I think you're the one who should "get a grip". If you actually read my post rather than instantly putting on your "flame anything that says anything negative at all with the word 'PostgreSQL' in the same sentence" hat then you would quickly realize that I am not making a "vicious attack on Postgres". I am making a vicious attack on the assholes out there who feel it is necessary to viciously attack MySQL for having the temerity to exist and actually be USED by, well, let me see, lots and lots of people. Very successfully, I might add, and for large installations, which I know must be a little galling for all you purists out there. Isn't it irritating how non-neat and tidy the real world is? Excrutiating that some people seem to be able to get by very well without subqueries or even (gasp) transactions and stored procedures? Wow, I sympathise. Life is hard.
If you read my post, you can see that nowhere do I slam PostgreSQL itself - just the stupid zealots who alienate real people (who might otherwise be interested in investigating the product) with their elitist and dismissive crap.
I don't need a treatise on how to select a database or consultants, thanks very much. Yet another smug post reeking of hubris and the usual blinkered thinking exhibited by zealots on a daily basis...
Comparing me to Osama (get the spelling right at least, don't you read?) and saying that I want to blow up anyone is in such poor taste that you pretty much invalidate anything else that comes after. Sorry, mate.
Nah, more likely it's the disgust that the more experienced developers feel with an obvious charletan that claims that well-proven techniques aren't necessary. Kinda like if ten years ago Yuga had insisted that air-bags, seat belts, and bumpers weren't needed "by most people in most situations". Yeah, right.
yep, I've occasionally developed trivial databases without [subqueries]. But I've seldom developed a serious database without them.
Wow, thanks! Another snide and elitist response! MySQL works very well for a very large number of people, thanks very much. Calling someone who challenges your viewpoint (that transactions MUST be used everywhere and if you don't do subqueries then you're somehow a charletan and not a "real" developer of "serious" databases) is just more of the same drivel I've been reading all this time...
You know, it's not the fact that either database has this or that feature that really grates on me here. It's the attitude of the people who make claims that you can only be developing "serious" databases if you require subqueries, or that "MySQL is a toy" and the classic "it's not a real database"... these are extreme and unnecessary statements that fly in the face of people's REAL WORLD experiences with this RELATIONAL DATABASE.
Sorry, but Yes, it IS a real database. And no, I truly don't appreciate the dismissive, elitist attitudes of people like you. You conveniently ignored the 8k row limits and the VACUUM blocking yet again, because it doesn't fit with your cozy world view. Oh well.
Nice try, but sorry, no dice. Don't you understand my main point, which is that by making your responses so repellent and insulting, you truly don't do yourself or anyone else in the PostgreSQL camp any favors. The zealots will just cheer (as usual) while the rest of us just sigh and discount any valid points you may have made because they are buried inside snide, intellectually elitist remarks.
I get so tired of reading all these "MySQL sucks" comments from the PostgreSQL crowd every time there is any mention of the two databases. Well, for whatever reason, this time I feel compelled to comment in return.
First of all, I remember back when I was trying to decide which of the two databases to use for my bicycle touring community website, crazyguyonabike.com, and I heard all the same arguments that MySQL is a "toy" that's simply an SQL interface to a flat file system, and how PostgreSQL was a "real" database with actual transactions, subqueries and so on... but then, digging further into PostgreSQL, I found out about the 8k row size limit, and the requirement to halt everything to do a VACUUM occasionally. That really set me back on my heels - 8k? That's really pathetic. Ok, so they've fixed that in the later versions, and the VACUUM also seems to be bit less blocking now, but still - it left me wondering just what else these zealots were conveniently choosing to forget about when recommending PostgreSQL. How about speed? How about all the anecdotal accounts of corrupted databases (again, probably improved somewhat more recently). What other major limitations are they conveniently neglecting to mention these days? I ask, because these people were making exactly the same noises back in the 8k rowsize limit, blocking VACUUM days as they are now.
Thing is, MySQL works really well right out of the gate for most things that people want to do in real life. Here are a few factoids for you:
MySQL *does* have real transactions. It has for quite some time now - just use the InnoDB table type. And for anyone who whines about this not being the default table - get a life! Choice is good, and I would say that for 99% of applications out there, you just don't need transactions. Go ahead and crucify me for saying this - I don't care. It works for a lot of people, many companies all over the world use it for heavy duty database work. I've used it for four years now, and it has *never* crashed on me, *never* lost any data. I use RAID and backup regularly too, but MySQL is ROCK SOLID.
MySQL is getting all these things like subselects and procedures, just with a different priority to PostgreSQL. So what? Again, I have never missed subqueries, and I have some pretty complex queries in my website. Again, go ahead and flame me, I don't care - it works for me, and it works for a LOT of other people too. Which brings me to another thing:
There is a definite sense of intellectual snobbery and elitism in many of the comments from PostgreSQL fans slamming MySQL. This reminds me of some of the people I used to know back at University - these are people who are so wrapped up in their own little cozy theoretical worlds that anything remotely practical or pragmatic in its approach seems to be extremely threatening. I think this is one of the big reasons for the bile - these people are simply threatened by the fact that MySQL is so popular and so much more used across the world. Could there also be a little tribalism creeping in here? You know, my team vs your team. They just can't seem to understand that BOTH databases work very well and BOTH databases have weaknesses, which BOTH development teams are working very hard to fix and make better. Getting religious about this sort of thing helps nobody. But, I guess, given the fact that Religion is in fact so widespread in the world, this would then also point to there being a somewhat large number of people who are so hung up on their own little certainties and fixed viewpoints that anything else just can't be left to stand. Like the Christian missionaries, they have to have everyone else adhering to THEIR point of view, regardless of the fact that there are, in fact, many "ways" to spiritual enlightenment (or data management). Again: Not everyone needs transactions. Many people just need simplicity and speed, and they get it with MySQL. And you know what, if they want transactions then they can have that too! Th
What about Debian books?
on
Linux Power Tools
·
· Score: 3, Interesting
I've asked this question before, but it is still bugging me, since I am about to switch my server and workstation from RedHat 7.3 (RIP) to Debian (just as soon as I'm clear about what I'll be losing by going back to Debian "stable" from RH 7.3, or what I'll be risking by going to "testing" or "unstable")... I'm afraid RedHat has used up all my trust by disenfranchising *paying* customers like me. So, good-bye Red Hat, hello... what? At present, Debian looks like the most viable candidate. Everyone seems to think it'll be around for the duration, apt-get rocks, etc etc.
