(by a mechanism as yet unexplained... I'd expect a time travelling species to be pretty much indestructible under normal circumstances).
Daleks can also travel in time. Hence it being a 'Time War'. Imagine the ending of the second Bill and Ted film - "We set up the key, and we set up the gun" - but with facistic death crazed pepper pots.
- Aidan (Staaaaatiiooon)
Carpet. Enterprise. Red.
Rearrange into an ALL POWERFUL SOFTWARE MANAGEMENT SOLUTION and rebrand to ZENworks.
I like it anyway, but I would say that wouldn't I, being a total shill.
That's just not true, apple incorporates gcc, emacs and other bits of the GNU system into OS X.
If you read the GPL, it states:
In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.
So it's perfectly safe to do this, and FreeBSD, OpenBSD and NetBSD distribute gcc as part of their base systems as well as things like cvs, awk and grep
The cisco 7940 IP phone sitting on my desk runs off the power from the ethernet cable plugged into it (it's the only thing plugged into it), so it must carry some current, whether that's enough to power anything more useful than a phone, I don't know, but usb hdds and scanners generally come with an external power supply anyways.
What about if the email is on the company's or university's Unix box, where you ssh to run pine
It doesn't really matter then if you encrypt the files or not, if they really give a toss about your mail then there's many ways of reading it (eg. attaching gdb to your program, having a second xterm pop up whenever you login that shows them exactly what you're doing). Ultimately if you don't trust your admins you're fucked.
Welcome to monopoly abuse. They figure that they have such inertia that they no longer have to be nice to developers. It's an interesting theory. What comes first the chicken or the egg = Do IT departments make decisions based on a suporvisory dictator ship, or do they consult their programers on what they would be most productive at.
There isn't anything suporvisory about the dictatorships in most IT departments....
The versions of GNAT that are shipped with Red Hat 7.0 don't work with Red Hat's GCC snapshot[*]. And since you can't recompile GNAT from sources without having a version of GNAT installed to bootstrap itself, that means GNAT is fundamentally and profoundly broken.
GNAT 3.12p is based on gcc 2.8.1, which is included in gnat-devel as gnatgcc.
I think this is unlikely, sending executables in mail can be handy ¥it's not the "right" way to do it, but it's sometimes easier than sticking it up on an ftp machine
Have you read the typical license agreements that most commerical software companies put out?
Yes, but when you buy a suppourt contract (which is really RedHat's product, not the distribution), you're basically buying an arse to kick.
You can buy suppourt contracts from M$, but they're not great (and terribly expensive).
The software you buy as an end user, and the software you buy as a corporation are the same. It's the level of suppourt (and liability) that the vendor gives you that's different.
Why would they need different interfaces, your still doing the same thing, just doing it in a different way.
It's called "modularity", and it's a Very Good Thing.
If introducing modularity requires re-writing code
that uses the module that code is Broken to start
with and should be fixed.
All the memory management stuff should be done in one place, and only exposed via necessary interfaces (such as malloc()), and kernel-internals should use those interaces instead of talking directly to the mm code.
Assuming, of course, well designed interfaces to start with.
here was a bit of
foreshadowed warning in that those GNU tools, and indeed the whole of Linux, were under Richard Stallman's GNU General Public Liscence,
and thus subject to his whim and the whim of the Free Software Foundation, but this was ignored as RMS was a freedom fighter, and it was well
known he would never take Linux away.
Umm, remember that software under the GPL almost always includes the phrase ("Licesnsed under the terms of the GPL, version 2"). He can't chnage the GPL to read "You must donate 10 dollars to the FSF whenever you distribute this program" and have it affect any code that he doesn't hold the copyright for.
If you don't like the GPL, don't use it. But nobody can take your copyright away from you. AT&T owned the copyright for UNIX, RMS doesn't own the copyright the linux kernel.
> Whatever packing is used, it needs to be kept simple. RPM isn't. I don't know the inner details of.DEB yet
Makes RPM look simple, that's what. It requires a whole *directory* full of miscellaneous spec type files, and about 4 different programs to build.
That's not true. RPM needs one spec file and one commaned (rpm -ba foo.spec) to build a package.
What next, should Redhat and the other successful companies run around buying closed source companies and release their source, maybe they should do this immediately before their market price crashes:-). Even if they go belly up afterwards they still will have achieve opening up the source. A cunning idea methinks
I was really hoping that when RH aquired cygnus, it would lead to cygnus being fully opened up. Sadly this hasn't happened (yet). dammnit...
You are also responsible to any third party to distribute source on request if you don't distribute it along with the binary.
I don't believe that this is true. From the GPL:
3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)
So if you wanted, under 3.a you could just bundle the object code and the source code onto a cd or two and give it to a customer. There, that's it, you're done. You don't have to give the source code to j-random user who asks for it...
