Debian Announces Sarge Will Include GNOME 2.8

El Cubano writes "A recent posting to the debian-devel-announce mailing list announces that Sarge will release with GNOME 2.8. From the announcement: 'After requests and a detailed proposal from the GNOME team, we accepted an upload of GNOME 2.8 into sid, and, via the usual mechanisms, into sarge. We should mention that the release team was running out of objections to GNOME 2.8 in unstable that the GNOME team hasn't satisfactorily addressed; this, and the fact that they have demonstrated good reaction times of late are the main reasons why we're approving it despite the timing.'"

Comments

Sarge Will Include GNOME 2.8

[gowen]

And its scheduled for release at roughly the same time as Saddam Hussein

Re:Sarge Will Include GNOME 2.8

[Smuttley]

If I had mod points I use them to smack who ever marked this post as troll around the head :)

Re:Sarge Will Include GNOME 2.8

[Anonymous Coward]

Don't be naive -- Saddam will have a "heart attack" shortly before his trial.

Re:Sarge Will Include GNOME 2.8

[kabloom]

Debian only has releases once in a blue moon. July was a blue moon, therefore the Debian project started the release process by freezing the base system. The developers now figure that the release will get out the door when Sarge freezes over!

Re:Sarge Will Include GNOME 2.8

[Anonymous Coward]

You mean debian hired a bunch of NYU Lawyers to work on the project?

Re:Sarge Will Include GNOME 2.8

[pretzelgod]

You joke, but we'll be lucky to see sarge in 2006 if this is any indication.

I have been waiting for this

[Dancin_Santa]

Currently, Gnome is simply too slow in reacting to interrupts like mouse movements or keyboard input. The lag between the physical action and the action on the screen is much greater than in KDE or (yech) Windows. 2.8 seems to have improved this a lot and I will be happy to finally have it on the base install.

Unfortunately, the problems with security that Gnome has (namely unchecked buffer overruns) are still lurking there. So despite the speed which the C implementation gives, the boundary checking taketh away.

So it goes.

But compared to Windows, Linux (and especially Debian) is really secure.

Re:I have been waiting for this

[Anonymous Coward]

So despite the speed which the C implementation gives, the boundary checking taketh away.

You'd think they'd compile all the test versions with compiler bounds checking turned on (will gcc do this? If not, electric fence), regress test the bugs out, and then turn off the bounds checking on the stable versions.

Re:I have been waiting for this

[wild_berry]

Are you sure this isn't a troll?

I've never had that problem with Gnome. My system is as responsive with Gnome as my previous operating systems: Win98se and WinXP. Are you sure your system is okay?

Re:I have been waiting for this

[Anonymous Coward]

Unfortunately, the problems with security that Gnome has (namely unchecked buffer overruns) are still lurking there. So despite the speed which the C implementation gives, the boundary checking taketh away.

You mean like the linux kernel?

Re:I have been waiting for this

[oddbudman]

I'm with you on this call dude. Gnome 2.6 and Gnome 2.8 seemed fine to me (responsive wise). It doesn't take a genius to have a poke at C lacking boundary checking. Last time I checked the kernel was written in C also- what a hack! Furthermore I thought that most Gnome's apps were written in python.

Re:I have been waiting for this

[yermoungder]

You can't prove the absense of bugs by testing - only their persense. If you want to _prove_ absense of run-time exceptions then you have to start using tools like those from www.sofcheck.com, www.polyspace.com or www.praxis-his.com

Re:I have been waiting for this

[Anonymous Coward]

Absence and presence, thank you very much.

Re:I have been waiting for this

[Anonymous Coward]

Unfortunately, the problems with security that Gnome has (namely unchecked buffer overruns) are still lurking there. So despite the speed which the C implementation gives, the boundary checking taketh away.

There is no boundary checking in Gnome.

Re:I have been waiting for this

[Short Circuit]

Run it on a slow enough machine, and you'll see a difference. How slow? Try, say, a 200MHz box.

Re:I have been waiting for this

[Bigman]

Yeah and I can't even get it to install on my Commodore PET.

Re:I have been waiting for this

[jrcamp]

Unfortunately, the problems with security that Gnome has (namely unchecked buffer overruns) are still lurking there.

Why don't you submit a bug of these supposed overruns that you've found then? This post is just a general troll anyway.

Re:I have been waiting for this

[barrkel]

In my experience, X is amazingly slower than Windows for all UI stuff - including percieved mouse and keyboard latency.

Re:I have been waiting for this

[Anonymous Coward]

save a file 2 ur desktop, now look at ur desktop, this is typical

Re:I have been waiting for this

[pe1rxq]

Ofcourse windows responds faster: you get a connection refused immediatly when trying to connect to port 6000.....

In reality I have seen X respond very well on systems were windows wouldn't fit even if you used a sledgehammer. On what kind of setup is your experience based? Are you sure it was really X that was slow or the tookit you used?

Jeroen

Re:I have been waiting for this

[gullevek]

X _is_ slower. This is a fact. It is slower on my sony 1.4 ghz laptop with a ati mobile 7500, and it is slower on my ibm 1.4 ghz box with a (crappy) nvidia onboard card. It's a sad fact. (at least XFree 4.3 is slower)

Re:I have been waiting for this

[yermoungder]

Well, there you go - I didn't have a spell-checker tool available to spot that bug! :-)

Re:I have been waiting for this

[pe1rxq]

Like I said before: name the thing you are measuring....
What was slower? Did it take longer to hit the floor when you dropped it? Did you drop X, or did you drop X+KDE+GNOME+OOo+JVM?

Re:I have been waiting for this

[zemoo]

From what I understand, this is due to the Kernel.

In Windows, the application that is in the foreground (with the currently active window) actually gets a scheduler priority boost from the operating system.

Under Linux, this is not the case.

What does that mean? In Windows, interactive applications are snappier, but the background programs lag more. If you're running Windows, right-click on My Computer and hit Properties.
Go to the Advanced tab and hit the Settings button under Performance. Go to the Advanced tab and the first option I believe controls this behaviour.

Re:I have been waiting for this

[Anonymous Coward]

Parent is absolutely right - not only does Windows give the foreground task greater priority, but it uses several 'dirty' tricks to acheive much greater responsiveness (provided your machine meets the recommended specs of course - I find Linux tends to cope a little better under extreme load).

I read somewhere (can't remember where) that Windows has windowing support at the kernel level which gives window movement/creation a significant speed boost.

Re:I have been waiting for this

[loginx]

The file shows up on my desktop immediately.
Famd works well with Nautilus.

Re:I have been waiting for this

[QCompson]

I'm really not trying to be a troll, but whenever I use linux exclusively for an extended period of time (for a few months), and then switch back to 2000/XP, I immediately notice how much more responsive Windows seems to be. This is a AthlonXP 2500, with 512 mb ram. Currently I'm using Ubuntu, but have also recently been using Debian, Fedora Core 2, Suse 9.0 and 9.1, and Mandrake 10. Also, I have experimented with Gnome, Kde, and Xfce. Perhaps it is more a matter of application start-up time than overall responsiveness, I'm not sure. I just know, that whenever I switch over to Windows, I am (rather unfortunately) impressed by its speed.

Re:I have been waiting for this

[Anonymous Coward]

Try running a WM like ratpoison or openbox and see if it is still just as slow.

If it is still slow, then X might be slower than windows. Otherwise, you are a troll.

Stop confusing X with your desktop environment... KDE and Gnome ARE _NOT_ X.

X is a small-footprint system that is able to squeeze in hundreds of simultaneous transparent TCP connections with very low overhead. Let's see how fast windows can do that... oh wait!

Re:I have been waiting for this

[ArsonSmith]

I have to agree with you. I don't really consider it unfortunate though. In fact I see it as good because it gives Linux something to shoot for, and hopfully beat.

Re:I have been waiting for this

[ajr_trm]

I just know, that whenever I switch over to Windows, I am (rather unfortunately) impressed by its speed.

Apparently when you switch you are reinstalling windows, right? Freshly installed windows with only office onboard is fast. Try installing some more apps work with it few months and it won't be so fast anymore.

Re:I have been waiting for this

[laughing rabbit]

Maybe it is a multi-boot machine.

Re:I have been waiting for this

[QCompson]

Actually, no. It is the linux install that is more often than not fresh (it is a dual-boot setup). I've had the same XP installed for over a year and a half, and because I only sporadically use it, it is full of clutter and applications that I haven't bothered to clean up (although I'm pretty good about adaware, spybot, virus scan, etc).

Re:I have been waiting for this

[drinkypoo]

I'm running XP Pro and I have the foreground application boost turned off. My interactive applications are still snappier than I'm used to X clients being. However, I have to admit that I've never put Linux on a machine as fast as my XP desktop. Linux is my server, Windows is my game console :)

Re:I have been waiting for this

[ssj_195]

Indeed. xorg have some good things planned, and progress, now that they have thrown off the shackles of XFree, is very swift. I read a post last month on their devel mailing list that says there is a functional X-server layered on top of Open GL - it's currently unstable, immature, and unoptimised, but hopefully it will be able to match or beat Longhorn's vaunted new graphics engine one day.

Re:I have been waiting for this

[Anonymous Coward]

Try running a WM like ratpoison or openbox and see if it is still just as slow.

Yehaaw... if I strip my system down to the bare essentials and run a window manager that works like something from the 70s it'll seem as fast as Windows with its full featured GUI. Yup pardner... you've convinced me! X certainly ain't slow!

Re:I have been waiting for this

[ssj_195]

Perhaps it is more a matter of application start-up time than overall responsiveness, I'm not sure

I understand that most Linux GUI apps do indeed start up more slowly than their Windows counterparts, due to the fact that they must be dynamically linked(?). Pre-linking apparently speeds start-up time considerably. Gentoo is supposed to have excellent support for pre-linking apps (which must be fully recompiled whenever one of their compent libraries is updated, in contrast with the dynamic(?) linking model).

Re:I have been waiting for this

[InfiniteWisdom]

Tell me about it! Also, I can't believe the time it took for just bash to load on that old ENIAC (running a x86 emulator). Clealy we need a hand crafted assembly OS with and interfact approximating "?", to ensure that it runs blazingly fast on every computer ever built

Re:I have been waiting for this

[Jason Hood]

Its not a troll, its a valid complaint for those who choose to not ignore the current problems in gnome2.8.