But for a while now I've been noticing that all the books on Debian seem to be out of print or else rather poorly rated. Why is this? Why does nobody seem to think it's worth writing new books on Debian, while there are tons on Red Hat? Is Debian becoming sidelined, or do people think that it has a really solid future alongside the "mainstream" distros such as RH, SuSe et al? Changing distros is a pain in the ass, all those little small tiny differences that eat up days when trying to get your box back to working the way it did before... I don't want to have to change again in a hurry. Debian sounds great, but I am at the same time a little disturbed by how long it seems to be taking to (e.g.) get a better installation program working. If I go with Debian, am I doomed to choosing between either extremely out-of-date packages (stable) or a newer-but-might-crash setup (testing), or spending a lot of time cobbling together a mishmash of my own from both?
I've already heard the arguments about "everything you need to know is somewhere online", but the simple fact is it's often *convenient* to just have a reference all in one place for the common stuff you might want to do with a particular distro. Yes, of course everything is out there and Google is my friend etc, but I know from experience that it can take days to get together the right search query that comes up with the relevant posting or doc for a particular issue. Sometimes the thing you want to do is maddeningly simple, and yet it takes forever to track down. A good book on any distro can only help. Also, when new users are looking at switching to a distro (even relatively intermediate-level ones like me), they often look for a book that will put everything together on paper. It's just human nature. A lack of decent books on a particular distro only hurts that distro, imho.
Is it because a new release of Debian is relatively close? Is anyone aware of anyone else working on a new Debian book that is more specific to the new stuff in Sarge (the upcoming next "stable" version, I think)?
This may seem extreme, I know, but I really think that a lot of things that are currently screwed up in this world are so because of our increasingly complex and impenetrable legal system. See the book "The Death of Common Sense" by Philip K. Howard for more on this subject. This excellent book talks about how our legal system of rules has become increasingly specific ("rationalistic") over time, to the point where it is like a huge, dysfunctional computer program (my analogy) that attempts to anticipate every scenario that might occur - all in the name of taking human discretion out of the equation. As a result, just the opposite effect is achieved, with small minded bureaucrats being able to pick and choose arbitrarily from the morass of rules as it suits them. In the old days, laws were more broadly written, to allow real people to use their own judgement ("common sense") in applying the rules to real world situations. Since the rationalistic approach assumes a finite number of rules can model a world with an infinite number of possible situations, it will only get more and more complex, with more and more loopholes.
But specifically regarding patents - one of the biggest reasons given for patents these days is that without them innovation would cease, since it would no longer be in any company's interest to bear the initial research and development costs. The reasoning is that if anyone else can immediately copy your new ideas (which cost you a lot of money to develop) then you are at a disadvantage, and thus won't do it.
Well, how about the Personal Computer? If IBM had made the PC architecture closed, would the industry have exploded in the way it has? Probably not - witness Apple's failure to grab much of the market due to their own architecture not being clonable. In the beginning, Apple was certainly in a good position to dominate the new market, but instead the cheap "PC clones" dictated the direction, and the industry expanded rapidly. Much innovation followed (in the hardware sense, if not in the operating system sense). Did IBM, the originator of the PC architecture, suffer as a result of all this? Perhaps in the short term, but in the longer term the company has thrived from the byproducts of the PC industry - and, more importantly, so has the whole computer industry. When the greater world benefits, then so does IBM, albeit in a more subtle and indirect way.
Another example: The Internet. Open standards, open source software. Would TCP/IP, HTTP, XML, CGI, and all those other foundation blocks of computing today have taken off if they were patented? I think not. Would anyone have believed that Open Source could work at all if it hadn't done so already? Obviously people are willing to produce innovative stuff just for its own sake. In the short term, perhaps, there is no direct profit to be seen - but it comes further down the road, to everyone. The whole world benefits.
So what about things like drugs? They don't work in the same way as computers, surely... but really, they do. If companies shared their "secrets" then the whole field would leap ahead so much faster than now, and there would be enough for everyone to thrive. Think of all the things that could be done with shared knowledge instead of hoarded secrets! Also, with a more open research community, research and development wouldn't cost as much as it does now anyway.
The whole IP industry is one based on fear, rather than sharing. Fear that I won't make enough money, and everyone else will steal my idea. But history shows that people innovate quite happily without the millstone of patents and IP law hanging around their necks. We have lived through most of history without patents and IP - why is it suddenly essential now? Answer: Greed and fear.
Patents these days stifle the small inventor and favor instead huge corporations who can afford to employ armies of attorneys for the sole purpose of getting "defensive" patents. The system has grown so complex that it simply doesn't do that it was originally
I had 2 licenses for RHN, paying $120 per year. I've used RH since 6.2, something like three years now... and here they are, purposely dumping the very people that put them where they are now. I mean, are they totally stupid or what? Don't they realize that the geeks who use RH at home are the ones who then go to work at the corporations and "enterprise" level shops and recommend systems for new projects... don't they realize who it is that brings things like Linux in through the "back door" to the workplace? Red Hat didn't "just happen", it was championed by geeks. Just because geeks often want to run their own servers at their own expense from home or colocated servers, doesn't mean these *same* geeks aren't in a position to determine what gets used at work. Dumping customers like me who are willing to *PAY* (hello? anyone home???) for stuff - ok so it's not $350 per seat, but on the other hand I don't WANT support, ALL I WANT ARE THE ERRATA. Just the security updates. The expensive part, if I'm not mistaken, is not backporting patches to 7.x, 8.x and 9.x and whatever, they have people who can do that sort of thing in their sleep. No, the expensive part is support. I was quite willing to pay RH for RHN and access to updates.
