow... this is the beginning of the end

[chris_mahan]

The GP is making the comparison to a free car. "There must be something wrong with it" attitude that most people have.

Business people in general are adverse to risk; If they can't demonstrate a fairly sure path to profitability, they won't go down that path.

What I am saying is that the "next big business" takes risks.

Re:Wow... this is the beginning of the end

[Locutus]

But I think that without a large organization pushing the distro( RedHat, Suse, Canonical, etc ) it is likely that the enterprise users are just going to be confined to smaller department users. Granted, the overall numbers might be larger but a company like MySQL AB can't make a profit supporting thousands of 5-10 user departments like they can supporting 1,000+ user installations.

I agree with you on were Debian fits and its capabilities but from a financial standpoint, going where the larger "bubble" of users are just makes cents.

Canonical and Linspire do put a public face behind Debian-based distros, and Canonical is pushing for the server market, but otherwise, it's like killing ants with a beebee gun. Not very efficient.

LoB

Solution

[Shawn is an Asshole]

Loudly drop support for MySQL. Here are two excellent alternatives:

PostgreSQL
Firebird

Still, Debian provides good MySQL packages. Use them instead. If you need support, I'm sure you could find someone to provide it for you.

Re:Solution

[PatrickThomson]

for "support" read "liability when it all breaks". That's what linux support is really all about. Would you want to be a technician personally responsible for downtime and several million of lost sales? Your bosses won't let it happen, because obviously you can't pay it back.

Re:Solution

[Anonymous Coward]

Ahh, the good old "who do you sue" chestnut. How's suing Oracle working out for you whenever you find bugs in their database, or if you got bad advice from their support techs?

Re:Solution

[virtual_mps]

you think that mysql support will buy unlimited legal/financial liability for costs incurred by downtime of your mysql installation?

really?

seriously?

hahahahahahaha

What your support contract buys you is the ability to call someone on the phone. If it makes your boss happy to have someone to call and yell at when shit breaks, well, ok.

Re:Solution

[Anonymous Coward]

I could, but it would make basically no change for us (because we're not in the business of selling dinky php scripts for blogs). 99% of our customers use and want MSSQL/Oracle/DB2, and the few left mostly want PostgreSQL. I doubt we'd get a single request for supporting MySQL.

Re:Solution

[Thomas Charron]

I have never in my entire life seen a softare company held financially liable for lost sales as a result of a database failure. Please, feel free to cite one single lawsuit if you can find one.

Re:Solution

[peragrin]

You know in the software industry that is a bunch of bullshite.

If that were true then MSFT wouldn't have any money at all as they would be responsible for billions in lost sales annually. Just one Virus through one product line(not even windows but MS SQL) a year would be expensive. Yet MSFT doesn't have to pay so why would Mysql, or IBM, or any other software company for lost sales or data?

Re:Solution

[saleenS281]

And that would buy you what? Neither of those solutions are offering enterprise support for Debian either...

Re:Solution

[Anonymous Coward]

Firebird? I thought they changed the named to Firefox?

Re:Solution

[Shawn is an Asshole]

So, how successful are you from getting money for downtime from Microsoft when a computer gets a virus? Or breaks due to an update?

Re:Solution

[spiritraveller]

for "support" read "liability when it all breaks". That's what linux support is really all about. Would you want to be a technician personally responsible for downtime and several million of lost sales? Your bosses won't let it happen, because obviously you can't pay it back.

Any system has the potential to cause damage if it goes down. But this doesn't create "liability" for the company's support staff. Being the support provider does not automatically make you liable for any problem. But it does make you the responsible party... the person that they can turn to to fix their problem. That's called "having a job."

Some companies are big enough to keep their own support staff on hand. Other companies need to outsource it to companies like Red Hat and MySQL AB. But Red Hat does not become liable for your downtime just because they are your support provider. You can be sure that they their support contracts state that will NOT be liable for resulting damages of lost sales and the like.

The same is true of Microsoft's EULA and pretty much any software that you buy or use and any service that you purchase. Even if it is not written in the contract, it doesn't matter, because contract law does not presume liability for resulting damages, only direct damages.

Re:Solution

[Anonymous Coward]

Why not stop using Dweebian? Its uncool.
Here are two excellent alternatives:
Redhat Enterprise
SUSE Enterprise

Re:Solution

[kilgortrout]

Have you ever read the GPL or any EULA? I have yet to read any software license, whether blessed by the FSF or not, that does not contain disclaimers in bold type for any significant consequential damages, eg. lost profits.

Re:Solution

[Sxooter]

This says otherwise.

Re:Solution

[saleenS281]

Those are a bunch of third party companies, and a search of that page returns nothing for "debian". As a poster above has said, nothing is stopping ANYONE from opening up their own third party company to support mysql on debian.

Re:Solution

[giorgiofr]

I think the virus writer is responsible for the damage caused. Otherwise, it'd be like blaming gun manufacturers instead of murderers. Or knife makers, for that matter.

Re:Solution

[Sxooter]

All support for PostgreSQL comes from third party companies. PostgreSQL is a community project not a commercial company.

Dozens of the companies there support Debian. Command Prompt and Enterprise DB for a couple of examples.

The problem with MySQL and support is that you have to buy a commercial version if you write non-open source code for it, because of the GPL libs. so, buying support from a 3rd party doesn't cover you.

Not so with Firebird or PostgreSQL where you don't need a commercial contract to write commercial software.

Re:Solution

[SamTheButcher]

Well, shit. If that's all you want/need, I'll sell you my "support" for 3k/year. Whenver anything breaks, call and yell at me.

Hell, I'll even stay with you and help you google the solution or search through the MySQL manual! And as an added bonus, I'll send you a basket of Chocolate and Cheese around the holidays as a thank you for choosing MyButcherSQLSupport.

Re:Solution

[styrotech]

Ahh, the good old "who do you sue" chestnut. How's suing Oracle working out for you whenever you find bugs in their database, or if you got bad advice from their support techs?

That's not the reason they want it. They want it so that if a screwup happens and the shareholders etc are baying for blood, they can divert the blame to the vendor.

Re:Solution

[Anonymous Coward]

If it makes your boss happy to have someone to call and yell at when shit breaks, well, ok.

Hmm.. my boss yells at me. And it does make him happy. I'm in the wrong business.

-M

Re:Solution

[j79zlr]

In America we do blame the gun manufacturers, knife makers, etc. Sue everyone and let the courts figure it out.

Re:Solution

[saleenS281]

And where in that commercial license does it say you can't use debian? Again, nobody is stopping you from using it and paying a third party for support. You never needed a commercial contract to write commercial software for Firebird or PGsql, this changes nothing....

So, Debian users can't...

[Rastignac]

So, Debian users can't COUNT() on MySql because they DROP support...
SQL can be funny.

Generic, huh?

[Just Some Guy]

I guess that's fair - my company migrated to supporting only "generic Red Hat Database", aka PostgreSQL.

Seriously, except in cases where you have no choice about database availability, I can't see a single reason to use MySQL these days. All of their cool features are owned by their competitors, and they're starting to pull desperate financing tricks like whittling away tech support and partnering with SCO. Are people still using it for new deployments, and if so, why?

Re:Generic, huh?

[LWATCDR]

Simple. Every nickel and dime hosting company uses MySQL so every CMS blog, and forum supports MySQL.
Up to and including Slashcode.
It is now catch 22. Everybody uses MySQL because everyone uses MySQL.
Heck I use MySQL for our CMS because not every module supports PostgreSQL.
I would much rather use PostgreSQL for everything but I don't have time to re-invent the wheel.

Re:Generic, huh?

[Just Some Guy]

Gotcha. Maybe I should have changed that to "for new development", as in starting-from-scratch projects where there are no database dependencies already in place. Is anyone still using MySQL in those situations?

Re:Generic, huh?

[syphax]

Are people still using it for new deployments, and if so, why?

Inertia and familiarity. Fortunately, the barriers to switching aren't all that great, so it'll be easy enough to jump ship when/if needed. I'm one of the 80-90% of users that just needs some tables and industry-standard SQL; for almost all of my needs, I could use just about any DB backend. Yes, I know, YMMV.

Re:Generic, huh?

[Anonymous Coward]

What part of "Every nickel and dime hosting company uses MySQL" did you not understand?

Re:Generic, huh?

[Anonymous Coward]

Yes, everyone

Re:Generic, huh?

[grahamm]

It is now catch 22. Everybody uses MySQL because everyone uses MySQL.
Everyone used WordPerfect, that is until almost overnight everyone was using Word.

Re:Generic, huh?

[Anonymous Coward]

MySQL has a lot of name recognition among the pointy-haired crowd. That alone will keep it going for years.

Re:Generic, huh?

[Jay Pipes]

Sorry to rain on your parade of myths, but this: "availability, I can't see a single reason to use MySQL these days. All of their cool features are owned by their competitors, and they're starting to pull desperate financing tricks like whittling away tech support and partnering with SCO" is nowhere near true. If by cool features, you mean InnoDB, sure Oracle does own InnoDB, but the largest growth MySQL has seen is in the area of scale-out replication, telecom's usage of MySQL Cluster, and in the expansion of pluggable storage engine partners. MySQL Replication and MySQL's NdbCluster aren't owned by anyone other than MySQL, and the growth in storage engine partners shows you that a lot of companies believe MySQL's growth in the commodity scale-out market is because of something substantial, not just from a "whittling away of tech support". Contrast this to GreenPlum owning the (not open source) Bizgres MPP clustering project, to EnterpriseDB's (not open source) Replication Server "fork". Please. Quit propogating complete fabrications. If you want to argue that PostgreSQL is more "Oracle-like" in its SQL-implementation, fine. But don't troll away with flames that don't have a lick of truth in them.

Re:Generic, huh?

[LWATCDR]

Unfortunately someone already replied to you in a rude and less than helpful way.
Yes they are because they want people to use what they are developing.
Just about every CMS, Blog, or forum will support MySQL first because it is available from the majority of web hosting services.
PostgreSQL isn't as available so it will play second fiddle for a long time. Frankly it is almost the same situation that we have with Windows. MySQL probably isn't technically the best solution but it is the most common solution so that is what you work with.
Add in that when people think of web development they think LAMP and the M is for MySQL. The P now means Perl, PHP, or Python but it used to just mean Perl or was it PHP?
Add in that MySql is used in most of the example code and books for Web development.
Add in that Bindings for MySQL are available for just about every language.
Then add in the really big fact that for most CMSs, Blogs, and message bases MySQL is good enough.
Add the fact that everybody and their dog kind of knows MySQL all ready and you can see that you are fighting a lot of inertia.

On a positive note a lot of small applications that used to require MySQL for no good reason are now using SQLLite. I always found it kind of dumb to run a SQL server to keep track of a CD collection.

Re:Generic, huh?

[Just Some Guy]

I apologize if you (mistakenly) thought I was trolling. However, the reality of the situation is that MySQL does not own the rights to its primary feature-laden backend. If Oracle decided to pull the plug on InnoDB - or worse, shift its direction so that it retains developer mindshare but becomes less useful for MySQL - then what's your fallback plan? Revert back to telling would-be users that ACID, stored procedures, etc. are unimportant and that they belong in client code?

If you have other plans, by all means let us know! In the mean time, the thought of building business logic on MySQL is growing scarier by the month. You can write me off as a troll if you want, but I think you'd be better off publically addressing these issues and making a liar out of me.

And one free bit of advice: for God's sake, distance yourself from SCO! That affiliation by itself is an enormous dealbreaker for a lot of people.

Re:Generic, huh?

[Just Some Guy]

Add in that Bindings for MySQL are available for just about every language.

Caveat: those bindings link against GPLed libraries. It's not possible to use MySQL as a backend to proprietary applications without shelling out some cash. Whether that is good or bad is another issue. Note that even Oracle allows restribution of their client libraries under those conditions; this restriction seems to be unique to MySQL.

Re:Generic, huh?

[Dewin]

Everyone used WordPerfect, that is until almost overnight everyone was using Word.

You can open WordPerfect documents in Word and they work. (Disclaimer: I haven't actually tried this, and only assume that this was the case when Word was 'new')

You can't open up your MySQL database and applications in PostgreSQL and have them automagically work. If so, I'd have tried moving to PostgreSQL long ago.

Re:Generic, huh?

[LWATCDR]

But most of the applications I am talking about are GPL or some other flavor of FOSS so that isn't an issue.
Don't have to convince me. I am just pointing out why people still use MySQL.

Re:Generic, huh?

[Slashdot Parent]

I can't see a single reason to use MySQL these days.

How about: "It's dead easy to stand up a MySQL server and use it to put your data into and get your data out of."