I like gnome2.8 and use it at home for the family. For that use, its unmatchable, even against windows. I use KDE3.3 at work for development.
Here is what I have personally seen though in gnome2.8:

When using the file chooser, it is possible for it to not use / as the root directory when browsing in some cases. It is also possible to click on your home directory on the left and completely lose access to the root filesystem. There is no back or up button so in that case the user must close the file chooser and reopen if they lose their spot.

If I do an AltF2 in gnome 2.8 to run a program, it takes 4-5 seconds to get the second character. I know its trying to autocomplete but that is ridiculous. In KDE, is instantaneous.

When I open firefox or even galeon, it takes 5-10 seconds to actually get a window. In KDE, konqueror loads in less than 2 seconds, everytime.

If I want to add a base item to the main menu, I am SOL in gnome.

There are three different screens configuring CD actions when CD/DVD is put in the drive. Each one does something different and only a specific combination of options across all of them will make gnome do what I want it to.

If you are using an app that utilizes the gnome filechooser and want to load a file from the local network, you are SOL unless you have previously created a network device/mount. There is no location bar that can be used to manually type in an address (for me ssh:// and smb://) on the fly. Sure grandma doesnt need this, but developers and power users sure do. Its not even an option.

Why does gnome still show http source when doing a http://www.cnn.com from nautilus? I would think that this would at least open up your default browser and give you the website. I dont know why an everyday user would want to see the source from a website instead of the rendered website. This is probably due tot he fact that gnome believes file browsers and webbrowsers are two separate applications I imagine.

Gnome needs to show either OSS or alsa, but not both (but support both). Its hard to determine which sound bar to use to get the appropriate sound level. When your kernel loads both modules. Sure I can disable one, but why cant gnome just choose one on the fly to use and stick with it.

Re:I have been waiting for this

[Reteo Varala]

I can name one instance where this could be an issue... albeit a mere annoyance, rather than anything resembling a showstopper.

I have one of those multimedia keyboards, with volume control buttons on it. Since the volume knob on my speakers isn't working properly, I tend to use the keyboard to adjust volume. Problem is, if the sound is too loud, I will try to immediately lower the volume, and it takes approximately 5 seconds for GNOME to respond.

Re:I have been waiting for this

[undertow3886]

I'm pretty sure you can hit Ctrl+L in the file selector to bring up a location box.

Re:I have been waiting for this

[Short Circuit]

Hehe. I'm still trying to get Linux working under Bochs under Linux on my 750MHz Duron. I think I'll try a Knoppix liveCD and let it boot over the weekend.

Re:I have been waiting for this

[Slack3r78]

The file browser does have problems, yes. As far as needing a location bar, tried pressing Ctrl-L? You *do* have that option.

The stutter in the run command is definitely a problem. No disagreement there.

I've never had particular issue with FF on any of my machines (including a P3 1GHz notebook), but it wouldn't surprise me if Konqueror seems faster as KDE precaches it on boot, the same trick IE has used for years. Personally, I'm not sure if I'd support hacking a similar feature into Gnome for Firefox/Epiphany/Galeon/whatever. So this is one that could go either way, really.

By base menu item, do you me the defaults in the Gnome menu? Generally program categories, etc, depending on your distro. Again, I'll agree with you there that it's annoying that there's no easy way to modify that. They did at least open up adding your own launchers for the individual categories.

The only menu I know of for configuring CD insert behaviour is Removable Storage. Could you elaborate? If it really is that bad, have you considered filing a bug report? The HAL stuff is very much in its infancy, so I'm sure there are still some major kinks in its behaviour.

Agreed on the Nautilus quirk, that one's always made me stratch my head as well. Same goes for the OSS/ALSA thing.

I'm not one to claim that Gnome isn't without its severe quirks here and there, but at the same time, I've found that it's the OSS DE closest to what I look for in a work environment. I used KDE almost exclusively until around 3.0 before giving Gnome 2.x a serious try.

Going back now, KDE feels cluttered and overwhelming compared to the simplicity of Gnome. This is the reason why I'll continue to use it as long as it continues to get better. The improvement it's seen in each release since 2.0 has been marked, and I'm sure that trend will only continue with future releases.

Re:I have been waiting for this

[walt-sjc]

Hmm. I have a multi-media keyboard too with a volume "wheel". It's very quick - a fast spin from full to zero or zero to full takes place in just a about a quarter of a second. This wheel has little "detent" feel which each click is a button press. There seems to be 17 steps, so each event is taking around 14ms, and that's to update mixer volume, the OSD pop-up, and the panel icon. I see no issues here.

Perhaps you have a configuration problem, a really slow computer, etc...

BTW, Windows on the same machine feels about the same - no noticable difference.

Re:I have been waiting for this

[walt-sjc]

Considering how delayed longhorn is, it may be out sooner.

Re:I have been waiting for this

[Anonymous Coward]

Linus Torvalds doesn't like sneaky entanglements between the UI and the kernel. He would prefer to avoid copying the way Windows does things in this regard.

For the 2.6 kernel, however, he came up with a brilliant hack that makes the system snappy without any sneaky entanglements. The scheduler rewards tasks that act like user tasks by boosting their priority automatically, with no communication from the UI.

If you run a user task such as a word processor, it will spend most of its time waiting for you to do something. Thus it will "yield" frequently (inform the system that it doesn't need to run right now and some other task can have the rest of the time scheduled for it). Linus's hack keeps track of tasks that yield frequently, and gives them a priority boost. If they stop yielding they use up the priority boost and go back to running normally.

There are other improvements in 2.6 that improve snappiness as well: kernel preemption, and the improved disk scheduler, are the two biggest ones but there are others.

For Linux kernels older than 2.4, it was often common practice to set the priority of the X server up really high (nice -10). For the 2.6 kernel, this is no longer a good idea; if you just let the kernel manage the priority boosts, you get a smoother running system.

For the highest possible performance, you should get a graphics card that X supports well, and you should try to run mostly KDE apps or mostly GNOME apps. I run a mostly GNOME desktop and my system is snappy.

Re:I have been waiting for this

[Anonymous Coward]

You can use openbox as a replacement for the default metacity gnome wm just issue a 'openbox --replace' then save your gnome-session on exit, restart gnome.
I'm using it. There are some problems with gnome keybindings, however I have noticed a vast improvement in response times(mouse,window selection movements,etc) on some older programs that I run that were unusually unresponsive with the default metacity wm.

Re:I have been waiting for this

[Reteo Varala]

*chuckle* I don't think a 2.4GHz Pentium 4 running Stage-1-based Gentoo is what one would call slow.

However, there is the possibility of a misconfiguration somewhere, since you say your volume wheel seems pretty snappy. I'll just have to hunt for it.

Re:I have been waiting for this

[Anonymous Coward]

Give QEMU a try. It's much faster than Bochs.

Re:I have been waiting for this

[Anonymous Coward]

Xorg won't make horrendously bloated stuff like OpenOffice.org or GNOME start up any faster, though.

It utterly pains me that my Win2k3 desktop (yeah, it's a server OS, but works fine as a desktop) boots up, starts apps and generally runs so much faster than any Linux distro I try.

We need REAL incentives to switch, people. An equally slow - or slower - OS doesn't impress people.

Re:I have been waiting for this

[Anonymous Coward]

That doesn't stop OOo taking bloody ages to load. It doesn't stop Linux taking longer, in my case three times longer, to boot than Windows.

Re:I have been waiting for this

[gullevek]

Seriously. If I can only run Openbox or similar on a 1.4 Ghz PIV box... Then something is wrong with Linux. Because Windows XP runs fine on them. Very fine. I think X is slow and it has nothing to do with KDE/Gnome/Xfce/fvwm2 or whatever. They are all the same slow. because all use the same code if you move a window from a to b.
I use mozilla, I use xterm, I use gaim, I use thunderbird, I use gkrellm. They as slow in KDE as in Gnome as in XFCE as in whatever. X _IS_ slow. Thats a fact.

Is Armagaedon upon us?

[AlanS2002]

Debian incorporating newly released software into stable in less than two years, who would of though.

Re:Is Armagaedon upon us?

[m50d]

You forget that it won't be in stable until Sarge is released as stable, at the moment it's just in testing. Expect the actual release just after Duke Nukem Forever.

Re:Is Armagaedon upon us?

[deminisma]

They aren't incorporating into stable though. They're incorporating it into testing, which by the sound of it, won't be released until next year, despite earlier projected release dates.

Re:Is Armagaedon upon us?

[Antique Geekmeister]

Maybe they ran out of people who bother to use "stable", since so many tools there are deprecated and a monster to maintain?

The Debian "stable" vs. unstable seems to match the new RedHat "Enteprise" releases vs. the new "FEdora". Maybe Debian can shorten their transfer time and testing enough to use "stable" for production servers? I know a bunch of people who'd like that.

Re:Is Armagaedon upon us?

[arose]

Stable is for people who need an unchanging system (no, I'm not one of them), please leave it alone.

Re:Is Armagaedon upon us?

[vrt3]

But AFAIK there isn't really an alternative for people who are willing to use a somewhat changing system. unstable changes too much; it's OK for a desktop, but not for a server. testing would be acceptable, but doesn't get security updates.

So I run stable on my little home server, eagerly awaiting the release of Sarge.

Re:Is Armagaedon upon us?

[Zach Garner]

The Debian "stable" vs. unstable seems to match the new RedHat "Enteprise" releases vs. the new "FEdora". Maybe Debian can shorten their transfer time and testing enough to use "stable" for production servers? I know a bunch of people who'd like that.

We [beware of the Marketing people's use of shockwave and flash] already do that. I'm sure many others do to. We count on stable being secure and reliable. We're ok with it being relatively slow moving, if that's what it takes. A few select applications, such as X11VNC are brought in from testing or unstable, or made into custom packages internally. It works great.

We would like to see some newer software make it's way into stable, such as subversion. Right now I run a mixed testing/unstable at work, and a mixed unstable/experimental at home. I've never had a problem, though I do take time to understand what the effects of an update will be on the unstable and especially the experimental applications.

