Talking about this, French government, that added a tax on CDR and CDRW since February 2001, is considering some tax on hard drives now...
See vachealait (french link) for details.
But then, would it still be Linux as we know and like it ? Decentralized development does not have *only* drawbacks, choice is good, competition is good, cross influence can be good,... Oh well, it has been said thousand times already...
Is the instability caused by the increase of bandwidth usage, or by the use of random target IPs (that must certainly include unallocated IPs, seldom used IPs, and even some non routable IPs, not seen in "normal" traffic) ?
>So you don't think a large software project (like an operating system) should be designed before the coding phase? You're not in software business, are you.
If you really think that software should be designed only before coding phase, then I guess you learned software engineering in the '80s or something. Modern courses usually teach "cyclic (re)design" instead.
Sure, Linux started as a hobby project, but after nearly 10 years, most of today's code is far from its origins. Obviously, you cannot manage millions line projects without good design/coding rules enforcement (ok, some code still sucks, and drivers for example have very heterogenous quality).
And I think you're trolling, because it's obvious you never had a look at the kernel code, nor seen how kernel hackers work, nor have the slightest idea of who most of those hackers really are; Some are computer students, other work in computer field for years, and the average competence is really high; That's also competence that dictates who manages subsystems code in the kernel. "a loose bunch of hobbyists" is just FUD. (and no, I'm not a kernel hacker).
>I guess most unix programmers have uniprocessor systems and therefore don't get to see the benefits a thoroughly threaded application yields on a multiprocessor machine. I had a good laugh at that one. Salesman want you to believe that threads are necessary for multiprocessors. It's fine for them, as long as it sells, but it's not true. I could as well say that most unix boxen usually had many processes, so they didn't feel the need to go for threaded apps, which is a more complex model to master (=> error prone). So you can have your share of laugh too, if you feel like it.
But that's just different models, with their own advantages.
>You can really see the difference between a uni- and multiprocessor machine if you're running Windows NT, OS/2 or BeOS. Unlike with Linux, you can really see that you've got another processor running stuff in parallel. Again, unix boxen were seldom used for monolithic apps. This is changing, slowly. Also, it's easy to scale when you're very slow on uniprocessor systems:) Just do the math.
>Sad, but that's what you get when you use an OS that has not been properly designed (unforgivable sin in software engineering!) in the first place and has been coded by a loose bunch of hobbyists. A troll if I ever saw one, no need to answer.
From one of the given URL, you can read that TUX and Apache could (not yet done, but the API is there ?) be bolted together, so you could switch from standalone Apache to Apache + TUX setup, then slowly transfer some load from Apache to TUX for everything TUX does better...
BTW, I tried doing something (barely) similar with Apache and thttpd (on different ports). The problem I've found is that it doesn't work well because so many people are now behind firewalls that only allow outgoing HTTP connections to port 80 only (not tried IP aliases yet). "Bolting" looks promising (not to mention TUX seems to be really, really fast...)
I just had a look at your code, and you're using the same "technologies" I used at first when I started a VHS covers database in September. Since then I'm almost 10x faster, and I also much less memory (you can have a look at the result : http://www.moviecovers.com/). The two main changes were :
I switched from CGI interface to Embperl environment (I could have used plain mod_perl instead, but I've found Embperl, which runs on top of mod_perl, to be easier to program with). Keeping compiled Perl script in memory gives a big speed boost.
I switched from Berkeley BTree databases to MySQL. It's much faster, simplifies locking problems, and reduces memory usage by a good factor. Optimizing SQL requests can also be much more rewarding than many Perl optimizations.
Re:If OO is so great, why is bulk of GNU/Linux in
on
Object Oriented Perl
·
· Score: 1
Linux kernel is written in C, but that doesn't mean it doesn't use OOP, only that is doesn't use OOP *language*.
It uses the standard way of doing OOP in C, struct* fulls of function pointer, etc. VFS is an obvious example, but most drivers can also be seen as implementations of a virtual class, etc.
And using OOP that way (that is, where needed) surely helps maintainance.
>Well, as the article referenced by this thread suggests, some of "the fancy stuff" missing in MySQL includes real triggers and things that would actually speed up many MySQL web apps.
No, the MySQL docs state that the lack of things like transactions and triggers makes MySQL around 3 or 4 times faster, even when you don't use those features. So adding features is not always a win, and that's why you got alternatives when implementing databases:
Speed daemon, but lacking advanced features (MySQL);
[all intermediate solutions];
Feature rich, but slow for simple queries (Oracle for example).
