Entered as a new post because of character limit
Thursday, October 23rd, 2003 01:27 amIn response to b0rg:
Risking to start another flame I can tell you one thing: microsoft have a luxury of huge amount of ppl testing their code for bugs, and publically announcing those bugs.
[Cut to the scene in Aliens where Ripley straps the flamethrower to the combined assault rifle / grenade launcher.]
I wonder why otherwise educated people take some strange (well for me) pleasure in slagging the presentation language of (surprise, surprise) MS CEOs?
"Otherwise educated"? You really do want to play with fire, don't you?
Actually, I'm not a "grammar nazi", I'm a semantics nazi. I can quibble about grammar points (in fact my job demands that I do) but I generally come down on the side of flexibility: I loathe made-up grammar rules (like not splitting infinitives or not ending a sentence with a preposition); I quite like the "so ... NOT" structure, even if I don't recommend it to my students; I start sentences with co-ordinate conjunctions. What I can't stand is sloppy semantics, especially when used to avoid saying anything definite. You are right that MS executives are not the only offenders; I picked those two examples because I read press releases by software companies, but I have no reason to read Fortune or Forbes. I imagine you're right, though - I've heard some terrible things from managers in my school. (I'm in the unusual position of working in a university, but in an English department which is staffed mainly by people from the commercial world of TESOL, which has adopted management-speak in an attempt to give business respectability to a profession that was founded by backpackers.)
microsoft have a luxury of huge amount of ppl testing their code for bugs, and publically announcing those bugs. Which in turn being fixed.
So why are there so many bugs? I'm not talking about bugs in applications - anyone can write a buggy application - I'm talking about the OS. I used the original, not the second edition of Windows 98, and it was enough to get me off my arse and install Linux. Netscape was crashing every twenty minutes. Sure, Netscape in those days was buggy, and it also crashed sometimes when I was in Linux. The difference was that when it crashed in Linux, I didn't have to reboot, because the OS was stable. Then Windows 98 Second Edition came out - a bugfix trumpeted as an improved version. I use it occasionally, and it's still buggy.
Recently I moved offices and had to deal with networking a computer which some poor sucker had installed XP Home on (apparently we even paid for it). I spent a whole day trying to get it to accept printing commands from Samba, only to find out that this was impossible in the Home version. I then tried to set it up so that I could virus-scan it from my Linux box, as I do with the Win98 computers in our office. No deal. Next week it got a virus - something trivial like Sobig. Er, hello? Bugfixes? Trustworthy Computing?
Being in software for about 12 years I can tell you another thing: every computer program (well apart from hello world) has bugs in it
Hello World and TeX ;-)
it's only a matter of finding and fixing them. Which they do.
Which they don't, or if they do, they make new bugs in the process. Last time I was home, I had a look at my mother's computer, and downloaded about a year's-worth of MS security patches and updates. The result was that Windows broke and I had to reformat and reinstall from scratch (that's not me being bloody-minded; I had my brother, who has almost thirty years' experience in programming, look at it). I only managed to save her data files by installing Linux on a different partition and copying them over. I wouldn't mind so much if they didn't make people pay good money for it. If I download some alpha software from Sourceforge or Freshmeat, I expect problems, but if I pay $100 for anything, I expect it to work. If someone sold me a microwave oven which burnt my food, I'd ask for my money back. Why should an operating system be different?
The biggest problem in software as I can see is not the software itself and not the programmers - it's the damn users who always moan for the reason "we want that next best feature", or just moan with no reason at all :-)
This is nonsense. I realise that users can sometimes be a problem (I had to work with one person who was death to computers - she once called me in to deal with a computer "problem" caused by her accidentally going into the BIOS settings). However, it is not users who demand the next best feature. New features are pushed on users by software companies who need to push the latest verison on them. The average Word user uses a tiny fraction of the available features, and usually doesn't even know about them: all they want is a smart typewriter (anyone who wants to do real typesetting uses something like Quark or LaTeX). The typical computer user wants to surf the Net, send emails, play games and write stuff. They do not want their computer to crash; they do not want to have to pay for anti-virus software or firewalls; they do not want an "experience". In other words, they have no reason to use Windows.
Risking to start another flame I can tell you one thing: microsoft have a luxury of huge amount of ppl testing their code for bugs, and publically announcing those bugs.
[Cut to the scene in Aliens where Ripley straps the flamethrower to the combined assault rifle / grenade launcher.]
I wonder why otherwise educated people take some strange (well for me) pleasure in slagging the presentation language of (surprise, surprise) MS CEOs?
"Otherwise educated"? You really do want to play with fire, don't you?
Actually, I'm not a "grammar nazi", I'm a semantics nazi. I can quibble about grammar points (in fact my job demands that I do) but I generally come down on the side of flexibility: I loathe made-up grammar rules (like not splitting infinitives or not ending a sentence with a preposition); I quite like the "so ... NOT" structure, even if I don't recommend it to my students; I start sentences with co-ordinate conjunctions. What I can't stand is sloppy semantics, especially when used to avoid saying anything definite. You are right that MS executives are not the only offenders; I picked those two examples because I read press releases by software companies, but I have no reason to read Fortune or Forbes. I imagine you're right, though - I've heard some terrible things from managers in my school. (I'm in the unusual position of working in a university, but in an English department which is staffed mainly by people from the commercial world of TESOL, which has adopted management-speak in an attempt to give business respectability to a profession that was founded by backpackers.)
microsoft have a luxury of huge amount of ppl testing their code for bugs, and publically announcing those bugs. Which in turn being fixed.