Re:Is Armagaedon upon us?

[arose]

What new software is needed on your server anyway? I too am waiting for Sarge, to replace the Slackware on my little home server.

Re:Is Armagaedon upon us?

[vrt3]

It happens regularly that something I'd like to be on my server is not yet in stable; I can't remember all of them, but things that I can remember are Python 2.3 (or even better, 2.4), MySQL 4, PHP 5.

Re:Is Armagaedon upon us?

[gullevek]

you can always use apt-pinning. run moste in stable and get what you need from testing.

Re:Is Armagaedon upon us?

[BrokenHalo]

I too am waiting for Sarge, to replace the Slackware on my little home server.

Why? I know Slackware might not be attracting too many new users at the moment, owing to Pat Volkerding's illness, but it is still an excellent server platform by any standards.

OK, I admit I'm a bit biased, since Slackware is also my primary desktop platform, where it fits very nicely.

Re:Is Armagaedon upon us?

[DataCannibal]

Hell is bad english

who would have thought

Re:Is Armagaedon upon us?

[Dionysus]

Use backports.org. Some of the packages (most?) are put there by the Debian developers. Backports.org basically give you a somewhat changing system (since you are only upgraded the packages you want to upgrade and not the whole system).

On my servers running stable, I've done this with packages like samba, ssh etc.

Re:Is Armagaedon upon us?

[Anonymous Coward]

In Korea, only old people use stable.

Old software for old people

Re:Is Armagaedon upon us?

[vrt3]

The packages I like never seem to be available on backports.org. Python 2.3 for example, and PHP 5.

Re:Is Armagaedon upon us?

[Dionysus]

I would probably just get the source from the unstable branch, and compile from there. Or, there is another site that might have it (but I can't remember it the name right now, something like apt.org or something...)

Re:Is Armagaedon upon us?

[Jason Hood]

Depends on what is changing for a server. gcc? glibc? kernels? Yeah those should rarely change for a server. X (some run X on servers for management), nmap, alsa? Those can be updated with little affect on a "server".

Lots of ports based distros allow a variety when it comes to updating. Gentoo for one, you can choose those critical pieces and make them stay the same where as you can also specify that nmap should always be the most current possible. You dont have to specify this every time either, you set it once and forget about it. Does debian have this functionality? I know you can pretty much install any unstable version of software at any time, but can you tell debian to always install "unstable" versions of particular software?

Re:Is Armagaedon upon us?

[MasterOfMagic]

On any "front-line" server, the first thing I remove is GCC. More tools for script kiddies to use if they ever break in. Compile on another box and SSH the results over.

Re:Is Armagaedon upon us?

[JerkBoB]

The packages I like never seem to be available on backports.org.

So build or backport 'em yourself. Builds character and stuff.

http://www.debian.org/doc/maint-guide/

There is a bit of a learning curve, but it's much more satifying to install your own packages than to just give up and throw a tarball into your nicely managed system.

Re:Is Armagaedon upon us?

[arose]

1) Consistency: I have Debian on my desktop. I know how to set up and maintain, Slackware was set up by a friend and I know just enough to run it.
2) Cron does not work and I'm to lazy to fix it.

Re:Is Armagaedon upon us?

[Anonymous Coward]

Hell is bad english

It is bad English to omit your punctuation and to forget about capitalization of proper nouns.

Re:Is Armagaedon upon us?

[mikefe]

This will work until the package in testing is compiled against a newer lib than in stable.

The solution in this case is backports.org.

Re:Is Armagaedon upon us?

[mikefe]

That would be apt-get.org

Re:Is Armagaedon upon us?

[Anonymous Coward]

", who would of though."

And who would have thought a brainfart could result in something like that...

Parent needs some sleep.

Re:Is Armagaedon upon us?

[MacJedi]

Correct me if I'm wrong, but as of this date, PHP5 is not even in unstable!

Re:Is Armagaedon upon us?

[gullevek]

it depends. there are two ways for apt-pinning. one where you just want the package und you get a "no go" if the package is compiled (== requests) new libs or where you see "get all you need" and the package tells you it will install "all libs".

so you are right, if you install a core package it might be equal to an update to testing

Is it weird?

[abuendia]

Good news?
Kinda normal to have Gnome on Debian.

KDE

[ultrabot]

How about KDE 3.3?

Not that it really matters anymore - many of whom have been waiting for Sarge have got with the program and switched to Ubuntu.

Re:KDE

[coekie]

Actualy reading the announcement would have answered your question: "In the meantime, we were also asked why we decided to go with KDE 3.2, and if it would be possible to go with KDE 3.3 instead. The main reason is that KDE 3.3 in unstable started with some RC bugs, and there was no proposal from the KDE team how to proceed. The door is only closed, but not locked for KDE 3.3. We are still open for proposals how to sort the KDE 3.3 issues out, and there has been some productive discussion of late about that - but no final decision yet."

Re:KDE

[AndrewRUK]

How about KDE 3.3?

RTFA, why don't you?
It says that they were asked about KDE 3.3, but felt it was too buggy ("The main reason is that KDE 3.3 in unstable started with some RC bugs".) They also mention that big upgrades of both KDE and Gnome at the same time would be harmful to their sanity...

Re:KDE

[ultrabot]

RTFA, why don't you?

Come on, this is slashdot after all. Articles are only read by Old People.

Re:KDE

[datadriven]

The main reason is that KDE 3.3 in unstable started with some RC bugs

They must be Debian bugs, not KDE bugs, because I've been running KDE 3.3 on slack since the day after it was released.

Does anyone have any news re: Pat's condition?

Re:KDE

[Mr.Ned]

People who have "got with the program and switched to Ubuntu" aren't the target audience of Sarge. Sarge is for users who want to use an unchanging base system for well over a year.

Re:KDE

[quinto2000]

Did you run it on all 15 of the architectures that Debian supports? Lots of Linux software runs fine on some architectures but not on others. Debian is committed to only releasing software that runs on all of them. The user-friendly Debian-based distributions may be nice in other ways, but they usually only worry about one architecture - ia32 or sometimes ppc.

Re:KDE

[BrokenHalo]

Hmmm... but let's face it, the latest Gnome releases have not been without its bugs and quirks, not least the dispute as to whether or not to install and use hal and dbus, and consequences arising therefrom (which in my case, incidentally, were mostly bad).

Before the flames start, I've been a big fan of Gnome for 7 years now, but of late I have been increasingly attracted by KDE.

Re:KDE

[MsGeek]

KDE 3.3.1 is stable enough to where I'm relying on it on my ThinkPad. Sid is your friend. Gotta love the improved Konqui.

Oh yeah, and KDE has none of these problems that people are reporting with GNOME. Snappy performance on a Pentium II Mobile 400MHz. I daresay even snappier than the install of Windows 2000 SP4 on the other partition.

There is no reason why people running personal Debian desktop systems shouldn't liberally add Sid packages to their system. What Debian.Org calls "unstable" is actually ready for prime time on non-critical machines.

If you run a critical server, go with Woody aka Stable. If you can live a little on the edge with your server, run Sarge/Testing/Release Candidate. If you are setting up a desktop for Grandma, use Sarge with no Sid packages. For everyone else, live on the edge, baby! ^_^

Re:KDE

[ultrabot]

Did you run it on all 15 of the architectures that Debian supports?

No, surprisingly I only have an x86.

Re:KDE

[hal9000]

"If you are setting up a desktop for Grandma, use Sarge with no Sid packages."

I disagree... A lot of the recent improvements in [Linux] [desktop] software have had to do with UI. Grandma had a hard time grasping some older GNOME UI crap. I understand the benefit of a stable computer for grandma, but why not install Debain unstable once, then run a dist-upgrade everytime you visit? It'll be fine as long as she doesn't upgrade it herself.

Re:KDE

[Janek Kozicki]

I pictured a grandma updating debian herself. ROTFL :)))

Re:KDE

[Phleg]

Oh yeah, and KDE has none of these problems that people are reporting with GNOME. Snappy performance on a Pentium II Mobile 400MHz.

I'm not trying to troll or be contrary, but I can't tell if you're trying to imply whether or not GNOME is slow, or if these two comments are unrelated.

If it's the prior, I really don't know what you mean; running GNOME 2.8 (and 2.2, 2.4, and 2.6, when I was using them) on my Celeron 1.5 laptop is perfectly snappy. This wouldn't normally be notable, but when I unplug the AC adapter, it throttles down to 500MHz--and I never notice the difference. I can realistically say that I can tell no difference between GNOME on a 500MHz system and a 1.5GHz system; it's fast, responsive, and feels light regardless.

If it's the latter, which problems are you referring to? I hadn't heard of any.

Re:KDE

[calc]

No, most of the bugs were in KDE itself. KDE 3.3.0 was one of the worst releases yet by upstream. Even KDE 3.3.1 still depends on several programs that have yet to be released by their upstream, of which I can recall at least gpg 1.9 and rdesktop 1.3+.

Chris Cheney
Debian KDE Maintainer

Re:KDE

[poofyhairguy82]

If you are setting up a desktop for Grandma, use Sarge with no Sid packages.

Or maybe... Ubuntu?

good for debian, good for users

[feepcreature]

Gnome 2.8 is stable enough in ubuntu, so it's good to see it being used back in debian.

Also, this might help combat the "Debian [stable] never includes new stuff" meme. Another good thing.

Re:good for debian, good for users

[Anonymous Coward]

Gnome 2.8 is great in Ubuntu. I'd tried Gnome before 2.8 and not really liked it, but I now use it all the time. Is KDE 3.3 as big a leap?

Re:good for debian, good for users

[Captain Large Face]

Also, this might help combat the "Debian [stable] never includes new stuff" meme. Another good thing.

"New" is relative. After stable switches from Woody to Sarge, it is highly unlikely that you will get a new version of GNOME until Etch becomes stable. So no, Debian stable will never be the place to look for new stuff, except when the stable version changes, but then that is the whole idea.

Re:good for debian, good for users

[Anonymous Coward]

Dear lord... Sarge will never be released if they keep adding new software to it...

