Slashdot Mirror
Sybase Releases Free Enterprise Database on Linux
Tassach writes "Sybase announced today that they are releasing a free (as in beer) version of their flagship database for Linux. The free version is limited to 1 CPU, 2GB of RAM, and 5GB of data, which is more than adequate for all but the most demanding applications. This release provides a very attractive alternative to Microsoft SQL Server, and gives developers and DBAs an extremely powerful argument to use against the adoption of Microsoft-based solutions. For those who are unfamiliar with the product, Microsoft's version of Transact-SQL is nearly identical to Sybases's. This high degree of similarity makes porting applications between the two platforms very easy. Sybase is supported by numerous open-source projects, including sqsh (SQL shell), FreeTDS, and SybPerl."
Comment removed based on user account deletion
People who just want a full featured database aren't really all that interested in the source.
Sybase have made a very smart move - by association people will see thier larger user base as a sign that they are more stable, and more widely used.
They will be the 'oracle' of linux. Of course this is first impressions, I haven't used Sybase, or Postgresql - only oracle, mysql, mkcoi and db2 (oh that toy database, from a company in redmond?)
Anyone had experience with Sybase ?? Anyone using Postgresql for really heavily loaded DB?
Any real differences in todays markets? (patch reliability, support)
I am not a db administrator.
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
Well I want a full featured database and would like to have the source. That makes it "people - 1".
they are usually interested in using more than 1 cpu and just 5Gb.
the sybase offer is useless.
Where I'm working, we use MS SQL Server an awful lot. And we write an awful lot of code that uses it. We don't spend any time reading source, or trying to find holes / improvements to it. We're too busy.
They're giving you the product for free, but you don't want it because you can't have the source.
I'm sure they'll miss you.
More free shit to play around with. It's great for them, I play around with it, I learn it.
Soon they will have another tech that knows how to operate it. A small business may end up using it so they can ditch the MS SQL stuff and move to a more robust enviroment. They hire a admin, he needs a assistant? I am aviable, and I trained myself enough to be familar with it.
Whoopie.
Then as the company grows, so will it's need. If it's a decent product then they'd definatly pay for it after using it for year or two for free.
Best damn advertising you can hope to get. Got to love it. Sure beats the snot out of ending up being another MS victom and another footnote in history:
"So and so company had a product similar to the insanely popular MS Widget. Although widely considured superior to MS's solution by a large part of the industry, MS's continued dominace of the desktop arena gave the leverage nessicary too".... blab blah blah
Did I mention I also get some free shit to play around with? (given a choice between free and Free, Free usually wins, but we'll see how it goes)
The limited database size of 5 gig makes it worthless to just about every Open Source web site or developer other than the very very small guys who would rather use something like MySQL or PostgreSQL instead.
How exactly is it helpful to release a free version that most people can't use in real world applications? The answer is, it isn't.
Move along people nothing to see here.
Shareware strikes OSS back!
What happenes when Sybase stops maintaining the `Free' version?
Microsoft's version of Transact-SQL is nearly identical to Sybases's.
Just how nearly is it? I'd like to know in terms of things just broken enough to make finding them absolute hell.
When they say it is limited to one proc, do virtuals count (ie, P4 HT?)
.
.
I've never looked at Sybase and have no clue how it works; especially their licensing . .
I'm assuming if I have a true multiproc system, it's only going to utilize one physical proc . .
Anyone have the dirt, I couldn't find a detailed link on the limitations other than the single blurb that was in the original post.
Closed source, no guarantees, too many limitations...
And btw, sybase sucks, especially if you're on your own without support.
MS, Sybase & Ashton Tate jointly developed the core engine up until the mid nineties at which point there was (I believe) a very acrimonious split due to some licensing argument. That was at version 4.2 and so the SQL syntax remains common between the two. Since then I'm sure there's been a certain amount of divergence (and then some!), but it theory porting should be easyish (famous last words).
I used to be a Sybase DBA and still dable with it a bit. It's a very nice db, and at one time was a real contender against Oracle. It still has a very strong footing in the Financial sector as it was deemed to be faster than Oracle. In todays world of cheap hardware and spare cpu cycles I don't think that's quite as important.
All your (Sy)BASE are belong to us!
I fact, I've been waiting for free-download Oracle/DB2 "personal database" or some limited opensource release of Oracle/DB2 for a while. This release will put much pressure to Oracle, IBM and of course, MS. This is one major strong point of Linux, which pretty much is ignored by the press. With MS solution every small piece of software is at least shareware, and while the cost might be nominal, you still have to go trough the process of buying/registering it. With Linux, you may have to buy some software, but most of the stuff you need can be found around the net, just couple of clicks away from being ready for you to use.
Auferre trucidare rapere falsis nominibus imperium, atque ubi solitudinem faciunt, pacem appellant.
Take an example of when an individual's age suggests this individual is an infant. In this case, I would immediately disable the widget that receives anything to do with children since an infant cannot have children. There is much more...all in the name of business logic. Cb..
anything to improve the current mysql replication situation..
The database caps of the free version are high enogh for the product to be usefull for web applications and smaller projects, a market that is completely dominated by free alternatives such as Postgres and MySQL. Almost everyone who shelves out $$$ for a database server run much larger systems.
I bet they are hoping that by giving away the product for free to people who would never buy it anyway, they get droves of people who are experienced at running their system who will eventually buy it for larger projects 'cause that's the system they know how to use.
Kind of like how SUN sells computers to universitys dirt cheap.
Try out fish, the friendly interactive shell.
I worked with both and, IMO, Sybase is NOT that much better to justify restrictions like "the free [Sybase] version is limited to 1 CPU, 2GB of RAM, and 5GB of data". Does it work when it rains?!
MS-SQL was licensed FROM sybase. Then Gray came on board and fixed it in the 6-7 era.
It isn't necessarily the speed - it's also the ease of maintenance compared to Oracle. Sybase is limited but it works, while Oracle is very powerful and fiddly. Fiddly is bad when you have a lot of dataservers.
Mr. Goatse is back with a vengance...
Don't click this link! It's a porn shock site!
Who modded this up?
my god. i've never seen anything so pathetic and obvious case of brainwashing in all my life.
"...extremely strong argument against Microsoft solutions" a baseless and deliberate assumption like this in injected into what seems like 75% of slashdot stories. This is stupidity in the extreme. if MS makes the more appropriate solution, you damn better well pick them, for your own server-monkey sake
if your motivation is just 'no microsoft at any cost', you are a tool of all these jerks riding high on the general sense of ill-will they cultivate and the work of those countless volunteers that built the friggin apps they are pimping (esr, i'm looking at you)
It's as if slashdot's mission is more about indoctrinating the anti-MS mindset than championing free software.
here's a good point to remember: use the right tool for the job. it's as simple as that. sometimes it's MS. Sometimes it's your favorite fanboy project founded on idealism and granola. sometimes it's also Redhat or IBM (think big $$$).
having your ability to assess these tools tainted by such fervid invective, knee jerk hatred as slashdot gots to offer, well, you are going to reap what you sow.
see comment
-- MartinG To mail me: echo kewyjlcxyzvjfxbqwh | tr bcefhjklqvwxyz
This is, IMHO, a very good thing. Sybase is an excellent DB (I've worked with some top developers who swear by it) and a much better option than MsSQL for small projects where 1 proc/5Gb is more than enough (and there are lots of projects like that). Considering that no client I've ever worked with has been willing to go for PostgresSQL/MySQL (preferring MsSQL), this will be a welcome victory against Microsoft, as Sybase has a pretty good name in the db game.
Other than that, Mrs Lincoln, how did you enjoy the play?
You are missing the point. It makes it easy to convert from Microsoft SQL. Imagine thousands of independent software developers with an alternative to MSQL within easy reach. Their entire solution cost is now reduced, and they will sell better. At least the ones that take the chance.
I'm not sure so many independent software developers use MS SQL anyway, but there has for a while been a light version of MS SQL, MSDE, available for a free download, with most of the features of MS SQL but with similar restrictions to this Sybase offering.
But this appears to be targeted mostly at Linux developers so it's competition for PostgreSQL and the Abomination That Shall Not Be Named.
postgresql officially supports Windows from version 8, then how can Sybase on Linux suddenly claim "a very attractive alternative to Microsoft SQL Server"?
depends who you are selling to ... my company sells to a few (UK) local government authorities and SQL Server is a "tick in a box" on their checklists. Sybase currently isn't, but being a "brand name" will probably help it there.
Postgres doesn't even come into the equation
...that microsoft sql server is sybase (albeit 1993 codebase)
This is good. Though people will complain, they always do, about everything.
:)
While I'm a pgsql myself, the more the merrier. As long as there are many differend dbms's we'll all be safe, because homogenicy is the root of all evil.
This will hopefully help Sybase stay in buisness longer thanks to the increased popularity it will give them, which therefor is good for me as a pgsql user.
Simply because improvements caused by competition and the lack of common ground for root exploits.
Now, if only MySQL would just die we would all be better off
Imagine thousands of independent software developers with an alternative to MSQL within easy reach.
I also still dont get it. we converted from MsSQL to postgreSQL easily. a simple program converted all data over a weekend (3 seperate databases with over 10Gb data in them) and the software changes were extremely minimal.. SQL syntax differences are not difficult.
Yuo cant simply point your app at the new database and let it rip anyways, changes have to be made to your apps no matter what DB you switch to.
Do not look at laser with remaining good eye.
If you're expecting to take a recent Microsoft database script and run it on Sybase without any problems you're dreaming.
I have worked with both a few years ago (migration from Postgres 6 to Sybase 11) and sybase performance was outstanding, compared to Postgresql.
I hope postgresql performs better now.
Moreover, this is not the first time Sybase makes this offer : Sybase for Linux 11.0.3 was free to use on Linux, with no limitation.
I personnaly used it for my Web shop, as this database is not only fast and secure, but also quite easy to program with ( especially compared to Oracle ).
The only drawback of Sybase is the lack of standard administration Tools. You have to use a product like (overpriced and windows-only) Emabarcadero DBArtisan.
I fact, I've been waiting for free-download Oracle/DB2 "personal database" or some limited opensource release of Oracle/DB2 for a while.
I may be way off-base here, so apologies if I've missed the point, but Oracle have allowed free-downloads for at least a couple of years: Linux version of Oracle 10g.
Not free-as-in-speech, and if you want to deploy it commercially it's not even free-as-in-beer, but it does seem to meet your "personal database" criteria: it's the reason I've more Oracle experience[1] that SQL Server experience (though MSDE briefly threatned to change that - to some extent).
I'd need to check, but a few years back DB2 was also a free download, with the no-commercial-depolyment caveat. I'd be surpirsed if it still isn't; it's a neat trick to get developers hooked on cheap/free versions so that their organisations then migrate.
