Re:Windows forces costly upgrades for home users.
on
Making Linux Beautiful
·
· Score: 2
most users upgrade their harder faster than [...] their OS
If we're talking about people who just want word processing and web access, then there are a lot of people who don't need new hardware. Gamers need new hardware; you mum shouldn't.
libc has gone through two major changes
No, glibc 2.0 was not meant to be a stable release. Some distributions heeded this; some didn't. Most propriatory binaries are still made available for libc5 even today. (netscape, rvplayer, etc)
If a printer works with Linux, it'll also work with Windows 3.1
This isn't true; the HP Deskjet 710 C is an example. (You can use a driver for an older deskjet, but you don't get the new features).
Completeness Theorem != Incompleteness Theorem
on
SSH v. SRP
·
· Score: 2
> I hope you have tongue firmly in cheek [...] > See Godel's Incompleteness Theorem
Ah, these are two different results. Godel's Completeness Theorem says that any *first order* statement which is true can be proved. "First order statements" are, essentially, mathematical statements which only talk about certain things. Amongst other things, first order statements can't talk about "For all sets such that blah".
On the other hand, there are statements which do include things like "for all sets such that blah " - called second order statements. Godel's *Incompleteness* Theorem says that there are some second order statements which are true but cannot be proven.
My comment was that FLT fits into the first category, so an automated proof is (theoretically) possible. Of course, it would take ages to run!
Sometimes I think windows [...] is the perfect tool for people who just want to write a letter and browse the web.
I beg to differ. Windows forces people to upgrade to the latest software, which costs a fortune for a home user. Your mum is happy with Word version 6 on Win 3.1? Oh yes, but her printer just broke. All those printers they're selling at PC World only come with drivers for the latest version of windows, so she'll have to buy Win 95. Someone sent her a Word 97 attachment? Sorry, she can't read that without upgrading. She needs to look at a website that uses HTML 4, CSS, etc? Sorry, she'll have to upgrade windows to be able to run a modern web browser. You can bet that the same thing will happen, in a couple of years, to people who are happy with Win 98 software.
In the free software world, change is more incremental. 1994 versions of Linux can run the latest software. Utilities in the operating system allow access to the latest file formats. And above all, if your mum did need a new version of Linux, perhaps to use that new printer she just bought, she could get it for free.
One OS is a ticking timebomb for the end user. The other will remain viable for years.
This is not correct. It is a statement about the integers which can be formalised in a first order language, so [by Godel's completeness theorem] there is either a proof that it is true, or a proof that it is false. You could write a program which exhaustively searches all possible proofs (and disproofs). Eventually the program must terminate.
He only has to open the source to anyone who can prove they downloaded binaries from him. He can stop distributing completely, and keep the source private.
> he'll put it on a 56k uplink on a windows box
That's not too bad - only one person has to download the source like this, then they can put it somewhere sensible (e.g. GeoCities) for everyone else.
Since most communications aren't encrypted, including a surprising amount of sensitive stuff, tcpdump | grep probably gets more data than any subsequent analysis. After that it'll be diminishing returns on computing time.
It's true that you can concoct a polynomial which will spew out any finite sequence of numbers, and that this means you can't *guarantee* you've got the formula for a sequence just by checking finitely many terms.
However, if you find a formula that "seems right", it may make it easier to prove that it *is* right, because now you're barking up the right tree.
An example of this happening in real life is the number 196884. It turned up in two seemingly unrelated places, in the character table of the Monster Group and in the expansion of the j function. This lead mathematicians to search for - and find - the connection between the two.
See Scientific American for a good article about this "Moonshine Conjecture".
cx is jointly shortest with the other 237 two letter domain names which represent each country in the world. The two letter abbreviations are defined by ISO 3166. The only exception I know of is that the UK has ".uk" instead of ".gb".
The capitalism of today is broken: see Gnome!
on
Gnome 1.1.4 Released
·
· Score: 2
What's up with all the people calling Microsoft fascist? [...]Any fair way for them to make money is fine. Their tactic of "embracing" and extending is a beautiful business plan.