Re:good for debian, good for users

[Anonymous Coward]

this might help combat the "Debian [stable] never includes new stuff" meme.

That rather depends on when sarge is officially designated "stable", doesn't it?

Re:good for debian, good for users

[Anonymous Coward]

Also, this might help combat the "Debian [stable] never includes new stuff" meme. Another good thing.

Debian stable never includes new stuff - that would make it /unstable/. Stable means "tested to death." Adding stuff would mean going through the whole testing cycle again (and those efforts go to the next stable..)

If you're in need of newer stuff for stable (currently woody - sarge is currently testing and will become the next stable) - go to a place like www.backports.org or search through apt-get.org. Keep in mind though, that backports are not part of the official debian base and you can't bug the debian people about them (though most backports have active maintainers - try them instead!)

Re:good for debian, good for users

[byolinux]

I have this weird little bug in GNOME under Ubuntu - when I log in, I get an XFree 'X' cursor in the middle of the screen, and it won't go away.

Apart from that, it's all really quite nice. I'm looking forward to running UserLinux though.

Re:good for debian, good for users

[_the_bascule]

try this

You are not alone with that one :)

Re:good for debian, good for users

[zerblat]

Google and ye shall find.

Re:good for debian, good for users

[BrokenHalo]

In other words, as far as Debian is concerned, stable == fossilised, i.e. too ancient to break.

For those of us who require something more than a basic server (which, of course can be done without any kind of screen at all, let alone an X server or Gnome/KDE/blackbox/whatever), something a little more current is necessary.

Re:good for debian, good for users

[HiThere]

You might look a little at that word "require".

Different purposes require different feature sets. So there's stable and unstable, Red Hat and Mandrake. And then there's Gentoo and Lycoris. And others. Each one serves a slightly different market.

FYI

[bionicyeti]

Garden Gnome 1.0 is scheduled for release on my lawn this spring...

Re:FYI

[kaszeta]

Garden Gnome 1.0 is scheduled for release on my lawn this spring...

If it's a Debian Garden Gnome, you'll have to leave it in the back yard until it's weathered and outdated, and then move it to the front yard after it's demonstrated its stability for two years...

Re:FYI

[Anonymous Coward]

Depending upon the user environment, weathering may lead to cracks and the need for patches. Be sure that your Garden Gnome is set to automatically apply patches.

Not suitable

[Anonymous Coward]

How about KDE 3.3?

Debian aims to be an international distribution, used by people of all races, genders and age groups throughout the world. Gnome fits in with this. Debian considered KDE 3.3 as well, but found that it wouldn't be acceptable on a global basis because they found that in Korea KDE 3.3 is only used by old people.

Debian Announces Sarge Will Include GNOME 2.8

[ModMeFlamebait]

... in 2008 :P

E?

[quamaretto]

Maybe they'll put E17 in there too, now that it's available. Except E17 is for old Koreans. I use GNOME, mainly because it's visually pretty clean and well-concieved among WIMPs; that said, I still occaisionally want to drop into evilwm. It's nice to see Debian being proactive about this stuff. Now, if only they would put Firefox 1.0 and Gimp 2.x (or have they? *doesn't check*)in there, we'd be getting somewhere.

Re:E?

[Shadestalker]

Firefox 1.0 has been in sarge for a while now. Gimp 2.0.5 is in, though I'm not sure when it was released for sarge.

Re:E?

[Anonymous Coward]

Aptitude goes :

gimp : 2.0.6-1
mozilla-firefox : 1.0-2 :-)

But you still find : enlightenment : 0.16.6-3

Re:E?

[drnlm]

Umm - E17 has been checked into public cvs, which means it's at the "code good enough for other people to see, hack on and improve" stage, not the "release this to non-developers of the package" stage. Trying to get it into sarge at this point might be considered a tad premature and would be a great disservice to both the users (stuck with buggy windowmanager until next release) and E (bad reputation becuase debian version is buggy and can't be fixed). Of course, E17 might stablise and release well before sarge does :). Firefox 1.0 is in. Gimp 2.0 is in, Gimp 2.2 is unlikely.

Re:E?

[quamaretto]

QFE:

Maybe they'll put E17 in there too, now that it's available.

Note the sarcasm.

Re:E?

[quamaretto]

Sorry for the confusion - I was referring to 'stable'. I use unstable, mainly because it has all of that stuff. (Excpet E17 :)

Seriously, I should put something about "inspect message for sarcasm before replying" in my sig.

Re:E?

[Malc]

"Seriously, I should put something about "inspect message for sarcasm before replying" in my sig.

That wouldn't help. Many people -- myself included -- don't see signatures. And thank goodness because some people write such drivel in them. ;)

Re:E?

[quamaretto]

Lemma: It takes a higher level of intelligence to turn off signatures than to recognize sarcasm.
Lemma 2: If you are not smart enough to turn off signatures, you probably read them.
Theorom: You are either smart enough to recognize sarcasm or read my signature to find out about it.

:P

applicable Star Wars quote

[for_usenet]

Luke: What is it Obi-Wan ?

OWK: I felt a deep disturbance in the force. It was as if Debian decided to be more current with their packages.

Re:The obligatory..

[templest]

In Korea, regular computers are for old people. They have everything they would ever need in their cell-phones. https://www.pcworld.com/news/article/0,aid,118327, 00.asp/

Almost time for regular users to run testing

[martinde]

I've been running stable (woody) since it came out, and it has served me well. I started using Adrian Bunk's backports, and then selected things from backports.org... Then I upgraded to KDE from downloads.kde.org, and then openoffice from some other backport collection. Amazingly enough, this collection of software worked well enough for me.

I recently took the plunge and converted a couple of machines to testing (soon to be sarge). First thing I will say is that even with all of the backports, the upgrade went very smoothly. And I'll also say that sarge is working well for me; so well that I've installed it on several other machines using the new debian-installer rc candidates, and that has worked flawlessly for me as well!

As soon as security update support is up and running for testing, anyone remotely interested in sarge should consider upgrading and filing bug reports as appropriate. This is how you can help speed up the "real" release of sarge!

And I do think that when sarge comes out, it's going to be an excellent platform. It is so much nicer about hardware autodetection, font handling, and about a million other things... Without losing any of the old things that you love about Debian.

Lets hope that the next stable release doesn't take too long, although given Debian's nature, it's hard to see how it won't... Assuming the official compiler moves to gcc 3.4 (or the upcoming 4.0), then there is going to be another painful transition for all of those C++ applications. Hopefully someday g++ will have a stable C++ ABI and those transitions won't be an issue for projects shipping C++ libraries... (This was one of the major issues for getting KDE into unstable earlier this year.)

Re:Almost time for regular users to run testing

[synthespian]

As soon as security update support is up and running for testing (...)

I didn't know they were planning this...Must of missed some Debian news. Any pointers?

Re:Almost time for regular users to run testing

[petteri_666]

I would recomend using unstable rather than testing. Testing don't even get any security updates or new software soon enough so it's pretty useless. And at least desktop users should be running unstable not stable or testing. It's not so unstable that it is useless. I have been running it for a while (couple of years) and I have had zero problems. Other people are making distros based on it (Ubuntu). It's really sad to see that people are avoiding it because it called "unstable", it should be renamed as desktop debian or something like that.

Re:Almost time for regular users to run testing

[martinde]

Read the article...

It's not support for security for testing in general, but in preparation for the release of sarge.

Re:Almost time for regular users to run testing

[martinde]

> I would recomend using unstable rather than testing.

In general I agree with you - there's usually not much point to running testing. But given the imminent freeze and upcoming release of sarge, testing is a good place to be right now. In fact, running it is one of the only ways to help debug the sarge release... Getting lots of people to do live upgrades and file bug reports when they have problems is important to having the best new stable release possible.

Re:Almost time for regular users to run testing

[synthespian]

I would agree with you guys. I believe this is a flaw in the Debian engineering cycle: nobody uses testing. martinde is right when he says *this* is the time to be in "testing." But this would be the *only* right time. As it was said, it's pretty much pointless sticking with testing all year long, because you don't get the fixes upstream made that get into sid.
I believe this calls for a change in practice: have a bipolar cycle (sid/stable) and in the *final* weeks move for testing. Either that or drop testing altogether, and then have something like the BSDs practice.

Re:Almost time for regular users to run testing

[pebs]

I would disagree and say avoid testing. Go with either unstable or stable. Testing is just that -- an area to test for the next stable release. Sometimes it is broken for weeks. It is not meant for people to use unless they are helping test for the next stable release. I learned this the hard way by switching to testing from unstable. It may work for the moment, but at some point you will run into many problems after an upgrade.

Re:Almost time for regular users to run testing

[koreth]

When was the last time testing was broken for weeks? I've been running it on two servers and a desktop machine since early this year and have never once seen anything break. Pretty impressive, really. What has broken lately?

Re:Almost time for regular users to run testing

[Dionysus]

That's been because testing (Sarge) has been on the verge of freeze for almost a year now (just waiting for the installer most of that time).

Wait a couple of weeks after a release, and you will see breakage galore in testing.

Re:Almost time for regular users to run testing

[DoctorPepper]

I don't mean to sound contrary, but I've been running Testing on my primary workstation now for over six months, since the last time I upgraded Sid and it trashed several of my apps.

Testing has been working great, no problems at all, and I got to see it getting better and better with each update I performed.

Re:Almost time for regular users to run testing

[killjoe]

The reason debian takes so long is that there so many packages and so many architectures. Just too much to test in a timely manner.

IMHO they should pare down the list some. Call it core, call the rest contrib. Test the hell out of the core and treat the contrib as testing. Maybe then they could release once a year. I honestly don't see a need for more then once a year release myself.

Anonymous coward is confused

[Anonymous Coward]

So which is Sarge?
Despite using debian on two servers I can never remember which is sarge, and also if unstable or testing is the one I want.

Re:Anonymous coward is confused

[Anonymous Coward]

So which is Sarge?

Situation now:
potato = obselete
woody = stable
sid = unstable
sarge = testing

Once sarge is declared stable,
woody = obselete
sid = unstable
sarge = stable
unknown = testing

So, sid will remain unstable and a new name for the testing branch will have to be decided (unless I missed something and that's already happened).