[1] Twice as much - a whole extra week ;)
This is where the serious fun begins.
I downloaded a free version of Sybase for Linux back in 2000. Wonderful product, arguably easier and better designed than a certain other big RDBMS player.
You are missing the point. It makes it easy to convert from Microsoft SQL. Imagine thousands of independent software developers with an alternative to MSQL within easy reach. Their entire solution cost is now reduced, and they will sell better. At least the ones that take the chance.
OR they could keep running their ASP.NET applications on IIS and use SQL Server 2005 Express which happens to have the 1cpu limit, 1gb memory and 4gb database and is free as in beer. You're right.. the solution cost is reduced but not in the way you intended. :)
Simon.
does everything that runs on Linux have to come with the source and an oss license???
I bought and paid for Textmaker for Linux and also Opera for Linux, both closed source programs. This Sybase move now means that I can download and play with a serious database. It's a smart move because it means that I will be gaining skills in programming for that database engine, skills which are seriously marketable.
Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
Most distributions can't ship the client libs for the closed source databased. Thats makes it's somewhat difficult if you like to use things like a distro shipped version of php.
We use sybase at work and I try to use freetds as client lib ehenever possible because it's easier to maintain (The ebuilds are alreay there in gentoo).
You can already download for free, Oracle for Linux,Windows and a few more platforms. All you need is an OTN membership. However its only for Non-production use i.e. you cant run your business off it.
As for Open-sourcing the DB engine, you can keep dreaming though..
5 gigs is real pittance for the amount of data being collected these days, almost making this useless. What happens when the 5 gigs fills up ?
Skype Me! username: john_allen_mohammed
cmon Firebird(Interbase) is much better than this strapped DB http://www.sphere-data.com/docs/ib_vs_ss.shtml
PostgreSQL 6.X is ancient. PostgreSQL 7.0 was released in May 2000, over 4 years ago.
I think I can safely say that PostgreSQL is dramatically better in every aspect since the 6.X days.
Social scientists are inspired by theories; scientists are humbled by facts.
On completely unrelated note, what ever happened to IPv5? We're all running IPv4, migrating to IPv6, and where did the 5 go?
Ah yes... but what if you are stuck on an application that you don't have access to the codebase? With this "most" ms-sql or Sybase SQLAnywhere applications can simply be told where the new datastore is and work.
We hae several POS and ERP applications on our campus that have been locked into MS-SQL or SQLAnywhere (bleh!). Yesterday after downloading sybase and getting it installed I was able to transfer and fire up test instances of 7 of the 9 applications without ever needing to ask the company that wrote it to make any changes for me.
Would I prefer these apps be FOSS... YES! We are slowly writing new versions as we get time... but it takes time and this gives us a way to save money now.
Telcos have alot of dark fibre in the States. Most people assume that's optical fibre...but it's actually moral fibre.
What is the problem with ANSI joins in Oracle?
I would say that's a very common mistake. I've spent a LOT more time downloading and building stuff and man it's hard.
Bleh, nothing is a couple of clicks away except total destruction of your boxThe fixed ABI has its own problems - see Opcode DB. (of course the problem's all due COM with the a.pVT->xhx() calls).
Don't delude yourself about anything in Linux being a click away. Shareware you pay with cash, Free Software with your time - I've had to hack proxy support into at least half-dozen things that has crossed my path.
Quidquid latine dictum sit, altum videtur
Definately beer, but beer they need a company name and phone number for, not everyone in this world has a job and a phone number.
:S
But then I'm bitter because my download took 2 seconds got corrupted and refused any furthur download attempts
What is the problem with ANSI joins in Oracle?
They work now, in my experience of working with Oracle 9i, but they *are* a relatively recent innovation (introduced in Oracle 9i I think). Certainly SQL Server supported ANSI joins well before Oracle did; perhaps grandparent is remembering an older version of Oracle.
Will a dual core chip be a Single CPU ?.
Is HyperThreading treated as a dual CPU ?.
And if you treat them differently , they are still a single socket chip, so why the discrimination ?.
Quidquid latine dictum sit, altum videtur
The thing is free whatever the motivation behind it. If you don't like it then don't use it so stop whining. What did you expect , them to say "Here , have our full database system for free, no restrictions! We've planted a money tree in our garden, we don't need sales anymore!".
Wake up to the realities of commercial life , its what keeps the worlds economy running.
If you think that "SQL syntax differences are not difficult", it's because you know neither MSSQL or Postgres. MS offers many features that once you start using them, are hard to move away from. In particular, stored procs that return multiple result sets with different schemas, nested transactions and distributed transactions with DTC.
Over time OSS dbs will have these features, but not yet.
The advantage with Sybase is that MS SQL Server is based on an earlier version of Sybase - though there are now some differences, the method of accessing the database server, of doing backups and the version of SQL are either identical or so similar that it is trivial to transfer to Sybase.
Don't delude yourself about anything in Linux being a click away. Shareware you pay with cash, Free Software with your time - I've had to hack proxy support into at least half-dozen things that has crossed my path.
No, shareware takes your cash and your time. Free software takes only some time -- today it's usually very easy to get working. And even if you pay, it still might not work. With free (as in speech) software, you at least have the source.
What you say might have been true for RedHat 6.x and related applications. Today, the claim that MS software takes less time to get working and/or administer is completely untrue.
Auferre trucidare rapere falsis nominibus imperium, atque ubi solitudinem faciunt, pacem appellant.
I fact, I've been waiting for free-download Oracle/DB2 "personal database" or some limited opensource release of Oracle/DB2 for a while.
So, err, maybe you wanted this?
DB2 UDB Personal Developer Edition
Cheers,
Toby Haynes
Anything I post is strictly my own thoughts and doesn't necessarily have anything to do with the opinions of IBM.
Did not they use to have this exactly product (ASE) several years ago free to use on Linux? (IIRC it was a version below their then-current product).
And anyway - who needs ASE? Had they offered ASA then that would be talking.
Huh? Please explain. I'm genuinely curious.
One simple rule for its versus it's
by slashdotting their site to a grinding hault.
Guess they are taking advantage of their own linux promo.
I fact, I've been waiting for free-download Oracle/DB2 "personal database" or some limited opensource release of Oracle/DB2 for a while.
Someone please mod parent down. Free-downloads of all kinds of Oracle products (including their full RDBMS database, or the Lite version, or the personal version, or almost any of their products) have been available for about 20 years. Anyone who didn't know this would have to go through this very complicated procedure:
1. Go to Oracle.com
2. Click on the "download" link at the top of the page.
I don't respond to AC's.
Right! Here's something else to consider...in the past, when you had an application people might want to evaluate before they committed time and money to buying the full product, what did you do if it required a database and you were concerned your clients may not have an existing database implementation? Exactly...you included the free MSDE engine so people didn't have to go out and spend money on MS SQL or Oracle for what was only an evaluation. If it worked out well and the customer bought the software, they now had a database which, if MSDE wasn't up to snuff for a full production deployment, could be painlessly migrated to MS SQL. The engine is exactly the same, so no translation is necessary.
This worked out so well, precisely because MSDE was free to redistribute and easy to migrate to MS SQL, that MSDE is now included with thousands of applications. And remember -- if you ever outgrow its limitations, it can be directly moved over to MS SQL.
Coincidentally, MS SQL (which, as everyone is ecstatic to be able to point out, used to be Sybase) continues to gain market share. Sybase (see above) does not.
The big three at the moment in terms of market share are Oracle, IBM, and Microsoft. Oracle is #1 but is slowly losing market share to IBM and MS. Sybase is #4 -- but that #4 translates to 3.6%. And it's static -- they're not gaining any of that market share being lost by Oracle.
Someone please mod parent down. Free-downloads of all kinds of Oracle products (including their full RDBMS database, or the Lite version, or the personal version, or almost any of their products) have been available for about 20 years.
I know -- I have a copy myself. I also have an evaluation CD of IBM DB2.
But these products are positioned for evaluation of for the professionals. They are not positioned positioned to compete against SQL Server, MySQL or Postgre.
But this is the whole point of SyBASE offering. And my point is that Oracle and IBM are soon forced to respond.
They are not likely to offer their full product as GPL/OpenSource or even a free download, because it would be too risky and might cost them much of their software licensing business.
But they are likely to produce a more limited version; either with a restrictive license -- or with less capabilities -- or a free download product with no source. But way or another, they must eventually respond.
Auferre trucidare rapere falsis nominibus imperium, atque ubi solitudinem faciunt, pacem appellant.
But they are likely to produce a more limited version; either with a restrictive license -- or with less capabilities -- or a free download product with no source. But way or another, they must eventually respond.
Oracle Personal is free.
I don't respond to AC's.
Thank you for clearing that up. I remembered that Syabase used to have teh same code base and engine as MS SQL. I wasn't sure where the similarites/differences were these days.
The Kai's Semi-Updated Website Thingy
One of the things that is missing from MySQL is stored procedures, and quite honestly, that's a bit of a pain for me. There's ways of working around it, but it's really useful.
See this link.
You may not: use the programs for your own internal data processing or for any commercial or production purposes, or use the programs for any purpose except the development of a single prototype of your application;
Please tell me, how this product is comparable in terms of licencing to the SyBASE offering?
Auferre trucidare rapere falsis nominibus imperium, atque ubi solitudinem faciunt, pacem appellant.
I happen to use Sybase at work. It's....OK. I tried converting our whole app (couple hundred LOC and several GB of data) to postgresql. It took me less than a week and it ran better. Converting from MS SQL to Sybase isn't going to be any easier than that.
Sounds an awful lot like Microsoft's new free version of SQL Server, SQL Server 2005 Express Edition, currently in Beta. Almost exactly like full-blown SQL Server, but only supports 1 CPU, 1GB of RAM and 4GB DB size.
The "developer" edition has been available for a couple of years, and comes with a lot of the "extra" features turned on (such as Java in the database), but is limited to one engine and 25 user connections. It's also a version that you can't use for production purposes. It is available for a number of platforms (Windows, Linux, Solaris).
The new "Express" edition is (AFAIK) only available on linux, does not have the 25 user connection limits but instead has a disk space limit, and is usable in a production environment.
Michael
A few years back I ported the Transact SQL stored procedure code for my company's application from Sybase 11.5 to SQL Server 7. There is some divergence between the launguages so the task was not trivial. The names of some of the built-in variables are different. Our application uses cursors extensively and the variable which checks for the cursor being finished is @@fetch_status in SQL Server and is @@sqlstatus in Sybase. They also return a different set of values so I had to go change all of places where cursors were used. There are also subtle differences in the way that cursors work that caused me to have to rewrite some of this code. There were other small things I had to change that I cannot remember. In short, don't count on your stored procedures from SQL Server to just load up and run in Sybase unless they are very simple.