I didn't see MySQL in any enterprise project I've ever done. It's all Oracle or DB2. But MySQL has cornered the "I've got some data that I need to put somewhere and eventually get it back" market, and for good reason: MySQL makes it so simple.

A complete database novice can go from zero to "I've got data in my database that I can get back by saying SELECT blah FROM blah WHERE blah" in under an hour. With PostgreSQL, a novice could spend an hour trying to figure out what the heck template1 is.

The vast majority of database applications just need to do simple CRUD. Why should anyone bother with the learning curve of Postgres? But I think you already know the answer: they don't.

Re:Generic, huh?

[Sxooter]

For an example of how truly outstanding the performance of high performance MySQL clusters are, read this forum entry from the mysql forums:

Test cluster slower than single machine

Re:Generic, huh?

[An ominous Cow art]

> You can open WordPerfect documents in Word and they work. (Disclaimer: I haven't actually tried this, and only assume that this was the case when Word was 'new')

For what it's worth (which I realize isn't much) the last time I tried this was around 1999, and Word made a mess of the very simple WordPerfect documents I was trying to open (it was Office 97, but I'm not sure of the WordPerfect version involved, anymore).

Re:Generic, huh?

[Jay Pipes]

You are proving that you don't really understand the technology or the business behind MySQL and are fine with throwing your lot behind myths that have been propogated for a long time now. You say:

Revert back to telling would-be users that ACID, stored procedures, etc. are unimportant and that they belong in client code?

Stored procedures have nothing to do with Oracle, InnoDB, or anything else here. They exist above the storage engine level at the language level and work with any storage engine, InnoDB or otherwise, as do triggers, views, functions, events, etc.

If you have other plans, by all means let us know! In the mean time, the thought of building business logic on MySQL is growing scarier by the month.

Innobase/Oracle is one of several storage engines that have/will have ACID-compliant transactions. PBXT, SolidDB, the upcoming Falcon storage engine, and MyISAM++ all have or will have ACID-compliant features. So, yes, we do have a "fallback plan" and have for a long time now.

And one free bit of advice: distance yourself from SCO! That affiliation by itself is an enormous dealbreaker for a lot of people.

Thanks very much for this piece of advice. I hadn't heard that before. Seriously, I don't know why I bother even responding on /. anymore. It's just a breeding ground for dogmatic idealogues to flame-bait.

Re:Generic, huh?

[Just Some Guy]

You are proving that you don't really understand the technology or the business behind MySQL and are fine with throwing your lot behind myths that have been propogated for a long time now.

Jay, if I'm wrong, then is it my fault (as a potential customer) that I have misinformation, or yours (as my potential vendor) for not bothering to publicize your case? This is the first time I've heard of the backends you mentioned, and I follow tech news very closely.

It costs nothing to submit a story to Slashdot, so whenever you add support for a cool new backend or add major functionality, why not try to publicize it, especially outside your regular channels? Maybe MySQL vendors and partners are already up-to-date with your newest developments, but you've already sold your product to them. Do your company a favor: walk over to the Public Relations department and explain that your product is garnering a bad reputation in public forums. In the absence of good news from you, all we have is the bad news from anecdotes to share.

Re:Generic, huh?

[Anonymous Coward]

By that logic you should just give up and go to M$ Frontpage and Access.

Re:Generic, huh?

[tokul]

Everybody uses MySQL because everyone uses MySQL.

Some people use database abstraction libraries and are not locked to MySQL.

Re:Generic, huh?

[Thundersnatch]

You can, almost. Of course it would never be a simple as converting a single file format, but if you're using a tool like this for conversion, and an "abtracted" database library like JDBC or your language's equiavalent, the transition could be pretty easy. Of course, you have to tweak all that non-standard MySQL code to be ANSI SQL compliant. Perhaps it's trivial, for your app, perhaps not.

I've seen an application that only uses ANSI SQL, usually the 1992 variant, and JDBC. It concurrently supports Oracle, DB2, Microsoft SQL Server, and PostGreSQL from the same code base with simple configuration file changes. It takes a lot of discipline to build an app to the lowest common denominator like that, when every DB vendor has a multitude of cool time-saving but not-yet-totally-standard SQL features you could use. But MySQL wouldn't even support the lowest common denominator of ANSI SQL-92.

Re:Generic, huh?

[Dewin]

For some fairly hardcore applications it's not really *that* feasible.

This is fairly tame compared to some of the queries done by my application (mix of SQL and PHP):

sql_query("
    INSERT INTO allocation (year, sid, term, start, end, fte, grade, amount, relevant)
    SELECT
        yci.year, s.id, yci.term,
        IF(ls.start IS NULL, yci.start, GREATEST(yci.start, ls.start)),
        IF(ls.end IS NULL, yci.end, LEAST(yci.end, ls.end)),
        ls.fte, ls.grade+yci.year-ls.gradeyear,
        ((
            (UNIX_TIMESTAMP(IF(ls.end IS NULL, yci.end, LEAST(yci.end, ls.end))) - UNIX_TIMESTAMP(IF(ls.start IS NULL, yci.start, GREATEST(yci.start, ls.start))))
            / (UNIX_TIMESTAMP(yci.end) - UNIX_TIMESTAMP(yci.start))
        ) * gi.allocation * ls.fte * 0.01)/numterms allocation,
        yci.relevant
    FROM
        ((
            SELECT yi.year, 0 term, 1 numterms, MIN(cal.startdate) start, MAX(cal.enddate)+INTERVAL 1 DAY-INTERVAL 1 SECOND end, '0000-00-00 00:00:00' relevant
            FROM yearinfo yi, calendar cal
            WHERE cal.year=yi.year AND cal.type='term' {$ywhere}
            AND NOT yi.allocation_periodic
            GROUP BY yi.year
        ) UNION ALL (
            SELECT yi.year, cal.code term, (SELECT COUNT(*) FROM calendar cal2 WHERE cal2.type='aterm' AND cal2.year=yi.year) numterms, cal.startdate start, cal.enddate+INTERVAL 1 DAY-INTERVAL 1 SECOND end, cal.startdate relevant
            FROM yearinfo yi, calendar cal
            WHERE cal.year=yi.year AND cal.type='aterm' {$ywhere}
            AND yi.allocation_periodic
        )) yci,
        student s,
        log_status ls,
        gradeinfo gi
        LEFT JOIN {$temptable} o ON o.year=yci.year AND o.sid=s.id
    WHERE
        ls.sid=s.id AND ls.status>0 {$swhere}
        AND (ls.start <= yci.end OR ls.start IS NULL)
        AND (yci.start <= ls.end OR ls.end IS NULL)
        AND gi.year=yci.year AND gi.grade=ls.grade+yci.year-ls.gradeyear
        AND o.year IS NULL
");

I'd love to be able to move it though -- a lot of it would make more sense as triggers and/or functions ran on the SQL side.

Converting a blog or message board, however, seems fairly trivial.

Re:Generic, huh?

[Thundersnatch]

You're not using ANSI-style JOIN syntax (except for the LEFT JOIN), which is a big no-no. It should never have been written that way to begin with, unless an old version of MySQL didn't understand INNER JOIN syntax. The "JOIN condition in WHERE clause" appraoch can be ambiguous in a lot of situations, which is why the ANSI JOIN syntax was created.

As for the use of the temp table... well, enough has been said about that by others.

Actually, to be ANSI-compliant, the code you have would need changes in the JOINS, the use of the temp table (make it a subselect), and the non-standard functions (UNIX_TIMESTAMP, INTERVAL, GREATEST, and LEAST), which have ANSI equivalents.

So really, you coded this to MySQL's spec, without regard for prtability. It's not surprising that it is not very portable.

Re:Generic, huh?

[chromatic]

It's not possible to use MySQL as a backend to proprietary applications without shelling out some cash.

If you don't distribute the application, why would copyright have an effect?

Re:Generic, huh?

[Just Some Guy]

If you don't distribute the application, why would copyright have an effect?

But the point is that you can distribute proprietary applications with bindings to just about any other major database without paying for client library redistribution rights. MySQL is the only one I'm aware of that won't let you do that.

Re:Generic, huh?

[chromatic]

I'm not really familiar with the licensing structure of other major databases, excepting PostgreSQL and SQLite (if it counts as major), but my point was only that there's a big difference between usage and distribution, with regard to copyright. It's important to maintain that distinction when comparing copyleft licenses.

Re:Generic, huh?

[Dewin]

You're not using ANSI-style JOIN syntax (except for the LEFT JOIN), which is a big no-no. It should never have been written that way to begin with, unless an old version of MySQL didn't understand INNER JOIN syntax. The "JOIN condition in WHERE clause" appraoch can be ambiguous in a lot of situations, which is why the ANSI JOIN syntax was created.

I'm aware of the issues with JOINs. It actually happens to be one of a few issues preventing me from upgrading to MySQL 5 as well -- and the project is too large to fixup.

It also doesn't help that I self-taught myself databases from MySQL exclusively (Postgres was crummy documentation-wise at the time, among other things) so it's still the habit I'm stuck in.

Actually, to be ANSI-compliant [...] the use of the temp table (make it a subselect)

The temptable happens to have a subset of the contents of the table being inserted into. MySQL (at least 4.x) does not allow you to select from the same table being updated from in a subquery, thus the temptable is required as a workaround. I'd be very happy if I could remove the few remaining temporary tables from my code.

[...] and the non-standard functions (UNIX_TIMESTAMP, INTERVAL, GREATEST, and LEAST), which have ANSI equivalents.

GREATEST and LEAST don't have equivalents that I've seen -- at least not in PostgreSQL (which is the only other database I've seriously considered moving to). They can be imitated, yes, but that method seemed overly clunky.

Oh well

[xenocide2]

I can't say for sure whether it's the same level of support, but there's always Canonical for Ubuntu and Progeny for Debian support.

And yet...

[merc]

They're more than happy to be a SCO/Canopy partner.

I know where I'll not be spending my IT budget next year.

Fork or Spoon

[Paulitics]

MySQL only lets me spoon it.

But Postgre lets me fork it all night long.

Get Ready...

[eno2001]

I see that a definite split of "Premium Linux" vs. "Unsupported Linux" is coming soon to a vendor near you. That doesn't mean that Linux will die, it's just going to smell funny (possibly like pee).

Re:Get Ready...

[Anonymous Coward]

Yeah, it used to all be "Unsupported Linux".

Re:Get Ready...

[bziman]

I see that a definite split of "Premium Linux" vs. "Unsupported Linux" is coming soon to a vendor near you.

The irony is that the "premium linuxes" of RedHat and SuSe are the ones that suck just as much as Windows... and the "unsupported linuxes" like Debian and Ubuntu just kick all sorts of ass.

Re:Get Ready...

[eno2001]

On that I certainly agree. I'm no fan of the commercial linux distros myself. I use Gentoo even with all the pain that's involved (only when you don't keep the box up to date which requires a ton of time that a lot of people don't have).

Re:Get Ready...

[CodeBuster]

Look at it this way...its like self-serve vs full serve gasoline (yes some places still offer full serve but you have to look around a bit to find it). In the full serve model the station employee(s) check my tire pressure, wash my windows, and fill up my gas tank while I sit in the comfort of my BMW listening to the [name of city here] philharmonic playing Vivaldi OR I pull up in my beat up pickup truck, get my butt out of the truck, check the pressure in my own worn out tires, wash the bug grime off my own windows, and fill up my own gas tank with the cheapest 87 octane gas that I can find. The gasoline still powers my vehicle, but I didn't get any service. There is a market for both types of gasoline service and there is nothing wrong with that.

Re:Get Ready...

[eno2001]

Heh. Nice analogy. But you forgot about my take on it:

Dr. Emmett Brown drives past the gas station in his modified DeLorean laughing wildly. He knows that his car does things that no other car currently does. Today he's going to visit the dawn of humanity, which essentially consists of some bearded proto-human walking out of cave and falling on his face. The proto-human's next action? Becoming Bill Gates roommate in college and growing up to be Steve Ballmer. ;P

Forking won't necessarily do anything

[iamjoltman]

I see there's already a few comments that the code should be forked. The thing is, what is forking going to do for it? They are dropping support for Linux distros, but that's not saying it won't run on other distros, just that it's not supported. The only way a fork would do anything is if the forked version had it's own support as well.

Re:Forking won't necessarily do anything

[Anonymous MadCoe]

You're right. Actually forking will lead to more fragmentation which is the problem in the first place. You can't expect a software vendor to support many different Distros (you could argue over it, but in general very few people see that commercially viable).

The problem would never have existed if things didn't get forked all the time and everyone would re-use what's out there. But then again, that would take the fun out of it ....