Oh, yes, and another thing - they actually had the nerve last week to AUTOMATICALLY renew my RHN subscription ($120, cha-ching, thanksverymuch) without my permission. I never said they could do that, but they went ahead anyway. I called immediately, and they have now refunded my money, and I am now in the process of preparing for the (one way) migration to Debian.
This is exactly what Netscape did when they turned their backs on the very people who put them where they were back in the late '90's - Netscape thought they could just forget all about the browser and focus on their Enterprise Server product line. Bad decision - buh bye Netscape!
And now this - the nerve of the guy! Saying that Linux is not ready for the desktop. Jeez. These people never cease to amaze me... I've been using Linux as my desktop for the last three years, and it has literally NEVER crashed on me the way that Windows does ALL THE TIME. There is nothing fundamental about Linux that would prevent it from being used by any user - with the added benefit that in Linux, you can't totally fry your machine settings (as long as you're not running the desktop as root, that is)... and isn't Red Hat the company that is in prime position to MAKE Linux usable on the Desktop? If anyone had the chance to do something here, it was them. But no, they blew it - and they even have thrown a fantastic branding opportunity totally out the window... I mean, Fedora doesn't even have "Red Hat" in the title. So just as they get almost universal recognition, they change the name. Duh. Not the sharpest tools in the shed, methinks.
Ok, end of rant. Sorry, but this whole thing has been annoying me for quite a while... Red Hat has officially lost it, in my opinion. Buh bye now, you'll never get my business again... or any revenue from businesses that I start in the future, either. Chew on that.
Will this mean that security updates will still be available for RedHat 7.3 after it is End of Life'd at the end of this year? If not then I will still be switching to Debian when that happens.
Anyone have any insight on that issue, which is the biggest one by far at present for me regarding RedHat?
Thanks a lot for the suggestions (and the other people who replied as well)... I will in all likelihood be switching over to Debian soon, and these links will come in very useful. I still can't get over how I actually bought a subscription to Red Hat Network for two servers, thinking that I was supporting a "good" Open Source company, only to have them turn around and demand more money for something I don't want - or else lose access to security updates. I am a paying subscriber! What the hell are they thinking? I don't want support, just security updates... how on earth they think they can "end of life" products after only one year, even for people who are *willing to pay for them* via RHN is beyond me... never mind, they've lost one paying customer. Their loss.
I have been using RedHat for the last 3 years, and am currently using 7.3. This is quite nice for me, and I don't want to upgrade to 9.0 (and every year thereafter) when RH end-of-life's 7.3 at the end of this year. I don't like any company forcing me to upgrade... I think a lot of other people feel the same way. I have looked at Debian (and have it installed on one of my partitions now) but to be honest I am a little disturbed by the lack of good Debian books. There just don't seem to be any really good ones out there, let alone recent editions. The most recent is the Debian/GNU Linux Bible, which is 2001, and gets tepid reviews on Amazon. There are, however, tons of Red Hat books, and I am wondering if this says anything about the longevity of Debian going forward? Surely if the publishers thought there was a market out there, then they would be commissioning new and better books on the subject?
I know all the documentation is "out there" but I've "been there, done that" with regard to rooting out all the distributed sources of documentation which exist on the various topics, and to be honest I don't relish the idea of making my life be "about Debian" for the same amount of time that it took to find out all the little tricks that I now know about my RedHat installation... Switching distributions will never be trivial if you have large pre-existing software packages running. Does anyone have any suggestions for moving away from RedHat, and any reasons why there aren't any good up-to-date books on Debian? I just like having at least one reference on hand - we have good books on Perl, MySQL, Apache, Sendmail - why not Debian as a whole?
Sorry if this seems negative - it's not really, I will in all likelihood be switching come November when my RH Network subscription expires. I can't get over how Red Hat is turning its back on the small users like me who can't afford at the moment to buy Advanced Server licenses, don't want or need support, but just need the errata updates! I mean, I am trying to develop a business here, and if/when I switch to another distro, I won't be coming back. It's just too much hassle (the small details and differences are the ones that kill you, as I'm sure everyone here will agree)... very short sighted on RedHat's part.
Red Hat's attitude reminds me of Netscape's just prior to Microsoft destroying them - Netscape seemed to lose interest in the very people that had made them successful, i.e. the small users out there who used their browser. Netscape thought they could instead focus on the corporate server market, and we all know what happened. I tried calling Red Hat about their policy of "end-of-life" for 7.3 and even 8.0, and all the woman I spoke to would say was that I could always buy the Advance Server edition. I explained that I am in that curious middle-ground position of running serious, production servers and yet not being big enough to be able to afford that, and she basically hung up on me. Unbelievable. If that's their attitude, then to be honest I really do hope that they go out of business.
Wrong. Netscape 4.x doesn't even come close to supporting web content as well as Mozilla. It simply doesn't do pretty much the same thing.
Yeah, I've heard all the hype. All I can go by is my day-to-day experience, which is that Netscape 4.x is more usable than Mozilla at the moment. I fully recognize all the improvements in adhering to Web standards, but at the end of the day it doesn't matter if the app runs like treacle.