From the user point of vue, choose the database that suit your needs.
And be sure of what you need, MySQL docs gives some real life examples where you can avoid using transactions by using auto_increment fields, conditional updates, etc.
>In the end, if you want to use package management, you really need to use it all of the time, or hassels will develop.
That is, either use precompiled packages (or source packages you compile yourself, not much difference), and learn to create packages out of not-packaged softwares... More easily said than done, but worth it I guess (you can then distribute your packages, install them on many boxes easily, etc.)
Stupid question, but does someone keep track of free software trademarks, so that they can be easily found and added whenever I mention one of them ? Is any GNU software trademarked, for example ?
The mathematical (topological) definition of the "diameter" of something, is the upper limit of the set of the distances between all pairs of elements taken from this something. Since we're talking about a discrete set (a directed graph), you can forget about "upper limit" and just say "maximum of".
So what they are measuring ("the average distance between two random pages") does *not* match the mathematical definition of a diameter, in spite of their claims.
I must admit I dunno what is the right term for what they are measuring, though.
Cache is an economic tradeoff... Memory used for caches is faster, yet more expensive than "normal" memory. On the other side, locality theorem says that you don't need a lot of cache to achieve a high hit rate (on average), and get a mean access time very close to the one of the cache memory alone.
If money is not a problem, you can design a computer without cache. That is, where the whole main memory is already made of the fastest memory available (done in some Cray computers, iirc).
A rule of thumb for a good hit rate is to get a cache size of 1/32 of the size of the main memory, so it is normal that cpus come with bigger caches since main memory sizes are expanding exponentially...
/. seems to me to offer decent coverage of other operating systems, but it in no has ever included as many articles on any operating system other than Linux as much as Linux (even Windows, and that rarely positively). Sure, but can it be blamed for that ? Linux started to get really 'hot' around that time, and talking about hot stuff is what a news site is all about...
It's never even really been much for intelligent posts about other operating systems A year ago, I was getting flamed on/. for posting pro-OS/2 comments, and topical/interesting OS/2 submissions passed over. OS/2 looks somewhat special, as I must admit I don't remember seeing lots of articles about OS/2 on Slashdot. I can't say for sure why (subject of the next poll ?:) ).
That said, there's still always been plenty of articles and comments about other operating systems; Mac OS X, BeOS, Amiga, QNX, OS/2, even Windows 2000. For short, lots of comments, but few clever ones ?:)
Perhaps our perceptions of the old times are different. . . (although I've been around longer, I think -- nyah nyah!:) (Some digging later...) The bookmark is dated April 8th, 1998, but I could find an access to Slashdot in my Squid access log dated from the January 27th, 1998... So it looks like it took me three months to get tired of typing "http://slashdot.org" every time;) Since the domain name has been registered the October 5th, 1997, those dates look right.
Slashdot Mirror
Must be the boomerang effect ? ;)
Noisy flash crap while I'm earing music on my computer, they lost a potential customer...
Talking about this, French government, that added a tax on CDR and CDRW since February 2001, is considering some tax on hard drives now...
See vachealait (french link) for details.
Falcon's Eye Nethack !
But then, would it still be Linux as we know and like it ? Decentralized development does not have *only* drawbacks, choice is good, competition is good, cross influence can be good,... Oh well, it has been said thousand times already...
Is the instability caused by the increase of bandwidth usage, or by the use of random target IPs (that must certainly include unallocated IPs, seldom used IPs, and even some non routable IPs, not seen in "normal" traffic) ?
Of course, I also took the time to fill their poll, to explain why I unsubscribed.
Also, look for MD5, Content-MD5 or ETags on the www.w3c.org, their silly patent doesn't fly for a second.
They also do lots of money out of patent-free informations, they teach them...
If you really think that software should be designed only before coding phase, then I guess you learned software engineering in the '80s or something. Modern courses usually teach "cyclic (re)design" instead.
Sure, Linux started as a hobby project, but after nearly 10 years, most of today's code is far from its origins. Obviously, you cannot manage millions line projects without good design/coding rules enforcement (ok, some code still sucks, and drivers for example have very heterogenous quality).
And I think you're trolling, because it's obvious you never had a look at the kernel code, nor seen how kernel hackers work, nor have the slightest idea of who most of those hackers really are; Some are computer students, other work in computer field for years, and the average competence is really high; That's also competence that dictates who manages subsystems code in the kernel. "a loose bunch of hobbyists" is just FUD. (and no, I'm not a kernel hacker).
latency ?