Re:Anonymous coward is confused

[peterpi]

The releases of Debian go through the following stages:

• Unstable
• Testing
• Stable
• Dead

And for most users, at any one point in time the Unstable one offers the best tradeoff between features and stability. The current situation is that Sid is unstable, Sarge is testing and Woody is Stable.

Real Soon Now, they'll all shuffle along one, Woody will die and Sarge will become stable. I run sarge on my home and work machines and it's completely rock solid.

Re:Anonymous coward is confused

[Anonymous Coward]

You forgot experemental, the dreaded wasteland where packages are born before being shipped off to unstable.

Re:Anonymous coward is confused

[Anonymous Coward]

> unknown = testing

it will be called "etch" (as in etch-the-sketch)

Re:Anonymous coward is confused

[Todesmetall]

So, sid will remain unstable and a new name for the testing branch will have to be decided (unless I missed something and that's already happened).

The release after sarge will be called etch.

Re:Anonymous coward is confused

[Cro Magnon]

Does that mean we'll have a long wait while the developers scratch their etch?

Re:Anonymous coward is confused

[ratamacue]

Perhaps "unsupported" would be a better term than "dead". I'm still using Potato for our Samba server (which is not exposed to the internet), and although there are no more updates, it's still alive and kicking, good as new.

Re:Anonymous coward is confused

[myowntrueself]

"Real Soon Now"

Sorry, but announcements from Debian that sarge will go stable 'real soon now' usually seem to mean 'next year some time'.

Re:Anonymous coward is confused

[Anonymous Coward]

http://catb.org/~esr/jargon/html/R/Real-Soon-Now.h tml

Re:Anonymous coward is confused

[mikefe]

I believe the next stable release after sarge will be called etch.

Re:Anonymous coward is confused

[myowntrueself]

Exactly.

I've seen projects go quite sour because they depended on sarge going stable according to the debian schedule and so-called plans.

My scepticism is warranted and I predict (fingers to temples, makes buzzing sound) that we woody will still be stable this time next year.

And when sarge *finally* goes stable, Gnome 4.something will be out but *not* in sarge and people who want an up-to-date desktop will *still* have to go with unstable or testing.

Am I the only one..

[Negaiss]

..who misread this as "Debian Announces Sarge Will Include GIMP 2.8" and freaked out?

Re:Am I the only one..

[Anonymous Coward]

Yes. Yes you are.

x.org

[sewagemaster]

anyone know whether x.org will make it into sarge as well?

Re:x.org

[Anonymous Coward]

Yes, it will not.

http://necrotic.deadbeast.net/xsf/XFree86/trunk/ de bian/local/FAQ.xhtml#debianplans

Re:x.org

[smoking2000]

No, x.org will not be in Sarge.

See the news item "Future of Debian X11 Packages" in this issue of Debian Weekly News

Re:x.org

[Anonymous Coward]

The Debian team will begin integrating X.Org into Unstable when Sarge is released. Except, they'll try and do it modularly, instead of monolithically...read more http://lists.debian.org/debian-x/2004/05/msg00431. html.

Re:x.org

[Mr.Ned]

X.org is top on the list post-Sarge.

Re:x.org

[Anonymous Coward]

Fixed link: http://lists.debian.org/debian-x/2004/05/msg00431. html

How about these?

[Anonymous Coward]

PostgreSQL 8,
PHP 5,
Tomcat 5.0.x (5.5 would probably be pushing it a little..),
Sun jdk1.5?

Re:How about these?

[RAMMS EIN]

Definitely not Sun JDK. It's not DFSG compliant.

PHP 5 probably still has some issues to be worked out, meaning it won't get into stable.

I don't know about the others.

Re:How about these?

[vrt3]

I would like PHP 5 in Sarge too, but it's not even in Sid now (and neither is PostgreSQL 8) so I guess I'll have to install it from source if I'm really going to need it.

Re:How about these?

[Anonymous Coward]

jdk1.5 will not be included since they can't distribute it. I've just been using make-jpkg (found in java-package) with to package it up after downloading it from Sun's website. The rest aren't even in unstable for the specified version

Re:How about these?

[Anonymous Coward]

Dude this is still Debian STABLE. It's one thing to advance to slightly more modern software, it's another to put in Postrgresql 8 when that hasn't even been released yet!

Re:How about these?

[Technonotice_Dom]

PostgreSQL 8 is only on beta 5, not ready for production use yet!

PHP5 is only on 5.0.1 and barely out of the door. It isn't in Sid yet, and Debian would rather ship a very stable copy of 4.3.9 instead of a x.0 release. You can install PHP5 from www.dotdeb.org though. No idea on Tomcat, but as others said, the JDK isn't free.

Re:How about these?

[Anonymous Coward]

Yeah Java isn't free as in freedom yet. They have the free as in register all your details and let us (Sun) know what deeds you plan to carry out with our only asset. But if Mono is mature-ish by then ikvm can be used, so Java can still come to the party although not with Sun.

Re:How about these?

[RdsArts]

Yeah Java isn't free as in freedom yet

Yes it
is.

Oh sure, not Sun's, but...

Re:How about these?

[SeaGK]

For Sun j2sdk 1.5 (or 5.0) as they call it, use the java-package from unstable. It works pretty well for me here (40+ Workstations, 6 servers)

wajig install/unstable java-package
download the sdk tar.gz file from java.sun.com
run java-package on the download folder:

make-jpkg ./sun-sdk-file-name

This will give you a perfectly functional .deb package.
Then just do:

dpkg -i ./my_new_sun_sdk_package.deb
(configuration will fail because an unresolved dependency .. proceed)

wajig install sun-j2re1.5debian

this will fix the unresolved dependency, if it doesn't, do a:
apt-get -f install
and voila! .. Debianised sun-j2sdk (mozilla plugins and all working)

Re:How about these?

[Anonymous Coward]

That's great if you want to run "Hello, World!", but I can't see it running JBoss.

Re:How about these?

[batkiwi]

Sun JDK is not DFSG compliant, but you can download fakeroot and javapackage, download the JDK 1.5, and create your own deb out of it using javapackage.

It's a great compromise. You get java, properly in your package managment, and they get to keep their principals.

I haven't been in the army...

[Buchenskjoll]

is Sarge Will someone I should know? I am familiar with General Failure, though.

Re:I haven't been in the army...

[vasqzr]

Sarge, Woody, Potato, Sid...

All characters from Pixar's Toy Story.

Previous versions were Bo, Hamm, Slink, Rex, etc

Re:I haven't been in the army...

[Anonymous Coward]

I only know major Mistake

Re:I haven't been in the army...

[SCOX_Free]

At least your aren't familiar with Colonel Panic...

Re:applicable Star Wars quote - yoda style

[Anonymous Coward]

deep disturbance in the force, i felt. Debian decided to be more current with their packages, it was as if.

Re: You like the cup, drink from the cup.

[andersa]

The X Strike Force

To sum it up, Debian is maintaining it's own tree of Xfree86, without any material that has the new license, but with some x.org and other patches. This is what will be in Sarge.

Actually, they switch names...

[Anonymous Coward]

...as they switch places. Woody will die. Sarge will become the new Woody. Sid will become the new Sarge. A new Sid will be born.
Counter intuitive, but true.

Re:Actually, they switch names...

[zifferent]

Uhmm... no. Sarge remains sarge, woody remains woody and sid remains sid. The new unstable gets a new name. I believe the pre-woody stables were called potato, slink, and hamm respectively

Re:Actually, they switch names...

[peterpi]

Yeah, that's what I mean. Perhaps I worded it vaguely :)

Re:Actually, they switch names...

[Anonymous Coward]

Uhmm... no. Sarge remains sarge, woody remains woody and sid remains sid. The new unstable gets a new name. I believe the pre-woody stables were called potato, slink, and hamm respectively

unstable will always be called sid. i.e.:

stable -> slink
testing -> potato
unstable -> sid

stable ->potato
testing -> woody
unstable ->sid

stable ->woody
testing -> sarge
unstable ->sid

stable ->sarge
testing -> etch
unstable -> sid
sid never changes.

Re:Actually, they switch names...

[mflinquin]

The new unstable gets a new name.

No, it doesn't. sid = Still In Development, this is always the unstable branch. A new testing branch (etch?) will be created.

Re:Actually, they switch names...

[Anonymous Coward]

Sid isn't an acronym for still in development. All the names of the Debian releases are from characters in 'Toy Story'. Sid is the one who broke everyone's toys. :-)

Re:Actually, they switch names...

[Anonymous Coward]

I have read somewhere that they were doing it differently this time. I remember reading Bruce Perens commenting that it was stupid on an unrelated Userlinux page. I can't find a link. Maybe I was mistaken.

Re:Actually, they switch names...

[Anonymous Coward]

"Sid is the one who broke everyone's toys" And that btw is why "Sid will never be released".

Re:Actually, they switch names...

[Xibby]

unstable will always be called sid.

Yes, go watch Toy Story...Sid is the unstable next door neighbor with a tendency to blow stuff up. Debian unstable is aptly named.

Re:Actually, they switch names...

[Anonymous Coward]

sid = Still In Development
yes, but also: sid is the neighbours mean unstable kid, who breaks toys. -now that makes sense!
see toy story

Re:Actually, they switch names...

[Anonymous Coward]

Was unstable callse "Sid" in the Slink days? I thought that name only came along when Potato was released.

Re:Actually, they switch names...

[calc]

Testing was not created until Dec 18 2000, which was after potato was released. Prior to testing unstable was just "frozen" for a while. However, there has been a "sid" dist for a lot longer than that. sid was previously used for stuff that was not suitable for release, such as unfinished arch ports.

19 Oct 2000 - RFC: implementation of package pools
13 Dec 2000 - package pools implemented on ftp-master
18 Dec 2000 - testing to be implemented on ftp-master
10 Feb 2001 - Changes to experimental, proposed-updates and orphaned distributions

Release date

[kbewley]

Yah! A Debian insider is reported to have said that the Sarge release will form part of a bundle with Duke Nukem Forever.

Sarge and Subversion 1.1.x Branch Fiasco Continues

[Anonymous Coward]

Well, gnome has succeeded where the subversion team has not been able to yet.