I'm not happy with Netscape 4.x, for what it's worth. It has a lot of problems with CSS and JavaScript, and it crashes these days more than it used to (probably because web designers are writing for standards which 4.x didn't know about at the time). However I'm more willing to put up with some flakiness in the standards dept than extreme slowness. Nothing more frustrating than clicking on a button and waiting a few seconds for a window to appear.
BTW - By "pretty much the same thing" I was speaking of the large picture of getting me around the Web, not regarding specific low-level functions. Of course Mozilla has a lot more going on. No question. Just wish it could have been implemented with more of an architectural eye toward performance. It's certainly possible to make cross-platform GUI's without resorting to stuff like XUL - just see wxWindows for one example of a C++ toolkit that runs on Windows, X, Mac etc.
Would you like to take time away from the programmer coding, say, the routines to resize text so that people with less-than-perfect eyesight can surf more easily, and instead assign him to optimising some code so that it starts up 5% faster for you?
I was expecting someone to say pretty much what you're saying, and there really isn't a "right" or "wrong" answer here. It's a matter of opinion. In my opinion, it's not a question of making something load 5% faster for me at the expense of visually challenged users. That is being way too black and white - and by the way, it's a pretty nasty way of making your point. If you're going to use that kind of argument, why not say that the developers are discriminating against poorer people who can't afford faster, newer, more expensive computers (e.g. many schools)??? And finally, we're not talking about a minor optimization of 5% here, we're talking about the application being actually unusable for someone with a computer that is perfectly adequate for all other apps on the desktop, including software development.
I am simply saying that an application (Netscape 4.x) that was written over five years ago, and does pretty much the same thing as Mozilla, is a lot, lot faster. As a software developer myself I can understand the desire to keep adding features, but I don't see why it should be mutually exclusive to have both performance and stability and (some) features.
Don't bellyache about the speed when you aren't contributing in any way to the project.
This is a crass statement. Are you saying that everyone should just keep quiet and not say what's on their mind? This reminds me of the current political climate in America, where if you criticize the government's policies then you are labeled as "unpatriotic" or "UnAmerican".
Let's be very clear: I am absolutely free to comment on something that I am not contributing to, just as other people are absolutely free to comment on stuff I do, but that they are not contributing to. The argument that criticism is only valid if you are a contributor is a classic knee-jerk playground response which is on the same level as "Well, if you don't agree with the government, then don't live in America". If nobody commented on stuff, then things would not improve. I can't do everything, I can't participate in every single thing in this world that I have an opinion about. However I do value input from other thinking people regarding the stuff I do participate in. Sometimes the most valuable input comes from people who are not deeply involved with a project, because they are not too close to the woods to see the trees.
I don't particularly agree with you about C++. It is certainly possible to write fast, efficient code using C++ (I've used it since 1989). As with any language, you can also create a real mess. I don't think C++ is the problem here, but rather the use of XUL for all user interface tasks. For some reason the implementation of this requires a lot of processing just to bring up a window or pretty much do anything significant. It's even more sluggish than Java.
Rather than being lazy, I would posit that the Mozilla developers all must have relatively newer, faster machines to develop on, because if they had a box like mine then I don't see how they could be satisfied with the performance of Mozilla. I guess you can argue that you need to move with the times, and I agree to an extent, but then I go and look again at Netscape 4.x, and how fast it is. And how it was written five years ago or more. And Mozilla really isn't doing all that much more.
I truly do not care about all the fancy features such as "skins" and tabs and "intelligent" browsing. I just want a fast, lean browser that does the job, adheres to standards and gives me the ability to customize stuff (up to a point - take that too far and you end up in slow, bloated XUL land)...
JMHO, of course. I am not slamming the developers of Mozilla, but rather speaking out loud about what is, for me personally, the biggest issue with Mozilla.
Something I don't hear talked about all that much is how slow Mozilla is compared to Netscape 4.x. I use a 450 MHz desktop, which is perfectly adequate speed-wise for every other application that I use (including software development on both Windows and Linux). I can't for the life of me understand why "progress" always has to involve slower, bigger, more complicated applications. It happened with Windows, and now it seems to be happening with Linux and Open Source too. What on earth is Mozilla doing so different to what the older Netscape 4.x did? They both are supposed to be just web browsers. But 4.x is so much faster and more responsive. Sure, 4.x is getting long in the tooth now with regard to standards, but come on - if they could write something this fast five years ago, why is the latest and greatest so demanding of CPU? From an end-user point of view, it really doesn't do anything different - it renders web pages and does email. Why should that be so CPU intensive? I understand that Mozilla uses XUL, and while it's a laudable goal (cross platform), in the end I judge software these days by how usable it is in real life. And I'm sorry to say that when it takes a good three to five seconds just to bring up a "new message" or "preferences" window, that is unusable to me. Are we moving away from compiled code and towards interpreted scripts for everything, no matter what the cost in terms of performance? And surely if so much is interpreted, shouldn't the code size go down, not up? I'm sad that nobody seems to be talking about this. Since when is a 450 MHz computer too slow for web browsing and email???? I can understand it being slow if I were trying to render 3D animations, but come on...
1. Custom software for the place where you work - this is the kind that most people get money for. Open Source is not really a good fit for this, because the application is usually very narrow, specific and often tied to proprietary business systems, databases and business logic.
2. General tool software which is applicable to lots of different situations - this is the type of software that Open Source is good for. Look at all OSS, it's general stuff. Tools. Utilities.
OSS can be very helpful in building the specific stuff. It's not all that complicated. People make their money developing (1), preferably using (2). In their spare time (and, if they're lucky, work time) they develop the Open Source stuff. If the open source stuff they develop becomes successful enough, then they maybe get a chance to do it most of the time, because some company will recognise the value, and dedicate some resources to paying the guy.
I really don't see what's so complex about this. People do Open Source software because they like doing it. They get peer recognition. They don't do it to get paid - not directly, in money. They get paid in reputation and self respect. Everybody benefits. People get their bread and butter from doing the specific, non-general, business stuff.