Yes, it supports stored procedures. Ini fact, you can use about a dozen languages to write them: plPG/SQL (which is similar to Oracle's procedural SQL), C, C++, Perl, Python, Java, Ruby, R, TCL, PHP and (I believe) Javascript.
Sybase 11.9.2 was also free for linux as well
yes
PostgreSQL has had support for stored procedures and triggers for quite some time now.
it's a DB that you really need to keep an eye on from time to time. Lots of people are still touting that it does not have stored proceedures and usually those people are simply talking without knowing.
it works quite well and the link above is the first one I could find on google that detailed it.
I can find more when I'm not surfing and posting from my Zaurus on the way into work.
Do not look at laser with remaining good eye.
Since when is 1 cpu , 2 gig of RAM and 5 gig of Data good enough for only the most demanding. Our smallest product easily uses more than that and limiting to 1 CPU really hurts. This seems more like a toy or the oracle personal database.
Postgres is much faster now than it was in the 6.x days. Back then it had a (deserved) reputation for being slow. There were many speed enhancements put into place before 7.x was released.
Postgresql is still not quite as fast (in my experience) as MySQL, but the comparison is not fair due to MySQL's lack of features.
I've read some benchmarks comparing oracle to postgresql and pgsql comes out close on most tests and beats oracle on a few. The benchmarks are gone, sadly, due to Oracle's "no benchmarking" clause in their EULA.
If you want a database that offers what Sybase does, but without those silly restrictions ("5GB ought to be enough for everyone" indeed), you might as well look at Ingres. Open Source but available with full support from CA should you want it.
This wasn't a shot at Microsoft this was a shot after being fired at by MS. SQL Server 2005 Express is replacing MSDE and its query governor restriction with almost identical restrictions to the Sybase announcement:
1 CPU
1 GB RAM
4 GB Database size (excluding log files)
No those same developers (myself included) will happliy stay SQL Server.
PostgreSQL has had support for stored procedures and triggers for quite a long time now.
it seems it is you that knows absolutely nothing about postgreSQL.
i suggest leaving the MS compound and learning before talking next time.
Where I'm working, we use MS SQL Server an awful lot. And we write an awful lot of code that uses it. We don't spend any time reading source, or trying to find holes / improvements to it. We're too busy.
You might not be reading the source code, but others will and their interests are probably more closely aligned with yours than a for profit institution that is more concerned about customer lock-in. But hey, do what makes you happy.
For anyone interested in Sybase ASE 11.0.3.3 for Linux, it's available for download at http://www.sybase.com/detail?id=1009516
A FreeBSD flavor is also available.
Also, I believe that Sybase ASE Express Edition for Linux is aimed more at potential MS SQL Server 2005 Express Edition users rather than PostgreSQL users.
There are highly capable DBMS available already. From the ubiquitous PostgreSQL and MySQL to the less familiar Firebird, SAPDB, and Ingres, I'd say there's again almost too much choice in the OSS world.
This is a noteworthy announcement from Sybase, but nothing more than Score: 3, Interesting.
All that being said, it would be different if Sybase literally were to open source their product. The reason for this being that while they have diverged since 6.x, Microsoft SQL Server and Sybase were once one-and-the-same. The divergence is, I'm willing to bet, still a minority of the codebase. Making Sybase a drop-in replacement for SQL Server in an OSS environment would be killer.
Great. What for?
To run on different flavours and architectures perhaps? The *n?x world isn't nearly as homogenous as Windows. Just one example.
Is that what he said? It sounded to me like he was asking if Sybase is significantly better. If the performance/features are the same, and you can get the source to one (not to mention use it on multiple processors or with databases large than 5 GB), then then why not use the latter? You get more. Was he demanding that Sybase open source their stuff? No.
But, by all means, let's all jump down his throat for asking if and what significant advantages it has over OSS solutions.
I've come for the woman, and your head.
I bet they are hoping that by giving away the product for free to people who would never buy it anyway, they get droves of people who are experienced at running their system who will eventually buy it for larger projects 'cause that's the system they know how to use.
... And they're doing something nice while they're at it, too.
And what exactly is wrong with this scenario? Sybase IS a company, and they have to make a profit
By the way, for those of you who are about to cry out 'But Sybase is worse than _____________ ', sometimes there are situations where it doesn't matter. An overwhelming majority of financial companies use Sybase, and will not hire you unless you have experience in it, no matter how good you are with Oracle.
The parent is stuck in nineties. Oracle 8(i) didn't support them.
Quite typical actually. Oracle 9 was at a kind of beta level for a long time. Now that it's getting stable, they're pushing 10g:)
Free is a great way to get people to try your products, if it's really free and not some crippleware lame ass substitute. But with unrestricted MySQL available, why would anyone use a crippled version of Sybase? I'd recommend a full version MS SQL Server before this free piece of crippleware somewhere in the misty flats of functionality. If you want me to try it fine, give me a full version and maybe we'll actually like it well enough to buy the next version. Though I suppose it's okay for people wanting to add Sybase to their database skills, no one with two neurons left to make a spark is going to use this in production.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
Its stuck in the 90s and the technology is out of date. I've used it, and it took me 1 week to do what it took 1 day using SQL Anywhere to do, and 4 hours to do with MSSQL (I was building a cross platform application example). Performance isn't bad, but it takes a hell of a time to get there. And it isn't compatible with MSSQL any more.
It doesn't come with source. If you want source and freedom, stick to postgre or Firebird. If you want productivity and power, get MSSQL or Oracle.
Sybase is great at replication and has had it for a long time. It can replicate data from non-Sybase databases too. I've used it in production. (Ironically, my company is moving to MS SQL Server to take advantage of better support from our hosting provider, but frankly, I'm not religious about it.)
Here's the product sheet, a white paper from Sybase, a quick intro, a slightly old quick reference guide, and a performance and tuning guide.
Note that Sybase isn't giving away the replication server for free though. So I'm not sure it'll solve your MySQL problem.
--LP
As I understand it, Sybase is actually a rebranded (forked?) Microsoft SQL Server.
Some of the things that didn't get mentioned was that you'll have to live with about 80% uptime if you are using Sybase on Linux. It's a crappy database and anyone who's used both it and MS SQL Server would never in a million years consider switching to Sybase no matter how free it is.
Software changes as you say were minimal, but you are assuming that everyone has the option of changing their software. Being able to swap out MSSQL for a free alternative with no changes might be the ONLY other option.
How did you convert all of the SQL Server Enterprise Manager jobs you had created? You know, the ones you have to create with the GUI'fied workflow editor and can never get a textual representation of.
Usually the problem with moving to a different relational database management system involves things like porting stored procedures that were written using the vendor's proprietary language, upgrading all of your applications and reports to work with the new database utilities the vendor provides, dealing with custom SQL extensions, etc. Conversion to different datatypes is the easy part.
I have used PostgreSQL, Sybase and MS SQL Server extensively.
Sybase is faster at properly indexed select statements than PostgreSQL (both are slower than MS SQL...sorry). Sybase offers two types of stored procedures languages (WATCOM and Transact) and the T-SQL is very nearly identical to MS SQL Server in terms of syntax and features (just watch your niladic functions and your join orders, Sybase can be sensitive about these, also it expects YYYY-MM-DD string date formats by default). It is much more robust than anything PostgreSQL has in this area. T-SQL is a more robust language than PostgreSQL, you don't have to use sequences (which could be good or bad) and you have a lot more options for triggers, views and cascading operations. Finally Sybase has incredible version compatibility...if you write a database for Sybase 7, you can relatively sure it'll work in 8 and 9. Postgres has a poor track record in this arena, at least on my server...we lost a customer because a minor version update broke our config for two days. That's a no-no.
I like PostgreSQL; I like it a lot. But it's no Sybase. And Sybase is no MS SQL Server. They are all better (in terms of stored procedures, complex queries, stability and tuned performance) than MySQL. Any of them are more than enough for what you need, but for my money, Sybase is the easiest to support. The command line tools are better than any of these, Sybase Central is crap compared to Enterprise Manager but it's still better as managing a SQL server than PGAdmin or some dumbass web client like PHPMyAdmin. They've maintained version compatibility so long that they're actually quite good at it; in fact, my only major gripe with Sybase is their Ole Driver, which last time I used it was more beta than beta. Also the resource governor, but you can work around that and it's a nice way to prevent older queries wasting your ram.
Oh, and the best thing about Sybase? It puts all the information about your database (tables, indexes, stored procedures, views, user functions, logins) into one file. Meaning that migrating the database to another server, even one on a different operating platform, is as easy as copying the file. I did this just last week...guy called asking if our system supported Linux servers. I installed Sybase on my Linux box, took a few seconds, copied my trusty db file and it worked. Slowly (P2 233), but it was identical to the same file running on Windows Server 2003.
Hey freaks: now you're ju
I also would like to know.
But what about the db Sybase released as open source I thought about 1 year ago?
Anyway you should be on your guard when anyone tells you they are being really nice but you "shouldn't ever need the code" and "that is plenty for any reasonable application". Come on, 5 gigs is not enough to hold one DVD. Sounds interesting as a microsoft killer for small companies but then again those companies should probably be going with postgres not mssql or oracle anyway. I for one would like to know what is so great about Sybase and is it worth getting into it.
Well, technically it's the BSD-like license with PostgreSQL I want - but the source availability is the mechanism that leads to that goal.
I wouldn't even mind paying for the product, so long as I knew the bug-fixes and other ongoing costs would be free.
I've worked mainly in big international banks. Of the 5 I last worked at, J. P. Morgan uses Sybase, and so does the investment-banking arm of UBS (formerly UBS Warburg). These are two of the largest banks in the world (UBS was world #3 in the latest rankings I've seen - April 2003 - and J.P Morgan Chase was world #8 and the biggest American bank). They both employ thousands of software developers who need database skills.
Just for the record, Oracle has always been available for free download in complete, unrestricted form. So for evaluation, people would just download and install it. Now, running production on unsupported and therefore unpatched Oracle instance - is the whole other matter.
But is it good enough to get Best Software to port Paradigm to Linux?
hahahaha, just kidding. They're going to dump Sybase for MS SQL. Whatta buncha dupes.
Paradigm is the premier mailing list manager for medium size ($1,000,000 annually) to tiny nonprofits. Lotsa bucks in this stuff, nerds.
Sybase Central is really coming along, esp after 12.5. I use DBArtisan too. Good product.
If you're just using DBArtisan for non-admin query building, you might want to look at Aqua Data Studio.
IIRC, the Oracle downloads are only licenced to be used for developing apps for Oracle. i.e. you're not supposed to use them (except for developing/testing), even for noncommercial use.
I downloaded 8i or 9i for Linux and tried installing it on Knoppix-installed Debian. Never quite got it working, but honestly I didn't try too much, and Oracle isn't supported on Debian in the first place.