Re:Forking won't necessarily do anything

[Bastian]

Even then, forking isn't necessary to provide support. All that's needed is for a company to come in and offer third-party MySQL support for other distros.

However, I don't see that happening; most likely dropping support for everything but RH and SuSE has something to do with the fact that those two distros dominate the enterprise marketshare so much that there just isn't any money to be made in providing support for MySQL on Debian.

All of my servers run Debian

[maynard]

While I don't currently have or need a support contract from MySQL, I wouldn't transition away from Debian within our machine room just for their sake. I can't say this is a mistake for them, as I don't know what sales numbers they see, but here's one potential customer that's gone as a result.

Re:All of my servers run Debian

[PHPfanboy]

> While I don't currently have or need a support contract from MySQL

I think this says it all for most Debian users. They are either in-house experts, testing the water for their app or don't have a culture of procurement (read: lower budget or just plain cheap). This is not a criticism, it's just a business reality.

MySQL is a business, unless we want them to go out of business and drop support for everything there's not much to complain about.

Re:All of my servers run Debian

[chundo]

I doubt that's the point. I'm sure they just decided that rom a cost/benefit perspective, money spent training their support staff on Debian wasn't worth the amount of business they were getting from Debian customers. Which makes a lot of sense to me - in my experience, people that run Debian servers have a more thorough knowledge of the system and administering it, and consequently have less need/desire for software support (yourself included, it sounds like). And assuming that's true, it's also not much of a stretch to assume that someone that interested in the guts of a system would choose something like Postgres over MySQL anyways if they had a choice, since it's had more advanced features for much longer than MySQL has.

Re:All of my servers run Debian

[maynard]

In point of fact, we are already running a PostgreSQL server as a backend for a call tracking system. So... yeah. :)

Re:All of my servers run Debian

[Jamesday]

We have people using just about anything. The issue isn't knowing the platform, it's compile and build-time bugs. We don't build the binaries Debian supplies, so we don't know what bugs they might contain. We do know that we have had to work around a wide range of build bugs on assorted platforms, from compiler glitches to dodgy libraries.

Ask a question about MySQL, and we'll answer it. Ask a question about some really weird behavior that causes us to believe it's a bug in the build and we'll tell you to use the generic build we provide to eliminate the build as a possible cause. Ask a question about how to do x in Debian and we might answer out of charity but it's not covered by the support contract: we don't sell support for the Debian OS.

What we don't do is say that we will rebuild the Debian build of MySQL to fix a problem in its build process - Debian owns and controls that, not MySQL. For the supported platforms we do say that we will fix such issues with our builds for them (but not those of the distro owner). This matters for our contractual obligations and what they do or don't require us to do. It's generally unimportant in answering questions.

The same applies to just about any distro or home build there is. We'll answer the MySQL questions but not deal with broken builds we didn't make ourselves.

The marketing people are writing a more official answer which will tell every Debian and MySQL lover just what they want to hear from MySQL. Managers involved, so it takes a few hours to arrange.

Not a lot of point in modding this up when an official answer will be arriving soon.

Why fork it?

[syphax]

I don't see this as a technical deficiency of the software. This is a business issue.

Do you have Debian and MySQL expertise? Find yourself someone business-savvy (hint: it's probably not you) and sell support for MySQL on Debian. Be your own boss (hint: make sure your business-savvy person isn't a PHB). I think MySQL AB has been pretty clear in the past that they are but a small (if central) part of the MySQL ecosystem, and they clearly want to focus on their high-margin customers. Might be a smart move, might not, but it sure opens the door to players who want to seize the other niches.

Re:Why fork it?

[dsci]

This is good advice overall, and I'll continue to take a middle ground approach to it. That is, I'll continue to support MySQL (on whatever distro) for my clients. I don't offer general MySQL support, but when I 'design' a system for my clients, I support whatever OSS stuff goes into that system. If MySQL is the best fit for their needs, that's what I recommend. If it needs support, they call ME.

So, I'm not out there selling MySQL support to anybody that wants it, but I do take care of MY clients. That said, I service ONLY small businesses - nothing remotely like the kind of market the MySQL AB team targets, but I think the OSS + Support ecosystem is big/diverse enough to handle a bunch more niche providers like me.

Re:Why fork it?

[Baba Ram Dass]

Exactly. If the demand for MySQL support on "non-generic" Linux is there, then the demand will be met by 3rd-party support companies. And if the demand ain't there, then it's a moot point. (God I love the magic of the free market...)

But I'm still trying to figure out how they defined "generic".

How to commit corporate suicide in OSS

[Anonymous Coward]

How to commit corporate suicide if you're an OSS vendor:

1. Kill support for Linux
2. DieDieDie

Re:How to commit corporate suicide in OSS

[Cigamit]

I believe you forgot a few steps...

3. ????
4. Profit!!

Its an odd business plan, but it always seems to pan out in the end.

Tragic

[certain death]

I suppose you could still compile it on any platform, and get support somehow. Maybe this is just their way of cutting costs, I mean, if I had some piece of software, I truly do not believe I could support it on EVERY Distro out there. MySQL is good software, I don't think they should be based for this...until we at least find out the truth why they did it.

Sounds like a business plan waiting to happen

[xantho]

MySQL just said, 'We don't think that your business is profitable to us,' for whatever reason they might have. Well, I'm willing to bet that MySQL support for Debian in the enterprise setting is plenty profitable for some other people.

The only thing that really happened is that MySQL cleaved off a part of their business and gave it away for free to anyone who wants it. And I'll bet plenty of people do.

Re:Sounds like a business plan waiting to happen

[bluephone]

Careful there, that's a reasoned and rational response. That's not going to go over well on Slashdot. ;)

Re:Sounds like a business plan waiting to happen

[ClayJar]

I believe what they said was merely, "We can make enough money on the easy stuff."

The people who are going to pay for the supported "Enterprise" linux distros have lots of money to throw around. Supporting only a couple well-defined distros also optimizes MySQL AB's investment in training and testing. They only need to know the specifics of a couple packaged distros, and the people who buy those distros have the money to pay for support as well as the corporate personality to want to pay for the value they perceive in commercial support.

Seems that people should be commending MySQL AB for realizing that it's a better business decision to "do one thing, and do it well" while opening up the market for others who would like to provide support options. They noted the inefficiencies inherent in trying to provide everything for everyone on every level, and they decided to concentrate on what they feel works best for MySQL AB -- i.e. providing the MySQL database and support to people who want to pay for it.

If they were forking a closed-source database and would only provide future versions for a couple distros, that would be one thing, but all they're doing is deciding who can pay them for support. The funny thing is that people think it matters, while they openly use other software that -- *gasp* -- doesn't even offer commercial paid support.

Re:Sounds like a business plan waiting to happen

[Anonymous Coward]

Well, I'm willing to bet that MySQL support for Debian in the enterprise setting is plenty profitable for some other people.

Okay name a couple. Doh, you just lost that bet. Yeah it might be okay for an individual to be a DBA but to organize a company around it?

Re:Sounds like a business plan waiting to happen

[xantho]

As far as I can tell, sole proprietorships and partnerships are still valid corporate structures in all 50 states, so for one DBA who has enough time to support enough clients, absolutely, organize a company around it.

But even before you challenged my nonexistant assertion that there already are companies offering 3rd party support contracts, I went looking to see if there were actually any. And there don't seem to be many. Why is that?

1. The business plan just isn't workable for a small company that isn't profitable in other areas.

Could be. I don't know nearly enough about the market to speak authoritatively on the subject.

2. MySQL AB was eating all the customers that 3rd party support companies would have been able to serve to survive.

Well, this is no longer an issue, because the biggest MySQL support company, MySQL AB is no longer supporting them. So this certainly opens up the possibility that the business of being a 3rd party MySQL support company is viable.

P.S. I bet it felt good to type "Doh, you lost that bet." Come on, you can tell us, it felt good, didn't it?

Re:Sounds like a business plan waiting to happen

[xantho]

I totally agree with you. If there are a couple of clear winners in the enterprise Linux market, then the opportunity cost of maintaining support for a bunch of loser distros (in that market, shut up Debian fanboys) is probably not high enough to justify going after.

It seems like support for MySQL on a particular distro is vastly different than simple support for the db itself. In that case, you can't just tell a customer, "Hey, you figure out why the db log is eating all your disk space and crashing your server." Now, you have to know about default log rotating and clearing policies on that distro and version, and how to modify the policies to fit a client's needs. Which is way more work than just supporting the db itself.

I like the idea of MySQL AB only supporting distro specific stuff on the most popular ones, or maybe none at all, and a strong, maybe localized 3rd party army that resells the support contract and provide a value add for support on a particular disto.

Linux

[Archangel Michael]

"Generic Linux"???

Isn't "Linux" "generic" almost by definition. The only differences between packages are choices and package manager and usually only a few homegrown eye candy pieces.

No really, I'm not trolling. I'm serious. I've used all sorts of different "distros", Redhat, SuSE, Debian, Slackware etc and I am able to quickly move between them because at the core of it, its all but the same. And I'm not a Linux expert by any stretch of the imagination, so if I can manage, why can't the big boys who do nothing but Linux?

Re:Linux

[Just Some Guy]

The differences are subtle but sometimes important. An example is where each distro puts startup scripts and how they are written. Some are even migrating towards Apple's launchd which is an entirely different animal from the customary SYSV or BSD scripts.

That's not a huge obstacle in and of itself, but multiply little issues like that by a few hundred and it's not so pretty. The Linux Standards Base was supposed to address a lot of that, but no one seems to be clamoring to support it.

Re:Linux

[Anonymous Coward]

I've used all sorts of different "distros", Redhat, SuSE, Debian, Slackware etc and I am able to quickly move between them because at the core of it, its all but the same.

RedHat is not a Linux at all, it is some perverted abortion of it, adheres to no standards whatsoever and is very cavalier about how they treat customers hardware. I would not touch RedHat with a 10 foot pole. Total crap. To top it off, they want money for it! Unbelievable gall.

Why all the drama?

[derrickh]

Why is this such a sore spot for so many people? Just because MySql no longer supports the flavor of the month distro of Linux, you all throw up your hands crying 'I never liked you anyway'.

The vast majority of mysql users will never buy a support contract, and those few who do, will probably be RedHat or Suse. (When was the last time a Debian user admitted he needed help for anything?)

Instead of having to support dozens of distros, Mysql is supporting the main two. It may be Open Source, but it's still a business.

D

Re:Why all the drama?

[NorbrookC]

Just because MySql no longer supports the flavor of the month distro of Linux, you all throw up your hands crying 'I never liked you anyway'.

In other news: Oracle announces they'll only support Oracle on Oracle's Linux, Red Hat is selling support for Red Hat Linux, and SuSe announces that it's selling support for SuSe Linux. Canonical announces support for Ubuntu, but not CentOS. Slashdot readers erupt in fury.

This is a business decision. I would bet that they looked at who was actually purchasing support contracts, and what they were running MySQL on. If 95% of your support contracts are running either one of two distros, then that's where you focus. It's not a slap at Debian as a distro, it's a decision reached because most people running Debian/MySQL weren't bothering with support contracts.

Re:Why all the drama?

[Anonymous Coward]

Debian isn't really a "flavor of the month". It's a distro for conservative people. And the vast majority is affected and influenced by what they run at work. It's like trickle down economics, but not imaginary.

Re:Why all the drama?

[Synn]

Why is this such a sore spot for so many people? Just because MySql no longer supports the flavor of the month distro of Linux, you all throw up your hands crying 'I never liked you anyway'.

First off, Debian stable isn't a "flavor of the month". It's been one of the most robust server distributions of Linux for over a decade. And people are pissy because these kinds of decisions very much affect what flavor of Linux goes into server rooms. I know companies that won't use CentOS over Red Hat because certain hardware vendors only support Red Hat.

Re:Why all the drama?

[bitspotter]

Unity Fallacy

"We all" aren't throwing up our hands about anything. I'm not really concerned about having one less vendor for support contracts I don't need to begin with, and I'm certainly not entertaining a big server migration away from Debian because of it.

Maybe I'm alone, but I really doubt it.

Almost there

[glwtta]

So now they just have to drop RedHat and SUSE and we are finally done? Great!

I've been getting kinda tired of the whole cult surrounding MySQL's substandard "RDBMS".

Re:Almost there

[newt0311]

MySQL's substandard "RDBMS". Amen to that. go PostgreSQL!!!

Of all the posts here

[SocietyoftheFist]

Who is actually running MySQL Enterprise?