From the article: The problem is that if Perl 6 were written in Ada--it would require people to bootstrap GNAT before they could even get to Perl. That's too much of a burden to put on anybody.
Basically Ada '95 didn't get used because a lot of people don't have it. Chicken, egg, egg, chicken.
But that may also be suffering from lack of current maintainers.
Errm? 3.11p was released in January, and 3.12p is suppoused to be out RSN.
AdaCore Technologies is the normal maintainer, but they were sounding very commercial when I talked to them a year ago.
Hmmm. I'm a member of the Ada-Linux team and we've got a very good relationship with ACT and ACT-Europe. They are a company, but they do do a lot to help both the Ada and GNU communities (there was an ALT/ACT meeting in Paris in June), both officially and unofficially (the Ada-mode for emacs is maintained by them, and several members of ACT-Europe helped create GtkAda)
Of course the problem of where to get his programmers couldn't have anything to do with his choice.:-)
Course, comp.lang.ada wouldn't be a good place to start would it?
- Aidan (dislikes disinformation, but is horrendously behind on/.)
Sorry, dont know what came over me there... anyway:
The problem I have noticed with CS graduates who have learnt high level langauges to the exclusion of all else, is that they can't then grasp `real world' programming concepts. This may only be a problem afflicting British CS grads, but it is
Surely every CS grad will have had to do some assembler at some point? You do it in 1st year here (University of Glasgow, officially best CS dept in the UK), and 2nd year....
C is a shit language, that requires you to do too much at a low level, in an unsafe manner, that could be better abstracted away.
There's no inherent difference between a compiled language and an interpreted language (and where to just-in-time compiled languages like Perl fit into this world view?), there are C interpreters and Basic compilers.
I've got an obvious choice for introducing people to programming, but it's an obvious choice for doing most programming in and i don't want to get into a language flame war.
Buffer Overflows are a result of a lack of bounds checking. This is a logic error. Logic errors are the one hardest error to detect in programming. The reason there are so many buffer overflows are because when you program, you dont
Buffer overflows could be avoided by using a language which has bounds checking built in.
movement is doing a lot in this direction. Cryptography is on top of the list. Free, easy to use, public domain cryptographic tools are a necessity. And with a few targeted public research grants they could become a reality rather sooner than later. An other
The Gnu Privacy Guard already provides freely available, easy to use public key cryptography. It's extremely simple to integrate it as a filter in eg. Pine or your favourite mailer. Version 1.0 is due out RSN, and 0.9.11 was released today.
Just a short response to LL's point... It seems you are criticising RedHat for simply living up to the GNU Manifesto (not that RH actually knows it is doing that). That is, software is not a commodity to sell, it is
I think that RedHat, out of all the main commercial distributions (not the ones based off of it, they probably get it as well, but the others) best understands the GNU Manifesto. Debian obviously does, but it's a slightly different beast. personally I use RedHat, because it was the closest morally acceptable one when i first installed and haven't had any reason to change to Debian. I do intend to run at least one box off of Debian though at some point in the future.
Slashdot Mirror
(by a mechanism as yet unexplained... I'd expect a time travelling species to be pretty much indestructible under normal circumstances). Daleks can also travel in time. Hence it being a 'Time War'. Imagine the ending of the second Bill and Ted film - "We set up the key, and we set up the gun" - but with facistic death crazed pepper pots. - Aidan (Staaaaatiiooon)
Carpet. Enterprise. Red. Rearrange into an ALL POWERFUL SOFTWARE MANAGEMENT SOLUTION and rebrand to ZENworks. I like it anyway, but I would say that wouldn't I, being a total shill.
Apple doesn't incorporate *anything* GPL'ed
That's just not true, apple incorporates gcc, emacs and other bits of the GNU system into OS X.
If you read the GPL, it states:
So it's perfectly safe to do this, and FreeBSD, OpenBSD and NetBSD distribute gcc as part of their base systems as well as things like cvs, awk and grep
- Aidan (typing on a FreeBSD machine)
Firewire can carry power, but no kind of Ethernet
The cisco 7940 IP phone sitting on my desk runs off the power from the ethernet cable plugged into it (it's the only thing plugged into it), so it must carry some current, whether that's enough to power anything more useful than a phone, I don't know, but usb hdds and scanners generally come with an external power supply anyways.
What about if the email is on the company's or university's Unix box, where you ssh to run pine
It doesn't really matter then if you encrypt the files or not, if they really give a toss about your mail then there's many ways of reading it (eg. attaching gdb to your program, having a second xterm pop up whenever you login that shows them exactly what you're doing). Ultimately if you don't trust your admins you're fucked.
Welcome to monopoly abuse. They figure that they have such inertia that they no longer have to be nice to developers. It's an interesting theory. What comes first the chicken or the egg = Do IT departments make decisions based on a suporvisory dictator ship, or do they consult their programers on what they would be most productive at.
