USL (first part of AT&T, later Novell) vs. BSDi about whether BSD contained illegaly copied sources from original Unix. IANAL, so I do not know how important it was for (usasian) law, but it certainly was for the Unix world.
More generally, if you are interested in the history of proprietary software, some classics are Brian Reid's Scribe programm, which he sold to Unilogic, and which was one of the first ones to technically enforce restrictive licensing terms (via a timebomb), and Bill Gates' "Open Letter to the Hobbyists", one of the first and - for obvious reasons - most well-known documents to call people violating software copyrights "thieves". Google should find a lot on both of them, although neither event was ever dealt with in court, AFAIK.
There are already easier conventions for markup in mail and news messages, like for/italics/, *bold* or _underlined_ text, for links like <URL:http://example.org> etc. Many clients also support +v/-v blocks to indicate "verbatim" sections for code and similar things and render them with a fixed-width font, don't substitute emoticons with icons in them and so on.
These are years, probably decades old. But I guess it is too simple to implement and use to become really popular in an Outlook dominated world...
I for one support slashdot's attempts to make money by generating hits with unethical journalism. Maybe VA Resarch/Linux/Systems/however they are called this week can use it to fix Sourceforge, anonymous CVS access for example. No sane person would expect (or want) unbiased, accurate reports from/. anyway, and sf going down because of VA going bankrupt would be a worse hit for the FLOSS world than everything SCO and Microsoft can even dream of.
(Morale of the story: only read/. when drunk, and don't host important projects on sourceforge. At least have decent backups and spare web/CVS space)
Boy, I want your job. If my experience with commercial, proprietary software development has taught me anything, then that politics and lack of interest screw up that just as well, probably more than it happens in FLOSS projects. Idealism is less of a problem, however, but I fail to see how it is a problem for FLOSS either.
Every Posix-compliant system has/bin/sh, but not neccessarily Python or any other interpreter or compiler. Which is the reason why Python comes with a ~20000-line configure script written in sh, not a python-based one even if it probably would be nicer.
It is the same language you use in an interactive shell. Once you figured out how to do stuff manually, you just have to put the commands you typed in a file together with a shebang line, and there is your script. Knowing shell scripting also helps when working in interactive mode, a loop or case/esac is also useful in one-liners.
It is generally faster to start a shell than another interpreter (although bash has almost fixed that, so this is not so much of an issue on GNU systems). If you have lots of small scripts, that matters (think init, for example).
You can source shell scripts, hence you can write a shell script that modifies the environment of your interactive session. You cannot do that with any other language.
On the other hand, nobody proposes to write something like Zope in sh. Both Shell and higher-level languages have their uses, as have lower-level languages like C or assembly. The good programmer/admin is the one who can use more than one tool productively, and knows when to use which.
I wonder, was the XFree86 core team elected, like for example the BSD core teams are?
I personally do not think that having some structure in a FLOSS project is a bad thing, and in the case of the BSDs it works quite well, IMHO. Of course there are often conflicts, but they would not go away if there was no defined way to deal with them, or by this way being to leave all decisions to a "benevolent dictator". Was the Xfree86 case just due to the wrong people in the core team? Why then were they in it in the first place?
Why not, it already works the other way around. RedHat, Debian and Gentoo can be installed on FreeBSD as ports (emulators/linux_base*) thanks to its Linux ABI support. Even using FreeBSD and various Linux distros (except the actual Linux kernel, of course) simultaneously on different virtual consoles is pretty easy, just a little 'chroot' here and there...
A lot of those delays stem from the simple fact that legal staff usually haven't got the vaguest understanding of how software is architected or compiled. They don't know that half the headers mentioned are part of ANSI and ISO C/C++ standards.
They don't know that every single platform with a C compiler since the early '80s has had an "errno.h" header file.
AFAIK, (most of) errno.h is not part of ISO or ANSI C, but of POSIX. Anyway, errno and the actual error numbers are likely to be only defined as "symbolic constants", which is the usual Posix way to ensure that the standard is useless for any halfway sane language^W^W^W^W^W^W^W^W^W^Wportability, so from a copyright point of view, what matters is actually whether EWOULDBLOCK is 1 or 19.184, not that it is defined at all.
While defining any errno.h is not exactly rocket science, or any harder than defining a hello-world implementation, I fail to see why copyright would not apply. It's not as if there were a rule that laws must not be stupid and pointless, mind you.
they'd have the sense not to override the default behaviour of the damn language, which would be to go to bignum if necessary. It would take effort to write a declaration to actually deliberately override the behaviour, and would be A Seriously Stupid Thing To Do. Doesn't mean that somebody, somewhere wouldn't do it, of course
Indeed, someone did, sort of. Namely the implementors of the SBCL compiler (and they probably inherited it from CMUCL) who, generally, definitely do not qualify as stupid.