Slashdot Mirror
I think you're the one who should "get a grip". If you actually read my post rather than instantly putting on your "flame anything that says anything negative at all with the word 'PostgreSQL' in the same sentence" hat then you would quickly realize that I am not making a "vicious attack on Postgres". I am making a vicious attack on the assholes out there who feel it is necessary to viciously attack MySQL for having the temerity to exist and actually be USED by, well, let me see, lots and lots of people. Very successfully, I might add, and for large installations, which I know must be a little galling for all you purists out there. Isn't it irritating how non-neat and tidy the real world is? Excrutiating that some people seem to be able to get by very well without subqueries or even (gasp) transactions and stored procedures? Wow, I sympathise. Life is hard.
If you read my post, you can see that nowhere do I slam PostgreSQL itself - just the stupid zealots who alienate real people (who might otherwise be interested in investigating the product) with their elitist and dismissive crap.
I don't need a treatise on how to select a database or consultants, thanks very much. Yet another smug post reeking of hubris and the usual blinkered thinking exhibited by zealots on a daily basis...
Comparing me to Osama (get the spelling right at least, don't you read?) and saying that I want to blow up anyone is in such poor taste that you pretty much invalidate anything else that comes after. Sorry, mate.
Thanks for nicely illustrating my point!
Nah, more likely it's the disgust that the more experienced developers feel with an obvious charletan that claims that well-proven techniques aren't necessary. Kinda like if ten years ago Yuga had insisted that air-bags, seat belts, and bumpers weren't needed "by most people in most situations". Yeah, right.
yep, I've occasionally developed trivial databases without [subqueries]. But I've seldom developed a serious database without them.
Wow, thanks! Another snide and elitist response! MySQL works very well for a very large number of people, thanks very much. Calling someone who challenges your viewpoint (that transactions MUST be used everywhere and if you don't do subqueries then you're somehow a charletan and not a "real" developer of "serious" databases) is just more of the same drivel I've been reading all this time...
You know, it's not the fact that either database has this or that feature that really grates on me here. It's the attitude of the people who make claims that you can only be developing "serious" databases if you require subqueries, or that "MySQL is a toy" and the classic "it's not a real database"... these are extreme and unnecessary statements that fly in the face of people's REAL WORLD experiences with this RELATIONAL DATABASE.
Sorry, but Yes, it IS a real database. And no, I truly don't appreciate the dismissive, elitist attitudes of people like you. You conveniently ignored the 8k row limits and the VACUUM blocking yet again, because it doesn't fit with your cozy world view. Oh well.
Nice try, but sorry, no dice. Don't you understand my main point, which is that by making your responses so repellent and insulting, you truly don't do yourself or anyone else in the PostgreSQL camp any favors. The zealots will just cheer (as usual) while the rest of us just sigh and discount any valid points you may have made because they are buried inside snide, intellectually elitist remarks.
I get so tired of reading all these "MySQL sucks" comments from the PostgreSQL crowd every time there is any mention of the two databases. Well, for whatever reason, this time I feel compelled to comment in return.
First of all, I remember back when I was trying to decide which of the two databases to use for my bicycle touring community website, crazyguyonabike.com, and I heard all the same arguments that MySQL is a "toy" that's simply an SQL interface to a flat file system, and how PostgreSQL was a "real" database with actual transactions, subqueries and so on... but then, digging further into PostgreSQL, I found out about the 8k row size limit, and the requirement to halt everything to do a VACUUM occasionally. That really set me back on my heels - 8k? That's really pathetic. Ok, so they've fixed that in the later versions, and the VACUUM also seems to be bit less blocking now, but still - it left me wondering just what else these zealots were conveniently choosing to forget about when recommending PostgreSQL. How about speed? How about all the anecdotal accounts of corrupted databases (again, probably improved somewhat more recently). What other major limitations are they conveniently neglecting to mention these days? I ask, because these people were making exactly the same noises back in the 8k rowsize limit, blocking VACUUM days as they are now.
Thing is, MySQL works really well right out of the gate for most things that people want to do in real life. Here are a few factoids for you:
MySQL *does* have real transactions. It has for quite some time now - just use the InnoDB table type. And for anyone who whines about this not being the default table - get a life! Choice is good, and I would say that for 99% of applications out there, you just don't need transactions. Go ahead and crucify me for saying this - I don't care. It works for a lot of people, many companies all over the world use it for heavy duty database work. I've used it for four years now, and it has *never* crashed on me, *never* lost any data. I use RAID and backup regularly too, but MySQL is ROCK SOLID.
MySQL is getting all these things like subselects and procedures, just with a different priority to PostgreSQL. So what? Again, I have never missed subqueries, and I have some pretty complex queries in my website. Again, go ahead and flame me, I don't care - it works for me, and it works for a LOT of other people too. Which brings me to another thing:
There is a definite sense of intellectual snobbery and elitism in many of the comments from PostgreSQL fans slamming MySQL. This reminds me of some of the people I used to know back at University - these are people who are so wrapped up in their own little cozy theoretical worlds that anything remotely practical or pragmatic in its approach seems to be extremely threatening. I think this is one of the big reasons for the bile - these people are simply threatened by the fact that MySQL is so popular and so much more used across the world. Could there also be a little tribalism creeping in here? You know, my team vs your team. They just can't seem to understand that BOTH databases work very well and BOTH databases have weaknesses, which BOTH development teams are working very hard to fix and make better. Getting religious about this sort of thing helps nobody. But, I guess, given the fact that Religion is in fact so widespread in the world, this would then also point to there being a somewhat large number of people who are so hung up on their own little certainties and fixed viewpoints that anything else just can't be left to stand. Like the Christian missionaries, they have to have everyone else adhering to THEIR point of view, regardless of the fact that there are, in fact, many "ways" to spiritual enlightenment (or data management). Again: Not everyone needs transactions. Many people just need simplicity and speed, and they get it with MySQL. And you know what, if they want transactions then they can have that too! Th
I've asked this question before, but it is still bugging me, since I am about to switch my server and workstation from RedHat 7.3 (RIP) to Debian (just as soon as I'm clear about what I'll be losing by going back to Debian "stable" from RH 7.3, or what I'll be risking by going to "testing" or "unstable")... I'm afraid RedHat has used up all my trust by disenfranchising *paying* customers like me. So, good-bye Red Hat, hello ... what? At present, Debian looks like the most viable candidate. Everyone seems to think it'll be around for the duration, apt-get rocks, etc etc.