There isn't anything suporvisory about the dictatorships in most IT departments....
- Aidan
The versions of GNAT that are shipped with Red Hat 7.0 don't work with Red Hat's GCC snapshot[*]. And since you can't recompile GNAT from sources without having a version of GNAT installed to bootstrap itself, that means GNAT is fundamentally and profoundly broken.
GNAT 3.12p is based on gcc 2.8.1, which is included in gnat-devel as gnatgcc.
checkout Ada for Linux for more info.
Reading the WP's story how the crackers got initial access, I wondered if this action will end the possibility to sent executables with email messages using MS software ¥as in: they'll patch the tools to get rid of this feature, as they should have done ages ago© I mean: the way the crackers got access wouldn't have possible with the lack of a way to send a person an executable by email ¥as a trojan©
I think this is unlikely, sending executables in mail can be handy ¥it's not the "right" way to do it, but it's sometimes easier than sticking it up on an ftp machine
From what I read, it wasn't auto-executed, but was deliberately run by the reciever, which means somebody's due for the kicking of their lives©©©
- Aidan
In general, there has been a move away from Usenet and towards other fertile discussion forums within the last four years© I expect this trend to continue well into the next five years© Today, Usenet is nothing like what it was ten years ago© It'll be even less so, tomorrow©
People keep predicting that usenet will die© People have been saying that usenet's been in decline for the last five years for at least as long as I've been on it ¥early 95©
Usenet won't die because it's easy to use ¥news readers are a very mature technology, it's informative and there's a real sense of community there© No matter what happens in terms of technological advances, you'll never get those communities shifting en-masse to somewhere else©
And ultimately that's what usenet is about, communities© I met my first girlfriend, my wife, a good portion of my close friends from one newsgroup© I learned Everything I Need To Know about programming from some uber-intelligent people on another©
I was off usenet for while recently, 6 months© I missed it terribly© I'd been checking usenet probably 355+ days a year© I missed the sense of community©
Until you kill that, until you remove all my friends internet connections, take away their newsreaders and burn the servers, news will survive© It's the most succesful form of online community, it always has been and it always will be© When IRC is ancient history, when ICQ, AIM etc no longer exist, I'll still be checking news©
- Aidan
GNU is just the compiler.
No, GNU is also the ls, the tar, the gzip, the /bin/sh... infact, most of you're operating environment.
- Aidan
Have you read the typical license agreements that most commerical software companies put out?
Yes, but when you buy a suppourt contract (which is really RedHat's product, not the distribution), you're basically buying an arse to kick.
You can buy suppourt contracts from M$, but they're not great (and terribly expensive).
The software you buy as an end user, and the software you buy as a corporation are the same. It's the level of suppourt (and liability) that the vendor gives you that's different.
Why would they need different interfaces, your still doing the same thing, just doing it in a different way.
It's called "modularity", and it's a Very Good Thing.
If introducing modularity requires re-writing code that uses the module that code is Broken to start with and should be fixed.
All the memory management stuff should be done in one place, and only exposed via necessary interfaces (such as malloc()), and kernel-internals should use those interaces instead of talking directly to the mm code.
Assuming, of course, well designed interfaces to start with.
- Aidan
here was a bit of foreshadowed warning in that those GNU tools, and indeed the whole of Linux, were under Richard Stallman's GNU General Public Liscence, and thus subject to his whim and the whim of the Free Software Foundation, but this was ignored as RMS was a freedom fighter, and it was well known he would never take Linux away.
Umm, remember that software under the GPL almost always includes the phrase ("Licesnsed under the terms of the GPL, version 2"). He can't chnage the GPL to read "You must donate 10 dollars to the FSF whenever you distribute this program" and have it affect any code that he doesn't hold the copyright for.
If you don't like the GPL, don't use it. But nobody can take your copyright away from you. AT&T owned the copyright for UNIX, RMS doesn't own the copyright the linux kernel.
- Aidan (damn FUD)
> Whatever packing is used, it needs to be kept simple. RPM isn't. I don't know the inner details of .DEB yet
Makes RPM look simple, that's what. It requires a whole *directory* full of miscellaneous spec type files, and about 4 different programs to build.
That's not true. RPM needs one spec file and one commaned (rpm -ba foo.spec) to build a package.
- Aidan
Any program which grabs a network socket and accepts connections from the outside world represents a potential threat from buffer overflows.
Any program written in C like languages may be vulnerable to buffer overflows.
Programs written in other languages (which don't let you do that kind of shit) won't.