While Subversion (svn) 1.1.1 contains NUMEROUS important bugfixes, Debian's misunderstanding about svn library compatibility has kept it from getting into testing.

So svn 1.0.9, the much buggier version is likely to be included with Sarge when it is released. Even more sad given that svn 1.0 branch is DEAD and will no longer be fixed.

So what are we end-users to do when a debian maintainer is unresponsive to us and to the programmers who create the apps we love?

Sadly amusing in some ways since Debian seems to be using Subversion:

http://www.debian.org/devel/debian-installer/svn

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[Anonymous Coward]

Here's a list of fixes/changes in svn 1.1.1 and 1.1 for people who think 1.0.9 deserves to be in debian testing while svn 1.1.1 is stuck in debian unstable limbo hell with NO NEXT STEP in sight to get the ball rolling.

Version 1.1.1
(22 October 2004, from /branches/1.1.1)
http://svn.collab.net/repos/svn/ tags/1.1.1

User-visible-changes:
- Client:
* fixed: 'svn status' win32 performance regression (issue #2016)
* fixed: 'svn ls' dying on non-ascii paths over DAV (issue #2060)
* fixed: allow URI-encoded colon or pipe on win32 (issue #2012)
* fixed: broken win32 UNC paths (issue #2011)
* fixed: memory bloat when committing many files over DAV (r11284, -321)
* fixed: eol-style translation error for 'svn propget' (r11202, -243)
* fixed: 'svn propedit' does EOL conversion properly (issue #2063)
* fixed: 'svn log --xml' shouldn't be locale-dependent. (r11181)
* fixed: 'svn export' of symlinks with 'use-commit-times' (r11224)
* fixed: 'svn export -rBASE' when WC has added items (r11296, -415)
* many translation updates for localized client messages.

- Server:
* fixed: 'svn ls' HTTP performance regression (r11211, -232, -285)
* fixed: make it possible to set "SVNPathAuthz off" in httpd.conf (r11190)
* fixed: fsfs validating revisions when accessing revprops (issue #2076)
* fixed: 'svn log -v' hiding too much info on 'empty' revisions. (r11137)
* fixed: encoding bug with 'svnlook log'/'svnlook author' (r11172)
* fixed: allow mod_authz_svn to return '403 Forbidden', not 500 (r11064)
* fixed: XML-escape author and date strings before sending (issue #2071)
* fixed: invalid XML being sent over DAV (issue #2090)

Developer-visible-changes:
* fixed: IRIX compile error (issue #2082)
* fixed: error in perl bindings (r11290)
* fixed: error leaks in mod_dav_svn (r11458)
* fixed: javahl should use default config directory (r11394)

Version 1.1.0
(29 September 2004, from /branches/1.1.x)
http://svn.collab.net/repos/svn/ tags/1.1.0

See the 1.1 release notes for a more verbose overview of the changes since
1.0.x: http://subversion.tigris.org/svn_1.1_releasenotes. html

User-visible-changes:
* new non-database repository back-end (libsvn_fs_fs)
* symlinks can now be placed under version control (unix systems only)
* cmdline client now supports psuedo-IRIs and autoescapes chars (issue #1910)
* 'svnadmin recover' no longer waits forever for a lock (new '--wait' option)
* new $Revision$ synonym for $Rev$ and $LastChangedRevision$
* new runtime option 'store-passwords = ' gives finer control (r10794)x
* fixed: working copies now shareable by multiple users (issue #1509)
* fixed: diff and other subcommands correctly follow renames (issue #1093)
- new 'peg' syntax for diff/merge: 'svn diff -r X:Y TARGET@REV'
- now able to compare working copy with URL: 'svn diff --old WC --new URL'
* new framework for localized error/info/help messages, initial translations:
- German, Polish, Swedish, Norwegian Bokmål, Traditional Chinese,
Japanese, Brazilian Portuguese.
* speed improvements:
- faster 'svn up' on complex working copies -- no more repos txns (r8840)
- faster 'svn status' -- fewer stat() calls (r9182)
- faster 'svn checkout' -- fewer sleep() calls (r9123)
- faster 'svn blame' -- new RA->get_file_revs() func (issue #1715)
* new switches added:
- 'svn blame --verbose' - show extra annotation info
- 'svn export --native-eol TYPE' - export using TYPE line-endings
- 'svn add --force' - recurse into version-controlled dirs
- 'svnadmin dump --deltas' - include binary diffs in dumpfile
- 'svnadmin crea

buffer overruns

[synthespian]

But isn't this due to a "cultural" problem? For instance, the fine OpenBSD guys have given us strncat and strncpy, and a patch has been applied to glibc in August 2000.

http://sources.redhat.com/ml/libc-alpha/2000-08/ ms g00052.html

GNU/Linux programmers have a bad track record on this issue. Not to mention safer languages. Consider the little attention a mainstream language like C# has gathered in the community (I won't even mention other languages that achieve C-like performance such as SML or Common Lisp). People use C for everything, not just systems programming. The result you get to read at CERT. And let's not even mention formal methods...

Re:buffer overruns

[Anonymous Coward]

...and let's not mention that saying that you're not mentioning something is, in fact, mentioning it.

Re:buffer overruns

[EsbenMoseHansen]

While I agree with the sentinent, I cannot but hasten to warn that strncpy is evil, and it should almost never be used. Why? It doesn't zero-terminate, so it might turn a C-string into a non-C-String, with hilarious results.

I assume several C-String libraries must have been designed and built over the years. But I none seems to have prevailed for whatever reason. Which is a bloody shame, because C's string (mis)-handling are a trategy, so messy and unclean in such a nice & clean language. I wonder why?

(And yes, I'm perfectly aware why strncpy does what it does, I just wish they had called it fillrecord or something sensible instead.

Re:buffer overruns

[Anonymous Coward]

I suggest you read the rest of the thread on libc-alpha whose parent you posted. Drepper rejected the patch.

Re:buffer overruns

[Anonymous Coward]

There are numerous C libraries that handle string data in a safe manner that run on GNU/Linux. There is no need to move to a different language just to achieve safety. That's a myth.

Re:buffer overruns

[metamatic]

It's not a big deal to use strncpy with a buffer size of 1 smaller than the actual buffer size, and always put a zero at the end of the actual buffer.

The problem is careless programmers. It doesn't matter if you build a safe alternative to strcpy; as long as the original is still laying around looking handy, careless programmers will use it. I mean, strncpy way predates GNOME, so frankly if there are any strcpy calls in GNOME's code base the offending programmer ought to be taken out and shot.

Re:buffer overruns

[caluml]

Calums radical idea. Why can't we simply remove strcat and strcopy from glibc (or provide a patch that does the same)?

Once the users start moaning that package-foo.1.2.3 fails with Unknown function: strcopy, or whatever, developers will soon moving to the new calls.

Re:buffer overruns

[EsbenMoseHansen]

The problem is that people forget to append that stupid 0 termination... and with long strings strncpy tends to fill buffers up with useless 0'es. I always code my own (it's only three lines) and use that. Safe, and efficient. strn?cpy should be deprecated, and replaced with stringcpy, which does the right thing and fillfield() which does with strncpy does.

Instead of shooting the poor programmers, including a "grep strcpy checkinfile; if [ $? != 0]; then send_evil_mail; fi" in the cvs checkin scripts is probably more effective. Works for KDE people, at least (not that that KDE would have problems with strcpy, obviously, but that kind of thing.)

Re:buffer overruns

[MasterOfMagic]

strncpy is in the C standard, according to my K&R second edition

Re:buffer overruns

[0racle]

The grandparent seems to have misspelt the function names. strncpy and strncat are the standard, 'unsafe' functions, OpenBSD replaced them with strlcpy and strlcat as safe functions to do the same thing.

Re:buffer overruns

[Type-R]

They also gave out strlcpy(3) , but the glibc crew decided to go with the weaker more broken strncpy, making many people reimpliment it themselves. Unfortunatly egos won out over common sense.

Re:buffer overruns

[Anonymous Coward]

Ulrich's argument was that the string functions shouldn't ever truncate strings. The programmer needs to malloc a buffer that is long enough to hold the results. The glib string functions work this way. They can't ever truncate a concatination or overflow a buffer because they allocate the buffer and return it.

Re:buffer overruns

[MasterOfMagic]

strncpy still is "checked", assuming that you use it right (though I'm not advocating its use). I know that OpenBSD has brought forth strlcpy and strlcat, but until they're in standard libraries and taught in basic classes on C, don't expect them to be widely used. People use what they're comfortable with (no matter how unsafe).

Re:buffer overruns

[EsbenMoseHansen]

Well, I'll be rolled in tomato source. Thanks! I never knew --- it's even made it's way to linux :) At least "man strlcpy" fetched a likely looking manpage, if nowhere as nice as the BSD one you linked.

Thanks!!

Now that's a first

[__aahlyu4518]

Gnome 2.8 isn't older than 12 months and it is already included !! WOW

Well... we'll still have to wait for Sarge to be released as STABLE. That could take another while... ;-)

object orientation

[Doc Ruby]

Maybe they're just looking for more "objections to GNOME 2.8 in unstable".

Re:object orientation

[myowntrueself]

You mean so that they can, YET AGAIN, delay the release of sarge?

I mean, COME ON debian, how many times have they said that its going to be released 'Real Soon Now'... I've been hearing that for over a year.

Re:object orientation

[Doc Ruby]

OSS will be unbeatable when we achieve Open Project Management. Since closed-source project management is so abyssmal, I expect this holy grail will remain lost forever - and OSS remain beatable.

Ignore parent

[peterpi]

Sorry, got confused with what message was replying to what. On a new machine and seem to need to set up my /. display perferences again.

Just run a chrooted Debian sid on stable

[synthespian]

Here's what I do: on top of a stable Debian, I run a chrooted sid. Pronto! I get the best of both (ideally, I should be using testing instead of sid, since I'm not a Debian developer).
My point being that you get the best of both worlds. It is ridiculously easy to set up a chroot jail in Debian. "Google and ye shall find."

Next branch will be 'Etch'

[Anonymous Coward]

...as in Etch-a-sketch.

As a Debian user myself...