"... and of course, CL transparently uses bignums when a numeric
quantity exceeds the range of machine words, so we don't get overflow
problems"
* (decode-universal-time (+ (* 86400 365 50) (get-universal-time))) debugger invoked on condition of type TYPE-ERROR: The value 2635786389 is not of type (SIGNED-BYTE 32).
This is because I didn't specify a timezone, so it asks unix for the
default timezone and DST settings, and unix needs a time_t, which is
32 bits on this box. Dan Barlow, SBCL and the Y2038 problem
So even if Lisp tends to not have overflow problems, Unix and C will come back and bite you if you give them a chance...
If only dumb Linux users would fill out this survey, their results would be that they should make an OS for dumb people. Which is what they do anyway. Not a big win for their marketing department, probably. But then I guess such uncontrolled web surveys are pretty useless anyway (as are marketing departments).
Commercial operating systems don't waste time with supporting different gui
Download Solaris some time, it's free, and definitly commercial. The current version comes with two desktop environments (CDE and Gnome) installed by default, and several optional ones.
For Mac OS X, while having a consistent default desktop, Apple "wastes time" with supporting completely unrelated graphics subsystems, their own and X11. They also have their own version of Java's Swing, as a second API for their look and feel.
It's just MS Windows that is that monocultural. It might still be a valid choice for UserLinux to only support one desktop, but it is not the only possible one.
Actually, that Trolltech GPLed Qt is a problem in itself. ISVs using KDE would have to open their applications or pay Trolltech for a commercial license, while GNOME-using ones don't have to do either, scince it is all LGPL.
Not that there are that many ISVs anyway, but I guess that some people want to change it. But all that has been discussed to death already.
Well, don't just look at MySQL. Maybe PostgreSQL, SAP DB or Firebird is an option for you, each of it has satisfied users that will assist you in evaluating them (well, I guess so. I never met any SAP DB user, satisfied or otherwise, but there must be some...)
Now, would please somebody write a Free clean-room reimplementation of Symbolics Genera, so that the FLOSS community can catch up with the operating system state of the art of the 80ies?
I mostly agree, Solaris is a great, reliable server OS. But it is slightly annoying as a desktop system, and what else would somebody use x86 hardware for, after all?;-)
Anectodal evidence: I never experienced ugly crashes due to bad server code (except of the bad server code itself, of course), however I just tried Gaim on my home Ultra5 recently. It took 15 seconds avarage after I tried to send a message to take down Gnome or CDE, and sometimes I ended up in text mode because the X server became unusable. Granted, Gaim is not a Sun-supported (or generally supported by anyone, really) product, but Gnome, CDE and XSun surely are, and they are not supposed to crash just because I run a buggy piece of shit as an unpriviledged user.
IMHO, the free OSes, especially the non-commercial ones like the BSDs, Gentoo, Debian etc., are more habitable, the trivial stuff tends to work out of the box more often. If you do non-trivial stuff, the situation is different.
Re:back when Solaris7 arrived Sun decided to...
on
Solaris 9 x86 Review
·
· Score: 1
CDE is the default in current Solaris 9, but that only means that it is the preselected item in a drop-down list in the login dialog unless the admin changed that or the user decided to use Gnome instead when he logged in last time. Gnome 2 is installed by default as of Solaris 9 8/03.
(Unfortunatly, the Gnome desktop sucks not much less than the CDE option, due to it being rather bare-bones, outdated (still 2.0) and not integrated as well as CDE is. Many Solaris-specific options are available only in a "CDE" submenu of the "real" Gnome menu, for example.)
Not to mention pkg-get, a really nice utility roughly similar to apt-get.
Another great option to make installing lots of free software packages painless on Solaris (disregarding the obviously superior strategy of LARTing all those l33t L1nux c0d3rz who think that "portable" means "compiles on both Red Hat and Debian" until they beg for a set of Coherent floppies to test their buggy code on it) is the NetBSD pkgsrc tree (what the other BSDs call ports), which happens to be actually platform independent. Not every package works, and it is completely non-integrated with the native Solaris package management, using its own package database, file format and utilities, but it's still great that it's there.
With these tools package management on Solaris can become nearly as comfortable as on a community-developed free OS; I'd say you reach about the same convenience level as with a commercial freenix distribution.