So, tried actually using something which supports these newfangled 'dependencies' lately?
They really need to fix their site.. I fill out the forms, and agree to the agreements, and try and download the server, and I get a 403 page..
Forbidden
Your client is not allowed to access the requested object.
Without problems, you're right. But the changes aren't that great. We do primary development on MS SQL Server (it's easier than Sybase, because Sybase doesn't tell you the syntax of the proc you just wrote is wrong or references non existant fields or tables until you RUN the script. MS has a pre-processor) and I'm the guy who makes sure things run on Sybase. Basically, I do all the compatibility work in a single Textpad Macro. It's actually sort of simple:
1) Strip out the SET statements referencing ANSI_NULLS
2) Convert Niladic function names (e.g. CURRENT_USER -> USER)
3) Add Set DATE_FORMAT mdy, because the default in Sybase is ymd.
4) Find strings unicode strings and strip off the N'
5) Make sure all JOINS have their ON clause directly after themselves...MS lets you nest them, which I think makes for a better looking statement
6) Make sure the retarded VB developer didn't declare all his variables "@foo AS Integer", illegal syntax with illegal datatypes that MS SQL Server would fix for you.
7) Fix the IDENTITY syntax (basically, removing the step and start-at values) on CREATE TABLE
8) Remove ADD CONSTRAINTS that are really defaults or primary keys, and move them to ALTER TABLE ADD DEFAULTs
9) UNION ALL statements don't have column names during parsing in Sybase, so you can't do ORDER BY id_name, you have to do ORDER BY column_number. I think this is cleaner anyway, and it lets you change the name of the column more easily (can be important with ADO.NET, when mapping datasets)
10) Table variables don't work so hot in Sybase. I just create temp tables with hashmark names, same idea with a little less performance.
And I think that's it. Not that bad, really, and the script you end up with is comaptible with both MS SQL Server AND Sybase! I just finished a program that (unlike MS SQL Server) doesn't add crap like this to its scripts, thus making it trivial for us to port our apps back and forth.
Hey freaks: now you're ju
This is only slightly interesting in that it provides a database product for Linux. I also doubt they are giving away any of their "enterprise" stuff like replication, etc in the "free" edition.
I just wanted to point out that this is more than likely a response to Microsoft giving away SQL Express (formerly MSDE) for free. But Microsoft allows you to redistribute SQL Express with your software (free or no) at no cost and without signing any special deals or licenses. Sybase does not allow distribution in any form.
It is true that Sybase ASE and MS SQL Server were once the same core engine but since the split Microsoft has taken off and created a better database product. It is more ANSI SQL compliant (lol! one area where Microsoft is just about the only vendor with strict compliance). It is faster. It has much better tools and features (Enterprise Manager, DTS, etc).
And starting with Yukon (SQL Server 2005) it supports things like writing stored procs in VB.NET/C# or any dotnet language, along with many other new features. But the biggest thing of all, supported by both SQL Express, Standard, and Enterprise editions is what has always set Oracle ahead of the pack: Versioned rows. By turning SET SNAPSHOT ISOLATION on when User A locks in a table and User B tries to read data, instead of blocking user B the database serves up the row versions as they existed prior to A starting a transaction.
Anyone who has worked on a large database before knows that blocking (can be) the biggest issue you face. This is what has always let Oracle stand in front of MS SQL Server for large installs, but no more. This is also something that MySQL, Postgres, Sybase ASE, and even DB2 do not support.
Natural != (nontoxic || beneficial)
I wonder if Micro$oft is going to port their free "express" SQL server to Linux anytime soon. HAHAHAHAHAHA
How many of the chucklehead companies out there who license software "per CPU" are getting kicked in the groin by Microsoft hyperthreading? (Hyperthreading makes one CPU show up as two.)
Anyway...2+ CPUs seem to be standard on many servers these days, so this Sybase "free database" stuff appears to be crippleware, perhaps even more so than Microsoft's crappy little MSDE engine.
It's free! It's a damn good dbms! I don't care if I don't have the source because I don't have time to look at it anyway.
Ever heard of the idiom "looking a gift horse in the mouth"
Sybase and Microsoft released V4.0 of SQL for OS/2 in 1988. In 1993 Sybase and MS broke their partnership, but due to the licensing the MS product had to include 3 copyright notices for Sybase.
Sybase at this time came out with v10 of their product. Why v10 and not v5? Because 10 is better than 7 which is what version Oracle was at.
Sybase continued to develop the core of their datbase server while MS focused on usability. The core of MSSQL changed very little until SQL2000. Sybase continued to be the fastest at processing even though Oracle was dominating the market. This is why Sybase is so large in the financial sector.
I do not know if Sybase has changed their pricing scheme, but before 1999 if you wanted to use your Sybase server to connect to an application that was Internet available (i.e. web server) you had to pay $20,000 per cpu license fee. I sure hope it has changed.
A development version has been available for free download for sometime for both Windows and Linux. It limits the ammount of memory used, number of engines(1), and number of concurrent connections(5).
BTW An engine is not a CPU. You can have 5 engines configured on a box with 10 CPUS or 1 CPU. Their pricing scheme is tied to the number of CPUS on the system.
A good site for help with using TDS with Sybase or MSSQL http://www.freetds.org/faq.html
If you need documentation for Sybase...Its free http://sybooks.sybase.com/asg1251e.html
Since I've used Sybase ASE 12.5 recently, I'd like to make Some considerations:
-It is very sensitive about the OS configuration, only runs well on supported platforms (RH Enterprise, Suse, etc). If it finds anything not conforming its requirements, displays the annoying "Proccess is infected with 11" (meaning Segmentation faults).
-CLI client totally crippled: the user has to type go after each command, no line-editing, no history, poor output formatting. Fortunately, sqsh, mentioned on parent, is a nice replacement.
-ODBC driver for Unix/Linux not available for free.
-Native C/C++ API, ctlibrary, implements a finite state machine, enforcing the programmer to ask twice if the result set is done.
-Conversion to Transact-SQL data types from MySQL or Postgres is very tricky, because it doesn't have sub-integer types (smallint, mediumint, etc). Also, it uses different escape characters: eg. '' (two apostrophes) to represent one apostrophe inside a delimited string.
From what I saw, it's not worth it migrating from MySQL or Postgres. If the migration is from MS SQL, though, it is a very interesting move, since the compatibility among the two is is still good.
Is it possible to create a clustered database based on nodes with clients accessing to a unique source?
Is the 5G limitation including indeces?
To reminisce a bit, I first started using Sybase when they were at version 4.8 and I was starting work with one of the fledgling genome database projects. They practically gave away licenses to such groups (80%) discount. They were known then, as now as a business-oriented RDBMS. They failed to continue in this vein, and the support costs started rising faster than our budgets. The Linux version cost less, and the annual maintenance was far less.
In the meantime Oracle was very agressive in the higher education market, and still is. These two companies have a lot to offer, even if they aren't Open Source.
A reason why a small company might want to go with a free version of a commercial enterprise level RDBMS might be found in the disaster recover features that each has. I am far more familiar with Sybase internals, having taken a course that told me more than I every wanted to know about how it keeps track of data. And I have also recovered data that otherwise might have been lost forever. Oracle's features are just as good, but I have not had the "pleasure" of testing them when the pucker factor is maxed out.
-----------------------------------------
Computeri non cogitant, ergo non sunt
One difference is that if you get another CPU, you had a legal copy.
--
What part of "A well regulated militia" do you not understand?
Next to that, MS SQL is much more intelligent when it comes to using inner joins with indexes.
;-)
We have Solaris + Sybase 12.5 and Win2k + MS SQL 2000. When you ask both for a cost estimate (time) for a query, MS SQL is much smarter in using the fast indexes where Sybase gives up and just performs the slow query.
So flame me...
But this appears to be targeted mostly at Linux developers so it's competition for PostgreSQL and the Abomination That Shall Not Be Named.
I didn't know MS Access ran on Linux.
-- I could tell right away that she was impressed with my HUGE Slashdot Karma.
Does the database includes full ship schematics with part numbers?
oh... what?
nevermind.
5 GB for an enterprise system, especially one that handles lots of blobs, is not enough.
I expect this to be popular with hobbyists and those running simple websites. It might compete with MySQL and Postgresql in this area.
Not in the enterprise though.
-- yawn. --
I have used DB2 on all 3 platforms, have Oracle experience, but most importantly I have used Sybase from v10 (ouch) to 12 over 5 years and MS_SQL Server from 6.5 to 2000 for 6 years. So I am certainly qualified to touch on this.
Sybase works - No doubt about it. We had much better performance on an 8-way Solaris box vs. DB2.
In terms of code compatibility with Microsoft - it is close - extremely close even to this day. That's the nice thing about Microsoft stealing the Sybase code all those years back - it takes a lot of originality to subvert the database.
Some posters have noted that it's not hard to port a DB from MS SQL Server to {database x} That's true. BUT - the real thing to note here is T-SQL (Transact SQL) and the stored procedure implementation. If your Microsoft database has a boatload of T-SQL code written up over time (and most do) then the compatibility betwen Microsoft and Sybase is very good. Any MS T-SQL developer will almost immediately understand Sybase syntax. In fact most of the syntactical differences I have seen in the past involve the fact that Sybase has more options on each command due to their UNIX implementation. This includes things like parallel processing, etc.
I would not have any issues recommending a migration from MS-SQL Server to Sybase from a code standpoint. it would be the most efficient in terms of ease of conversion IMHO.
Now, WRT market share - that is true. Sybase made some bad missteps with (a) trusting Microsoft and (b) putting out a really bad release about 6 months too early (I think it was release 10). The 3-4 ish percent market share number is true. An important note though is that Sybase has about 40-50% of the Japanese and Chinese market share (this number is off the top of my head - corroborate it yourself) - so they are certainly not going to go away soon.
Just my opinion. Every database has its strengths and weaknesses - Sybase has a small footprint, solid performance, has run on *Nix forever, has mediocre marketing, and poor marketshare.
Bill.
The only *reason* your company can even USE mySQL right now is because it's open source. If it wasn't, nobody would have been using it for the past years, nobody would be working to improve it..
It's popular and as good as it is now because it's open source. And I forsee mySQL just continuing to improve until it's right up there with the top enterprise DB's.
So no, I don't need to crack open the source code to use my database, but I use mySQL because I *could*, because other people *do*, and because mySQL AB will never be able to squash mySQL just because they want to.
- It's not the Macs I hate. It's Digg users. -
This move by Sybase is an interesting one but pretty much follows along with their corporate strategy, as far as I can see.
Disclaimer: We're a Sybase shop, programming in PowerBuilder and deploying on Sybase ASA (Adaptive Server Anywhere, the 'lite' version of their DB engine) but we're small, only a few hundred systems a year.