So why are there so many bugs? I'm not talking about bugs in applications - anyone can write a buggy application - I'm talking about the OS. I used the original, not the second edition of Windows 98, and it was enough to get me off my arse and install Linux. Netscape was crashing every twenty minutes. Sure, Netscape in those days was buggy, and it also crashed sometimes when I was in Linux. The difference was that when it crashed in Linux, I didn't have to reboot, because the OS was stable. Then Windows 98 Second Edition came out - a bugfix trumpeted as an improved version. I use it occasionally, and it's still buggy.
Recently I moved offices and had to deal with networking a computer which some poor sucker had installed XP Home on (apparently we even paid for it). I spent a whole day trying to get it to accept printing commands from Samba, only to find out that this was impossible in the Home version. I then tried to set it up so that I could virus-scan it from my Linux box, as I do with the Win98 computers in our office. No deal. Next week it got a virus - something trivial like Sobig. Er, hello? Bugfixes? Trustworthy Computing?
Being in software for about 12 years I can tell you another thing: every computer program (well apart from hello world) has bugs in it
Hello World and TeX ;-)
it's only a matter of finding and fixing them. Which they do.
Which they don't, or if they do, they make new bugs in the process. Last time I was home, I had a look at my mother's computer, and downloaded about a year's-worth of MS security patches and updates. The result was that Windows broke and I had to reformat and reinstall from scratch (that's not me being bloody-minded; I had my brother, who has almost thirty years' experience in programming, look at it). I only managed to save her data files by installing Linux on a different partition and copying them over. I wouldn't mind so much if they didn't make people pay good money for it. If I download some alpha software from Sourceforge or Freshmeat, I expect problems, but if I pay $100 for anything, I expect it to work. If someone sold me a microwave oven which burnt my food, I'd ask for my money back. Why should an operating system be different?
The biggest problem in software as I can see is not the software itself and not the programmers - it's the damn users who always moan for the reason "we want that next best feature", or just moan with no reason at all :-)
This is nonsense. I realise that users can sometimes be a problem (I had to work with one person who was death to computers - she once called me in to deal with a computer "problem" caused by her accidentally going into the BIOS settings). However, it is not users who demand the next best feature. New features are pushed on users by software companies who need to push the latest verison on them. The average Word user uses a tiny fraction of the available features, and usually doesn't even know about them: all they want is a smart typewriter (anyone who wants to do real typesetting uses something like Quark or LaTeX). The typical computer user wants to surf the Net, send emails, play games and write stuff. They do not want their computer to crash; they do not want to have to pay for anti-virus software or firewalls; they do not want an "experience". In other words, they have no reason to use Windows.
no subject
Date: 2003-10-24 03:53 pm (UTC)[Pauses to think what a non-geek would make of that last sentence out of context. And to think all this started with my post about mangling the English language!]
The Mac analogy is good, especially considering that OSX is essentially a pretty GUI on top of Unix.
no subject
Date: 2003-10-24 08:07 pm (UTC)As for ACPI - the version of ACPI in 2.4 is rudimentary, and doesn't support suspend states - some people have tried to wedge them in, but they generally don't work very well. And, just as a point of reference, I work for Intel, and I work about twenty yards away from the guy who wrote the version of ACPI that went into the 2.4 kernel, and the version that's in 2.6. With the 2.6 kernel, ACPI will be much improved, but it will still be nowhere near the functionality provided by WinXP.
thedward is quite correct in saying that the problems of Windows are not inherent in the niche it fills, but I think the problems are inherent in the way in which Windows came about filling that niche.
x86 architecture has been more or less the same, and public, for about the last twenty years. Win16/Win32, is only slightly younger. Each of them have to maintain at least a level of backwards compatibility with each release, or they get support calls from a couple of billion people - when I put out a new platform, I still have to test to make sure we didn't bust DOS. They also have to support a couple thousand individual component vendors who are all making products targeting those releases, and past releases. Linux doesn't bear this burden. Apple doesn't bear this burden. Windows is a behemoth. It dwarfs both Linux and OSX by orders of magnitude, if you measure lines of code. And yet, with each release, it does get better, and it does get more stable - there haven't been real stability problems since the first version of Win98. Both Linux, and OSX had the luxury of starting from scratch to try to support the features of a modern OS.
no subject
Date: 2003-10-25 01:49 am (UTC)Software is more problematic. I see plenty of posts along the lines of "I want to use version n of application x with distribution y - will it work?" The answer is usually "Maybe" (an even less helpful answer is "sure, if you compile it from source" - as though someone who needed to ask that question would know how to compile a program).
no subject
Date: 2003-10-25 01:59 am (UTC)It depends on how you define "stability problem". If stability problems only include the OS crashing of its own accord without your having to go near the keyboard, then I suppose Windows 98 First Edition was the last to do that (not counting servers, since they are doing crashworthy things all the time). If stability problems include anything that makes you have to reboot, then we stil have them with XP. Note the example I gave of MS Update killing my mother's computer (Win 98, Second Edition) - that wasn't just a reboot, that was a reformat.
To be fair, I've had to reboot Linux occasionally, though that's usually because I can't be bothered to find out what was causing the problem.
no subject
Date: 2003-10-25 05:17 pm (UTC)At any rate, I think we're going around in circles - if you want to talk stability and streamlined code - Linux wins, hands down. But, if you want to talk consumer-level functionality, and ease of use, I think Windows still wins, not to mention the pretty do-it-yourself support structure around Linux. My only real point was that Windows isn't the way it is through poor planning, poor development, or lack or trying, it's just a monster of a problem to solve, and they've done it piecemeal, over the last decade or so, whereas Linux could come along and benefit from everyone else's trial and error.