The browser is not suited for traditional applications in many realms. It can be leveraged into one - it's what I do all day at work - but it's just not suited for somethings. Complex GUIs are one of those things (drag & drop, for example, is a hassle to implement and looks crappy in a browser. Integeration with other applications is basically impossible).
Oh, and if you're going to write thin client apps like that, IE is a far superior platform to Mozilla (assuming that it's a closed environment and not a public use application, of course). HTML behaviors and extensions make obnoxious tricks like your hidden window totally unneccesary, and you can even expand the functionality of tags with compiled code if you want.
For what it's worth, the Microsoft documentation calls them "attributes", so he didn't just make it up. Selectors are described as such in the CSS2 specification.
Markup for a basic table layout is less than that used by the CSS in most cases. Feel free to prove otherwise.
CSS is great for applying styles. It's layout mechanism sucks ass.
Re:They still haven't fixed the a huge issue
on
Mozilla 1.4RC2 Released
·
· Score: 4, Informative
The other reason is that the markup is simpler, more portable, and less bandwidth intensive. How about that?
The Wired site loads and renders slowly, does wierd things when sized very small, and is much heavier on markup than slashdot (when balanced agasint the larger size of a slashdot page).
I agree that using tables for layout is a crappy way of doing things. On the other hand, it's well known and commonly supported (all modern browsers render tables more or less identically, the same cannot be said for CSS markup, especially level 2), but CSS layout semantics are crappy, overly verbose, and lend themselves to pixel-width positioning. Try reproducing all the built in features of table layout in CSS - it's very difficult. And your newly marked up pages will be noticably heavier than the table layout.
More of a case than you'd think. See if that page works in a screen reader or a braille converter. If it rejects non-IE user agents, odds are that it doesn't.
Does it really matter if we're "rewarding" them when "we" own them? Wouldn't even need 51%. Just get enough and start a shareholder suit against McBride.
You can't brute-force reverse an MD5 in any meaningful way. It's basic information theory. The MD5 checksum would, as mentioned, allow people to find matching code in the Linux kernel - and thats all.
b) Even if they did, the community they're accusing has a right to know what they're supposed to have done wrong. This is well codified in our legal system, by the way. SCOs public statements are on the verge of slander against Linus (and all kernel developers), and barratry in the threatning letters they've sent out.
It's entirely possible for someone to sue them to shut them up - in fact, it's already been done in Germany. However, it's harder to do that here (1st Amendement reasons) which is one reason it's not been done.
In any case, "challenging" them doesn't neccesarily mean "forcing them through legal means". On top of that, SCO is very mouthy in public, not simply in it's court filings. The people who it's accusing and harrasing have a right to respond.
I'm not sure why you think that the US legal system shouldn't be in the media anyway. I kinda prefer to know what goes on in courtrooms.
No, it's simpler than that. They just want all software development to be done in a corporate environment, where there's a dedicated legal team responsible for evaluating all patches before they're passed onto the project manager, who reviews them before passing them onto the project maintainer, etc.
It's a shot at the open source methodolgy, rather than at Linus per se. Although I'm sure they'll attack him directly as well.
Theoretically, a patent should have all the information you need to recreate the invention - that being the whole point, after all. Someone with reasonable skill in the field should be able to duplicate your invention with just the information in the patent.
Now, because the patent system has been horribly abused and twisted, and because capitalism is stupid about some things, the idea of the patent lawyer came into existence, and the MAIN PURPOSE of a patent lawyer is to write your patent in as obfuscated and broad a manner as possible. IE, to totally subvert the intent and goal of the patent system.
This is one of the reasons the patent system needs an overhaul.
Well, then, obviously your organization is going to have to trade the enhanced security offered by such a script for the convenience of 24/7 access from any IP. Other organizations, without the need for such broad access, can still use it.
The books are famously about nothing in particular. All the CG and action in the movies? Less than a tenth of the books. If that. The fast majority is things like long expositions on the history behind various races, and people talking about dinner. It's kinda like a tour guide. It's the sort of book you'd expect from someone who wore tweed all the time.
Which is not to detract from it's importance as a ground breaker and basis for most modern fantasy. But it's hardly the best fantasy around, either (for a given value of best, naturally).
For what it's worth, when I needed to install Gnome on my debian system, I just went to debian.org, found the first gnome app, and installed it. Since the app depended on gnome, and all the things gnome needed, it "just worked". Now, naturally, if you had a bad mirror or there was a bad package or something, that happens. I don't know alot about and can't speak to the technical merits of apt over RPM as far as the packages go. But auto-calculation of and then retrieval of dependencies is something that I pretty much expect. I was shocked to learn that RPM doesn't have it.