Sybase is making a move (and has been, for quite some time) to the mobile market. They want their databases and utilities running on every mobile device out there. CEO John Chen has stated this is their mission for a long time, even while everyone kinda nodded and said, "Sure, John, whatever." Small things keep happening in this area to promote their move. They keep improving their development platform (PowerBuilder) but have cut some features and released the PFC (PowerBuilder Foundation Classes) as an open source project. Now the database is released to the public. The runtime version of ASA has been available for a long time for a $99 one-time fee (a deal, if you ask me) and many tools are cheap or free.
I think this is just another move to get out into the space. If their quest is to conquer the mobile market, they're not concerned about Joe Blow using their DB to power an internal time-tracking application, they want the name recognition so when you decide to go mobile, they'll be there to wag their tails happily.
Good luck, Sybase. Nice products, but it'll be a struggle.
Blog,Twitter
And if MS ever makes a change into the SQL server that breaks some of your programs or discontinues the product, what will you do ? You either port an awful lot of code that you've written to the new version / alternative product, or get stuck with an aging database product which won't have any bugfixes anymore and will cease working with newer operating systems (and processors - was it the AMD or Intel 64-bit processor that can't run 16- and 64-bit code at the same time ?) eventually. And it's even worse if the code is for selling, as opposed to just internal use - then you won't have the option of staying in the older version, and definately have to rewrite at least parts of it.
On the other hand, if you use PostgreSQL, you simply hire someone to maintain the old version and backport any helpfull new features. And if you want to sell your code, you just bundle your own version of PostgreSQL with it.
Having the source code is like an insurance: unneccessary most of the time, but if you don't have it when you need it, you will be sorry.
The article refuses to load, so I can't check to make sure; but this does have all the characteristics of a demo version. Get people used to using a certain product line, and they are more likely to choose it in the future. Also, the size of the installed base directly affects the likelihood of the third party tools supporting this database, which in turn directly affects the likelihood of this database being purchased over the competitors.
Why did you think Sybase decided to release a free version anyway ? Corporations do nothing unless they think they will benefit from it; therefore, they tought they would benefit from releasing this version free. And the most obvious benefit is the one explained above.
So yes, I'd say they will miss him.
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
Oh well, back to PostgreSQL then :-)
Is everyone else getting this page when downloading it in Mozilla?
s p?error=Sorry%2C+the+system+is+unable+to+process+y our+survey+submission%3A+%0ARule+Evaluation+error+ invoking+RA_SetOpportunityOwner2%28%29%3A+java.lan g.NullPointerException
http://crm.sybase.com/sybase/www/sybsurveyerror.j
What a bummer.
ASE and such are great... Personal SQLAnywhere is what we often get with these applications... which that is just a kick in the crotch from the vendors. 8 users and wham you get hit with having to buy a bigger DB from them (like ASE).
:}
A good chunk of these apps use SQLAnywhere 5.x and do not provide an upgrade path because they used to many table items that are reserved words in later versions.
In summary.. ASE is fairly nice... SQLAnywhere is a joke...
Telcos have alot of dark fibre in the States. Most people assume that's optical fibre...but it's actually moral fibre.
Come on, 5 gigs is not enough to hold one DVD.
Um, so what? Are you planning on somehow stuffing a DVD into your DB and querying it? 5G is a lot of data - certainly enough to evaluate the product with.
"We returned the General to El Salvador, or maybe Guatemala, it's difficult to tell from 10,000 feet"
Largely due to Transact SQL, Sybase has always been a niche product for the City of London and Manhatten. I've used it a lot in both financial districts.
Apart from the usual sucky things about SQL crippling the relational model, ( optional primary keys, no user defined types,etc ) it isn't bad.
Are there any other areas that use it as heavliy?
There are probably businesses here in my town that have 5GB of indexes alone. The guys a few floors down from me digitize a GB or so of aerial photos every day, and the results go into a database. Our piddly little print accounting system has several hundred thousand rows in it.
Still it's a thoughtful gift, as well as a good foot-in-the-door for Sybase. Lots of tiny app.s will be able to use it. [applause]
BTW, I recall that IBM is giving away a version of DB2. Someone else mentioned PostgreSQL. There are lots of good, rugged alternatives to MSSQL if you want one. My last database application was developed on PGSQL ('cos it's easy to get and easy to use), then put into production on MSSQL ('cos that's what they gave me to do the job).
Most companies I know barely want to pay enough salaries for the guys to write the code to use the database, now you want to hire an additional guy just to keep the Database software itself up to date? Most companies will just pay the license fee. Much cheaper. That one guy is going to cost, because you need someone very very goood. Someone writing bad code into the database software is a *bad thing*.
PostgreSQL got significantly faster in the 7.x release series. With the up coming release of 8.x series, it recieved another signifnicant boost in performance.
The 6.x series was known to be slow. That is very much a thing of the past.
Sybase has serious share in some specific markets. Here in NYC there are a lot of financial firms that use it heavily. I've also read it's big in healthcare and telecom, and I would imagine there are some hooks between it and SQLAnywhere, which is theirs, too.
It might not be attractive to you if you need to decide on a RDBMS from scratch. If, however, you have an existing Sybase system deployed, or you have a Sybase DBA, or you like the performance, tools, expertise, or vertical focus, than this might be wonderful for you.
He're the thing:
There was a time when there was a _big_ performance advantage in using Sybase/SQL server over Oracle for certain applications. As a result a bit base of code got built up in organizations with deep pockets. IMHO what is _really_ needed for Postgresql:
An api that emulates the Sybase SQL Server dblib api(this is the higher priority)
an SQL interpreter that closely emulates transact SQL
At that point, a bunch of code can be moved fairly easily. The other thing that would be helpful is
getting an interface to SQL Server from Postgresql(similar to the one that Microsoft did for Oracle). That would be useful for stuff like data migration.
Banks are nearly 100% sybase turf. You cannot get a job in a bank as a SA or developer if you do not know it.
While at it, the people who will be interested in it will be the people who want to get into that sector and need a toy to play. The demand for sybase skills outstrips supply around 10 to 1. Sybase is releasing it for only one reason - so that PHBs do not start considering alternative databases because of lack of staff.
Otherwise at least for me it is in the "not interesting" category until it gets a decent working DBI module compatible with the most recent version.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Postgresql is still not quite as fast (in my experience) as MySQL, but the comparison is not fair due to MySQL's lack of features.
Well, for single user experiences, MySQL probably is faster. PostgreSQL tends to scale much, much better than MySQL. In other words, if you have a DB where you expect to have lots of active users with a diverse set of concurrent activity (selects, updates, inserts, deletes), PostgreSQL traditionally zooms way ahead of MySQL. It's a question of how you expect to use your database.
Basically, it's a myth that MySQL is faster than PostgreSQL. A myth, I might add, which MySQL themselves help propagate. Simple fact is, MySQL probably is faster in the developer's eyes. Once it's deployed, PostgreSQL is more than likely the faster of the two. Granted, there are some corner cases where MySQL is still faster than PostgreSQL, but those are the exceptions and NOT the rule.
the postgres java driver I used was buggy and it's not to widely supported
so many people still use mysql and it's so horrible. I just hope those forums switch from mysql to this sybase stuff so that I dont see those "mysql connection error" any more
PostgreSQL doesn't need versioned rows or to turn anything on. Its implementation of MVCC gives the performance benefit of versioned rows automatically all the time.
(ANSI joins in Oracle? Hah)
: F4 950_P8_DISPLAYID:3512484632553
By searching for ANSI Join Oracle via Google, the first article reveals:
"Starting with Oracle 9i however, Oracle have now included support for many ANSI SQL/99 features including ANSI compliant joins, and there are several advantages in using this new syntax, one of which is the separation of the join condition from the WHERE clause."
Huh... Seems you are wrong.
To start a religous war on the subject, here's a thread from Tom Kyte comparing the two databases:
http://asktom.oracle.com/pls/ask/f?p=4950:8::::
Sybase is a damn good database (superior in many ways to Postgre IMHO) and they are *giving away* a perfectly usable version.
Now you're bitching and moaning that the source isn't open.
The source code for Postgre is available, but I, for one have never bothered looking at it. Have you?
Some people are never happy I guess.
And no, I'm not missing the point. The Sybase people have done a good thing. Give them some credit!
Okay, so Sybase finally released a new version of their database for linux. They had released Adapative Server Enterprise 11.5 and 11.9.2 *years* ago for Linux. Yes, it was under a developer's license back then, hence you couldn't legally use it for production. But, honestly, I don't remember the database being crippled in any way. My former employer was a Sybase shop, and they used Sybase on Linux for our development work and then used Sybase on Solaris and HP/UX for production. My personal test machine, at the time, was a dual Pentium Pro with a half a gig of RAM--I had that due to acting as the primary DBA.
There's also a real reason that Micro$oft's SQL Server uses Tranact-SQL just like Sybase's Dataservers....who do you think M$ bought their database technology from? In fact M$ bought the name "SQL Server" from Sybase too, hence why Sybase's dataservers are called Adaptive Server Enterprise.
"If you insist on using Windoze you're on your own."
It's all personal and professional preference, of course. But for me... the answer is "no - but it helps."
I'm no Free Software purist. I run Linux systems at home and work (right next to Windows and Solaris systems). I've bought proprietary Linux software for both home and work (mostly games and enterprise apps). But when making a choice, I tend to weight heavily on the side of FOSS. Why? Freedom.
I've been bit plenty of times in the past by intentionaly induced limitations, technical incompatabilities, and agressive licensing. Some vendors are better than others. But with proprietary software, the more one becomes dependant on the product, the more risk one runs of being unable to migrate from it.
Granted - there are no guarentees in IT. But an infrastructure designed on Open Source systems tends to allow a greater degree of freedom. Data formats are documented (if not in documentation itself, in the code - which makes migration possible if not easy). Functionality tends to be limited by technical issues rather than marketing. And if the primary developers of a particular project take a turn that conflicts with your environment, there is a good chance that there are others with the same view - migration utilities are developed or oft-maligned fork keeps the project in a favored direction.
That doesn't make FOSS the magic bullet. There are certainly times where particular examples of proprietary software offer advantages that makes it attractive. And that's where this line of questioning comes in. Sybase's offering lacks the freedom that other FOSS databases offer. So what advantages does it have that would make it attractive?
I'm no database expert, but what you describe sounds an awful lot like PostgreSQL's Multiversion Concurrency Control.
Sybase sucks. Then again, so does every other database product I've ever worked with (Oracle 7i-9i, MS SQL Server 6.5-2000, Informix, Postgres 7.1 - 8.0, *and* MySql). They all just suck in different ways.