The problem with capitalism has always been that it's bad at preventing monopolies. One reason capitalism has been the "least worst" system is because other systems are even worse at preventing monopolies. However, for the software industry, "bad" should be changed to "appalling". Because the marginal costs of software are zero, and pieces of software have to interoperate, the "networking effect" becomes extremely powerful and practically forces each part of the industry to be dominated by a single standard. And here's the crunch: with today's intellectual property laws, the dominating standard is often controlled by a single entity. This means that the industry gets dominated by an oligopoly of strong players, and it's virtually impossible for newcomers to break in, unless the oligopoly makes a serious mistake. This means that the free market cannot work its magic in the software industry. A product will win if it is from a big company which controls standards, even if there's a better product somewhere else. A single company has a grip on the standard, and nobody else can challenge that standard because of compatibility issues. A bunch of (mainly) volunteers using copyleft licenses are capable of challenging the rest of the software industry, because it is fragmented, inefficient and anti-competitive. The fact that it's possible for things like GNU, Linux and Gnome to capture market share should show this quite clearly. Until IP law is reformed, the mainstream software industry will remain as abysmal as it is today.
> "RedHat and Debian" - Hah! Redhat may be the current > marketing winner, but no one has any clear title > to number two.
There are at least three distributions which are based upon Debian. One of them, Corel Linux, has been aimed squarely at the desktop market, and has the marketing machinery of a fairly large software company behind it. Whilst I agree that there's no clear "number two" distribution, Debian is shaping up to be the biggest "metadistribution".
> The BSDs' scanf conforms to... ANSI C... > so it's not their fault if some Linux software > won't compile or run because of GNU weakening > standards.
I stand corrected about/dev on BSD. My post wasn't supposed to say whose fault the incompatibilities were - I was just describing them to the best of my knowledge!
> GNU does the same as Microsoft when they add their > own "features" to an official standard. I don't accept this. For one thing, GNU extensions are very well documented in the libc info page. The words "this is a GNU extension" are everywhere. The GNU sed info page is a good example; it tells you the maximum width of lines according to POSIX, and also the maximum width which various systems will use. Compare this to the MS J++ manual which doesn't even make it clear that J++ is not Java. Remember how the ANSI standard came about; the stuff in it was originally part of people's extensions to K&R C. Do you wish we'd sticked with K&R C? Personally I have no objections to people extending a standard per se. The thing that bothers me is when they don't make it clear that their library/browser/whatever is more forgiving than the standard allows, thus encouraging people to write unportable code. glibc can hardly be accused of such sneaky extensions - the libc info page is very clear about what is part of the standard and what isn't.
OpenBSD and NetBSD are almost exactly the same OS, but configured differently. They are more similar to FreeBSD than any of the three are to Linux.
BSD/dev is quite different to Linux. BSD doesn't have/proc.
The biggest cause of differences is differing versions of the standard C libraries. All GNU/Linuxes, and GNU/HURD, use glibc. BSD has its own libc. This means that a fair few standard functions work quite differently. In particular, GNU extensions, like the %a flag for scanf, will be missing from BSD.
Porting between unixes is often a matter of spending a bit of time tweaking bits of code here and there. Porting to/from something else, like Mac, Windows or VMS, is usually a major task and probably requires a complete rewrite. But of course this depends upon the program. hello-world.c will work anywhere.
How does BSD "steal from linux"? Anyone selling non-free BSD software will probably port it to linux, because the linux market is so much bigger. On the other hand, any free software which is written for BSD can be adapted to Linux.
Free unices are (pretty much) source-compatible. By expanding the free unix market, BSD attracts more free software development, which *helps* linux.
> I *want* air traffic controllers to speak in a > highly structured, syntactically and lexically > limited, and stereotypical fashion!
While I would agree with this statement about air traffic control, I don't think it holds in other "precision" areas. A mathematician can grapple with concepts which seem hopelessly impenetrable, because of the flexibility that mathematical notation gives him to express his problem in a way which just features the important points, and not the irrelevant information. A good programmer can tackle a complex problem in an elegant and maintainable way if his programming language gives him the same flexibility.
Of course, many problems are not very complex and/or are being tackled by mediocre programmers. In this case, a language with inflexible notation is more appropriate because using esoteric syntax won't lead to a more elegant or more maintainable program.
>> a dual p3 with MySQL outperforms a 30 cpu sun E6500 >> with oracle by over 500 times, with more features >> and better reliability
This statement is completely and wholly untrue. It's like saying a skateboard can go faster than a train, there's no matter of opinion. The poster wasn't saying MySQL is a better choice for some situations, he was saying that it is several thousand times faster than Oracle. The rest of the comments were equally crazy. Perhaps the original poster was being sarcastic, or perhaps he was just trolling. But he sure as heck wasn't being sincere.