But for a while now I've been noticing that all the books on Debian seem to be out of print or else rather poorly rated. Why is this? Why does nobody seem to think it's worth writing new books on Debian, while there are tons on Red Hat? Is Debian becoming sidelined, or do people think that it has a really solid future alongside the "mainstream" distros such as RH, SuSe et al? Changing distros is a pain in the ass, all those little small tiny differences that eat up days when trying to get your box back to working the way it did before... I don't want to have to change again in a hurry. Debian sounds great, but I am at the same time a little disturbed by how long it seems to be taking to (e.g.) get a better installation program working. If I go with Debian, am I doomed to choosing between either extremely out-of-date packages (stable) or a newer-but-might-crash setup (testing), or spending a lot of time cobbling together a mishmash of my own from both?
I've already heard the arguments about "everything you need to know is somewhere online", but the simple fact is it's often *convenient* to just have a reference all in one place for the common stuff you might want to do with a particular distro. Yes, of course everything is out there and Google is my friend etc, but I know from experience that it can take days to get together the right search query that comes up with the relevant posting or doc for a particular issue. Sometimes the thing you want to do is maddeningly simple, and yet it takes forever to track down. A good book on any distro can only help. Also, when new users are looking at switching to a distro (even relatively intermediate-level ones like me), they often look for a book that will put everything together on paper. It's just human nature. A lack of decent books on a particular distro only hurts that distro, imho.
Is it because a new release of Debian is relatively close? Is anyone aware of anyone else working on a new Debian book that is more specific to the new stuff in Sarge (the upcoming next "stable" version, I think)?
TIA, sorry if this is a little off-topic.
This may seem extreme, I know, but I really think that a lot of things that are currently screwed up in this world are so because of our increasingly complex and impenetrable legal system. See the book "The Death of Common Sense" by Philip K. Howard for more on this subject. This excellent book talks about how our legal system of rules has become increasingly specific ("rationalistic") over time, to the point where it is like a huge, dysfunctional computer program (my analogy) that attempts to anticipate every scenario that might occur - all in the name of taking human discretion out of the equation. As a result, just the opposite effect is achieved, with small minded bureaucrats being able to pick and choose arbitrarily from the morass of rules as it suits them. In the old days, laws were more broadly written, to allow real people to use their own judgement ("common sense") in applying the rules to real world situations. Since the rationalistic approach assumes a finite number of rules can model a world with an infinite number of possible situations, it will only get more and more complex, with more and more loopholes.
But specifically regarding patents - one of the biggest reasons given for patents these days is that without them innovation would cease, since it would no longer be in any company's interest to bear the initial research and development costs. The reasoning is that if anyone else can immediately copy your new ideas (which cost you a lot of money to develop) then you are at a disadvantage, and thus won't do it.
Well, how about the Personal Computer? If IBM had made the PC architecture closed, would the industry have exploded in the way it has? Probably not - witness Apple's failure to grab much of the market due to their own architecture not being clonable. In the beginning, Apple was certainly in a good position to dominate the new market, but instead the cheap "PC clones" dictated the direction, and the industry expanded rapidly. Much innovation followed (in the hardware sense, if not in the operating system sense). Did IBM, the originator of the PC architecture, suffer as a result of all this? Perhaps in the short term, but in the longer term the company has thrived from the byproducts of the PC industry - and, more importantly, so has the whole computer industry. When the greater world benefits, then so does IBM, albeit in a more subtle and indirect way.
Another example: The Internet. Open standards, open source software. Would TCP/IP, HTTP, XML, CGI, and all those other foundation blocks of computing today have taken off if they were patented? I think not. Would anyone have believed that Open Source could work at all if it hadn't done so already? Obviously people are willing to produce innovative stuff just for its own sake. In the short term, perhaps, there is no direct profit to be seen - but it comes further down the road, to everyone. The whole world benefits.
So what about things like drugs? They don't work in the same way as computers, surely... but really, they do. If companies shared their "secrets" then the whole field would leap ahead so much faster than now, and there would be enough for everyone to thrive. Think of all the things that could be done with shared knowledge instead of hoarded secrets! Also, with a more open research community, research and development wouldn't cost as much as it does now anyway.
The whole IP industry is one based on fear, rather than sharing. Fear that I won't make enough money, and everyone else will steal my idea. But history shows that people innovate quite happily without the millstone of patents and IP law hanging around their necks. We have lived through most of history without patents and IP - why is it suddenly essential now? Answer: Greed and fear.