[Cyn]

I have to ask, who the hell runs stable on a desktop - and who would want to?

If you want stability, testing provides plenty of it. If you don't want to update often - just don't update (often). If you really need rock solid core stability, but want newer desktop software - then run stable with apt-pining for testing or unstable, and only install what you know you want from testing/unstable.

If you want a Debian desktop with frequent releases without all this crap, use Ubuntu :P

Stable is supposed to be the rock solid hardware, and the only things that should change should be when there's a bugfix or security fix. The point of it is you can basically rest assured that when running updates, shit won't ever break. I don't mean just "PAM broke!" break, I mean config overwritten, changed options, etc. break. The system for all intents and purposes could be set to automatically grab updates and run for years.

Disclaimer: I've never used Ubuntu, I'm a Debian man who suffers the trials of using apt-pining just like everyone else who wants this should have to!

Re:As a Debian user myself...

[ultrabot]

If you want stability, testing provides plenty of it.

Testing is rarely usable. In fact, it's often more broken than Sid, or at least the broken state lasts longer with Testing. YMMV.

If you want a Debian desktop with frequent releases without all this crap, use Ubuntu :P

I agree. Ubuntu has been very solid for me, and I wouldn't be surprised if even Hoary Hedgehog went stable before Sarge.

Re:As a Debian user myself...

[koreth]

Testing is rarely usable.

Really? I've been running it on several machines (desktop and server) for most of this year and it has been rock-solid for me. What has broken for you lately?

Re:As a Debian user myself...

[Technonotice_Dom]

I have to ask, who the hell runs stable on a desktop - and who would want to?

If I was rolling out a set of desktops for corporate usage, stable is ideal. You don't want testing, which is prone to breaking dependancies (unless you're near freezing time), nor do you want unstable - you don't want to be troubleshooting brand new packages, you need them to just work.

With stable, you have rock solid security support, very few if any changes, and all package upgrades are tested far more than any other repository before being uploaded. A corporate desktop wouldn't matter if it was still using Gnome 2.8 in two year's time, providing there aren't any huge revolutions...

Re:As a Debian user myself...

[BenjyD]

But what is Debian Stable's policy on non-security bugs?

If you stick with stable, will you be stuck with bugs in packages that are fixed in a minor point release of the pacakged software? This could get annoying if so: the users either have to live with the bug or you have to create custom packages, defeating the point of using the distro in the first place.

For example: OpenOffice 1.1.2 has some annoying bugs in it that are fixed in 1.1.3. If I installed a stable debian release with 1.1.2, would 1.1.3 ever be released in the official stable repository, or would I have to add it manually?

Re:As a Debian user myself...

[Technonotice_Dom]

I believe that they fix RC bugs in packages, but otherwise it's frozen. You'd have to apt-pin that package through.

Re:As a Debian user myself...

[Unit3]

Testing is rarely usable. In fact, it's often more broken than Sid, or at least the broken state lasts longer with Testing. YMMV.

I'd say it varies quite a bit. I've used testing regularly for the past 4 years on multiple desktops and servers, and never had any major problems. I can't even say the same thing about the Mandrake, Redhat, and SuSE boxes I've had to administer, when even in their stable versions they seem to be quite a bit more buggy than Debian/testing.

Re:As a Debian user myself...

[cuerty]

They will patch the bug in OO (for this example) but mantain the 1.1.2 release. That way no new features are implemented but the bug es fixed.
I'm not saying that that's bad. I dislike it so I use other distro, thats all.

Re:As a Debian user myself...

[ultrabot]

What has broken for you lately?

Dependencies between packages have been broken - I've been unable to install various packages because their deps have not been available. This rarely happens with Sid, which is what I'm running on my "server" role machine.

Re:As a Debian user myself...

[SeaGK]

Parent should be modded up. That is exactly the way Debian should be used.

Stable == Servers (we Have 6 of them runnig our basic infrastructure: LDAP, PostgreSQL, /home over NFS, Apache, BIND, DHCP, apt-proxy etc)

Testing == Business workstations (we have 40+ with a hand selected set of packages .. including a couple from unstable like Evolution 2.0.2)

Unstable/Experimental == Home or testing machines at work.

Debian testing probably beats the heck out of most distros in stability ... We've been on Debian for almost a year now, and we are very happy with it.

Re:As a Debian user myself...

[gnalle]

I am using Woody at work (physics Ph.D.). Most people at the institution use programs like emacs, (which doesn't evolve so fast anymore) and Matlab (which installs nicely), so we don't lose much by using debian stable.

My only grudge is that I am getting a little tired of icewm. I mean it's a really fast wm, and the gui is clean, but I use GNOME 2.6 at home and I like all the bells and whizzles.

The problem with Debian is they set no deadlines

[synthespian]

Debian dosn't follow the common software engineering practice of setting deadlines for relase cycles. They claim this is because it relies on volunteers. But this is "no excuse", because OpenBSD and FreeBSD rely on volunteers. OpenBSD has always had 6 months release cycles and this is what Scott Long (From FreeBSD) has said recently:
"By the middle of 2002 is was very apparent that we needed to start focusing on getting 5.0 released. Unfortunately, we fell into the trap of wanting to finish more features in order to feel good about 5.x. (...) New -STABLE branched will be made on a calendar-based time line (...) While as engineers we all tend to hate timelines, this does have a lot of positive aspects. First, it increases the predictability of the development both for our users and for our developers. Users can plan effectively for upgrades and testing/validation knowing that there will be major and minor releases at fixed times of the year. Developers can judge when to start new projects and when to focus on bug-fixing because there will no longer be the temptation to delay a release by a month in order to slide 'one more thing' in. This is not unlike most commercial OS vendors, and we've received a _LOT_ of feedback that this method of planning is desperately needed.
Link for the above here
These practices should be adopted in Debian too. If BSD development can do it, and it involves kernel developments, Debian can do it too (mostly userland hacks).

Gnome's display manager GDM IS BROKEN!!!

[urbieta]

I guess I should learn how to submit bugs, since i have not done it; but there are complaints all over the forums.

I run multiuser ltsp.org 4.1 with mandrkake 10.1 and after a couple days of operation GDM 2.6 gives me a gray screen on all new display managers requested, gdm-restart and gdm-safe-restart simply don't fix the problem, it just worsens and to fix it I must restart the whole server!!!!

ltsp's author James McQuillan simply opted to use KDM, and that's too bad since it's not as themeable as GDM, heres an example:
http://www.themedepot.org/showarea.php4?area=11

I stick to GDM because it correctly updates wtmp and because of that the '# last' command works correctly with multiple users, it's very important for me to monitor computer use, so a little stability would be nice as well.

Re:Gnome's display manager GDM IS BROKEN!!!

[Daengbo]

I thought that Jim's main reason for using KDM for everything was automatic login for different users across different machines, which GDM isn't capable of doing.
BTW, I used GDM from Gnome 2.6 on a network of about 8 computers and never experienced the problem that you are having, despite being on for weeks/months. How's your .conf file look?

Re:Gnome's display manager GDM IS BROKEN!!!

[Anonymous Coward]

I have to disagree with you.

We've been using GDM in a LTSP environment for years and I've never seen the problem you describe. We have no issues with GDM at all. It just works. (currently running 2.6)

Re:Gnome's display manager GDM IS BROKEN!!!

[urbieta]

Autologin does work in GDM:

http://members.porchlight.ca/charm/wclp/enable-gdm .html#autologin

A workarround to THE bug is documented here:

http://www.k12ltsp.org/phpwiki/index.php/WrkArndGd mBug

I will play with the mentioned bug, maybe a couple configurations will make me breathe more calmly when Im away from the lan heh

Just ask Jim at irc, he's often logged in there, instead of offering help to fix GDM, the solution is always use KDM and get on with life.

Id be happy to use KDM only if it updated wtmp correctly, I even tried applying this:

http://www.ltsp.org/contrib/utmp_updating.txt

But it never worked for me.

Someday, X terminals will be perfect, but for now, Im still struggling :S

Re:Gnome's display manager GDM IS BROKEN!!!

[Daengbo]

Autologin does work in GDM:
You misread me. I said different users across different machines. For example:

• ws001: login as user001
• ws002: login as user002

...
I have been with LTSP since Jim was using XDM (maybe close to five years?), so I've talked to him opn IRC before. I think he just recommends KDM because that's what he always uses and is much more flexible.
Anyway, you stated that this was Gnome 2.6, right? This message indicates that the bug you're referring should've been fixed since June 2003, so I'm confused.
I'm not saying that you are wrong, because I've been out of thin terminals for a few months and I don't often read my LTSP mailing list, these days, but I'll say again that I never had that problem with Gnome 2.6.
What are you running that you get this? Email me off-line, and maybe I can make some recommendations. If not, just implement the workaround as you referenced.
If you really are having serious problems, just drop in K12LTSP on the server, and you should be ready to run. Eric does a wonderful job with that distro , and has for years. Thin-clients out of the box!

Re:Gnome's display manager GDM IS BROKEN!!!

[Daengbo]

BTW, Mandrake likes to patch systems, so maybe you've got a problem there. Any other distros with the same issue?

Re:The problem with Debian is they set no deadline

[BenjyD]

Both GNOME and Ubuntu have switched to this approach too. It seems a good idea with volunteer projects to set solid deadlines like that.

Re:The problem with Debian is they set no deadline

[cmbofh]

KDE does it in a similar way, too, and has for a long time:

Release Schedules / Development Plans

Don't forget MySQL

[lamber45]

I'm runnig Debian stable on a server, but the current version of MySQL is 3.23.49, so I installed the official 4.1-gamma binary as soon as it was released. I'll probably install PHP 5 and MySQL 5 as soon as they're both considered stable by their respective developers. A distribution is a base system, but it won't necessarily have everything needed for a specific application.

BTW, I recently had a less-than-stellar experience with RedHat Enterprise WS; I ended up downloading a lot of packages from Debian and converting them with Alien just to get decent functionality.

--
DLL

Software vs. Security

[boldi]

Stable is for anyone who needs security support.
Stable is for people who need an unchanging system.

What to do if I both want security support and less-than-three-years-old stuff in my box?