Re:Of all the posts here

[Anonymous Coward]

I actually purchased it a while back because I wanted to have the option of support if I had problems with it. Problem was that the versions of the "Certified Server" were much older(er.. stable)

Regardless, I removed it about 6 months later and moved my data to the regular old Community Edition. Happy as can be.

The main thing I saw that the Enterprise subscription got you was that you could receive automatic notifications ever time new versions / fixes / etc.. were released but I can get that anyhow. That and a couple phone calls. Nothing to get too excited about but as someone else put it here, if it makes your boss happy that they can call someone if it breaks then go for it.

Re:Of all the posts here

[Anonymous Coward]

My company is pretty darn close. We're running MySQL, we want to pay the company for support, we want to use their high availability features.

Then again, this is posted anonymously for commercial reasons. We've done the TCO comparison against Oracle and MySQL is still coming out cheaper.

If you need support...

[Sabalon]

Chances are that if you need the support they offer, then you are not just running some little fan site using MySQL to store what avatar's people choose. Most likely if you have support for the db, chances are you probably have some sort of support contract in place for the OS as well and the rest of your critical infrastructure. You are probably already playing by their rules using certain OS releases, etc...

That would be my guess at least.

Opportunity for Postgres

[JoeCommodore]

This looks liike an opportunity for Postgres to come out with some better documentation on installation and configuration of Postgres and attract some new users.

Re:Opportunity for Postgres

[Stumbles]

Wadya mean? Postgresql is pretty easy to compile from source and I've had zero problems installing it from RPMs, etc. As for it's documentation I have found it to be very useful. What do you mean by configure anyway? You got your conf files that normally live in /var/lib/pgsql and their annotations are pretty clear. So I think your just blowing smoke.

Re:Opportunity for Postgres

[tcopeland]

> This looks liike an opportunity for Postgres

Right on. And with the excellent performance of the newly-released PostgreSQL 8.2, it's a good time to make the switch.

Re:Opportunity for Postgres

[Anonymous Coward]

Install and "get started" with PGAdmin3 instructions for Windows can be found right here:

http://www.geocities.com/operationsengineer1/

it is very easy if you know what to do. it is not so easy if you don't.

the traffic limit on the pdfs linked from the page above is pretty limited, so bookmark it and come back in a few days if you can't get it now.

good luck.

Re:Opportunity for Postgres

[JoeCommodore]

Installing is a breeze (you just have to remember to use su properly at one point. it's the docs for the configuration of access with pg_hba.conf which is kind of bad, but also stated as real important.

Re:Opportunity for Postgres

[newt0311]

the HTML docs for pg_hba are actually really good and practically walk you through setting up a ssl equiped server for remote access.

Did anyone catch the relationship?

[Bright Apollo]

SUSE and RedHat are also the only IBM supported distros. Is IBM going for MySQL, ala Oracle grabbing Innobase and Sleepycat?

-BA

Re:Did anyone catch the relationship?

[LochNess]

Why would IBM be going for MySQL, when they already have DB2, which is a big money maker for them?

Re:Did anyone catch the relationship?

[davidsyes]

Holy FRACK, Apollo. May the Lords of Kobol shine a light ahead for us..., but I am sure MySQL will quote your father Adama and say, "Sometimes you gotta roll the hard six."

(Now, if BOOMER said that to me, I'd quot Paul Lekakis with "BOOM BOOM BOOM, let's go BACK to my ROOM, so we can DO IT all NITE and u can MAKE me feel RIGHT!")

(DOH! Slash captcha: "stench" sheesh, this captcha generator is too much sometimes...)

Re:Did anyone catch the relationship?

[Cederic]

Because DB2 doesn't have penetration in the same market as MySQL?

Re:Did anyone catch the relationship?

[MintyGreenMedia]

As previously mentioned (and I would have said anyway), IBM already has DB2, so they wouldn't have much incentive to buy another database. However, it might be worth noting that Dell offers both RHEL and SLES on some of its higher-end servers. Maybe you're looking at the wrong manufacturer?

(Not that I particularly believe Dell is going for that power play, either.)

Re:Did anyone catch the relationship?

[Bright Apollo]

By your logic, Oracle should have avoided Sleepycat and Innobase, and yet they bought both. Market consolidation is one factor, but IBM making a strong push to co-opt open source to provide more fuel for its services arm is the real driver here.

IBM will help create the storage engine for MySQL, then buy it, then give it away and sell support.

DB2 has its mainframe niche well established, but I think we can all agree that mainframes aren't the growth industry they once were.

-BA

What does this say for OSS as a business model?

[dsci]

MySQL (the support company) no longer sells support for Debian.

It seems to me that this decision must be driven by sales or market research indicated there is no market for support contracts on Debian based systems. So, does this challenge the notion that OSS can work in a capitalist world when the real "product" is support?

Debian based distros are a significant chunck of the Linux market|mindshare. This decision essentially means the combination of Debian + MySQL is doomed in the business setting.

On the other hand, this does seem to show that there IS a market for support on RH based distros.

In fact, as I think about it, I think what this is really saying is that they want to support MySQL, NOT the underlying OS. Perhaps they have some data that shows that many of their support calls are really for the OS or other parts of the system. In making this decision, they don't rope themselves into having to support anything but MySQL. They can answer a non-relevant (to them) call with "oh, that's an OS issue - call your OS support provider." I'd say that's fair.

It also helps them when there is a problem with MySQL on a client system...THEY can call RH (or whomever) support to make sure everybody gets things 'right.' No, the more I think about it, the more I think this actually strengthens the "give away the software, sell support" model.

Re:What does this say for OSS as a business model?

[stry_cat]

Except that Red Hat and Suse are not entirely open source.
What happens when MySQL depends on the closed part?

Huh? I didn't know this about RH (don't know anything about Suse). Is this really true? Wouldn't CENTOS have some serious problems in making a RHEL rebuild if there were some close source things in it?

Give us some examples please.

Re:What does this say for OSS as a business model?

[Wudbaer]

Now you just have to show me the closed parts both in Suse and especially in RedHat.

Re:What does this say for OSS as a business model?

[dekemoose]

Where is the mod for -1 Random made up bullshit?

Re:What does this say for OSS as a business model?

[codemachine]

It seems to me that this decision must be driven by sales or market research indicated there is no market for support contracts on Debian based systems.

Not necessarily true. It could mean that their research told them that it would be cheaper for them, and/or more profitable, to support just those 2 distributions. And that other systems are not profitable enough for them to bother with, given how much more profitable providing Suse and RHEL support is.

That doesn't mean that providing support for MySQL on Debian couldn't be a viable business. In fact, given the nature of OSS, it is quite possible for someone else to do so. Profit margins may not be as good as MySQL's own support business, but that doesn't mean it couldn't be done.

Re:What does this say for OSS as a business model?

[Anonymous Coward]

Hardware drivers. But if MySQL is dependent on certain proprietary hardware drivers, I think lack of official support is the least of their problems.

Re:What does this say for OSS as a business model?

[Orange Crush]

Where is the mod for -1 Random made up bullshit?

This is Slashdot. That would be a "+1 Random Made Up Bullshit."

Varying Levels of Support

[Dareth]

MySQL has varying levels of support for different versions and architectures.

The linked support list was to the Enterprise version, but check out Cluster and MaxDB versions.
Oddly enough, they claim FS - full support for Debian 3.0 on the PowerPC architecture.

QUIETLY?

[Viraptor]

The real problem?
"MySQL Quietly Drops Support..." ?
Ok - so what should they do? Place posters all around your city saying "WE DROP SUPPORT FOR DEBIAN USERS!!!"? Yeah - that would be a great marketing move. Get real - they don't want to go on with Debian support dept., then it's their choice. They're creating a place for a new company, that will do support for those who want it.

Re:QUIETLY?

[uglyduckling]

The real problem? "MySQL Quietly Drops Support..." ? Ok - so what should they do? Place posters all around your city saying "WE DROP SUPPORT FOR DEBIAN USERS!!!"?

I think the point is that they haven't made it clear, even on their website that they have made a business decision to ignore everything but Red Hat and Suse. From the story: "We learned of this when MySQL declined to sell us support for some new Debian-based servers. Our sales rep 'found out from engineering that the current Enterprise offering is no longer supported on Debian OS.'". So a company got bitten by using a generic (Debian) Linux then asking for support and finding out that "generic" means anything but.

They really should make some sort of statement, even if it's market spun, e.g. "...for the benefit of our enterprise customers we are concentrating on supporting the two most popular commercial distributions... we expect third-party support companies and the active MySQL community to continue supporting less popular and non-commercial distributions". (P.S. for the benefit of anyone flicking through, I made that up!)

Re:QUIETLY?

[Viraptor]

Hmm... browsing through it, I see, they list RH and Suse separately and Generic Linux somewhere at the bottom... For me it seems, that they added specific names and forgot to take down Generic thing.
Some kind of editing mistake probably. We'll see about it in a couple of days, when they react in some way.

Re:QUIETLY?

[Sexy Commando]

Probablly a victim of an injection attack
; DROP SUPPORT debian

Re:QUIETLY?

[NitsujTPU]

Actually, it's just a return of an old meme where everything is more interesting when it's done quietly. Like it's some conspiratorial move.

No need to fork!

[Builder]

There are a lot of calls here to fork the code. I'm a bit wary of calls to fork a project by people who lack the reading comprehension to understand the project. These may not be the best people to direct a project :)

Just to clarify the crappy summary, MySQL are not saying that their software won't run on Debian or Ubuntu or whatever... It will still run on most OSs and distros, but if you are using Linux, MySQL AB will only sell you a support contract for MySQL if you are running on Dead Rat Enterprise Linux (RHEL) or Novhell (SLES?).

Get it? Got it? Good!

Demographic question

[LaminatorX]

Do many mid-size to large (read shopping for juicy support contracts) enterprises use Debian?

No Free alternatives....

[Chanc_Gorkon]

SO, essentially they are giving you no Linux distros that are totally known for their freedom. Only Red Hat and SuSE for Linux flavors and Solaris, AIX and Windows for the others. Really dumb guys, but not really that much of a concern. Someone else can step up and support MySQL. No big deal in the long run I guess except it gives people less choice initially if their job requires them to have a support contract and believe me alot of companies require this as silly as it sounds. What I see happening is some other Linux company will step up and support MySQL as well as their OS.

Re:No Free alternatives....

[Anonymous Coward]

SO, essentially they are giving you no Linux distros that are totally known for their freedom. Only Red Hat and SuSE for Linux flavors and Solaris, AIX and Windows for the others. Really dumb guys, but not really that much of a concern. Someone else can step up and support MySQL.

Why is this dumb? Seriously, I'm reading a lot of uninformed shit-talking here. Maybe, just maybe their bean counters and finance folks ran the numbers for profit optimization and it turns out it isn't in their interests to support the other stuff. The time last I checked they were still a business and not an open source software foundation. As you so eloquently pointed out, it's wide open for other players to come in and support MySQL. They're happy, the third party support folks are happy, and (hopefully) the end users are happy. It's not like MySQL AB came down and said "not only are WE not going support MySQL on the other distros, but if anybody else does we're going to come over there and run their private bits over a cheese grater."

Indeed...

[Svartalf]

The "who do you sue" line's as old as the hills and, largely speaking, irrelevant because you're never
going to get to first base unless it's a screw-up of epic proportions. Even then, it's more likely to
be a colossal waste of your time and merely an exercise of fattening your lawyer's wallet.

Re:Indeed...

[Anonymous Coward]

Marcus Ranum wrote about this on his web site.
http://www.ranum.com/editorials/software-lawsuits/ index.html

A great quote from the article.

"So, if I understand the logic of my friends' senior management, they refuse to allow use of free/open source/unsupported software, so that they can have all the benefits of something that they never have the guts to take advantage of."

Re:Indeed...

[lyz]

I was reading a Microsoft EULA yesterday and saw something about this. Basically, it said that you were entitled for damages for up to US $5. It's most likely not worth the lawyer fees.

Moo

[Chacham]

Wow, i am very happy about this!

As a Database Programmer (and erstwhile DBA) i am saddened by the haphazard mySQL being called a database. For a while it didn't even support transactions. It's actually more of a storage system with a quai-SQL front end.

By dropping that facade from Debian people might be more inclined to use a real database such as PostgreSQL, which has been in the background for much too long.

For the quality that Debian stands for, from my PoV, this is a very good thing.

I would talk of progress here, but Progress is by far the absolute worst database system i have worked with. :)

Profitability

[Colin Smith]

I doubt it. And more important than my opinion, MySQL doubts it and has the sales figures to show it. Companies don't normally kill off profitable products and services, not even evil/stupid corporations. Just because one person can't do something profitably, doesn't mean that someone else can't do it profitably.