7 20775
If you spend enough time with any product, you'll find little quirks that drive you insane. A couple off the top of my head for Sybase:
1. Chained & Unchained modes. Sybase supports a SQL 92-compliant transactional mode, and a hacked up "autocommit" mode, with optional transactional support. The hacked up mode is default, and the SQL 92-compliant mode has some severe limitations.
2. No DDL inside transactions. So what? That includes creating temporary tables. You want to call a stored proc from within a transaction? It better not touch the tempdb.
3. Table-level locking by default. This one just blows me away; Sybase didn't support row-level locking until somwhere around version 11, and table-level locking is still the default. If you're DBAs aren't on top of things, you'll have deadlocks all over the place. They still haven't enabled it for the system tables, so make doubly sure you don't do any long-running code that touches them, or you'll have deadlocks for sure.
If you think I'm bullshitting, check out a quasi-white paper (grey paper?) I posted a while ago to the iBATIS support forms - it's got a lot more detail about some of the problems, and some Java-based work-arounds: http://sourceforge.net/forum/message.php?msg_id=2
So, would I choose Sybase over the competition? Maybe about 3 years ago, before other DBs got decent replication support - that was one of their claims to fame. Performance doesn't seem to be that big of an issue - hardware is often cheaper than engineering around db limitations.
If I had to rank what I've used in order of preference, it would be:
1. Postgres
Maybe because I've only used it for about 8 months, but Postgres has not *yet* disappointed me. Transaction support has been perfect, and no major performance problems. Then again, I haven't done any stored proc work, so maybe I should give it time.
2. MS SQL Server
I cringe to say this, but MS's developer tools push their DB up on my list. Query Analyzer has excellent "show plan" support, and their management tools are great. I'm generally pretty happy, although their JDBC driver could use some work, and DTS was pretty weak last time I used it.
3. Oracle
Cost knocks this one down a bit, and I'm a bit rusty as well. Last time I ran Oracle on Linux was shortly after it was released, and their install procedure was a *bitch*. However, nifty features like data partitioning were definitly worth the extra money.
4. Sybase
See above. It's decent, now it's free for small projects, but I'm annoyed.
5. Informix
I'm out of date on Informix, but I have bad memories, mostly of constantly overfilling the transaction logs, then having the DB stop working with an unclear error message. I understand the need for a DBA to monitor this on a production environment, but it was a pain in the ass in development.
6. MySql
OK, I'm going to get bashed on this one. The old limitations left a sour taste in my mouth, and too many critical features are brand new. I will reconsider, though, after it has a little more time to mature.
Scott Severtson
Senior Architect, Digital Measures
We have customers running ASA with 4,000+ simultaneous users, 50 GB+ databases.
I strongly suggest you consider taking a look at newer versions of ASA. There is a Developer Edition of 9.x available (the engine can run a 5.x database file without modifications!)
BTW: SA 5.x also can support much bigger loads that you let on...however the applications and/or schema "design" may be where the limitations you experience are coming from. I can write bad code that will cripple any RDBMS...it ain't that tough:->
I'm sure by now you realize they were talking about MS SQL, not MySQL, right?
>certainly enough to evaluate the product with.
Exactly. Like another poster said, they are doing this so people get familiar with their database, then decide to move to it later.
By the way,
Did some consulting for a company. It's a 30 people company. They mssql database is 7 gig now.
Any non-trivial database job will involve an enormous amount of data.
"Piter, too, is dead."
Microsoft SQL Server is in fact Sybase. This is why they're so similar (Transact SQL, etc). Way back when, when MS needed a SQL Database for the initital releases of Windows NT 3.1, they basically licensed the code from Sybase and ported it to Windows.
:)
That said...
If you do need to spend money on a databse, stick with Oracle.
bash-3.00$ uname -a
SunOS panda 5.10 Generic sun4u sparc SUNW,Ultra-2
If you can, throw it away.
Cliché perhaps, but the point is still valid.
Take it easy? I'll take it anyway I can get it . . .
Microsoft SQL Server Express is free to use and redistribute. It supports 1 CPU, 1 GB addressable RAM and 4 GB database size.
It's based on the core SQL Server 2005 Database Engine, including an advanced query optimizer and the new snapshot isolation level. It also supports the complete SQL Server programming model including T-SQL and CLR integration.
http://lab.msdn.microsoft.com/express/sql/
Did I mention that it's free to use and redistribute?
Yea, so I saw my instead of ms, but the point is still valid - the reaons someone would want the source code, not because they might want to change it, but because someone ELSE can.
I guess I read a few posts related to using mySQL in this context and I just hit reply on the one where he said MS. Oh well.
- It's not the Macs I hate. It's Digg users. -
How are these stats counted? Is it total number of users? Total gigabytes? Total number of hosts? VMs? Servers? Clusters? Enterprise contracts? Individual unit sales?
I know which markets use Sybase, Oracle, and DB/2 and I don't think most of them are feeling much threat at all from SQL Server. If nothing else, all three of the real major players scale on hardware that can crush SQL Server with a single node, never mind the whole cluster.
I do not fail; I succeed at finding out what does not work.
I've been using Postgres for this.
-- Stephen.
Specifically SQL Server 6 was based on Sybase 10. It really hadn't changed much by version 7 but I'm sure it has by now.
Last I heard, SQL Server was the fastest RDBMS around, Sybase was second, Oracle was third, and DB2 was last place. However, DB2 was most scalable, Oracle second, and so on back down to SQL Server being the chump in that department. It's been a while so I dunno how true all this is now.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
On the other hand, if you use PostgreSQL, you simply hire someone to maintain the old version and backport any helpfull new features. And if you want to sell your code, you just bundle your own version of PostgreSQL with it.
Would it not be more cost effective to hire someone to convert the data to a new format? Getting relational data between different databases is mostly trival. Code changes not so much, but if the codebase accessing the database was written in house my guess is that it's much quicker to change that known code to adapt to the nuances of the new database than to port over unfamiliar database guts.
Is MSDE client server? I thought it was more like access then SQL server.
evil is as evil does
According to IDC here is how the database market breaks down.
Oracle is #1 overall with 39.8% market share
IBM has second place with 31.3%
Microsoft is third with 12.2%
So as you can see although MS is third it's way behind the other two.
I don't expect MS to make too much progress in the database market though. Oracle has cut it's prices so that it costs the same or less then MS sql server for every product and there is fierce competition from open source databases.
When the next version of postgres comes out all the open source databases will run natively on windows. If I was MS I'd be thinking about another round of price cuts.
evil is as evil does
Better than waiting around to have ms stepping all over them, too...
What Sybase should do is foist this upon colleges, and maybe even high schools.
If enough companies get students into doing SOMEthing productive, challenging, and interesting, such as making people more SOFTware and hardware literate before they reach college, then some of these non-microsoft (lower-casing/deprecation of ms by me is intentional/perpetual) vendors can at least confuse or fragment the market and try to keep mshaft from penetrating too many vertical and horizontal markets.
MySQL, and PostgreSQL,for that matter ALSO should jump in on this. Anything to help people see the world from a "databasable" perspective (as I do) means it chips away at patenting aspects of database schemas or such. After all, the more that becomes "obvious", the more only TRUE innovations should be allowed to obtain or receive patents, much less even get more than a cursory prequalification review.
Unfortunately, there are a LOT of smart teachers, but they are so swamped in homework and mired in petty district politics that many of them cannot recall or get remedial training to improve their skills. Hence, as much as school districts and teachers would be thrilled to get new computers, they generally don't have the time, budget, or other resources to keep at bay mischievous students, much less cannot avoid being embarrassed by precocious or prescient students who can redesign some schoold district's database environments.
Anyway...
David Syes
Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
The next edition of MS SQLServer (2005) will include an Express Edition which is free subject to similar restrictions - 1 CPU, 1 GB RAM, 4 GB data.
From a product-line point of view, this replaces the MSDE. From a functionality/power point of view, it's the real SQLServer 2005 code. Last I checked it was at Beta 2.
So it seems all of the "oh great, we can throw out MS and now use free Sybase" arguments - at least, those that are based on the "free" part - don't hold much water. One might even say that this move on Sybase's part is a response to MSFT.
Why? Because chances are most people don't have the experience inhouse to debug and solve the problem even if they discover a bug. Yes, having the source is a nice thing but most companies don't want to mess around with it. If they discover a bug they want to call the vendor and get it fixed. Sometimes I have to laugh at the mentality of some /. posters. This is the _real_ world people.
"Thanks to the remote control I have the attention span of a gerbil."
...This release provides a very attractive alternative to Microsoft SQL Server, and gives developers and DBAs an extremely powerful argument to use against the adoption of Microsoft-based solutions. ...
I'd say VI outfitted with a few data retrieval regex scripts is already an extremely attractive alternative to anything with the words "Server" and "Microsoft" in it's name. No need for this Sybase thing to get all worked up about being an alternative to MS SQL Server. Sorry.
Besides that, I've long ago stopped being impressed about products that are said to be an alternative to anything from Microsoft.
We suffer more in our imagination than in reality. - Seneca
In what way is DBD::Sybase lacking?
Michael
At the bottom of the
Oracle's Free (as in speech) software
If you saw Chuck Rozwat's LinuxWorld keynote (2 years ago, I think) you'll know that Oracle uses Linux PCs for its base development. Not just for "back-office apps", mind you, I mean a gigantic development environment with THOUSANDS of Linux PCs. The resulting inevitable patches to coreutils, etc, are all on the oss.oracle.com site above, as are Oracle's (GPLed) Clustered Filesystem.
Part of the Second American Revolution!
The big changes arrive when you try to convert the stored procedures, views, constraints, subqueries, functions, etc.
;-).
I've worked with databases that have thousands of those, and porting the code has been very tedious and time consuming.
I can infer that your application had most of it database logic in the app, not in the database. Perhaps I used the wrong approach and need to learn a lot from you
Cheers
Adolfo
The article refuses to load, so I can't check to make sure; but this does have all the characteristics of a demo version.
No, it's not. It's a lite version of Sybase's flagship SQL database (the one that was licensed to Microsoft way back, and which they kept as SQL Server).
Why did you think Sybase decided to release a free version anyway ? Corporations do nothing unless they think they will benefit from it; therefore, they tought they would benefit from releasing this version free. And the most obvious benefit is the one explained above.
While Sybase isn't doing it purely for magnanimous reasons, it's primarily to get Linux users to try it. It does have limits (5Gb size limit on a database, and only one seat), but it's a good way to try it without committing the funds to purchase it. And as I said, it's not a demo. It's a lite version with not all of the extra bells and whistles.
No.
MSDE is the same as MS SQL Server, minus the admin and client tools and plus some hard restrictions on number of users, concurrent threads, database size, etc. I'm pretty sure they ripped the replication features out of it too. Same database engine, same protocol, same file format.
Scary. Back to hiding the money under the mattress.
Great. What for? [...] We're too busy.
