Slashdot Mirror
Debian Release Mgr. Proposes Dropping Some Archs
smerdyakov writes "In this story posted by Andrew Orlowski of the Register Debian Release manager Steve Langasek has announced that support will be dropped for all but four computer architectures. Among the reasons cited for doing this are improving testing coordination, 'a more limber release process' and ultimately a ('hopefully') shorter release cyle. The main architectures to survive will be Intel x86, AMD64, PowerPC and IA-64." Actually, the story says clearly that this is only a proposal at this point, but it's definitely something to watch.
Seeing as they're the major systems out there. But IA-64? I've barely heard of that, and TFA says Microsoft dropped XP for that. Can anyone elaborate as to why this one was kept?
Obligatory Soundbite Catchphrase
Thanks a lot! This was about time, or else we would never get a new stable release. Lets just hope thats it gets further then just beeing an proposal...
So the question becomes, who will bother supporting non-mainstream hardware? They are still functional machines for me...
Well, I'm sure Debian has their reasons, but I suspect they're suffering due to some of their fans dropping it for other distros. Late releases, stupid politics and aged packages isn't doing this distro any justice.
As for their decision to drop SPARC, good.. I ran Debian on my SPARC boxes for a few years, and it was garbage. Slow, clumsy and at times a few bad packages got in causing problems. Debian for SPARC made Solaris look like a rocket ship.
For all you SPARC users, switch to Gentoo (Running it and loving it) or support one of the other SPARC distros like Splack (Slackware-based SPARC distro).
It's called Ubuntu.
PowerPC is stuck with a crappy old pre-NPTL glibc
because of the feature freeze. Making PowerPC be
unofficial would allow this to get fixed.
Heck, drop every port but x86. It's not nice how
the x86 port drags around the others by the
release cycle.
Perhaps Debian isn't trying to address the embedded segment.
Engineering is the art of compromise.
'a more limber release process' and ultimately a ('hopefully') shorter release cyle.
You mean like Gentoo? And they still support Sparc.
an ill wind that blows no good
I wouldn't say that IA-64 is similar to RISC chips. If anything the VLIW paradigm is exactly the opposite of RISC.
Last release was 19 July, 2002. While one can apt-get his way to modern times, I have to believe an annual release (or more frequent) will only help bring in fresh users.
FWIW, I run Gentoo with a 2.6 kernel. I have issues from time to time, but they get ironed out with a little patience. There's always someone in the community that has an answer and very often, a solution.
It seems Linux and its distributions are at a minor crossroads where stable releases and unstable, bleeding edge releases meet. On one hand you want to get new features out to users so they can test them and the software can be refined, but now that Linux is finding its way into production environments and a few desktops, bugs can be real backbreakers.
"Play is the only way the highest intelligence of humankind can unfold." -- Joseph Chilton Pearce
In the long run, Debian may well have to concentrate more on some architectures than others, but a radical step such as the one proposed will probably not fly well with the community. Since our users are our top priority, you can expect many more emails on the topic before anything will happen.
As a long time Debian user, I'm all for it, but that's probably because I'm only interested in x86 and AMD64. I think having multiple arch's is a great idea in principle, and I'm not overly keen on the idea of stomping on the minority, but it's been pretty obvious for a long while that Debian is struggling get all this stuff together into a stable release. No other distribution seems to have anywhere near the long release cycle that Debian has. Interestingly none of the others have anywhere near the number of arch's to support either. The correlation seems fairly obvious to me.
A lot of ebedded users start with deb and then modify it.
I think the nslu2 hackers start with deb.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
I've always felt that one _major_ factor in Debian's ridiculously long release cycles was supporting lesser-used architectures. Glad someone up the food chain feels the same way.
Nosce Te Ipsum
Debian runs pretty well as the Familiar distro for iPaqs, on the ARM platform. It's that kind of cross-platform support that makes Linux so interesting, and keeps the embedded platforms such exciting targets for development: recompile apps developed by such a large, general-purpose community. Embedded apps are a much more exciting platform for developers, because of the huger market and wide-open opportunities as it gets started. Debian, don't blow it!
--
make install -not war
But you are not using the "great quality duarantee Debian's development process (tm)". If you go to Sid or anything that doesn't came in a .deb package for the stable tree you are using untested or not fully tested software, for that go to Gentoo or something like that, with more ports and more testing that Sid.
Debian, like anything in the Matrix is a choose, stability and testing or this century software.
