Slashdot Mirror
Kernel Feature freeze in 2 weeks?
According to an email written by Linus, "a feature freeze in about two weeks is the current plan", so people who wants their patch included (a hint to ISDN dev. guys) should work/write faster. I guess kernel 2.4 may be out this fall after all.
The easy solution might be that 'fall' is a lot
easier to spell than 'a*m'.
I don't know about SuSE or RedHat, but Debian tends to release releases when they're ready, which is usually after enough new features have accumulated (eg, new libc, new X, ...) The kernel is only one part of the system, there's no reason to make a special release just for it. People who want the new version can install it themselves, it's not that hard..and I suspect that if any packages need to be upgraded a 2.4-compatability APTable area will be set up. No big problem.
Daniel
KOffice....
He was talking about 2.2 not 2.3
Cool, I've got one too... would you mind emailing me, I've been having some problems with mine, especially the clock, getting it to turn off, and getting my g200 working
Michael
mjb@resnet.gatech.edu
Excuse me, that's a just a myth amongst some Linux users: There are other operation systems besides Linux and Windows, you know. And compared to many of them Linux doesn't shine that bright.
Stable? Yes, 2.0.3x is/was quite stable. But for sure not as stable to qualify it "legendary".
Ok, I know you guys are all hyped about new features and what not.. but I happened to be playing with some older hardware, and added in support for a really old xt controller card... (from 1987!) I was wondering what is the proper procedure for like maybe submitting it?
And?
Many thousand ISDN users have used the isdn code from ftp.suse.com for a _long_ time. The code in the main kernel is outdated, buggy, and supports fewer cards than the current i4l code.
If you tell someone to not "to futz around any more," I'd call that bitching. To use the analogy of the 1st poster: It's like a manager who says to a working group: "If you want to get your features in, work 48 hours a day."
Face it, Linus is not perfect. I also wouldn't call that good management style.
Don't forget that XFree86 4.0 is trying to solidify. Adding kernel 2.4 plus XFree86 4.0 might be enough to justify a whole digit change. For that matter, what is the KDE 2.0 release date? I don't believe GNOME is near a 2.0, but a point release on it is getting close.
Why did you have to do that? Now we have to track you down and mind-wipe you, along with everybody else who read that post.
Remember: There is no alien invasion. There is no alien invasion. There is no . . .
It's about 7 months away here...
Alan Cox is doing the 2.2.11pre series from pre2 on upward (current is pre4) so you'll need to check in his directory on ftp.xx.kernel.org (/pub/linux/kernel/people/alan).
There's something special about a brand new road...
Fall is an American phrase. Hence, autumn in America. Which is in about 3 months.
Autumn is ONE month away (Starts in September AFAIK)
I can't wait for it. My /dev is a mess!!!
Would be helpful on our 40GIG+ HD's...
Also what's the main repisotory for DVD work? we have been doing some of our driver work and would like to merge with what ever else is out there
--
JC
Do you mean this ?
Sometimes I wonder if Linux Torvalds could scratch his nose without having the event posted on /.
The best page for ATA66 patches:
h tml
http://www.linuxhelp.org/linuxide/
The best page (the only page) for pset patches:
http://isunix.it.ilstu.edu/~thockin/pset/
The up to date mirror for the SMP FAQ that
will tell you what pset is and why its a good
thing:
http://www.phy.duke.edu/brahma/smp-faq/smp-faq.
I have been running quite stable on my overclocked
dual celeron with the pset patches on 2.2.5. I
wrote the author and he said he would produce
2.2.10 patches fairly shortly and is taking
alpha testers.
You heard it here first.
*start singing of tune gerschwin song*
hey was it gerschwin who wrote this song? autumn leafs. well one thing for sure, it's not a british song.
In fact, it's a common phenomenon that emigrants are much more conservative in language issues, because they often "freeze" language changes at the point in time where they leave their country.
I've heard examples of that, with emigrants that emigrated from a country in their early teens, and that are now close to 100 years old, and that still speak their mother tongue perfectly, with the minor issue that they use words and sentence structure that was common around the turn of the century...
That phenomenon can in many cases be used to trace the usage of words in certain periods of time, because a word that is common among emigrants from a country are likely to have been common in the country they emigrated from at the time they emigrated, even generations later.
In the case of the US, you can't do that safely, because as the population in the US exploded, the British emigrants at a point weren't the largest English speaking part of the population anymore, and the language started changing more again.
Also, much of the reason that British English and US English hasn't become even more different is of course that the amount of communication between countries has been steadily increasing, and thus more and more of the changes influence that occur find their way to both places.
With regards to antialiased fonts etc., I agree with you, though.
And in general, I find X to be way to bloated.
When it comes to performance, that's VERY dependent on the GUI toolkit.
One thing I'd like to see though, was partial server side GUI toolkits as a X protocol extension. Being able to handle extremly simple basic widgets on the server side has the potential of removing lots of network traffic...