Bullshit. The classic Unix model system is completly insufficient to build secure systems, one of the best examples of the worse-is-better approach of "the Unix philosophy" and how this will always come back and bite you. We badly need to get away from stuff like a pure uid-based permission system and, more than anything, from the almighty root.
Even Windows is better than that, if it weren't for the many implementation bugs, the braindead default configuration and the fact that most app developers simply refuse to use it properly, out of sheer ignorance I guess.
Fortunatly, there are projects trying to overcome the Unix limitations, like SELinux, the TrustedBSD project (mostly merged into FreeBSD 5), Trusted Solaris (which will be standard functionality in Solaris 10) etc. However, this stuff is not widely know, hard to set up and just doesn't feel native to a Unix system. But they will surely help until better OSes, like Eros for example, are ready to use.
Well, I know that what you said sounds nice, but I'm willing to bet a large ammount of money that RedHat, SuSe, etc. had ulterier motives.
[...]
BUT, if there was an exploit in FreeBSD would the linux community come to it's rescue? Not a chance in the world.
But that's just the trick: Of course, RedHat and SuSE are in it just for the money, not because they share common idealistic goals with Debian. Or FreeBSD. Or give a fuck about "the community" whatever that means. But they all share a lot of code, and the sharing works.
I honestly have no idea who found the latest vulnerability in Bind 8, but even if it was RedHat or SuSE or some other company that would become the next SCO if they thought it would pay off, it is likely that the FreeBSD security advisory would be just the same. They have nothing to win with keeping such stuff secret, only to lose goodwill from the community and their competition.
Oh puleeese, keep you conspiracy theories to yourself, will ya? Debian is one of the highest-profile projects on the net, with many many users. It will either have been somebody who feels like he has been personally pissed off by Debian, or some script kiddie in it for the fame. Or just somebody who wanted to try it "just because".
The attack will not have been done on behalf of an official Microsoft, SCO, RedHat or *BSD representative. At least, if it was, the offender deserves to be laughed out of existence. This is not how the real world works.
Slashdot Mirror
More generally, if you are interested in the history of proprietary software, some classics are Brian Reid's Scribe programm, which he sold to Unilogic, and which was one of the first ones to technically enforce restrictive licensing terms (via a timebomb), and Bill Gates' "Open Letter to the Hobbyists", one of the first and - for obvious reasons - most well-known documents to call people violating software copyrights "thieves". Google should find a lot on both of them, although neither event was ever dealt with in court, AFAIK.
There are already easier conventions for markup in mail and news messages, like for /italics/, *bold* or _underlined_ text, for links like <URL:http://example.org> etc. Many clients also support +v/-v blocks to indicate "verbatim" sections for code and similar things and render them with a fixed-width font, don't substitute emoticons with icons in them and so on.
These are years, probably decades old. But I guess it is too simple to implement and use to become really popular in an Outlook dominated world...(Morale of the story: only read /. when drunk, and don't host important projects on sourceforge. At least have decent backups and spare web/CVS space)
Boy, I want your job. If my experience with commercial, proprietary software development has taught me anything, then that politics and lack of interest screw up that just as well, probably more than it happens in FLOSS projects. Idealism is less of a problem, however, but I fail to see how it is a problem for FLOSS either.
- Every Posix-compliant system has
/bin/sh, but not neccessarily Python or any other interpreter or compiler. Which is the reason why Python comes with a ~20000-line configure script written in sh, not a python-based one even if it probably would be nicer.
- It is the same language you use in an interactive shell. Once you figured out how to do stuff manually, you just have to put the commands you typed in a file together with a shebang line, and there is your script. Knowing shell scripting also helps when working in interactive mode, a loop or case/esac is also useful in one-liners.
- It is generally faster to start a shell than another interpreter (although bash has almost fixed that, so this is not so much of an issue on GNU systems). If you have lots of small scripts, that matters (think init, for example).
- You can source shell scripts, hence you can write a shell script that modifies the environment of your interactive session. You cannot do that with any other language.
On the other hand, nobody proposes to write something like Zope in sh. Both Shell and higher-level languages have their uses, as have lower-level languages like C or assembly. The good programmer/admin is the one who can use more than one tool productively, and knows when to use which.I personally do not think that having some structure in a FLOSS project is a bad thing, and in the case of the BSDs it works quite well, IMHO. Of course there are often conflicts, but they would not go away if there was no defined way to deal with them, or by this way being to leave all decisions to a "benevolent dictator". Was the Xfree86 case just due to the wrong people in the core team? Why then were they in it in the first place?