Companies kill off profitable lines all the time

[brokeninside]

If a company holds that it can make a 50% ROI on one product line and a 25% ROI on onther product line (and all other things being held equal) they will put their resources into the line with the 50% ROI until such time as the law of diminishing returns brings the marginal ROI for additional resources being added to under 25%.

For example, when I was a kid a local pizza delivery chain started delivering breakfast pizzas. They made money hand over fist. But after a few months, the calculated that the additional cost of maintaining a third shift of workers and an expanded breakfast menu would bring in more money if put into opening additiona stores serving the traditional lunch, dinner, late night crowd with the normal pizzaria menu.

Most likely what is happening is that the MySQL corporation finds that if it spends the same number of dollars training a support tech, those dollars bring in more money if the tech is dedicated to Redhat and/or SuSE than if the tech is also trained on Debian. This doesn't mean that there is no market for Debian support. It means only that MySQL has a higher relative profit from supporting just two databases. The calculation may be different for another company that has a different resource pool. For example a company that already supports Debian Linux, may have a very low marginal cost for adding MySQL on Debian support and, consequently, have a far higher ROI for supporting MySQL on Debian.

The Business Case

[sterno]

It makes sense though when you think about it. How many companies are out there looking for a support contract for MySQL but aren't using RedHat, etc? Considering that supporting Debian could entail supporting several different specific flavors, it doesn't really seem like the revenue for it would be worth the complications.

Presumably if there's enough of a business for such support, somebody will come in and fill the gap. That's the beauty of open source, non? You can actually get support from somebody other than the originaly developer. If it was Microsoft refusing to support some old version of their software, you'd be up a creek.

UBUNTU ! Why Hath Thou Foresaken Me ?

[Anonymous Coward]

UBUNTU ! Why Hath Thou Foresaken Me ? Have I no creed ? Have I no substance ? Am I not a ...woman ?

MySQL is a ``real'' database

[brokeninside]

It may not be an ACID compliant transactional DBMS. It certainly isn't a fully relational DBMS. But the definition of a database is pretty much a system for storing data which MySQL certainly qualifies as. And if we're going to start being pedantic, we also have to consider that none of the mainstream enterprise databases are truly relational. While they may have many relational features, none can consistently enforce proper relational behavior. In fact, any database that fully supports SQL cannot be a fully relational database. The only difference between Oracle, SQL Server, DB/2 and MySQL is one of extent, not of kind.

Re:MySQL is a ``real'' database

[Infernal Device]

My response is: Use what works. When MySQL fails to work for whatever situation I'm in, then I'll consider switching to something else. Probably PostgresSQL, but maybe not.

Firebird is out, regardless. Configuration is difficult, and I'll never forgive them for their pissing and moaning over branding. It's not just the project devs that have long memories.

Re:MySQL is a ``real'' database

[Chacham]

The only difference between Oracle, SQL Server, DB/2 and MySQL is one of extent, not of kind.

You obviously have never really worked with them then.

I was mid-level DBA of Oracle for nearly a couple years, programmer for both SQL Server (Microsoft and Sybase), and currently use DB2 (LUW) (DB2, not DB2, unless you are referring to OS2's DB2, which was called DB2/2), and they are worlds apart. The only way to consistently understand the difference between them is to understand the mindset, otherwise they are just "differences", and the user will most likely not know whow to take advantage of those differences.

Every time i use MySQL i have to hold my nose. Yes, it does the job, and it does it fast and easily, but for someone who cares about DBs, good design, and all that, MySQL falls very far from the tree.

Re:Suse, Red Hat and ??

[mpapet]

Besides the obvious Suse and Red Hat who's the third "premium" linux? I'd say Debian is not it, but Ubuntu sure has the resources to be #3.

Who do you think will be the top 3?

The Linux Meme Knows All

[broward]

http://www.realmeme.com/roller/page/realmeme/?entr y=linux_meme

My interpretation? In the United States, Linux is being marginalized as a specialty niche server. I didn't include the worldwide graphs, but Linux appears to have only a slightly better future. Predictably, Vista is ramping up and gaining mindshare and buzz.

Re:The Linux Meme Knows All

[Anonymous Coward]

Go away Troll.

Who cares

[houseofmore]

I've been using Mysql for many years, through several companies, small and large. Never once has mysql support ever been requested / needed -- it's rock solid. What does support conist of anyway, help with sql syntax?

I doubt most Debian users will care.

Re:Who cares

[GigsVT]

I assume you mean only in terms of "the daemon doesn't crash"... because MySQL still has a very long way to go in terms of catching up with just about any other database when it comes to speed of complex queries.

Re:Who cares

[houseofmore]

I'm not sure I agree (I have great results with 5.0), but even so, how is that related to support?

Debian is the second largest GNU/Linux distro

[Anonymous Coward]

From the Netcraft's GNU/Linux distribution share stats:

RH - 34%,
Debian - 25%
Suse - 11%

--
In any case, our company runs Postgresql so we are not likely to loose any sleep over this decision.

Re:Linux is not enterprise quality

[Count Fenring]

As opposed to.... what? Windows? VMS? Solaris? Your Mother?

I'm pretty sure that all operating systems, at some point, are hacked together by collections of individuals. How "craptastic" they are may have some variation, but I'd guess that, out of the hundreds of people who worked on Linux, some were "craptastic." Same for Windows. Same for VMS, the BSDs, etc.

I can't tell you if, by and large, the people who worked on any particular OS are "craptastic."

But you sure are!

Ba-ZING!

it takes money

[adaminnj]

I know of one company who will suffer from not having MySQL on Debian support.
A rather large ISP that is a Debian shop.

The guy who designed the email system is a friend of mine and told me that he is using Debian for licencing reasons. (something about the GPL)

I think that it's because it keeps costs down.

When I hear from him about new things he is doing, I always ask, did you send at least a donation to the project?

I think it is money that will keep companies like MySQL supporting many flavors of Linux.

I am a Debian user myself and I use MySQL

Yes, and No

[camperdave]

There are some significant differences between distributions:

• Fedora (based on RedHat) uses /etc/rc.d/init.d/rc<runlevel> to store start and stop scripts. Ubuntu (based on Debian) puts these scripts in /etc/init.d, and doesn't have the /etc/rc.d structure.
• Fedora has a program called service to start and stop services: eg. service mysqld start. Ubuntu users have to type /etc/init.d/mysqld start.
• Ubuntu has hidden the root user. You never log in as root. You never become root. root effectively does not exist. Everything is handled through sudo. Fedora, of course, has a root user.
• Runlevels are different. Fedora has runlevel 5 as the multi-user graphic runlevel. Ubuntu uses runlevel 2
• On Fedora, the default runlevel is set in /etc/inittab. Ubuntu doesn't even have an /etc/inittab. It uses a different process based on a /etc/events.d directory hierarchy. I have no idea where the default runlevel is set
• Package management is different as well. Fedora uses a program called yum to manage packages in rpm files. Ubuntu uses apt-get to manage packages in deb files. AFAIK, the packaging schemes are incompatible. (You can't apt-get an rpm package, and you can't yum a deb package.)

... and I'm sure there are others. Add on top of those, that linux can be running either Gnome, or KDE, or Xfce, or nothing at all, and linux support quickly becomes the proverbial "maze of twisty little variants, all different."

I don't really blame them for focussing on enterprise level versions. It fits the "Do one thing, and do it well" philosophy. However, I also can't help feeling that they're shooting themselves in the foot.

Re:Yes, and No

[grahamm]

Yet none of those differences have any effect on supporting users of the application. Things like different versions of glibc might make a difference...

Re:Yes, and No

[camperdave]

Funny... I would think that knowing whether mysqld was running, and how to set it up to run automatically on startup would be a *BIG* part of supporting mysql.

Re:Yes, and No

[Constantine Evans]

Here are a few responses:

* Ubuntu used to (in Dapper) put all scripts in /etc/init.d, and then link to them in the /etc/rc*.d directories for each runlevel. However, this is in the process of being moved over to upstart, where everything will be in /etc/event.d.

* In future versions of Ubuntu, start and stop will do this, since /etc/init.d won't exist.

* This isn't entirely true. One can become root via sudo with sudo -i or sudo -s. There is no root password. In reality, I think this is of somewhat dubious value for security.

* Runlevels in Ubuntu are a problem. Some Ubuntu developers apparently believe that Linux without a graphical login is "broken" (at least one has told me this), and therefore crippled runlevels so that one cannot choose between a text and graphical login via runlevel.

* I believe that Dapper has an /etc/inittab. Edgy, as it was intended to be rather experimental, shipped with upstart in a state that wasn't complete. When it is done, I believe that runlevels will be kept around only for compatibility reasons.

* deb and rpm packages can be converted with alien, though the results are not always very usable. The method does tend to work well with third party packages that are largely without dependencies, however, and those tend to be the many of the cases where this is necessary.

Re:Yes, and No

[JayAEU]

There are some significant differences between distributions:

        * Fedora (based on RedHat) uses /etc/rc.d/init.d/rc to store start and stop scripts. Ubuntu (based on Debian) puts these scripts in /etc/init.d, and doesn't have the /etc/rc.d structure.
        * Fedora has a program called service to start and stop services: eg. service mysqld start. Ubuntu users have to type /etc/init.d/mysqld start.
        * Ubuntu has hidden the root user. You never log in as root. You never become root. root effectively does not exist. Everything is handled through sudo. Fedora, of course, has a root user.
        * Runlevels are different. Fedora has runlevel 5 as the multi-user graphic runlevel. Ubuntu uses runlevel 2
        * On Fedora, the default runlevel is set in /etc/inittab. Ubuntu doesn't even have an /etc/inittab. It uses a different process based on a /etc/events.d directory hierarchy. I have no idea where the default runlevel is set
        * Package management is different as well. Fedora uses a program called yum to manage packages in rpm files. Ubuntu uses apt-get to manage packages in deb files. AFAIK, the packaging schemes are incompatible. (You can't apt-get an rpm package, and you can't yum a deb package.) ... and I'm sure there are others. ... yet I fail to see how any of them would actually matter in terms of supporting a running MySQL installation. Whatever the init scripts being used, once the daemon is installed and running, it's all a matter of customizing the configuration.

Re:Yes, and No

[Archangel Michael]

If you are supporting a LINUX system where the MAINTAINER of that system doesn't know how to "start" or "stop" a daemon, with the proper incantation for that system, then you probably shouldn't support it. That is basic linux. Even "I" know that there are differences in how startup scripts are used in the various systems listed in my original post.

I would expect someone who is supposed to support MySQL would know the same. Even if I couldn't remember, I could look it up or ask (or gasp ... RTFM).

Re:Yes, and No

[camperdave]

Don't get me wrong. I agree with you. I was just pointing out that linux isn't as generic as some people think. I recently switched from Fedora to Dapper Ubuntu, and was shocked when the install completed without mentioning a root account at all. Then to see the runlevels being used differently was an eye-opener. I upgraded to Edgy Eft Ubuntu last week after a hard drive crash, and when I went to check on default runlevels for my previous post, I couldn't find /etc/inittab. It doesn't exist! So, right now I don't really know how to start or stop a daemon, or how to adjust runlevels. Anyone who does a version upgrade on their linux distro could easily find themselves in the same boat.

Re:Yes, and No

[Archangel Michael]

One more comment on this: I think much of the type of differentiation you are seeing, is a linux distro TRYING to be different, not that it NEEDS to be. Some of the "trying" may be in the areas of improving usability, but I am 100% sure that some is "trying" to be different.

This level of differentiation is troublesome at best, because it varies from the psuedo standards that have been in place for the last 25 years or so.

Re:Yes, and No

[camperdave]

Well, I've done a bit of digging, and this new system (called upstart) definitely has potential. It is an event driven script launching system. Events can be:

• the system has started,
• the root filesystem is now writable,
• a block device (USB drive) has been added to the system,
• a filesystem has been mounted,
• at a certain time or repeated time period,
• another job has begun running or has finished,
• a file on the disk has been modified,
• there are files in a queue directory,
• a network device has been detected,
• the default route has been added or removed.
• ...plus many more

One particular application I can think of is what I call a "knock" system. Basically, a machine is set up on the net and sshd left unlaunched (thus it is immune to attacks. When a special ping is received, sshd is launched, and remote users can log in. I'm sure such systems already exist, but this jumped to my mind when I started reading about it.

My dream Database

[Anonymous Coward]

*sigh* Could the lack of vision of MySQL prompt for the birth of my dream DB?