Patents these days stifle the small inventor and favor instead huge corporations who can afford to employ armies of attorneys for the sole purpose of getting "defensive" patents. The system has grown so complex that it simply doesn't do that it was originally
I had 2 licenses for RHN, paying $120 per year. I've used RH since 6.2, something like three years now... and here they are, purposely dumping the very people that put them where they are now. I mean, are they totally stupid or what? Don't they realize that the geeks who use RH at home are the ones who then go to work at the corporations and "enterprise" level shops and recommend systems for new projects... don't they realize who it is that brings things like Linux in through the "back door" to the workplace? Red Hat didn't "just happen", it was championed by geeks. Just because geeks often want to run their own servers at their own expense from home or colocated servers, doesn't mean these *same* geeks aren't in a position to determine what gets used at work. Dumping customers like me who are willing to *PAY* (hello? anyone home???) for stuff - ok so it's not $350 per seat, but on the other hand I don't WANT support, ALL I WANT ARE THE ERRATA. Just the security updates. The expensive part, if I'm not mistaken, is not backporting patches to 7.x, 8.x and 9.x and whatever, they have people who can do that sort of thing in their sleep. No, the expensive part is support. I was quite willing to pay RH for RHN and access to updates.
Oh, yes, and another thing - they actually had the nerve last week to AUTOMATICALLY renew my RHN subscription ($120, cha-ching, thanksverymuch) without my permission. I never said they could do that, but they went ahead anyway. I called immediately, and they have now refunded my money, and I am now in the process of preparing for the (one way) migration to Debian.
This is exactly what Netscape did when they turned their backs on the very people who put them where they were back in the late '90's - Netscape thought they could just forget all about the browser and focus on their Enterprise Server product line. Bad decision - buh bye Netscape!
And now this - the nerve of the guy! Saying that Linux is not ready for the desktop. Jeez. These people never cease to amaze me... I've been using Linux as my desktop for the last three years, and it has literally NEVER crashed on me the way that Windows does ALL THE TIME. There is nothing fundamental about Linux that would prevent it from being used by any user - with the added benefit that in Linux, you can't totally fry your machine settings (as long as you're not running the desktop as root, that is)... and isn't Red Hat the company that is in prime position to MAKE Linux usable on the Desktop? If anyone had the chance to do something here, it was them. But no, they blew it - and they even have thrown a fantastic branding opportunity totally out the window... I mean, Fedora doesn't even have "Red Hat" in the title. So just as they get almost universal recognition, they change the name. Duh. Not the sharpest tools in the shed, methinks.
Ok, end of rant. Sorry, but this whole thing has been annoying me for quite a while... Red Hat has officially lost it, in my opinion. Buh bye now, you'll never get my business again... or any revenue from businesses that I start in the future, either. Chew on that.
Buh bye, Red Hat.
Will this mean that security updates will still be available for RedHat 7.3 after it is End of Life'd at the end of this year? If not then I will still be switching to Debian when that happens.
/Neil
Anyone have any insight on that issue, which is the biggest one by far at present for me regarding RedHat?
TIA
Thanks a lot for the suggestions (and the other people who replied as well)... I will in all likelihood be switching over to Debian soon, and these links will come in very useful. I still can't get over how I actually bought a subscription to Red Hat Network for two servers, thinking that I was supporting a "good" Open Source company, only to have them turn around and demand more money for something I don't want - or else lose access to security updates. I am a paying subscriber! What the hell are they thinking? I don't want support, just security updates... how on earth they think they can "end of life" products after only one year, even for people who are *willing to pay for them* via RHN is beyond me... never mind, they've lost one paying customer. Their loss.
Thanks again!
-Neil
I have been using RedHat for the last 3 years, and am currently using 7.3. This is quite nice for me, and I don't want to upgrade to 9.0 (and every year thereafter) when RH end-of-life's 7.3 at the end of this year. I don't like any company forcing me to upgrade... I think a lot of other people feel the same way. I have looked at Debian (and have it installed on one of my partitions now) but to be honest I am a little disturbed by the lack of good Debian books. There just don't seem to be any really good ones out there, let alone recent editions. The most recent is the Debian/GNU Linux Bible, which is 2001, and gets tepid reviews on Amazon. There are, however, tons of Red Hat books, and I am wondering if this says anything about the longevity of Debian going forward? Surely if the publishers thought there was a market out there, then they would be commissioning new and better books on the subject?
I know all the documentation is "out there" but I've "been there, done that" with regard to rooting out all the distributed sources of documentation which exist on the various topics, and to be honest I don't relish the idea of making my life be "about Debian" for the same amount of time that it took to find out all the little tricks that I now know about my RedHat installation... Switching distributions will never be trivial if you have large pre-existing software packages running. Does anyone have any suggestions for moving away from RedHat, and any reasons why there aren't any good up-to-date books on Debian? I just like having at least one reference on hand - we have good books on Perl, MySQL, Apache, Sendmail - why not Debian as a whole?
Sorry if this seems negative - it's not really, I will in all likelihood be switching come November when my RH Network subscription expires. I can't get over how Red Hat is turning its back on the small users like me who can't afford at the moment to buy Advanced Server licenses, don't want or need support, but just need the errata updates! I mean, I am trying to develop a business here, and if/when I switch to another distro, I won't be coming back. It's just too much hassle (the small details and differences are the ones that kill you, as I'm sure everyone here will agree)... very short sighted on RedHat's part.
Red Hat's attitude reminds me of Netscape's just prior to Microsoft destroying them - Netscape seemed to lose interest in the very people that had made them successful, i.e. the small users out there who used their browser. Netscape thought they could instead focus on the corporate server market, and we all know what happened. I tried calling Red Hat about their policy of "end-of-life" for 7.3 and even 8.0, and all the woman I spoke to would say was that I could always buy the Advance Server edition. I explained that I am in that curious middle-ground position of running serious, production servers and yet not being big enough to be able to afford that, and she basically hung up on me. Unbelievable. If that's their attitude, then to be honest I really do hope that they go out of business.
Suggestions welcomed, and sorry for the rant.
Yeah, I've heard all the hype. All I can go by is my day-to-day experience, which is that Netscape 4.x is more usable than Mozilla at the moment. I fully recognize all the improvements in adhering to Web standards, but at the end of the day it doesn't matter if the app runs like treacle.