I had a good laugh at that one. Salesman want you to believe that threads are necessary for multiprocessors. It's fine for them, as long as it sells, but it's not true. I could as well say that most unix boxen usually had many processes, so they didn't feel the need to go for threaded apps, which is a more complex model to master (=> error prone). So you can have your share of laugh too, if you feel like it.
But that's just different models, with their own advantages.
>You can really see the difference between a uni- and multiprocessor machine if you're running Windows NT, OS/2 or BeOS. Unlike with Linux, you can really see that you've got another processor running stuff in parallel. :) Just do the math.
Again, unix boxen were seldom used for monolithic apps. This is changing, slowly. Also, it's easy to scale when you're very slow on uniprocessor systems
>Sad, but that's what you get when you use an OS that has not been properly designed (unforgivable sin in software engineering!) in the first place and has been coded by a loose bunch of hobbyists.
A troll if I ever saw one, no need to answer.
BTW, I tried doing something (barely) similar with Apache and thttpd (on different ports). The problem I've found is that it doesn't work well because so many people are now behind firewalls that only allow outgoing HTTP connections to port 80 only (not tried IP aliases yet). "Bolting" looks promising (not to mention TUX seems to be really, really fast...)
$ less /usr/src/linux/REPORTING-BUGS
It uses the standard way of doing OOP in C, struct* fulls of function pointer, etc. VFS is an obvious example, but most drivers can also be seen as implementations of a virtual class, etc.
And using OOP that way (that is, where needed) surely helps maintainance.
No, the MySQL docs state that the lack of things like transactions and triggers makes MySQL around 3 or 4 times faster, even when you don't use those features.
So adding features is not always a win, and that's why you got alternatives when implementing databases:
- Speed daemon, but lacking advanced features (MySQL);
- [all intermediate solutions];
- Feature rich, but slow for simple queries (Oracle for example).
From the user point of vue, choose the database that suit your needs.And be sure of what you need, MySQL docs gives some real life examples where you can avoid using transactions by using auto_increment fields, conditional updates, etc.
That is, either use precompiled packages (or source packages you compile yourself, not much difference), and learn to create packages out of not-packaged softwares...
More easily said than done, but worth it I guess (you can then distribute your packages, install them on many boxes easily, etc.)
Stupid question, but does someone keep track of free software trademarks, so that they can be easily found and added whenever I mention one of them ?
Is any GNU software trademarked, for example ?
By the way, did you ever noticed that Microsoft's website organization is pathetic at best ?
Geeks don't all use computers...
Since we're talking about a discrete set (a directed graph), you can forget about "upper limit" and just say "maximum of".
So what they are measuring ("the average distance between two random pages") does *not* match the mathematical definition of a diameter, in spite of their claims.
I must admit I dunno what is the right term for what they are measuring, though.
If money is not a problem, you can design a computer without cache. That is, where the whole main memory is already made of the fastest memory available (done in some Cray computers, iirc).
A rule of thumb for a good hit rate is to get a cache size of 1/32 of the size of the main memory, so it is normal that cpus come with bigger caches since main memory sizes are expanding exponentially...
Clever people buy UPS anyway.
What's the use of intelligence in soccer anyway :))
Sure, but can it be blamed for that ? Linux started to get really 'hot' around that time, and talking about hot stuff is what a news site is all about...
It's never even really been much for intelligent posts about other operating systems /. for posting pro-OS/2 comments, and topical/interesting OS/2 submissions passed over. :) ).
A year ago, I was getting flamed on
OS/2 looks somewhat special, as I must admit I don't remember seeing lots of articles about OS/2 on Slashdot. I can't say for sure why (subject of the next poll ?
That said, there's still always been plenty of articles and comments about other operating systems; Mac OS X, BeOS, Amiga, QNX, OS/2, even Windows 2000. :)
For short, lots of comments, but few clever ones ?
Perhaps our perceptions of the old times are different. . . (although I've been around longer, I think -- nyah nyah! :) ;)
(Some digging later...) The bookmark is dated April 8th, 1998, but I could find an access to Slashdot in my Squid access log dated from the January 27th, 1998... So it looks like it took me three months to get tired of typing "http://slashdot.org" every time
Since the domain name has been registered the October 5th, 1997, those dates look right.