That would be a merge of PostgreSQL and Firebird, two excellent projects that are actually, hmm, real databases, and maybe, pooling the efforts of both communities, could actually make a big splash... c'mon peolple... merge!!! (* crosses fingers *)

Thinking of MySQL? Don't.

This just in...

[SQLz]

Debian has quietly lost all its users to Gentoo and Ubuntu.

Re:This just in...

[D4rk Fx]

Debian has quietly lost all its users to Gentoo and Ubuntu.

Sacralige! As a Debian Zealot, I am forced to respond in kind.

Gentoo is for Ricers, and Ubuntu is for hippies!

Interestingly enough, the CAPTCHA word for this was 'thefts'

Re:This just in...

[alienmole]

I've never seen Gentoo being used in a business setting. I suspect that the characteristics which attract people to Gentoo don't lead to success in the business world. What's the ROI on that mythical extra performance?

I've seen plenty of Red Hat and to a lesser extent, Debian, used in businesses, on servers.

I don't see much Suse, but then I'm not in Europe. That may change with Novell's involvement now.

No Ubuntu, but then Ubuntu hasn't been focusing on servers.

Screw this

[Yvan256]

I'm going back to flat text files.

j/k

Re:Screw this

[jusdisgi]

I love it! Most underrated post of the story. No doubt.

Re:UBUNTU ! Why Hath Thou Foresaken Me ?

[nutznboltz2003]

Am I not a ...woman ?

Only in WoW...

Yes, I know, there goes my Karma.

Re:Suse, Red Hat and ??

[eno2001]

Huzzah! Let me rephrase the question: "Besides the obvious Microsoft, who is the second or third premium Windows vendor"? There is no law or rule that says there must be three or even two premium Linux vendors used by the IT industry. In the end, Novell Linux will likely get more help from MS to be the number one leader. RedHat might die off or just become so irrelevant that it won't matter. Whatever the case, the last thing Microsoft really wants to see is a strong and unified Linux community. The way I see this playing out is that businesses begin to tepidly embrace Novell Linux which increases their market share. The developers who work on high profile projects (like Gnome, KDE, MySQL, etc...) are more and more driven by the business needs than the original "itch" that needed scratching. So there are some forks on major projects... However, the non-premium versions that come out of these forks have a lot of difficulty in attracting talented developers as they are mostly busy working on the premium versions that were part of the old guard Linux camp. There are good developers who would work on some of the forks, but not as many as there were previously. Therefore the forks are buggier, more prone to security holes, and in general don't work as well as the MS blessed versions from Novell.

This is part of Microsoft's campaign against what they term "hobbiests". I use Linux both at work and at home and although I find the term hobbiest insulting, that is what people would probably consider me. I find the uses that I apply Linux to at home to be quite serious. Calling a professional IT guy who uses Linux at home for day-to-day stuff a hobbiest is akin to calling an electrician or plumber who does work on his house an amateur. The fallout that I see is that potentially in another four or five years, I may find it very hard to use Linux at home unless I want to buy into the commercially blessed versions. And if I do buy into them, I'll have a second rate Linux that makes Windows look good. (You know that MS won't allow any MS blessed Linux to outperform or outdo Windows in certain key arenas) If I continue to try and use the non-premium Linux distros I'll probably find that support for new hardware and functionality is just as bad as it was in the early days because the developer mindshare will not be there. At least that's what I'd term a worst case scenario.

In reality it probably won't be THAT bad, but it will hurt. Even though the code is free/open for many of these projects, I've seen what a lack or very low count of talented developers can do to slow down or kill an otherwise decent project. We all have. It's likely that I'll be able to use non-premium Linux at home in the future, but not without even more headaches and hassles than I experience today. The premium versions will likely offer a better experience but always at the expense of being a step or two behind Windows (which is not the current situation). MS is likely doing this because they see that Linux has already surpassed Windows on many fronts. It's more clever maneuvering from MS. If only the FOSS world could think that way sometimes...

That may or may not be the case

[brokeninside]

I think it fair to say that Debian is the second most popular Linux distribution for web servers surveyed by Netcraft. It's also probably fair to extrapolate this to mean that Debian is the second most popular Linux distribution among all web servers. But I think it it quite the unwarranted stretch to take that to mean that Debian is the second most popular Linux in any market segment. I have no reason to believe that use of Linux in the business desktop, development database, database server, or application server follow the same trends. They may very well follow the same trends, but until someone studies the question and offers a cogent analysis that question has to remain unanswered.

Too many linuxes

[pr0nbot]

Not sure what the big deal is. The Linuxscape has always been a fragmented land-of-a-thousand-distros, and contrary to received wisdom, they're not all compatible.

Having just wasted a few days trying to get one bit of "Linux-compatible" software to work on another subtly different flavour Linux (thankyou RedHat for your borked gcc 2.96), I have some sympathy for saying "we only support flavour X".

Is there such a thing as one Linux distro to suit all users? Probably not. Could there maybe be 2 or 3 that would suit virtually all users? I think so. This move from MySQL seems like a step towards that.

Simple; Linux brings too much overhead.

[Anonymous Coward]

The way Linux is currently going simply leaves no other options for companies. Its nearly impossible to support Linux as a whole the way it is now. That is; if you also wish to make some money out of the process and still maintain reasonable fee's.

The whole problem I see (which is only a small portion) is the continously incompatibilty which keeps popping up. Compiling a driver for a Debian sarge kernel on your current Ubuntu distribution may very well turn out to be completely impossible, since the 2.6.8 kernel source depends on things like binutils 2.12. The current distributions ship with 2.16 which isn't fully backwards compatible with its ancestor. Since this is a basic component you can't simply install the other version next to it or above it, not without much effort that is (which is also a key element). So this leaves you with some major problems.

Your best bet is to setup a whole new environment (vmware for example) which you can then use to do this kind of development. All nice and well, but now realize the amount of linux distributions out there! And since most Linux components are simply not backwards compatible you might very well need to maintain a dozen versions only to keep your product running on Linux, IMVHO this is an insane development when looking at it from a professional (or perhaps 'Enterprise') point of view.

And things get worse every time. The same can be seen with the 2.6 kernel tree. New updates only make it into new releases, so if you wish to maintain a stable version (say 2.6.15 for example) you'll have to backport all the patches there are. Most vendors and distributors don't work this way, so they'll simply implement the new updates in the form of new versions. Now check up my previous point about backwards compatibility and you see the same problem surfacing again.

This development doesn't surprise me one bit. This way MySQL support doesn't have to worry about all this massive overhead which makes their lives miserable and only focus on the things which really matter: their product and making or keeping it working. And the only way to achieve this is using the product of a distributor which will make sure that its functionality will remain the same over a massive period of time and won't change after every new update.

As a side note; this is also exactly why I dumped Linux over Solaris on my servers. I don't have the financial resources to pay for a distribution which can keep things stable and compatible in a way both SLES and RHEL do, CentOS might be an option but is by no means one on one the same as RHEL. And as such I went for Solaris. Drivers / software build for Solaris 10 3/5 (first release) still work on Solaris 10 11/06 (current release). I can't say the same for SuSE 10.0 vs. SuSE 10.2.

Re:Simple; Linux brings too much overhead.

[TheAwfulTruth]

Boy that's for sure :(

USB subsystem changes between SUSE 10.0 and 10.1 produced some spectacular driver failures. New elements inserted in the middle of USB data structs in a point upgrade of a "stable" kernel?!?!? What is stable about that?

The Linux development and distrribution process has a LOT to learn about system stability. Expecting EVERYONE to ALWAYS be 100% current and recompile EVERYTHING for EVERY distro and then NEVER updrade an installed kernel or libs again (you know to fix bugs or security holes?) without chancing having to rebuild the entire universe or suffer random breakages is completely and utterly wrong headed. :(

This may have been fine in the good old days of "install and forget". But these days with the need to be CONSTANTLY up on security patches, it's become quite a nightmare to maintain a linux box for any length of time without having to do a complete reinstall because of unresolvable incompatibility problems between the Kernel, libs and software. Doing it by hand is a major recipe for disaster, but even keeping up with a distro's precompiled sets of upgrades is a crap shoot and has resulted in serveral system failures.

Linux needs stability in a BAD way.

Re:Simple; Linux brings too much overhead.

[JayAEU]

This may have been fine in the good old days of "install and forget". But these days with the need to be CONSTANTLY up on security patches, it's become quite a nightmare to maintain a linux box for any length of time without having to do a complete reinstall because of unresolvable incompatibility problems between the Kernel, libs and software. Doing it by hand is a major recipe for disaster, but even keeping up with a distro's precompiled sets of upgrades is a crap shoot and has resulted in serveral system failures. I take it that you have never actually used Debian before. It is the only distribution I know that really can be installed once and be upgraded from then on, without any nightmares or unresolvable incompatibility problems whatsoever.

Re:Simple; Linux brings too much overhead.

[Anonymous Coward]

Read the OP's post. This isn't about a distribution but the software it contains. And much of that software isn't backwards compatible. Sure, you can upgrade, but you have no guarantees what so ever that your stuff keeps on working.

they're still supporting me!

[destiny_uk]

well - interesting news. i have an existing support contract and they answered my support request today. having said that, i've never had a support issue with (the excellent) mysql support guys that was at all related to the OS.

fyi: if you run anything like a large site (we sustain 4000qps most of the day), i would highly recommend a support contract with them. it's very cheap for 24/7/365 access to the devs.

Re:Suse, Red Hat and ??

[codemachine]

Mandriva is probably the next largest distro to offer an enterprise distribution with a long support cycle. I had heard that the next version will be based on Debian, though their current ones are not (Mandrake Corporate Server).

Ubuntu's dapper also has an enterprise worthy product lifetime. Isn't it something like 5 years of updates being offered? Quite impressive for a free download.

Seems a Debian based distro will indeed be #3 one way or another. However, it may continue to be a distant 3rd.

Solaris is probably where it is at as far as having another choice of open source UNIX with enterprise support.

Not firebirt, though

[Anonymous Coward]

Since everyone seems to have forgotten the spat over the trademark dilution the Web Browser "Firebird" would have over the Database "Firebird". Y'know, people phoning up FirebirdDB support and asking "how do I open the interweb on this program of yours"...

Pah.

sqlite

[fyoder]

While it wouldn't be suitable for something like slashdot with a lot of concurrent reads and writes, a lot of little things for which mysql is really overkill could use sqlite. It's built into php5.

X not R!

[chaeron]

Why bother with RDBMSes at all?

Skip a generation and go with a good XML DBMS. Something like eXist perhaps? ;-)

Re:X not R!

[Anonymous Coward]

Why bother with RDBMSes at all?

Skip a generation and go with a good XML DBMS. Something like eXist perhaps? ;-)

You have got to be kidding. That is the most absurd concept I have ever heard. I'd be better off with a self designed binary or text flat file database, guaranteed.

Re:X not R!

[Bacon Bits]

I wish he was kidding. I've seen it tried. "Dog slow" does not even begin to describe it. Typically it goes like this: App v1.0 with SQL backend; a few years later App v2.0 is released with an XML DB; a few *weeks* later App v2.5 is released changing back to SQL backend. XML is good for data serialization. It was never meant as a data storage model.

The relational model survived ODBMS proponents by forcing people to use ORMs to get any kind of production grade system. It will survive this, too, and for similar reasons.

Wait for Ubuntu...

[mikelang]

...to join the list of supported distros. I hope it happens soon...

SCO

[Amazing Quantum Man]

Regarding SCO: Here's what Marten Mickos said.

Take with as much or as little salt as you desire.

sudo bash

[tinkerghost]

works on almost every Ubuntu system

Re:UBUNTU ! Why Hath Thou Foresaken Me ?

[StartCom]

Haha...that's the best one ever!

There is no fork!

[Dareth]

But if you start out spooning and things heat up a bit, it can quickly turn into a good "sporking"!

Gentoo Trap

[Dareth]

Gentoo is a trap for the overly technically zealous...

Gentoo was created to keep them out of the "proper" distributions. They will spend the vast majority of their time compling and recompiling their software hoping for that mythical extra 1% of performance that they never get to take advantage of before the next time they recompile.

Re:Gentoo Trap

[Chainsaw]

The real power behind Gentoo is not the ability to recompile all your packages with the -fzomgwtffast flag, but the ability to disable and enable specific features for all packages. Let's say I want a simple server for data storage and printing from Windows clients - Samba with Cups support. Gentoo allows me to install this without X11 and 140 different Gnome packages for some stupid control panel that will never be used. Want KDevelop but only support for C++ because you will never ever touch Fortran? Just specify "-fortran" and no application will ever add support for it.