BTW, I've never had any problems with apt not updating old packages (except in a very few cases with broken packages, and even then never from official mirrors), but I do always do a apt-get update, apt-get dist-upgrade before I install anything new.
UNix != SMP scaling. Or any of the other technologies needed for scaling to very large servers. I believe that SCO's Unix scales to 32 processors (although I'm far too lazy to do research to back that up).
A previous poster posted a link to a horribly auto-translated page from a German (programmer? researcher?) who claims through a foulup that he was given access to the code without an NDA. The linux side is mainly cut & pasted posts to LKLM. The SCO side is bare code, without context. He says there's quite a bit of similarity but nothing very signfigiant except one function in the scheduler which appears to be identical (roughly 60 lines). There are areas that have the same comments but different source (perhaps this is common code that was replaced?)
As a final, very strange note, all dates were removed from both sets of sources, even from the comments. I find this very strange and suspicious.
Copy of parents link to the Google translated page:here
"real" routers do (see some of the higher up posts in this article re: cisco machines). There still needs to be an OS of some sort to handle things like configuration. The basic reason is that CPUs are cheap, while special purpose ASICs are not. It's not a full linux distro, mind, and it's a stripped down embedded kernel.
Because they DIDN'T work hard for all the rest of the stuff they developed - it's pretty simple. Try this one - "We spent lots of money developing our Windows application, why should we have to pay for Windows?"
It's the same thing - the cost in GPL software is compliance. If the cost of compliance is greater than the benefit of the code, don't use it. Just like you wouldn't use a proprietary library if you had to pay more for it than it would cost you to develop it yourself.
the nVidia driver (and, I believe, other binary drivers) use a GPL stubload that just forwards calls to the binary. The binary itself isn't linked to the kernel (I'm not 100% up on all the arcane linking issues, but my understanding is that this is common, and while people prefer source drivers, they accept that binary ones are better than nothing). I believe some distros ship with the nVidia drivers. Suse? Lindows?
I would disagree, as long as what you provide is what you use to develop - for example, I have GPLed code that uses extensions specific to VC++, and the codebase includes the project files (or nmake files) needed to compile the source.
The clause about the "preferred form" is obviously there to prevent people from using obfuscated source - I'd prefer something along the lines of "in the form you used to create it". What's the preferred form of a makefile? There's only about 5 thousand different versions. What's the preferred compiler? A strict reading could, as you state, prevent you from writing GPL code that couldn't be compiled with freely available tools (I'm sure alot of people would love that). As an interesting aside, I, as the copyright holder, don't have to provide makefiles, but anyone re-distributing my code would.... that's clearly an overly-onerous request.
You can even extend this argument farther - for example, many embedded devices use special compilers (as well as special, minimialist C runtimes, etc) in order to target/optimize for that environment - this interpetation of the GPL would mean you could't place code under the GPL unless a) you used common, ANSI C b) provided makefiles for all common compilers or c) were able to provide a compiler for the source. That's utterly ridiculous.
... and they're going to lose. This gets spread around enough, but it's just obviously and provably false. One - free distribution certainly does not revoke your copyright, period. Two - a judgment is not limited to the cost of the product, or even to actual damages suffered. And even if it was, losing the case doesn't mean you just have to pay money - they'd be barred from continuing the sell the product, as well, at least until they can get a valid license.
Slashdot Mirror
The border attribute in the table element is valid HTML 4.1.
Oh, and if you're going to write thin client apps like that, IE is a far superior platform to Mozilla (assuming that it's a closed environment and not a public use application, of course). HTML behaviors and extensions make obnoxious tricks like your hidden window totally unneccesary, and you can even expand the functionality of tags with compiled code if you want.
For what it's worth, the Microsoft documentation calls them "attributes", so he didn't just make it up. Selectors are described as such in the CSS2 specification.
CSS is great for applying styles. It's layout mechanism sucks ass.
The Wired site loads and renders slowly, does wierd things when sized very small, and is much heavier on markup than slashdot (when balanced agasint the larger size of a slashdot page).
I agree that using tables for layout is a crappy way of doing things. On the other hand, it's well known and commonly supported (all modern browsers render tables more or less identically, the same cannot be said for CSS markup, especially level 2), but CSS layout semantics are crappy, overly verbose, and lend themselves to pixel-width positioning. Try reproducing all the built in features of table layout in CSS - it's very difficult. And your newly marked up pages will be noticably heavier than the table layout.
People need to get over themselves with all this bothersome new markup. HTML (even old HTML) isn't going anywhere. Deal with it.