This has been floating around the debian-sparc mailing list all day... there seems to be quite a few pissed off users...
I suppose that if the Debian devels are pig headed enough to have a meeting like this without inviting anyone from the sparc community, it really says something about what users they care about.
I've been saying for years that Debian/GNU is _the_ Server OS. A look at the proposed Arch support would leave one to believe that they want to re-vector themselves as _the_ desktop os.
The slow and steady release schedule that debian has stuck to is great for server and other enterprise uses but does leave a bit to be desired for the desktop, look at the void being filled by ubuntu, progeny and mepis on the desktop.
I suppose I knew this day was coming, #gentoo-sparc currently is a better source of tech support than #debian-sparc is on freenode....
I really didnt want to switch to gentoo, but unless I want to go the BSD route that seems to be the only viable option.
Can someone send me a 4x5" gentoo sticker that I can use to cover the red swirl sticker on my truck?
Words are only yours until someone else uses them...
As an embedded linux developer (who has worked on both ARM and MIPS), I can tell you that for a production, shipping system, it doesn't matter. You'll almost always end up rolling together your own thing. However, when a vendor (e.g. Cirrus Logic) has an evaluation board (e.g., EDB9315) that comes with a hard drive with Debian loaded on it and you can see that X11 works with the framebuffer driver and USB keyboards and mice work and network apps work, it's very impressive. Most imporantly though it verifies that the drivers (framebuffer, usb, ide, serial, network, pcmcia, CF, et cetera) are implemented in a standard way and will work with "off the shelf" linux apps. This makes things amazingly easier than with other companies whose linux ports are not as complete or functional. And if you're a small company doing an embedded Linux project, it's much better to go with a System-on-Chip processor from a vendor that provides a good Linux port and good Linux drivers than it is to either do your own or write your own drivers.
However, it is sometimes very useful to use a full system like this to do native compiles of your applications (instead of cross-compiling) and native debugging. Of course, when you move to your custom hardware, you usually have to drop all that nice stuff.
(By the way, I am really a big fan of the Cirrus Logic 93xx series system-on-chip processors. After working on two other ARM SoC systems and one MIPS system, the Cirrus 9315 was by far the best supported.)
My other first post is car post.
That's not exactly true either.
The real paradigm shift of commercially released risc processors wasn't a simplified instruction set (they may have once been simple, but that's definately no longer true). The real difference is a consistent addressing schema and a load/store architecture. EPIC, the instruction set architecture of the itanium, does this also.
In fact, if you read each instruction sequentially out of ia64 bundles, each could be an instruction on a hypothetical risc processor. This defeats some of the purpose of the ISA, but is technically valid. I have to agree with the previous poster who suggested that the itanium is risc-like. It is. It's a rather-wide risc processor whose pipe-line control logic is part of the compiler, rather than embeded in hardware. Everything else in the itanium could be added to a risc processor except for the back-wards compatibility thing. (rolling register window, predicated execution, speculative loads, etc)
You're kidding, right? Pretty much everyone involved in IA-64 is pulling out; all the IA-64 workstation vendors have stopped making workstations, Windows for IA-64 has been officially put out to pasture. The hardware (what there is) is still so expensive, it's ridiculous. No one's developing for it - everyone's using x86_64 ("x64", as Sun and MS are calling it). I really would have to agree that SPARC support would be more worthwhile.
Sam: "That was needlessly cryptic."
Max: "I'd be peeing my pants if I wore any!"
I'd like to install it too?
;-) but fit into an A1000 and can be bought cheap at ebay..
I didn't get around, I just installed an mtec 500/030 board, they're not fast (3 bogomips
Maybe it would make sense to have different types of Debian GNU/Linux for tiny devices, desktops and servers.
After all, it makes as little sense to have KDevelop running on m68k as having a Gaim package for s390.
http://www.dieblinkenlights.com
I would be a victim of that decision - I run Debian on my Amiga3000 just for the fun - but I still say "go for it".
That proposal aims for stable releases. I see no problem seeing an unstable m68k debian popping up after some time. Right now even the stable m68k-Debian is a rotting piece of shit not working at all so why bother with stable at all?
"Life is short and in most cases it ends with death." Sir Sinclair
So all I have to do is use an installer for Debian that's not on the Debian download/install page, and not mentioned in the Debian install guide, and then I'm all set?
If I wanted to do that, why wouldn't I just use Ubuntu?
(And if it's easier for people who've been using Debian for close to 10 years, like me, to switch to Ubuntu when installing, you've got a problem.)