Well, ok, but if you manually keep track of dependencies it works. Not as good as automatic dependency tracking, but better than having to fish out 00s of individual files.
> When you go to download some Windows software you get a single.exe to download and install. [...] > For linux software we already get to choose from half a dozen different packages
This is a non-issue for the end user, because nearly all popular [freely-distributable] software for linux is available on your distro's CDs / ftp site. The user doesn't need to worry about the format, because the distro handles it cleanly.
Of course, things aren't that simple if you want software which isn't freely-redistributable. But AFAICS there's no way to clear this up without abandoning shared files altogether, or risking the kind of corrupted mess which is possible with Windows packages.
Slashdot Mirror
If we're talking about people who just want word processing and web access, then there are a lot of people who don't need new hardware. Gamers need new hardware; you mum shouldn't.
No, glibc 2.0 was not meant to be a stable release. Some distributions heeded this; some didn't. Most propriatory binaries are still made available for libc5 even today. (netscape, rvplayer, etc)
This isn't true; the HP Deskjet 710 C is an example. (You can use a driver for an older deskjet, but you don't get the new features).
> I hope you have tongue firmly in cheek [...]
> See Godel's Incompleteness Theorem
Ah, these are two different results. Godel's Completeness Theorem says that any *first order* statement which is true can be proved. "First order statements" are, essentially, mathematical statements which only talk about certain things. Amongst other things, first order statements can't talk about "For all sets such that blah".
On the other hand, there are statements which do include things like "for all sets such that blah " - called second order statements. Godel's *Incompleteness* Theorem says that there are some second order statements which are true but cannot be proven.
My comment was that FLT fits into the first category, so an automated proof is (theoretically) possible. Of course, it would take ages to run!
In the free software world, change is more incremental. 1994 versions of Linux can run the latest software. Utilities in the operating system allow access to the latest file formats. And above all, if your mum did need a new version of Linux, perhaps to use that new printer she just bought, she could get it for free.
One OS is a ticking timebomb for the end user. The other will remain viable for years.
>Fermat's Last Theorem is unprovable.
This is not correct. It is a statement about the integers which can be formalised in a first order language, so [by Godel's completeness theorem] there is either a proof that it is true, or a proof that it is false. You could write a program which exhaustively searches all possible proofs (and disproofs). Eventually the program must terminate.
> he'll be forced to open up the source
He only has to open the source to anyone who can prove they downloaded binaries from him. He can stop distributing completely, and keep the source private.
> he'll put it on a 56k uplink on a windows box
That's not too bad - only one person has to download the source like this, then they can put it somewhere sensible (e.g. GeoCities) for everyone else.
Since most communications aren't encrypted, including a surprising amount of sensitive stuff, tcpdump | grep probably gets more data than any subsequent analysis. After that it'll be diminishing returns on computing time.
These are the first few perfect numbers, i.e. numbers whose proper factors (including 1) sum to the original number.
It's true that you can concoct a polynomial which will spew out any finite sequence of numbers, and that this means you can't *guarantee* you've got the formula for a sequence just by checking finitely many terms.
However, if you find a formula that "seems right", it may make it easier to prove that it *is* right, because now you're barking up the right tree.
An example of this happening in real life is the number 196884. It turned up in two seemingly unrelated places, in the character table of the Monster Group and in the expansion of the j function. This lead mathematicians to search for - and find - the connection between the two.
See Scientific American for a good article about this "Moonshine Conjecture".
cx is jointly shortest with the other 237 two letter domain names which represent each country in the world.
The two letter abbreviations are defined by ISO 3166. The only exception I know of is that the UK has ".uk" instead of ".gb".
> "RedHat and Debian" - Hah! Redhat may be the current
> marketing winner, but no one has any clear title
> to number two.
There are at least three distributions which are based upon Debian. One of them, Corel Linux, has been aimed squarely at the desktop market, and has the marketing machinery of a fairly large software company behind it. Whilst I agree that there's no clear "number two" distribution, Debian is shaping up to be the biggest "metadistribution".
> The BSDs' scanf conforms to ... ANSI C ...
/dev on BSD. My post wasn't supposed to say whose fault the incompatibilities were - I was just describing them to the best of my knowledge!
> so it's not their fault if some Linux software
> won't compile or run because of GNU weakening
> standards.
I stand corrected about
> GNU does the same as Microsoft when they add their
> own "features" to an official standard.