More of a case than you'd think. See if that page works in a screen reader or a braille converter. If it rejects non-IE user agents, odds are that it doesn't.
Does it really matter if we're "rewarding" them when "we" own them? Wouldn't even need 51%. Just get enough and start a shareholder suit against McBride.
You can't brute-force reverse an MD5 in any meaningful way. It's basic information theory. The MD5 checksum would, as mentioned, allow people to find matching code in the Linux kernel - and thats all.
b) Even if they did, the community they're accusing has a right to know what they're supposed to have done wrong. This is well codified in our legal system, by the way. SCOs public statements are on the verge of slander against Linus (and all kernel developers), and barratry in the threatning letters they've sent out.
It's entirely possible for someone to sue them to shut them up - in fact, it's already been done in Germany. However, it's harder to do that here (1st Amendement reasons) which is one reason it's not been done.
In any case, "challenging" them doesn't neccesarily mean "forcing them through legal means". On top of that, SCO is very mouthy in public, not simply in it's court filings. The people who it's accusing and harrasing have a right to respond.
I'm not sure why you think that the US legal system shouldn't be in the media anyway. I kinda prefer to know what goes on in courtrooms.
It's a shot at the open source methodolgy, rather than at Linus per se. Although I'm sure they'll attack him directly as well.
Now, because the patent system has been horribly abused and twisted, and because capitalism is stupid about some things, the idea of the patent lawyer came into existence, and the MAIN PURPOSE of a patent lawyer is to write your patent in as obfuscated and broad a manner as possible. IE, to totally subvert the intent and goal of the patent system.
This is one of the reasons the patent system needs an overhaul.
Well, then, obviously your organization is going to have to trade the enhanced security offered by such a script for the convenience of 24/7 access from any IP. Other organizations, without the need for such broad access, can still use it.
Which is not to detract from it's importance as a ground breaker and basis for most modern fantasy. But it's hardly the best fantasy around, either (for a given value of best, naturally).
BTW, I've never had any problems with apt not updating old packages (except in a very few cases with broken packages, and even then never from official mirrors), but I do always do a apt-get update, apt-get dist-upgrade before I install anything new.
UNix != SMP scaling. Or any of the other technologies needed for scaling to very large servers. I believe that SCO's Unix scales to 32 processors (although I'm far too lazy to do research to back that up).
Isn't that considered insider training?
As a final, very strange note, all dates were removed from both sets of sources, even from the comments. I find this very strange and suspicious.
Copy of parents link to the Google translated page:here
It was raised and resolved quite some time ago. I didn't pay attention to the details but remember the brief flurry of controversy.
"real" routers do (see some of the higher up posts in this article re: cisco machines). There still needs to be an OS of some sort to handle things like configuration. The basic reason is that CPUs are cheap, while special purpose ASICs are not. It's not a full linux distro, mind, and it's a stripped down embedded kernel.
It's the entire Linux kernel. I'd say thats pretty complicated :P
It's the same thing - the cost in GPL software is compliance. If the cost of compliance is greater than the benefit of the code, don't use it. Just like you wouldn't use a proprietary library if you had to pay more for it than it would cost you to develop it yourself.
the nVidia driver (and, I believe, other binary drivers) use a GPL stubload that just forwards calls to the binary. The binary itself isn't linked to the kernel (I'm not 100% up on all the arcane linking issues, but my understanding is that this is common, and while people prefer source drivers, they accept that binary ones are better than nothing). I believe some distros ship with the nVidia drivers. Suse? Lindows?
The clause about the "preferred form" is obviously there to prevent people from using obfuscated source - I'd prefer something along the lines of "in the form you used to create it". What's the preferred form of a makefile? There's only about 5 thousand different versions. What's the preferred compiler? A strict reading could, as you state, prevent you from writing GPL code that couldn't be compiled with freely available tools (I'm sure alot of people would love that). As an interesting aside, I, as the copyright holder, don't have to provide makefiles, but anyone re-distributing my code would.... that's clearly an overly-onerous request.
You can even extend this argument farther - for example, many embedded devices use special compilers (as well as special, minimialist C runtimes, etc) in order to target/optimize for that environment - this interpetation of the GPL would mean you could't place code under the GPL unless a) you used common, ANSI C b) provided makefiles for all common compilers or c) were able to provide a compiler for the source. That's utterly ridiculous.
... and they're going to lose. This gets spread around enough, but it's just obviously and provably false. One - free distribution certainly does not revoke your copyright, period. Two - a judgment is not limited to the cost of the product, or even to actual damages suffered. And even if it was, losing the case doesn't mean you just have to pay money - they'd be barred from continuing the sell the product, as well, at least until they can get a valid license.