Maybe he wants the source because then you will end up with hundreds of developers working on improving it. Making it faster, more secure, more robust, etc. There are plenty of other benefits to software being open source, if you cannot understand that then I don't understand what you are doing at slashdot.
I am sure that you are also to busy to look at the source code for Apache or Linux yet do you think that it wouldn't matter if they were closed source?
You might not be reading the source code, but others will and their interests are probably more closely aligned with yours than a for profit institution that is more concerned about customer lock-in. But hey, do what makes you happy.
Your point is a typical specious argument offered by OSS supporters: an for-profit organization doesn't care about customers as much as people who are just interested in developing software. My experience tells me otherwise. If you give engineers a free hand, they will most likely pursue issues of interest to them, like sophisticated algorithms, intricate coding techniques, and large-scale architectures, which are not directly related to what customers want, a solution to their problem. A paralell in another profession is college professors. Since most universities are not for profit and many professors don't have to worry about losing their jobs, a lot of them spend more time on their research work, something of interest to them, instead of on improving teaching skills, something of utter importance to their customer - students. On the other hand, the lifelihood of a for profit organization completely depends on customers. So they, out of necessity, established a set of mechanisms to focus their engineers' attention on their customers' immediate needs. I am not saying they are always successful at pleasing their customers, but they sure try very hard.
PostgreSQL gives you that benefit automagically at the expensive of increased administrative load by forcing you to perform a regular vacuum of frequently updated tables in order to expire "dead" (ie: non-visible to any existing or future transaction) rows. Not that huge a pain if you know PostgreSQL at all (or use the autovacuum daemon) but it constantly bites newbie users in the @ss.
And I can say, for sure, that Sybase is a damned good DB. I originally wrote my application to run on Postgres, and I ran into crushing performance problems on a torture test that I concocted - one 'simple' query ran for a few minutes.
Since I was having performance problems, I figured I'd try to move over to MySQL. It was hard, but I managed, and it ran fine. Very fast.
Finally, I decided I needed to use a 'Commercial' database. After evaluating costs and everything else, I went with Sybase, version 11.0.3, which was (and is, if you can find it) free to develop on and even deploy software based on it. It's run solidly since. Never a hiccup. Then again, I never had any reliability or crashing problems with Postgres or MySQL, so that's not as telling. But what is telling is that it's _fast_.
Now that they've come out with this new version, I think I may upgrade to it - I don't think my data is larger than 5 GB anyways, nor do I intend to use more than 1 CPU. But until now I figured I was just going to have to wait until I could afford the pay version, or move again to something else.
I'm figuring their real hope is for people to try this one out, develop on it, and when they outgrow it, buy the pay version. Which sounds pretty reasonable to me.
they are usually interested in using more than 1 cpu and just 5Gb.
the sybase offer is useless.
Hardly. You can start out with this Sybase system as a small business, and by the time you outgrow the 5 gig limit, you can afford to buy Sybase. That sounds to me like a rather obvious gain for both the users and Sybase.
Yes, we'll be in trouble if Microsoft ever discontinues SQL Server. Somehow, I'm not terribly concerned.
MS, Sybase & Ashton Tate jointly developed the core engine up until the mid nineties
Wasn't Ashton Tate dead by the early 90's? Or did Borland just buy dBase, rather than the whole company?
I'm one of the heretics that believe that free (as in beer) has helped spread Linux, Apache and MySQL more than free (as in speech), based on nothing more than the fact that the free beer part was what made me start using those products.
But I don't think this announcement will be met by anything but a shrug from most Linux and open source DB users, whetever version of free they believe in. The thought of a product having limitations at all will stop them from even trying it.
It can be relevant for some people, though, and that's for those who're evalutating commercial DB's in the first place and have the budget to buy them.
That should be true. The nature of postgres is always to go for completness and correctness (full featured SQL complaince and less defects) vs. optimization.
In the early days I even remembering the SQL developers apologizing (on the WebSite or in some README, I can't remember and couldn't find the reference) about the slowness. They reason: We want to make it work flawlessy first, then we'll start working on the optimization.
Which I find great. Specially since most of the features that are long ago into postgres are starting to be optimized (or already are) and you can practically use any SQL code you want on it and you won't be surprised by features missing or incomplete (except maybe some really obscure instructions which basically are there cause IBM or Oracle added them for a single customer or two).
Your point is a typical specious argument offered by OSS supporters: an for-profit organization doesn't care about customers as much as people who are just interested in developing software.
I seem to remember being taught that companies are in the business to make money, not necessarily do right be their customers. I'm sure that your experience / stat data is correct in some instances, just not all.
Also, you can remove most of the restrictions from MSDE using techniques such as this. And a lot of the missing functionality is contained in the service packs for SQL Server as .dll files that you can just add to your MSDE Binn directory (e.g. SQL Mail).
I have to say that I have used Sybase over the past five years and I am definately not a Sybase fan. I think that I'd rather be drawn and quartered than use Sybase. I am not a huge fan of databases to begin with but Sybase (11.9.2) was just aweful from my point of view.
The real silver bullet to good programs is caffeine; lots and lots of caffeine! *twitch, twitch*
Sounds like an awesome macro. How about posting the source on your site?
Why is it that the proponents of "one nation under God" are so eager to get rid of "liberty and justice for all"?
I'll consider it, but a Textpad macro wouldn't be a great format. If I did post something, it'd be this database migrator I wrote...which does all that I mentioned and more.
Hey freaks: now you're ju
"On the other hand, if you use PostgreSQL, you simply hire someone to maintain the old version and backport any helpfull new features. And if you want to sell your code, you just bundle your own version of PostgreSQL with it."
BEcause every company running a DB wants to additionally take on the burden of coding, testing and porting features to it all the while hoping they don't mess it up and can still remain competative.
The kind of money you would have to pay, on an ongoing basis, to hire someone (or more) who can intelligently and correctly port to, tune, test and enhance something as complex as Postgress is shall we say "non trivial".
Your money woudl be much better spent porting to a supported DB system.
--> Fight tyranny and repression.... read
Since most universities are not for profit and many professors don't have to worry about losing their jobs, a lot of them spend more time on their research work, something of interest to them, instead of on improving teaching skills, something of utter importance to their customer - students.
You obviously don't know much about what Universities are for. Students are not the customers. Big business and governmental funding agencies are. Research is the goal, and purpose of the University system, not teaching. Many professors would rather just be teaching, but know that they will not keep their jobs if they don't publish research (because that is their primary function).
Students don't pay tuition to get a service (teaching). They pay to be part of a program, where they have the unique opportunity to learn from some of the brightest minds. Professors don't owe students anything, grant money pays a big part of their salary, not tuition.
I'm a little surprised that people on hosting packages provide MySQL over PostgreSQL (i've only used MySQL) when it looks like PostgreSQL has a whole bunch of other benefits that MySQL doesn't.
If you are a small business then you can likely afford the $500 that a the most basic deployment of Oracle SE would cost you. 5GB is simply a joke. The alternatives aren't as expensive as you think they are for the domain under discussion.
A Pirate and a Puritan look the same on a balance sheet.
Hardly. A 5GB dataset will only give you familarity with the products most basic features under the most trivial conditions. When you go to do something more serious, you will likely find that all of your assumptions are wrong and that you've developed many nasty bad habits.
The "eval" copy of Oracle EE running on my Debian laptop is not crippleware. If I wanted to test it with a 30TB SAN, I could.
A Pirate and a Puritan look the same on a balance sheet.
Unfortunately, MSDE has even more limits than Sybase's offering. Database size is limited to 2GB (real easy to reach) and contains a performance "governer" that will throttle the database engine's performance down when the number of concurrent connections exceeds eight. The actual number of concurrent connections is in excess of 32,000 but the engine will delay processing of I/O operations by a few milliseconds once the eight-concurrent connection limit has been reached.
0 0.aspx
More info here: http://weblogs.asp.net/gad/archive/2004/03/10/872
Ryosen
One man's "Troll, +1" is another man's "Insightful, +1".
I just followed that link, and Oracle wants me to provide information about myself and take a survey before they will let me get the database. That does not count as free to me.
This has nothing to do with "microsoft niftiness". All stored procedure languages are quite proprietary and a hinderance to platform migration.
A Pirate and a Puritan look the same on a balance sheet.
My parents are college professor. And I spent a lot of (too much?!) time laboring through the school system (top student as well). So I do know a thing or two about universities. I understand the importance of research funding for universities. But you can't deny that universities were established first and foremost to educate students. Without the regular and guaranteed tuition income from students year by year, most universities will disappear. If professors only want to do research to begin with, why don't they go to pure research organizations.
I don't understand your logic. When you pay money, you are the customer of someone. What do you mean that students are not customers? They are the customers. They pay dearly for a service - education. Unfortunately, they are not at the top of the priority list for professors, their service provider. Smartest people don't necessarily mean great teachers. There is no direct relationship between a good researcher and a good teacher. You can't learn from the smartest mind if that mind can't communicate very well. Unfortunately for the student customers, they don't have a choice. Some are lucky to have a great advisor, some are lucky to have it figured out themselves, but most simply go through the system, parted with tons of money, in the hope that they may get their investment back sometime in their lives.
I don't understand your logic. When you pay money, you are the customer of someone. What do you mean that students are not customers? They are the customers. They pay dearly for a service - education. Unfortunately, they are not at the top of the priority list for professors, their service provider. Smartest people don't necessarily mean great teachers. There is no direct relationship between a good researcher and a good teacher. You can't learn from the smartest mind if that mind can't communicate very well. Unfortunately for the student customers, they don't have a choice.
Why is it unfortunate that students are not on the top priority list? Research benefits all of us. Teaching benefits a handful of students recieving the teaching.
You are right on about the communication problem. Alot of professors aren't that proficient at teaching, but then that isn't what they are meant to be doing there. Most lower level classes are taught by graduate students anyway (even worse teachers in my oppinion, and I am one).
Some are lucky to have a great advisor, some are lucky to have it figured out themselves, but most simply go through the system, parted with tons of money, in the hope that they may get their investment back sometime in their lives.
That is why I think there should be more done to let students know ahead of time what they are buying, which is an opportunity for a self-motivated enriching educational experience. If they want teachers who are only there to teach, they should go to a community college, or a trade school.
I'm all for open source, but would you even be able to understand it? When SAP released their database as open source, the complaints about the code quality ( or lack thereof) were widespread.
Pain is merely failure leaving the body
I work at a medium sized real estate management company. Our financial data going back to 1994 is no more than 2 gigs max.
5 gigs is plenty for most, unless you are a hardcore pr0n freak. A heavy duty pr0n fanatic probably needs about 1 to 3 gigs a day to keep up with their fetish.
Go burn a bra, hippy.