For AIX, it would be close to impossible, because there are so many factors that are dependent on other software, hardware configurations, timings, etc. in any OS. You could certainly benefit from proof for small critical subsystems, though.
A 2.2 machine still doesn't reliably read an exported directory from a 2.0 machine, lots of timeouts and stuff.
Maybe 2.4 should just be 2.2, without the bugs.
It's a Brittish word. Autumn is an American (or at least non-Brittish) word taken from French.
Bend over and let me hit you with a clue stick.
If you want stuff faster, help. Otherwise, quit your bitchin'. Your cute little comment makes you sound like a pointy-haired manager who thinks yelling at the development team will lead to faster development.
Here's a hint: Most of the effort expended in software development is THINKING. People don't THINK faster just because you tell them to.
Sorry to flame, that just made me mad.
even siller (sillies?) (wo)man,
.. cough.. patents.. I just Copyrighted the word patent and would sue anyone who uses this word without my permission. Thank you. Please forward all mail to my secetary, she's female. I am not a sexist.
There is no such thing as a Transmeta Prime. I can vouche for it.. seriously. So where did they come from? Best guess is from somewhere under the Atlantic Ocean.
And about these
I only have two things on my wish list for 2.4:
The ATA66 support that only exists in patches
to 2.3.12. The pset support exists only in
patches to 2.2.5 right now and allows you to
bind processes to specific processors, but is
still highly experimental. I have the Abit BP6,
can you tell?
Actually I would be happy to see both of these
fixes in even an experimental kernel (2.3.*).
The newbies are forgetting how long it took to get from 0.99 to 1.0 (anybody remember Ygdrassil, and 0.99 patchlevel 14-R?)
I know the last post was in jest, but it seemed just as nice a thread as any to reply to :-).
I've already seen a few threads starting up saying "Bah, american snobs." I have news for you people, Linus is in the Northern Hemisphere last time I checked. He's going to announce something for the "fall" or "autumn" season, he's talking about the hemisphere he grew up/lives in.
Can we stop this paranoid "Oh maybe to you american snobs" bullshit? Sure, some americans are patriotic to the point of being arrogant to the extreme. But people aren't trying to be arrogant here -- the only arrogant people are the ones seeing conspiracies behind every corner!
The non-american AC who's sick of crybabies. Shaddap already, eesh!
Here comes Redhat 7.0! ;>
Yes, but what is more embarassing.. having your fresh new untested kernel go down or your boxers? :-)
So the last major update took 2 1/2 years... Yeah, that's a long time, but there were TONS of changes... 2.2 is far ahead of 2.0... 2.4 won't be quite as big a leap. I would like to see PnP, USB, TCP/IP fixes all working before a 2.4 though....
They only thing I can think of, is that maybe they concentrating on stability now... add a few features, stabilize, lather, rinse, repeat... It could also be that they're saving 3.0 for the major changes.. whatever those might be.
Man.. I still know of boxes running on 2.0.35 (with currently 309 days uptime)... I started on 1.3.something... Our little baby's all grown up!
Wow, warm boxers! That sounds like a great idea....hmmm....
-David
I don't think you understand the term 'feature freeze' very well. The whole point is that nothing new gets added so it can be better stabilised.
Support for a full, robust journalling filesystem would be worthy of a major number increment (Linux 3.0). Filesystems (especially modern journalling ones) are *very* complicated and *very* fundamental to the system. If your filesystem only works correctly 99.9% of the time, your system is *wholly* unusable.
The changes that you mention are in the planning stages, but will take a very long time to implement and an equally long time before they are deemed "stable". 2.4 is supposed to be a quick cycle, so such changes do not fit in its scope.
As for what *is* new in 2.4, Pranevich put together a nice piece called the "Wonderful World of Linux 2.4" (just as he did for Linux 2.2). It is on Linux Today, and I suggest you check it out.
--Lenny
I wish I had one of those Neuralizer's From Men in Black. I'd flash my Boss.
"Huh? You won't give me raise?!?
Uh, yeah, I'll take that raise and the Promotion. Oh? and we're not installing NT?? Linux, Firm-wide you say???
Do not read this
While we're on the subject of badly needed pronouns, I'd like to add an addendum for the legitimization of "ain't" as a contraction of "am not." We've got second and third person, singular and plural, covered in all applicable tenses. It's time we spoke out for #1.
I'm mad as hell, and I ain't going to take it any more!
A few questions about kernel development:
;)
a) when do they decide that the next version should be for instance 2.4 instead of 3.0?
b) what constitutes a feature freeze? When everything is considered stable? if everything is stable, why do they keep adding new versions that must be debugged?
Just curious, Ive always wondered things like this
-Dave
--
Dave Brooks (db@amorphous.org)
http://www.amorphous.org
When you're -truly- speaking globally, why don't you just say 'In Q4'? The fourth Quarter. Then there's -never- any confusion with seasons. Whoever thought of using seasons as a timeframe should be shot. Seasons are for -farmers-.