Why would a successor for XFree86 be needed? As I understand it, this is only a change in the "political" structure of the project, not its end.
Why not, it already works the other way around. RedHat, Debian and Gentoo can be installed on FreeBSD as ports (emulators/linux_base*) thanks to its Linux ABI support. Even using FreeBSD and various Linux distros (except the actual Linux kernel, of course) simultaneously on different virtual consoles is pretty easy, just a little 'chroot' here and there...
While defining any errno.h is not exactly rocket science, or any harder than defining a hello-world implementation, I fail to see why copyright would not apply. It's not as if there were a rule that laws must not be stupid and pointless, mind you.
If only dumb Linux users would fill out this survey, their results would be that they should make an OS for dumb people. Which is what they do anyway. Not a big win for their marketing department, probably. But then I guess such uncontrolled web surveys are pretty useless anyway (as are marketing departments).
Try "M-x doctor" sometimes. More intelligent than many slashdot trolls.
For Mac OS X, while having a consistent default desktop, Apple "wastes time" with supporting completely unrelated graphics subsystems, their own and X11. They also have their own version of Java's Swing, as a second API for their look and feel.
It's just MS Windows that is that monocultural. It might still be a valid choice for UserLinux to only support one desktop, but it is not the only possible one.
Not that there are that many ISVs anyway, but I guess that some people want to change it. But all that has been discussed to death already.
Well, don't just look at MySQL. Maybe PostgreSQL, SAP DB or Firebird is an option for you, each of it has satisfied users that will assist you in evaluating them (well, I guess so. I never met any SAP DB user, satisfied or otherwise, but there must be some...)
Oh wait, there is, albeit in a suboptimal state.
Now, would please somebody write a Free clean-room reimplementation of Symbolics Genera, so that the FLOSS community can catch up with the operating system state of the art of the 80ies?
Anectodal evidence: I never experienced ugly crashes due to bad server code (except of the bad server code itself, of course), however I just tried Gaim on my home Ultra5 recently. It took 15 seconds avarage after I tried to send a message to take down Gnome or CDE, and sometimes I ended up in text mode because the X server became unusable. Granted, Gaim is not a Sun-supported (or generally supported by anyone, really) product, but Gnome, CDE and XSun surely are, and they are not supposed to crash just because I run a buggy piece of shit as an unpriviledged user.
IMHO, the free OSes, especially the non-commercial ones like the BSDs, Gentoo, Debian etc., are more habitable, the trivial stuff tends to work out of the box more often. If you do non-trivial stuff, the situation is different.
(Unfortunatly, the Gnome desktop sucks not much less than the CDE option, due to it being rather bare-bones, outdated (still 2.0) and not integrated as well as CDE is. Many Solaris-specific options are available only in a "CDE" submenu of the "real" Gnome menu, for example.)
Another great option to make installing lots of free software packages painless on Solaris (disregarding the obviously superior strategy of LARTing all those l33t L1nux c0d3rz who think that "portable" means "compiles on both Red Hat and Debian" until they beg for a set of Coherent floppies to test their buggy code on it) is the NetBSD pkgsrc tree (what the other BSDs call ports), which happens to be actually platform independent. Not every package works, and it is completely non-integrated with the native Solaris package management, using its own package database, file format and utilities, but it's still great that it's there.
With these tools package management on Solaris can become nearly as comfortable as on a community-developed free OS; I'd say you reach about the same convenience level as with a commercial freenix distribution.
Even Windows is better than that, if it weren't for the many implementation bugs, the braindead default configuration and the fact that most app developers simply refuse to use it properly, out of sheer ignorance I guess.
Fortunatly, there are projects trying to overcome the Unix limitations, like SELinux, the TrustedBSD project (mostly merged into FreeBSD 5), Trusted Solaris (which will be standard functionality in Solaris 10) etc. However, this stuff is not widely know, hard to set up and just doesn't feel native to a Unix system. But they will surely help until better OSes, like Eros for example, are ready to use.
I honestly have no idea who found the latest vulnerability in Bind 8, but even if it was RedHat or SuSE or some other company that would become the next SCO if they thought it would pay off, it is likely that the FreeBSD security advisory would be just the same. They have nothing to win with keeping such stuff secret, only to lose goodwill from the community and their competition.
The attack will not have been done on behalf of an official Microsoft, SCO, RedHat or *BSD representative. At least, if it was, the offender deserves to be laughed out of existence. This is not how the real world works.
(A: Slashdot weenies)