Interseting? +5? This guy is obviously a twat! Only on /. could this be considered +5 Interesting. -1 Flamebate I could accept.
Wow! Don't tell me you work in IT. Nobody in their right mind is going to cram a binary image of a DVD into a database. A fool might. A smart person will not.
The stored procedures in Postgres stink. Sybase's stored procedures are magnitudes better than Postgres. Also, Sybase has, arguable, the best documentation in the business. Postgres' documentation is somewhere between mediocre and marginally bad. Postgres is just getting around to replication and clustering. Sybase has had it for a while.
When working on M$'s SQL Server, it's not unusual for me to pull out my Sybase documentation. This is especially true when working with Transact SQL and stored procedures. M$ added some junk to TSQL in SQL2000 that's incompatible with Sybase but for the most part jumping from one to another is largely transparent.
I just checked, and Oracle SE appears to be $366 per user, which still strikes me as too steep.
It does, of course, depend on your business, but the only business I've had interest in starting would be an ASP, which would require at least a processor license, for $15,000.
Oracle SE is $150 per CAL. The minimum is 5 named users. So I did get the price wrong ($750 vs. $500).
For a robust RDBMS that is safeguarding key corporate data, that's a pittance. If you don't need the level of robustness that most commercial RDBMS products provide then you don't need a commercial RDBMS of any sort.
Either your data is worth paying for your RDBMS, or you don't need to bother with payware to begin with.
A Pirate and a Puritan look the same on a balance sheet.
Sybase's license seems much more useful to me. Oracle says I can use their free version only for developing a prototype and excludes any production use. Sybase will let me use it broadly for development use, which is much better. I can use it for day to day operations in a development environment, gaining familiarity until I am comfortable enough to develop the real paying applications.
Admittedly, PostgreSQL's license is even better, but I want to learn on a commercial DB that can lead to paying work.
If Oracle were substantially more reliable than Sybase, you would have a point, but we're not talking about MySQL and PostGreSQL here.
Yes, the data may be worth paying for, but you don't always gain anything by paying.
So lie. I've given false info to several sites over the years. What's the big deal?
Pain is merely failure leaving the body
Correct me if I'm wrong but, I believe that MS SQLServer was made from code and experience gained from Microsoft's collaboration with Sybase a few years back.
Much to Sybase's loss and Microsoft's gain. (Such is the fate of those who get in bed with Microsoft.)
This is good news for Linux and Sybase. It gives MSSQLServer shops a migration path to Linux and back to Sybase.
Codifex Maximus ~ In search of... a shorter sig.
The comparitive evaluation is kind of hit and miss. You have a point on replication -- I would certainly trust another database software more (although I would prefer DB/2 or Oracle over Sybase, certainly) in those kinds of configuration. As for the documentation, actually, I personally prefer Postgres's documentation. What do you think is done better in Sybase documentation? As for stored procedures, I think Postgres wins there too. This is primarily because Sybase offers, as far as I understand, only SQL-like stored procedures, not *too* different from Oracle's PL/SQL (but actually inferior to Oracle's language). Postgres offers something similar to this as well, but also (as Oracle and DB/2 do, to a lesser extent) lets you write them in other languages, like Java or Perl.
As for the original poster's concern about loading large amounts of data into a database, while it is unlikely that anyone would want to load an entire DVD into a single BLOB in a database, it is concievable that, for database/website/whatever designs that use large amounts of data, that much *worth* of data might be loaded into seperate BLOBs. It's an open topic in what constitutes best practice if heavy use of BLOBs is a good idea, but simply calling largish databases a stupid thing to want seems a bit shortsighted to me.
Of course, you're on Windows, and if that platform is a necessity for you for some reason, I wouldn't suggest Postgres to you for awhile -- while 8.0 will have a Win32 port, it's probably best to avoid trusting your data to something that new. If I were in your situation, and had a lot of money, I'd go with Oracle, otherwise DB/2. Then again, DB/2 is looking better and better these days, so perhaps looking at Oracle is a waste of time. I certainly wouldn't go with MSSQL or Sybase.
For every problem, there is at least one solution that is simple, neat, and wrong.
I would gain quite a bit actually.
Oracle SE isn't crippled. Sybase is.
With Oracle SE I am not limited to rediculously small datasets, single cpu systems or an amount of RAM currently common on desktops and workstations.
The same is true for postgres and mysql. That is why they are relevant in this discussion.
A Pirate and a Puritan look the same on a balance sheet.
You could equally keep running the old version (in the case if $closed_source_app) if you wanted.
Besides I think you want to pick a better target. SQL Server has consistently been backwards compatably. If that ain't enough SQL Server 2000 also includes a backwards copatability mode.
This does not ensure that future behaviour is going to be the same, but it does suggest it is pretty likely.
meh
Banks are nearly 100% sybase turf. You cannot get a job in a bank as a SA or developer if you do not know it.
The last three gigs I've had have been at Banks; 'super regionals' is what they're called. So they are by no means teeny-tiny local operations. I've been a dev and now a dba. Haven't seen the Sybase. I've seen it quite a bit at the big brokerage houses and master servicers in the eastern US, though. Based on my banking jobs: Oracle, DB2, MS SQL Server would be the RDBMSs in in use. But big banks have all kinds of technology pockets. What one sees at banks probably depends on one's area of expertise.
I work for Afilias. We sponsor Jan Wieck, a member of the Postgres core team, to work on Postgres full time.
So unless your plan is to port from a very expensive proprietary database to Postgres and you're looking at budgetary numbers over about 5 years, it is NOT more cost effective to just change platforms.
Your problem is with the vendors who lock you into SQL Anywhere 5 and don't permit enough concurrent users on it. If that's your basis for judging SQL Anywhere, it's not really fair for you to judge it. Considering ease of administration, ease of development, efficient use of hardware, platform availability, SQL compliance, features, performance, cost, documentation and support, SQL Anywhere is a terrific product, and generally my choice for small businesses and department-level applications.
Except that Oracle is evil incarnate for a developer.. it is great if you have a FT Oracle DBA on staff, otherwise, oracle is a *REALLY* poor choice compared to Firebird, Postgre and now Sybase... of which, imo firebird and sybase are easier to get up and running... I also think DB2 shouldn't be counted out either. ..but not at all a fan of administration tasks of an oracle database.
Beyond all of this, oracle's *REAL* performance, is when you are developing oracle solutions.. when using oracle as an rdbms for other systems, it really loses its' shine.
Michael J. Ryan - tracker1.info
I think Sybase is a fine alternative to MS-SQL, on commercial licenses, not sure of the limits, if you have a limited access website, the SQL Express versions are pretty capable, even if limiting the odbc pool to the 10 connections, it works well, and has even more to offer for .net apps.
I would pick DB/2, Firebird, or even mySQL over Oracle though... unless you can afford a FT dba to administer oracle, it's overkill, and odds are your company doesn't need it. for non-windows would go firebird, or postgre most likely.. and depending on the data being stored, would go mysql, simply because it works well for relatively flat data (articles, logging tables, etc.. not great for complex relations, queries, and doesn't have sp's yet though)...
I say use what's best for the job, and just say that if you don't have the need for a full time dba on staff, odds are you don't need what oracle has to offer.
Michael J. Ryan - tracker1.info
I think you can change the batch delimiter in isql or in the login sequence/handshake via the API's.
Sybase doesn't require a stmt delimiter, so the "go" is just at the end of a batch - I usually get all the data for a web page and its components - headers, titles, blurbs "Found 500 users with a last name of Smith", data, footers in one batch and so I only need one "go".
What I could never figure out is how people get anything done without sp's that can return multiple results sets or one result set for that matter.
Michael
Oracle is not a problem unless your developers are SQL illiterate or simply dumb as rocks. In this case, it's not going to matter which engine you choose. Lack of saavy is going to bit you in the *ss sooner or later.
Even mssql has it's arcana.
A Pirate and a Puritan look the same on a balance sheet.
No, I very much doubt any entity, company or a Real Human Being, likes to take on additional maintenance burdens. Nor have I ever claimed they would.
What I did say, was that with an open-sourced database, you have the option of maintaining the old version, instead of having to port everything to a new version/system or living with the old version and it's bugs. With closed source you have no such option. And I'd imagine that companies would rather have an option than not having it, since they can always choose to not take that option when the time comes.
Yes, but is more nontrivial than the money you'd have to pay to port all your applications to a new system and test them ? The original poster said he had "awfull lot" of code that uses MS SQL server.
Supported in what sense ? If you mean commercial support, then there seems to be quite a few companies offering it.
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
not talking about development side, sql syntax etc. I'm talking about configuration, setup, backup, recovery, optimization, etc... Administration of Oracle is a big job.
Michael J. Ryan - tracker1.info
The basics are not particularly Oracle specific. Some of the knobs might be more arcane. However, that is what manuals are for. Administration of any production system is non-trivial. Administration of any high performance system is non-trivial.
Even mssql can perform like a dog if you don't know what you're doing.
A Pirate and a Puritan look the same on a balance sheet.
Not arguing that, but the fact is that most development doesn't need a "performance system" .. my original point, I think still stands, in that oracle is best served when you have the resources to have a dedicated dba... if you don't, then oracle is definately overkill.
Beyond this, ms-sql is pretty easily setup, and deployed... I do see benefits to a system like oracle for *very* large systems, but the fact is, most situations don't call for this.
Michael J. Ryan - tracker1.info
I have used and been a DBA for a large database application. We first used Sybase ASE on Linux for our platform. The application worked great, but the DB was a bit difficult to administer (mostly because of the dump format/devices and the lame isql client. Sybase ASE was good. The support people did not know Linux and suggested we move to Windows NT/2000. UGH. We ported the application and greatly expanded it using PostgreSQL and found that PostgreSQL was easy to code in and easy to maintain. About the only thing we missed was nested transactions (Available in the 8.0 beta. PostgreSQL has performed extremely well in our large database application (86 tables, 90 stored procedures, 22 million records on average) The ASCII dump format is great and porting between platforms is a breeze. I have used Oracle, currently use MySQL but PostgreSQL is my fave.
Since there's been some mention of ASA (Adaptive Server Anywhere, aka SQL Anywhere), just thought I'd mention (sorta late) that ASA users are encouraged to fill out a survey put up by the iAnywhere database engineering team to improve the product: http://www.ianywhere.com/promos/sql_survey/index.h tml
Thanks...
Eric
Thank you for the information. No need to post nastiness on the web for posterity though. Yes I work in IT but no I don't do massive databases. Biggest to now was 1GB. No I am not going to put a DVD on a database obviously, I was reacting to the quote, "..which is more than adequate for all but the most demanding applications", which sounded much like Gates' famous quote. Sybase sounds great though, and I just got a job offer to work with it even. Will take your advice, and very happy to hear about the documentation and stored procedures. Thank you for your reply.