Buffer overflows should have been eliminated in 1980, along with GoTo and C itself[1]
Aidan
[1] If you want assembly, you know where to get it. If you want a programming language, try here
What next, should Redhat and the other successful companies run around buying closed source companies and release their source, maybe they should do this immediately before their market price crashes :-). Even if they go belly up afterwards they still will have achieve opening up the source. A cunning idea methinks
I was really hoping that when RH aquired cygnus, it would lead to cygnus being fully opened up. Sadly this hasn't happened (yet). dammnit...
- Aidan
The bill has not yet passed, and is not yet made law. It is, as yet, still legal to store encrypted data on our computers.
IANAL, but I don't think that this law makes it illegal, just that it means you have to hand over the key when asked...
the obvious thing to do is not to be asked... >;)
- Aidan
You are also responsible to any third party to distribute source on request if you don't distribute it along with the binary.
I don't believe that this is true. From the GPL:
So if you wanted, under 3.a you could just bundle the object code and the source code onto a cd or two and give it to a customer. There, that's it, you're done. You don't have to give the source code to j-random user who asks for it...
IANAL, etc, etc, etc.
- Aidan
Better than (for that matter) Gnome/CORBA (which is Linux only currently).
?
CORBA runs on many different architectures, both unixlike and non-unixlike. Gnome runs on most unicen IIRC....
- Aidan
the entire open government campaign are run on five Dell 2300 Dual Pentium II 450 machines, each with 512k of RAM and 27 gigabytes of hard disk space.
512k! 512k! That's *far* too much... a real sysadmin would be able to run a webserver in a much tighter config...
i was running NFS, INN and a very early version of Apache on my 48k Speccy in 1983 and *still* had plenty of space left to play Quake in...
- Aidan
OTOH, I wonder why he didn't choose GNAT.
From the article: The problem is that if Perl 6 were written in Ada--it would require people to bootstrap GNAT before they could even get to Perl. That's too much of a burden to put on anybody.
Basically Ada '95 didn't get used because a lot of people don't have it. Chicken, egg, egg, chicken.
But that may also be suffering from lack of current maintainers.
Errm? 3.11p was released in January, and 3.12p is suppoused to be out RSN.
AdaCore Technologies is the normal maintainer, but they were sounding very commercial when I talked to them a year ago.
Hmmm. I'm a member of the Ada-Linux team and we've got a very good relationship with ACT and ACT-Europe. They are a company, but they do do a lot to help both the Ada and GNU communities (there was an ALT/ACT meeting in Paris in June), both officially and unofficially (the Ada-mode for emacs is maintained by them, and several members of ACT-Europe helped create GtkAda)
Of course the problem of where to get his programmers couldn't have anything to do with his choice. :-)
Course, comp.lang.ada wouldn't be a good place to start would it?
- Aidan (dislikes disinformation, but is horrendously behind on /.)
(User Info) http://killingmiranda.pair.com/
Oi! Gofffik! Morrrrttttiiiissssshhhhhaaaa
Sorry, dont know what came over me there... anyway:
The problem I have noticed with CS graduates who have learnt high level langauges to the exclusion of all else, is that they can't then grasp `real world' programming concepts. This may only be a problem afflicting British CS grads, but it is
Surely every CS grad will have had to do some assembler at some point? You do it in 1st year here (University of Glasgow, officially best CS dept in the UK), and 2nd year....
C is a shit language, that requires you to do too much at a low level, in an unsafe manner, that could be better abstracted away.
There's no inherent difference between a compiled language and an interpreted language (and where to just-in-time compiled languages like Perl fit into this world view?), there are C interpreters and Basic compilers.
I've got an obvious choice for introducing people to programming, but it's an obvious choice for doing most programming in and i don't want to get into a language flame war.
- Aidan
Buffer Overflows are a result of a lack of bounds checking. This is a logic error. Logic errors are the one hardest error to detect in programming. The reason there are so many buffer overflows are because when you program, you dont
Buffer overflows could be avoided by using a language which has bounds checking built in.
- Aidan
movement is doing a lot in this direction. Cryptography is on top of the list. Free, easy to use, public domain cryptographic tools are a necessity. And with a few targeted public research grants they could become a reality rather sooner than later. An other
The Gnu Privacy Guard already provides freely available, easy to use public key cryptography. It's extremely simple to integrate it as a filter in eg. Pine or your favourite mailer. Version 1.0 is due out RSN, and 0.9.11 was released today.
- Aidan
Just a short response to LL's point... It seems you are criticising RedHat for simply living up to the GNU Manifesto (not that RH actually knows it is doing that). That is, software is not a commodity to sell, it is
I think that RedHat, out of all the main commercial distributions (not the ones based off of it, they probably get it as well, but the others) best understands the GNU Manifesto. Debian obviously does, but it's a slightly different beast. personally I use RedHat, because it was the closest morally acceptable one when i first installed and haven't had any reason to change to Debian. I do intend to run at least one box off of Debian though at some point in the future.
- Aidan