Gentoo is about flexibility, and the ability to create a system with the features you want and need - nothing else. I use a very modest set of flags (-march=k8 -O2 -pipe) for my system, simply because I know that quadruping the compilation time to gain 1% performance probably won't generate a speedup anyway, since the slowest part of my system is me.

Support Through OS Vendor?

[Voline]

Well the OS vendor could offer support for MySQL, if they don't already. Debian isn't a company so you couldn't get it there, but Canonical could offer support for MySQL on Ubuntu. They could charge more for the extra service.

Progress? Wonderful database

[MrBoombasticfantasti]

I would talk of progress here, but Progress is by far the absolute worst database system i have worked with. :)

I've used Progress for the last eleven years (started with 6.x, don't quite remember) and I just love the way it just keeps working. No need for a DBA to keep the database happy. And the 4GL, oh, the wonderful 4GL. It really doesn't get the credits it deserves.

That said, I'm available for freelance Progress work... ;-)

Re:Progress? Wonderful database

[Chacham]

And the 4GL, oh, the wonderful 4GL. It really doesn't get the credits it deserves.

credits in being evil. :P

I had to deal with it on one job, and disliked it greatly. It was the text based version. I guess no complaints on the DB itself, more on the textual UI coding.

Re:Progress? Wonderful database

[topham]

You haven't by chance had the 'pleasure' of using Progress 4GL with Oracle as the database, have you?

It makes you want to slit your wrists.

And then there are the issues I have with the analysts who keep looking at the Oracle data and telling me I have to set a value based on what they see there; when the table name, and value in Progress are potentially completely different.

It's so much fun.
Really.

Debian might as well make a trademark-free policy

[Kjella]

I got a feeling it's exactly the same every time Debian Legal and trademarks meet:

"We want to put trademarked software $foo in Debian"
DL: "Cool, we'll package it up and apply any patches we need"
"Umm, what patches?"
DL: "Well, whatever we feel like, security patches for example"
"You'll have to run them by us"
DL: "No, we won't or at least we won't accept that as a requirement. DFSG."
"Look we got to control our trademark, but what if we gave Debian a license"
DL: "No, it has to be free for everyone. DFSG."
"Look we can't let everyone do whatever they want, because then there'd be no trademark."
DL: "DFSG violation. We need to rip out everything about $foo and rename it $bar."
"..."

Of course, if they really followed through on that they'd miss a few things, like a kernel etc. but it seems that's how they'd *really* like it to be.

Actually, that's Debian GNU/Linux

[massysett]

That's the name of the distro, not Debian Linux.

Well it take about 20 minutes more.

[mistralol]

Install RedHat on a spare machine / vmware.
Move your database to the temporary system.
Reproduce the problem.
Contact their support team.

Since when debugging things your going to be using the offline system anyway ?

The real questions are...

[DrHex]

MySQL is an Open Source project, correct? Yes.

that being said, is it the "Open Source" project which is only supporting these distributions?
The link takes us to a link for MySQL Enterprise. Since MySQL AB provides contract based service for MySQL, this is one way MySQL generates corporate cashflow which allows them to continue to work on MySQL.

Unfortunately, corporations have limited resources called humans who aren't always as zealous or passionate as people involved in an Open Source project. Therefore, they pick and choose which path to follow the money trail not their passion.

The concept of Open Source is not the same as the traditional monetary based business model.

Re:QUIETLY? Why not a message on their site

[walterbyrd]

That reads something like: "We proudly support Redhat, Suse, and Scox."

It would reasonably assumed that they don't support anything else.

The real question is down the road.

[Irvu]

The real question is what this will mean down the road. While the support contracts are soley for paying customers (who as it has been noted would probably be on supported enterprises anyway) that still directs the future of the DB.

The Mysql support staff are still a major force in the open-source aspects of the Database submitting bugs, patches, etc. If they cease doing Q&A on other distros then either the rest of the community will have to pick up testing their patches. Over tim I expect that this will, at least, bias the systems patchsets in favor of the "supported" distros.

The real issues won't show up today but 2 to 3 years down the road.

Their loss, not ours. It makes sense though really

[v3xt0r]

I personally do not agree with the 'Support Licensing Scheme' business model or services offered by these vendors trying to capitalize from the open source software development model.

However, some Managers do (generally those who talk more affectively than they manage IT), and most of those Managers will be using SuSE and/or Red Hat, as they would much rather pay some vendor for support, rather than face accountability issues when the ish hits the fan.

I had a similar debate w/ our CTO regarding the matter, and of coarse, he chose RHEL over Debian. Most of it came down to the issue of vendor support, whereas I felt that I am paid to be the support tech, so why pay them as well.

I guess the real question here should be: Why would a debian admin need to pay for support when he/she is (usually) experienced enough to find most of the answers on their own??

I think it would be funnier if

[bberens]

MySQL began supporting Unbreakable Linux.

Since it won't show up in the tags...

[singingjim]

...like whenever something negative about Windows gets posted, I'll have to do it in a post.

> haha

Apparently fewer and fewer people are actually giving a shit about linux. It's called thinning the herd. Natural Selection. Get over it.

Is it really a problem?

[element-o.p.]

I suspect that MySQL (the support company) took a look at the number of customers they had and compared that to the number of customers using various distributions, then made a business decision to offer support to the customers that made up the lion's share of their business. If they see that 90% of their customers use RHEL or Suse, then perhaps it simply cost them more money than it was worth to continue offering support for the small fraction of their (support) customers who run MySQL on Debian, Slackware, Gentoo, Ubuntu, etc.

IMHO, it isn't really a problem because, even though I've used MySQL personally and professionally for something like six years, I've *never* needed support on it. Let's face it, my Slack and Gentoo servers aren't "supported" either, nor is my Apache installation, my Postfix installation, my Bind installation, etc., etc., etc.--except by me, and so far, that's been good enough.

Re:Their loss, not ours. It makes sense though rea

[SoulRider]

I guess the REAL question is why would a business rely on one person to support their entire network? What would happen if you got into a car accident on the way to work tomorrow? The next admin may not be as knowledgable as you, are they suppose to start all over again with a different distro? One the new admin knows? By using a product with support included they are assuring that if anything happens to you they can still get the support they need. It's better to have a support contract and not need it than to need a support contract and not have it.

Why, yes I have...

[MrBoombasticfantasti]

I've used that interconnect monstrosity that pretended to transparantly couple Oracle to Progress. It did this by duplicating (or somesuch) the Oracle schema in the Progress dictionary. I'm not sure if it is still available, but I hope to never, ever encounter it again. Never saw such non-deterministic behaviour in a database as in that piece of work.

However, I really like 4GL for business logic. I really hate PL/SQL for that. I just finished a project building a large data warehouse in PL/SQL and that's something I'm never going to do ever again.

Next project will be Perl + Postgresql, quite look forward to that! ;-)

Excellent rebuttal

[brokeninside]

You start with an ad hominem attack, then appeal to authority (your own), then offer your opinion as a factual assertion.

Further, your post reveals an astonishing arrogance combined with a complete lack of comprehension. I'm just not certain whether it was a complete lack of comprehension of my earlier post or a complete lack of comprehension of what a database actually is. My guess is that it's the former.

A database is no more (and no less) than a container for the storage of data. MySQL is no less of a ``real'' database than latest and greatest from Oracle and IBM. The variation lies only in the features needed for a given application. You're basically arguing that supporting certain types of enterprise class features makes databases from the big vendors more of a database than other databases. That simply isn't true. That's like saying that a Cray supercomputer is more of a computer than my venerable old Palm Pilot. The truth is that both are equally computers but that they were engineered to meet entirely different sets of needs. They vary in complexity and capacity but they don't vary in their computerness. The same is true of MySQL and Oracle.

Re:Excellent rebuttal

[Chacham]

You start with an ad hominem attack

The funny thing is, while that is not true, you are actually commenting on your own comment.

I did not use an ad-homenem attack. I was stating how anyone who can make the remarks you made, obviously has never used the systems mentioned. For, if you had, you would clearly realize the difference. I hold this truth to be self-evident, at least to anyone who has actually used the different systems.

Now, by claiming that that was indeed an ad-hominem attack, you have actually done it yourself! Touche! :)

then appeal to authority (your own)

Replace "authority" with "experience" and it would be an accurate statement.

then offer your opinion as a factual assertion.

I offered my first statement as inferring the factual inference, that anyone who has used the systems would know. Afterwards, the reply is subjective in nature, and never did i try to pass them as facts.

Further, your post reveals an astonishing arrogance combined with a complete lack of comprehension. I'm just not certain whether it was a complete lack of comprehension of my earlier post or a complete lack of comprehension of what a database actually is. My guess is that it's the former.

Again, empty words. You are talking more about your own comment than mine.

A database is no more (and no less) than a container for the storage of data.

No. We usually call that a file. A database uses files (or their equivalent) to produce a more complex function.

MySQL is no less of a ``real'' database than latest and greatest from Oracle and IBM.

<ad homimnem attack>
Bwahahahahahahahahahahahahahahahahahahahahahahahah ahahahahahahahahahahahahahahahahahahahahahahahahah ahahahahahahahahahahahahahahahahahahahahahahahahah ahahahahahahahahahaha!

I feel better now.
</ad homimnem attack>

The variation lies only in the features needed for a given application.

If the end justify the means, that statement would be accurate. If not, it doesn't.

To an application programmer, the database is just the means for their more important end, and MySQL often supplies it, so its being a database or not is largley irrelevant. To a DBA or DB programmer, however, the Database is more important than the ends, and thus just getting the job done is not enough.

You're basically arguing that supporting certain types of enterprise class features makes databases from the big vendors more of a database than other databases.

Do you consider my example, which was the erstwhile deficiency of transactions, to be an enterprise feature?

Re: Completely and totally misleading

[Tilex]

My company has a MySQL Network Platinum support contract on 2 Debian servers (which I guess is now called MySQL Enterprise). I contacted MySQL today, and they told me that they are NOT dropping support for Debian. All this means is that they do not have .deb packages MySQL Pro on Debian or Ubuntu -YET-. Their generic Linux tarballs will still install with no problem in Debian, and they will still continue to support users on the Debian/Ubuntu platform.

But here's the best part...

He also indicated that MySQL is in the process of creating Pro .deb packages for Debian AND Ubuntu.

So not only is this article inaccurate, it's actually 180 degress opposite of accurate. They're not dropping support, they're increasing it.

Here it is straight from the MySQL reps email:

"It is not yet fully supported which simply means that we do not yet have official
enterprise binaries for this platform. We will continue to provide the same level
of support and we hope to have both Debian and Ubuntu packages in the near future.
I will let you know when we have enterprise .debs, until then you can simply use
the 5.0.30-enterprise generic linux tarball package."

I tend to believe MySQL.

Build clusters?

[msobkow]

I know IBM and other vendors have clusters available for developers to do test ports.

Is there any sort of facility for mapping a project source tree to multiple distribution builds? Sort of a meta-ant for the different vendor packagings. If the efforts to standardize the Linux distributions was effective, the build and bundle should be the only real difference anymore. Especially if you were sticking to POSIX APIs for the core OS services.

DRM can be turned on it's head. Each developer in charge of an OSS project could use such a build cluster, maybe hosted or supported by various vendors or a .org. The build manager for the project would effectively sign the source and binaries, then the various distros would be responsible for integration test and rollout.

A lesson here

[synthespian]

Let this be a lesson regarding the GPL and the dual-licensing trap some companies set up, such as MySQL. What we see is the unfolding of another development of the loophole that the GPL license creates.

It just proves there is no dual-licensing choice. This is effectively a proprietary licensing scheme (or scam). It's just another form of making the customer fall prey to the vendor. Now we see yet another facet of this loophole: the company ties support to vendors that charge per-seat licenses. A perfect scheme, a +/+ game for the vendors, both of the software and the OS. You loose, sucker.

The BSD license does not have this loophole, and leverages the playing field for everyone. You want to "close" the BSD solution, and package it as a proprietary solution? Do it. You want it as free software? It's there. The GPL, on the other hand, by a flaw in design is used for the type of maneuvering we see in this case.

Is it any wonder Google has chosen non-GPL licenses for a lot of their released open-source code?

what do you expect

[cinnamon colbert]

first, people say forking is good, then they complain that vendors won't support unlimited forks. YOu can't have it both ways.

Re:Suse, Red Hat and ??

[Anonymous Coward]

This was one of the same arguments used back in the mid/late-90's by Sun. Didn't go far then, won't now.

What you're forgetting is that there will always be more developers than "premium" Linux distro development companies, meaning not all of them will work for "premium" companies. Add to this the urge to create and solve problems, and I don't think you'll see things change all that much. If things go too commercial for a package, some other one will come along and replace it.