At least, an 'alert' list for 'testing' security related holes should be available, checking bugtraq twice a day and trying to analyze wether my system is affected is _not_ a long-term solution.

Re:Software vs. Security

[MasterOfMagic]

debian-security-announce tracks these issues. Most of the time, if there's a problem with the version in unstable, it affects the (earlier or same) version in testing, and you know about it.

Re: no good alternative

[digitalride]

I had the same problem of debian stable being too stale, testing not including security updates, and unstable changing too much. I started a company http://opensensesolutions.com/ and we sell linux machines that use a frozen repository of debian testing with key security updates and bug fixes added on. Everything is designed for a select subset of hardware to make testing manageable.

You cannot prove the nonexistence of something

[Anonymous Coward]

I have to disagree.

It is mathematically impossible to prove the nonexistence of something. You can only prove the existence of something.

Is therefor impossible to prove that a software is bugfree. You can only prove that a certain bug exists.

Re:You cannot prove the nonexistence of something

[ssj_195]

It is mathematically impossible to prove the nonexistence of something. You can only prove the existence of something.

I can prove that there are no integral solutions to x^n + y ^n = z ^ n for n > 2. Sadly, this post is too small to contain it...

Re:You cannot prove the nonexistence of something

[E_elven]

My programming language has two operators: dowhatiwant and nonotthatyouidiot.

I assert that operator+ does not exist in my programming language. See above for proof.

It is possible to prove the nonexistence of something, but it requires that the entire context is known. For real-world problems (like does God exist) the entire context is not known. One problem, of course, is knowing at which point the entire context is known :)

Re:You cannot prove the nonexistence of something

[Anonymous Coward]

Well, I can probe that there is not such a natural number a=2*b so a=2*c+1, being b and c natural numbers also.

I would be able to do it not only on a Slashdot message but even on a short book margin too.

(For those retard enough to see it, it is the demonstration that an even number can never be an even number plus one, since an even number plus one is always an odd number).

Your sig (Re:good for debian, good for users)

[FuzzyBad-Mofo]

Array

I'm curious, does your sig somehow trick Slashcode into interpreting it as an array, or did you just type 'Array' there? :)

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[drinkypoo]

So what are we end-users to do when a debian maintainer is unresponsive to us and to the programmers who create the apps we love?

Opinions are like assholes: Everyone has one, and no one wants to look at the other guy's. This is why it's such a good thing that we have so many distributions. When someone tells me they need a Linux-based server, I send them to debian. Any other use, and I send them to gentoo. If you want up-to-date packages, gentoo should be your first choice. Debian is for people who need stability over all else. If you need something in between, well, that's what /usr/local is for. Build subversion yourself.

Re:The problem with Debian is they set no deadline

[metamatic]

Right, the way all software development should be done is to make frequent regular releases with some internal meaningless version number.

Then at some interval determined by market requirements, marketing approach you and say "We need another release to call 5.0". You take a look at your recent internal releases, pick one that has a good blend of reliability and features, and hand them that to label "5.0" and give to QA. If QA fail it, you can pick another one to try. If QA pass it, it gets a "5.0" sticker put on it and it's shipped to users.

Re:WTF is Sarge?

[metamatic]

I *am* using Debian and I can never remember what the stupid code names refer to. When I read the headline I had no idea whether it meant that GNOME was going into stable, unstable, or what.

Then again, since tried GNOME and switched to KDE for performance and stability reasons, I also didn't give a flying one...

Debian w/ 2.8 before Mandrake?

[bach37]

Whoa this is crazy! DEBIAN has Gnome 2.8 in one of their releases before Mandrake? Come on, Mandrake folks- give us Gnome 2.8! Puh-leese! :)

Re:Debian w/ 2.8 before Mandrake?

[narrowhouse]

Gnome 2.8 is in Mandrake cooker if that helps :). I expect that soon Debian will reform the way it does things, it was one thing when they fell behind the corporately driven SUSE, Redhat, Mandrake, etc. but now that there are several "Debian based" distros that are trying to keep packages up to date Debian risks losing developers to faster moving community driven distros like Mepis and Ubuntu.

Re:WTF is Sarge?

[Anonymous Coward]

The codenames are references to Pixar's "Toy Story". Hence "Woody", "Sarge", "Sid", "Potatoe" ... (oops, lapsed into Dan Quayle mode there) ;-)

Free Ubuntu CDs?

[Anonymous Coward]

Speaking of Ubuntu, they were supposed to send out free CDs of Ubuntu by the end of October (when Warty Warthog or whatever was released).

Did anyone ever get theirs? Or does anyone know anyone in the company who can say whether they'll ever send, and if so when?

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[myowntrueself]

"Opinions are like assholes: Everyone has one, and no one wants to look at the other guy's."

So... a philosopher would kinda be like... a proctologist?

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[grumbel]

Rule of thumb is really to not use Debian stable on system where you acutally want to work on. Debian stable is worth thinking about if you are running a bunch of servers or desktop systems and want to keep them in sync or a router which should only run one or two programms, but even then having a unstable testsystem and rsync might do a better job, since you can't really use a stable without inserting a bunch of unofficial and self backported packages first which in turn make 'stable' not so much 'stable' any more.

Debian stable has a long long history of being full of horrible outdated packages, the pure stable user is still running Gimp1.2 and Kde2.2 after all. I personally have simply given up on Debian stable, it wasn't much useable in the past and probally won't much useable in the future. Hopefully Ubuntu will do a better job at providing a system that I can actually recomment to people, but Debian stable, a few isolated cases asside, is really not much usefull.

The worst of all is really that Debian is completly working against their target audience, sure there are a handfull of people which are using Debian stable and where it makes sense, but I guess there are a hell of a lot more people which are using testing and thus don't get any security updates much in time. The current Debian release cycle does really not work much at all, and simply redirecting people to testing or even unstable is not the solution its the PROBLEM.

WTF is Sarge?

[mad.frog]

For those of us who are geeks but not linux users:

someone want to define (or at least provide a link) to wtf "sarge" is in this context?

Yes, I could google for it, but (1) I'm too lazy and (2) that's really the editor's job, isn't it?

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[thomasweber]

Rule of thumb is really to not use Debian stable on system where you acutally want to work on

Funny. I use it on my desktop at home exclusively (as does my girlfriend). You know, I have a life besides computers and don't want to update my system every now and then.

(Recently switched from StarOffice 5.2 to 7.0, because it handles .doc better - that and a recent Opera are probably the only not-outdated packages on the system.)

Re:WTF is Sarge?

[alc6379]

WOW! That's lazy! In a nutshell, there are 3 branches to the Debian Linux distribution:

Stable, aka "Woody"
Testing, aka "Sarge"
Unstable, aka "Sid"

Sarge is about to take Woody's place as the Stable distribution, basically.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:WTF is Sarge?

[toganet]

How can you call yourself a geek and not know that?

How to go from Sarge (testing) to Sarge (stable)?

[sublimespot]

I am using Sarge (testing). When it is released as (stable), Im wondering if I will auto-jump to sid? I ask because in my apt sources.list I have apt sources of "testing".

testing not recommended for security

[Michael Wardle]

Check it a little closer and you'll notice that that distribution is empty.

My understanding is that the testing-security package designation is seldom used, instead the packages are uploaded to the unstable distribution, presumably with a priority of high, which means it should enter testing two days later, depending on a few factors which you can read about on How Testing Works. In a sense, therefore, unstable is in some ways more secure than testing.

Also note that the testing distribution is not officially supported by the security team and that use of testing on a server for this reason is not recommended.

Re:How to go from Sarge (testing) to Sarge (stable

[GrimReality]

I am using Sarge (testing). When it is released as (stable), Im wondering if I will auto-jump to sid? I ask because in my apt sources.list I have apt sources of "testing".

Don't worry about that, because, the next testing won't be Sid, but Sarge itself. In other words, a particular testing version begins its life as a copy of the stable version. [For an illustration, check the following quote from the Debian Web pages describing the current testing distribution (which is to be called 'sarge'):]

This release started as a copy of woody, and is currently in a state called 'testing'.

But, just don't expect it to stay that way for long, because, I think, they remove the freeze on packages moving in from unstable.

Re:How to go from Sarge (testing) to Sarge (stable

[Anonymous Coward]

Just replace "testing" with "sarge" and everything will be fine

Testing and boundary exceptions

[oo_waratah]

Valgrind on X86 is a tool that will monitor access just as you described and provide this feedback. It simply hooks itself in at runtime so no compile needed (much more info if compiled -g of course)

Regarding testing, yes you cannot ever prove that a bug does not exist but we should always try.

Re:Sarge and Subversion 1.1.x Branch Fiasco Contin

[Anonymous Coward]

So what are we end-users to do when a debian maintainer is unresponsive to us and to the programmers who create the apps we love?

There's only one way to get a package from unstable into testing: Fix RC bugs.

Submit the patches to the bug tracking system (BTS, http://bugs.debian.org/). Find a sponsor to make a non-maintainer upload (NMU). Participate in the next bug squashing party, maybe you'll find a sponsor there.

Come on, this is Free Software For Beginners. Quit bitching about the problem, that doesn't help anyone. If you're capable, fix it yourself. If you're not, pay (or otherwise motivate) a capable person to do so. Otherwise, you still get more than you've paid for.

If all else fails, build the package from unstable on a stable system or in a stable chroot and use it yourself.

Debian Release Schedule

[Anonymous Coward]

If this is any indication, we'll be lucky to see sarge in 2006.

Re:WTF is Sarge?

[Anonymous Coward]

Sarge is the testing version of Debian GNU/Linux.
It is very stable but not as stable as Woody (Debian stable).
And I for one have not had one single problem with broken packages in it.

(See RPM hell and .deb)

I hate Ubuntu

[thegnu]

This is not to say it's bad, but I had a horrible experience installing a Samsung ML-1740. I tried:

1. Using gnome-cups-manager
2. Enabling the root account and using the installer off the CD, it wouldn't authenticate.
3. Installing as an ML-1710 and replacing the ppd with the 1740 ppd.
4. Using the CUPS interface, which is disabled.

This took me a good 3 hours to exhaust all my possibilities before switching to Fedora Core 3. RPM hell be damned, Ubuntu hell is worse.