I don't accept this. For one thing, GNU extensions are very well documented in the libc info page. The words "this is a GNU extension" are everywhere. The GNU sed info page is a good example; it tells you the maximum width of lines according to POSIX, and also the maximum width which various systems will use. Compare this to the MS J++ manual which doesn't even make it clear that J++ is not Java. Remember how the ANSI standard came about; the stuff in it was originally part of people's extensions to K&R C. Do you wish we'd sticked with K&R C? Personally I have no objections to people extending a standard per se. The thing that bothers me is when they don't make it clear that their library/browser/whatever is more forgiving than the standard allows, thus encouraging people to write unportable code. glibc can hardly be accused of such sneaky extensions - the libc info page is very clear about what is part of the standard and what isn't.
Doesn't seem too difficult to me. Or do you mean how Tux ended up being the Linux mascot? In which case I'm not sure anyone really understands why.
Wow, what an unusual reason to change OS. Which BSD are you using? How do users of other BSDs react? D'ya still use GCC?
I forgot to say: I've never done any nontrivial BSDLinux porting, so this is all stuff I've picked up second hand.
OpenBSD and NetBSD are almost exactly the same OS, but configured differently.
/dev is quite different to Linux. BSD doesn't have /proc.
They are more similar to FreeBSD than any of the three are to Linux.
BSD
The biggest cause of differences is differing versions of the standard C libraries. All GNU/Linuxes, and GNU/HURD, use glibc. BSD has its own libc. This means that a fair few standard functions work quite differently. In particular, GNU extensions, like the %a flag for scanf, will be missing from BSD.
Porting between unixes is often a matter of spending a bit of time tweaking bits of code here and there. Porting to/from something else, like Mac, Windows or VMS, is usually a major task and probably requires a complete rewrite. But of course this depends upon the program. hello-world.c will work anywhere.
How does BSD "steal from linux"? Anyone selling non-free BSD software will probably port it to linux, because the linux market is so much bigger. On the other hand, any free software which is written for BSD can be adapted to Linux.
Free unices are (pretty much) source-compatible. By expanding the free unix market, BSD attracts more free software development, which *helps* linux.
Who's Arkady? It rings a bell with the name "Darrell" but I'm not sure why!
> I *want* air traffic controllers to speak in a
> highly structured, syntactically and lexically
> limited, and stereotypical fashion!
While I would agree with this statement about air traffic control, I don't think it holds in other "precision" areas. A mathematician can grapple with concepts which seem hopelessly impenetrable, because of the flexibility that mathematical notation gives him to express his problem in a way which just features the important points, and not the irrelevant information. A good programmer can tackle a complex problem in an elegant and maintainable way if his programming language gives him the same flexibility.
Of course, many problems are not very complex and/or are being tackled by mediocre programmers. In this case, a language with inflexible notation is more appropriate because using esoteric syntax won't lead to a more elegant or more maintainable program.
>> a dual p3 with MySQL outperforms a 30 cpu sun E6500
>> with oracle by over 500 times, with more features
>> and better reliability
This statement is completely and wholly untrue. It's like saying a skateboard can go faster than a train, there's no matter of opinion. The poster wasn't saying MySQL is a better choice for some situations, he was saying that it is several thousand times faster than Oracle. The rest of the comments were equally crazy. Perhaps the original poster was being sarcastic, or perhaps he was just trolling. But he sure as heck wasn't being sincere.
Well, ok, but if you manually keep track of dependencies it works. Not as good as automatic dependency tracking, but better than having to fish out 00s of individual files.
The song is mocking the parents in the film, who tenuously blame Canada for the fact that their children swear.
I think this does what you want. (It temporarily deconfigures packages which depend upon things replaced by foo, then reconfigures them)
I think the original poster meant "to (handle dependencies like Debian does)" not "(to handle dependencies) like Debian does".
But a package manager that didn't handle dependencies would still be useful, to do clean uninstalls.
> When you go to download some Windows software you get a single .exe to download and install.
[...]
> For linux software we already get to choose from half a dozen different packages
This is a non-issue for the end user, because nearly all popular [freely-distributable] software for linux is available on your distro's CDs / ftp site. The user doesn't need to worry about the format, because the distro handles it cleanly.
Of course, things aren't that simple if you want software which isn't freely-redistributable. But AFAICS there's no way to clear this up without abandoning shared files altogether, or risking the kind of corrupted mess which is possible with Windows packages.