I'm not happy with Netscape 4.x, for what it's worth. It has a lot of problems with CSS and JavaScript, and it crashes these days more than it used to (probably because web designers are writing for standards which 4.x didn't know about at the time). However I'm more willing to put up with some flakiness in the standards dept than extreme slowness. Nothing more frustrating than clicking on a button and waiting a few seconds for a window to appear.
BTW - By "pretty much the same thing" I was speaking of the large picture of getting me around the Web, not regarding specific low-level functions. Of course Mozilla has a lot more going on. No question. Just wish it could have been implemented with more of an architectural eye toward performance. It's certainly possible to make cross-platform GUI's without resorting to stuff like XUL - just see wxWindows for one example of a C++ toolkit that runs on Windows, X, Mac etc.
I was expecting someone to say pretty much what you're saying, and there really isn't a "right" or "wrong" answer here. It's a matter of opinion. In my opinion, it's not a question of making something load 5% faster for me at the expense of visually challenged users. That is being way too black and white - and by the way, it's a pretty nasty way of making your point. If you're going to use that kind of argument, why not say that the developers are discriminating against poorer people who can't afford faster, newer, more expensive computers (e.g. many schools)??? And finally, we're not talking about a minor optimization of 5% here, we're talking about the application being actually unusable for someone with a computer that is perfectly adequate for all other apps on the desktop, including software development.
I am simply saying that an application (Netscape 4.x) that was written over five years ago, and does pretty much the same thing as Mozilla, is a lot, lot faster. As a software developer myself I can understand the desire to keep adding features, but I don't see why it should be mutually exclusive to have both performance and stability and (some) features.
Don't bellyache about the speed when you aren't contributing in any way to the project.
This is a crass statement. Are you saying that everyone should just keep quiet and not say what's on their mind? This reminds me of the current political climate in America, where if you criticize the government's policies then you are labeled as "unpatriotic" or "UnAmerican".
Let's be very clear: I am absolutely free to comment on something that I am not contributing to, just as other people are absolutely free to comment on stuff I do, but that they are not contributing to. The argument that criticism is only valid if you are a contributor is a classic knee-jerk playground response which is on the same level as "Well, if you don't agree with the government, then don't live in America". If nobody commented on stuff, then things would not improve. I can't do everything, I can't participate in every single thing in this world that I have an opinion about. However I do value input from other thinking people regarding the stuff I do participate in. Sometimes the most valuable input comes from people who are not deeply involved with a project, because they are not too close to the woods to see the trees.
I don't particularly agree with you about C++. It is certainly possible to write fast, efficient code using C++ (I've used it since 1989). As with any language, you can also create a real mess. I don't think C++ is the problem here, but rather the use of XUL for all user interface tasks. For some reason the implementation of this requires a lot of processing just to bring up a window or pretty much do anything significant. It's even more sluggish than Java.
Rather than being lazy, I would posit that the Mozilla developers all must have relatively newer, faster machines to develop on, because if they had a box like mine then I don't see how they could be satisfied with the performance of Mozilla. I guess you can argue that you need to move with the times, and I agree to an extent, but then I go and look again at Netscape 4.x, and how fast it is. And how it was written five years ago or more. And Mozilla really isn't doing all that much more.
I truly do not care about all the fancy features such as "skins" and tabs and "intelligent" browsing. I just want a fast, lean browser that does the job, adheres to standards and gives me the ability to customize stuff (up to a point - take that too far and you end up in slow, bloated XUL land)...
JMHO, of course. I am not slamming the developers of Mozilla, but rather speaking out loud about what is, for me personally, the biggest issue with Mozilla.
Something I don't hear talked about all that much is how slow Mozilla is compared to Netscape 4.x. I use a 450 MHz desktop, which is perfectly adequate speed-wise for every other application that I use (including software development on both Windows and Linux). I can't for the life of me understand why "progress" always has to involve slower, bigger, more complicated applications. It happened with Windows, and now it seems to be happening with Linux and Open Source too. What on earth is Mozilla doing so different to what the older Netscape 4.x did? They both are supposed to be just web browsers. But 4.x is so much faster and more responsive. Sure, 4.x is getting long in the tooth now with regard to standards, but come on - if they could write something this fast five years ago, why is the latest and greatest so demanding of CPU? From an end-user point of view, it really doesn't do anything different - it renders web pages and does email. Why should that be so CPU intensive? I understand that Mozilla uses XUL, and while it's a laudable goal (cross platform), in the end I judge software these days by how usable it is in real life. And I'm sorry to say that when it takes a good three to five seconds just to bring up a "new message" or "preferences" window, that is unusable to me. Are we moving away from compiled code and towards interpreted scripts for everything, no matter what the cost in terms of performance? And surely if so much is interpreted, shouldn't the code size go down, not up? I'm sad that nobody seems to be talking about this. Since when is a 450 MHz computer too slow for web browsing and email???? I can understand it being slow if I were trying to render 3D animations, but come on...
1. Custom software for the place where you work - this is the kind that most people get money for. Open Source is not really a good fit for this, because the application is usually very narrow, specific and often tied to proprietary business systems, databases and business logic.
2. General tool software which is applicable to lots of different situations - this is the type of software that Open Source is good for. Look at all OSS, it's general stuff. Tools. Utilities.
OSS can be very helpful in building the specific stuff. It's not all that complicated. People make their money developing (1), preferably using (2). In their spare time (and, if they're lucky, work time) they develop the Open Source stuff. If the open source stuff they develop becomes successful enough, then they maybe get a chance to do it most of the time, because some company will recognise the value, and dedicate some resources to paying the guy.
I really don't see what's so complex about this. People do Open Source software because they like doing it. They get peer recognition. They don't do it to get paid - not directly, in money. They get paid in reputation and self respect. Everybody benefits. People get their bread and butter from doing the specific, non-general, business stuff.
Or am I missing something here?