The other side of the coin is that Microsoft may not have long anyway. The reports I'm receiving (both net-based and private) are of no-joy for Vista.

Just staying competitive

[waTR]

They are simply trying to make a balance in the industry to Oracle Linux + DB by concentrating on partnering with RedHat & Novell as those are the only two other businesses that are competing with Oracle Linux. Together they are forming a united front so as to force a balance of power in the DB market (including OS for DB).

-eof

RedHat doesn't do closed source

[jmorris42]

> Except that Red Hat and Suse are not entirely open source.

Suse has had a spotty history, having several closed components for most of its history, becoming mostly Free after being bought by Novell and now with the Novell-Microsoft deal who the hell knows where they are going next.

RedHat on the other hand has always had a Free core distribution. In the past they have experimented with selling closed addons (Motif, Applixware, etc) and even had a couple of closed things in their boxed distros (a browser called Red Baron and a commercial X server) but long since dropped all that nonsense. No version of RHEL has any dependencies on closed source components. Paid media sets do include an "Extras" CD with the usual suspects (JDK, Flash, Acroread, etc) and there are a fair number of vendor supplied closed device drivers for RHEL, but nothing closed is required to install unless you have one of the pieces of hardware that require a driver, such as a SCSI controller.

There isn't anything with a RedHat copyright that isn't released under a license that passes the DFSG. They don't officially make binary packages for RHEL available but every SRPM is up on the mirrors and full installable binary rebuilds are available from a variety of 3rd parties. What more are they expected to do to avoid the "Microsoft of Linux" moniker?

Bitch about their engineering decisions, the quality of their support, etc vs other vendors. Reasonable people can disagree on any of those subjects, but the non-free rumors need to be put to pasture.

Are we surprised?

[petrus4]

I can feel my karma taking another bruising here, but what the hell.

Debian want to fork all and sundry, or at the very least create internal patchsets for pretty much everything they get their hands on, and then people don't understand why they don't get support from upstream vendors?

Said people might support their own applications, but they are under no obligation to support Debian's own non-standard, patched versions of their apps.

On a related note, I had my last negative compiling experience with Ubuntu the other night, and as a single individual at least am hereby boycotting Debian and its' derivatives entirely. The reasons are too numerous to count, but are both technical and political.

As a positive alternative, I advocate a rennaisance of Slackware. It's a clean, sane, non-fragmented distribution. It doesn't use any seriously concrete form of package management by default, meaning you're free to choose your own...and it also by extension doesn't use a particular, unholy form of perverted evil known as subpackaging. It also doesn't see Linux's heritage as a UNIX clone as something to be ashamed of, or a hindrance to the goal of creating a perfect imitation of Windows.

Going back to the parent topic, Slackware would also likely be great to use as a standard for vendors such as MySQL, *because* the Slack developers largely abstain from downstream patching, (at least AFAIK) and as mentioned are package management agnostic. Hence, it'd be a very easy distro to support.

If anyone here hasn't tried Slack themselves, I thoroughly suggest it. You can look forward to a level of transparency and reliability you'll scarcely find anywhere else. It's a form of Linux which isn't afraid of being itself, and that tragically is a very rare thing these days.

Bad title

[noz]

MySQL Quietly Drops Support For Debian Linux

*yawn* Another poorly named /. article. Troll? Maybe. Does this happen too often? Yes.

PostgreSQL support.

[Anonymous Coward]

Sun Microsystems: Sun Microsystems provides uptime support, including 24x7 enterprise support, for PostgreSQL 8.1 and above running on Solaris 10.
Hours: 24/7. Number of employees: 30,000 world wide

Fujitsu: Fujitsu provides a full range of value-added services to PostgreSQL business-users including support, training, migration services, consulting, hosting, performance optimisation and system integration.
Hours: 24/7. Number of employees: 150,000 world wide

EnterpriseDB Corporation EnterpriseDB Advanced Server is an enterprise-class relational database management system based on PostgreSQL, the world's most advanced open source database.
Hours: 24/7

Command Prompt, Inc: Command Prompt, Inc. is the oldest and largest dedicated PostgreSQL services company in North America. With over 10 years of profitable debt free operation.
Hours: 24/7

It is several others. Botrh large and small with different specialities. There is a partial list on the PostgreSQL website:
http://www.postgresql.org/support/professional_sup port

Re:PostgreSQL support.

[Jamesday]

Yes, I know about those useful options, it's why I mentioned the third party support.

Do those selling PostgreSQL support have service level agreements with, and support backup and software indemnity from PostgreSQL?

Those are some of the things that distinguish the level of support MySQL can provide via third parties, which, so far as I know, isn't and never has been equalled by any support offering for PostgreSQL.

Then there's MySQL's own direct support available if you don't want a third party. How about the ability to phone up whoever they need in the team at home 24/7 if necessary, not just the odd one or two they may have hired?

I'm not suggesting that there's a lack of good support available for PostgreSQL, what I am doing is highlighting some of the differences that are of interest to those who have 24/7 needs for critical business systems. Not everyone needs this, but for those who do, it's available for MySQL and isn't available for PostgreSQL.

Re:PostgreSQL support.

[mccoma]

Just to be fair to the PostgreSQL support companies, it should be mentioned that the MySQL people don't own all of the code you use - Oracle owns two major pieces.

Re:PostgreSQL support.

[Jamesday]

True. BDB is largely a non-event on the support side - it's extremely rare for us to get questions about it, since it's largely been replaced by InnoDB. Now it's being removed from the newest MySQL version.

For InnoDB, owned by Oracle, we have the service agreement and phone numbers. Though I know InnoDB so well that I haven't had any need to call Heikki about it. Another person in the support group has regular chats with him, discussing priorities and developments and such. Works well and the rumors of doom when Oracle bought them just haven't materialised: we retain an excellent working relationship.

Just another reason...

[tuxtastic]

...to use PostgreSQL

XML FTW

[furbearntrout]

Store your data tree as nested tags in a flat file. This should help with searches and sorting.
(going for + funny, not -1 troll)

MySQL response: MySQL's Commitment to Debian

[Jamesday]

Here's a copy of MySQL's official response to this story:

-----

MySQL's Commitment to Debian
December 13, 2006

MySQL AB apologizes for any miscommunication that may have implied that the
MySQL database does not run on the popular Debian Linux operating system, or
that the company does not offer technical support for MySQL Enterprise
subscribers using Debian.

We have a strong commitment to Debian and other forms of Linux - for both
open source community developers and corporate enterprises.

The Debian Linux operating system is an active, growing and successful
platform for the MySQL database to run on.

Our company offers freely-available downloads of the MySQL Community Server
in source code and binary format at
http://dev.mysql.com/downloads/mysql/5.0.html for Debian and other flavors
of Linux -- including Red Hat Enterprise Linux, SUSE Linux Enterprise
Server, Fedora, Ubuntu, etc. -- as well as Microsoft Windows, Macintosh OSX,
Solaris, FreeBSD, HP-UX, IBM AIX and SCO OpenServer.

For paying customers, our company also offers 'MySQL Enterprise', a
comprehensive set of production-tested software, proactive monitoring tools,
and premium support services.

Since its official launch in October, we have delivered versions of the
MySQL Enterprise Server software for RHEL, SLES and a general-purpose
version that runs on other forms of Linux -- including Debian. Starting in
Q1 2007, we will also deliver regular software updates for the Debian and
Ubuntu platforms as well.

As in the past, MySQL AB continues to offer paid technical support for
customers running MySQL on Debian and other versions of Linux. This is
available as part of our MySQL Enterprise subscription service. A complete
list of MySQL Enterprise supported platforms is available here:
http://www.mysql.com/support/supportedplatforms/en terprise.html

We will continue to monitor the popularity of other operating systems and
user requests when considering extending our platform support in the future.

Again, MySQL AB regrets any inconvenience this misunderstanding may have
caused.

-----

James Day, Support Engineer, MySQL AB

Stop spreading FUD

[Local Loop]

I've had apps hammering hard on MySQL since 2000. I've never seen any data corruption. Give an example or go home.

Re:Stop spreading FUD

[J.Y.Kelly]

I've had apps hammering hard on MySQL since 2000. I've never seen any data corruption. Give an example or go home.

I suspect the parent was talking about stuff like this (slightly reformatted because of the bloody lameness filter!):

mysql> describe test;

| Field | Type | Null | Key | Default | Extra |
| text | varchar(10) | YES | | NULL | |
| date | date | YES | | NULL | |

2 rows in set (0.00 sec)

mysql> insert into test values ('This text is too long','10000-01-01');
Query OK, 1 row affected, 2 warnings (0.00 sec)

mysql> select * from test;

| text | date |
| This text | 0000-00-00 |

1 row in set (0.00 sec)

Sure there are ways to mitigate this, but I've been bitten by this in applications. Having said that I still use MySQL for pretty much all of our database needs.

PS: Go on - ask me why I needed to store dates beyond the year 10,000...

Re:Stop spreading FUD

[Jamesday]

There have been occasional issues; I'm actually a keeper of an internal list of bug fixes for corruption issues in MyISAM. It's getting really tough to find and reproduce them these days, as the bugs have largely been beaten back to seldom-used combinations of situations over the years. Still, we do occasionally get new reports (generally of obscure ways to corrupt an index) to track down. I don't recall anything resembling what was described. InnoDB checksums the data pages and hence works well as a dodgy RAM and hard drive detector and we see that regularly.

If it happened it was probably either a really ancient version or someone ignoring upgrade instructions. Or, worse, downgrade instructions.

Like you I've really hammered MySQL in production, with a billion or two queries a day and a few hundred gigabytes of data. Generally speaking, it simply works. Which is in part why Wikipedia and most of the other (Alexa) top ten sites on the net are MySQL users. Not at all bad for a database you have to pay exactly nothing to use.

Still, reality applies, MySQL is software, so MySQL will always have bugs, and sometimes someone will discover a cute new one.

Re:Stop spreading FUD

[Local Loop]

Well I don't really count that as corruption, because the behavior is by design and well documented. Personally I'd prefer to get an error...

OK, I'll bite, why do you need to store dates beyond 10,000?

Re:Stop spreading FUD

[Anonymous Coward]

Not that i like MySQL very much, being a pgDude..
But where is error?
That the text is clipped? That the date rolls?
I don't get this one..

Re: Completely and totally misleading

[tinkertim]

Your exactly right. At first, nothing really is changing. MySQL will continue to build binaries and Debian will continue to package them appropriately (albeit slowly) into stable release trees, and it looks like some new packages (and changes in meta packages) are in sight 'down the road'.

The same thing applies to Ubuntu, and every other Linux distribution that actively maintains its packages. Its up to the packagers (as always) to make things available.

Yum / aptitude / emerge (whatever) do not download packages from MySQL , they download packages from whatever you put in their list of sources.

Amazing what people will do just to get a story posted on /.

This is fud. The story here is that soon better packages will be available as package maintainers begin to include better things from MySQL.

This is FUD. Nothing more to see here, move along .. move along ..

debian

[hyperstation]

does anyone even really use debian anymore? isn't it "functionally extinct", like that poor dolphin species?

Re:Their loss, not ours. It makes sense though rea

[v3xt0r]

No arguments there! =)

That's the cover-my-ass (to avoid accountability for my lack of management oversight) way to do it.

A) Managers should enforce strict oversight of documentation practices for network admins and software developers, which a lot of places don't do.

B) Developers and Administrators should document their work as they go, so that others can pick-up where they left off, which a lot don't do.

If these processes are already in place, which is usually the case at most structured IT shops, then there should be no reason why you have to rely on the support of some 3rd-Party vendor.

But I agree, it's not a perfect world, and sometimes it's good to cover your arsch! =)

God bless DFSG.

[jetxee]

Well said. And this is the reason why I prefer Debian over any suse. DFSG.

Python support for MySQL on Windows is dead.

[Animats]

Just recently, support for Python's connection to MySQL, "MySQLdb", stopped working on Windows. See this discussion in the MySQLdb help forum on SourceForge.

Unlike Perl and PHP, the standard Python distribution doesn't support MySQL. There's a third-party add-on on SourceForge for that. It has one developer, and he's not interested in maintaining the Windows version. The Python 2.5 update apparently broke the Windows build.

Some help is being provided by a World of Warcraft guild, which has managed to build MySQLdb for Windows. But that hasn't been tested by anyone else.

Also, although the current MySQL understands Unicode, and the current Python understands Unicode, the MySQLdb module in the middle is reported to crash on Unicode.

I'd thought something as basic as a database connection for a language used primarily on web servers would be a solved problem, but for Python, it's not.