Comics:
Sluggy.com - It rocks my nads.
Schlock Mercenary.
Depending on which of the Big Two dictionaries you ask, that *has* been a valid contraction for a few years now.
--GAck
3. Profit!
2. ???
1. On Soviet Slashdot, a Beowulf cluster of alien Natalie Portman overlords welcomes YOU!
What, in particular is wrong with X? I use even some of its more obscure features on a regular basis. What I would like to see in kernel graphics support is KGI, with good device support, and then XGGI on the userspace side, but that probably won't happen until GGI is ready.
Why? Because they cant program.
Well, in that case, please show us exactly how its done, Mr AC. Obviously you know what you're talking about, don't you.
silly man,
:-)
everybody knows that transmeta is just a front for the aliens. they are in constant contact with the motherbase back on Transmeta Prime, and are busily building a spacetime portal in the basement. You laugh now, and engage in idle speculation about these "patents", but see who's laughing when they pour through their portal and enslave us all.
--sam
--sam
Any technology distinguishable from magic is insufficiently advanced.
Looks like Linus learned from the last development cycle, where it took 2 1/2 years to get a new stable version out the door.
Of all the comments I've ever posted, this is definately one of them
just wishful thinking, really. I've got a Yamaha PCI Soundcard, an XG-DG or some such...sounds GREAT in Window$, but the only support via Linux is thru OSS from 4front. So for now, I'm stuck with my SB16 PNP. Oh, anyway, I'd like to see more support for multimedia hardware. Like my card. Esp. my card (:
Northeast USA Computer Show Schedule
http://www.vermontel.com/~vengnce/shows
Actually, a feature freeze will probably help stability.
Instead of spending more time adding new features than on refining existing code, the kernel folks will spend more time on polishing the code, having stopped adding new features (the word "stop" being a bit relative).
That would give, at the very earliest, assuming that fall in California starts around the 20th of September, about a month of bugfixing. And that's a minimum.
It doesn't seem like such a hurry from where I sit. *shrug*
--
QDMerge -- data + templates = documents.
how to invest, a novice's guide
In short, people who think they have major requirements had better get their act together. That means that if ISDN people actually want to try to get into a real release one of these years, they don't have all that much time to futz around any more.
Hm. I don't know if Tove will let you get near him with a clue stick.
----------
mphall@cstone.nospam.net
----------
mphall@cstone.nospam.net
"A horse laugh is worth a thousand syllogisms"
Personally all the hardware on my computer is supported well (except for an old scanner I had). Unless there is a speed improvement or memory management change I will stick with 2.2.10!
---------------------------
^_^ smile death approaches.
Kernel 2.4
XFree 4.0
XFS
IA-64
Some days it's good to be alive.
--
I think KDE 2.0 is a long way off. When it gets here it should be really nice. With KParts and KOffice, it really sounds like something that will be very comprable to windows.
-matt
Uhm...that's the point of a feature freeze. No new features, only bug fixes. Btw, if you think that the kernel programmers can't code, I'd love to see what you can do.
-matt
A parallel question, does anyone have any idea when we're going to see 2.2.11? 2.2.11-pre1 is on the kernel mirrors, but I'm not keen on running a devel kernel on my system...
-Cheetah
gee there are some bugs in 2.2 that should be fixed first. I hope that they still support 2.2 and make improvements to it and do not just plan on stopping at 2.2.11 or 2.2.12. What is the rush for 2.4? why not just clean up 2.2 or is that not possible?
Only 'flamers' flame!
Sure, we want to look like we are developing quickly, but people will ask "What's new in 2.4?" What will we be able to say?
Of course, some people could make the argument that y'all is singular second person, and all y'all is plural second person.
I'm not one of those people. I'll stick with you and y'all
Fall doesn't start in California until about mid October :)
San Francisco values: compassion, tolerance, respect, intelligence
The OED has citations for both back to the 14th century, which kind of pre-dates any American English usage. It states that 'fall' was short for 'the fall of the leaf', which probably explains why people still say 'in the fall' rather than 'in fall', while 'in autumn' is fine.
Good thing he posted as AC, not many ppl would want to get all the flame mail from criticizing Linus.
/. readers have enough background to make the judgement that the ISDN ppl are or are not "futzing around" but from what's been presented isn't it about time that they synch their schedule up with the next kernel?
Since you're so big on logic... Isn't it sad that when someone says "that's a bad attitude" the first response was "well Mr. Authority said it!" and that's supposed to be a valid argument?
A management mistake is just that; no matter who makes it. Management may not be #1 but it matters.
I doubt very much that most
That's exactly my point and the post you were responding to was either not ad homonym or your complaint of ad homonym against the AC was off topic (and therefore feeding a troll).
Your post can reflect a value judgement on the point; you named an informal fallacy in an attempt to refute that point. By presuming that such a method of management is "good enough for our/my purposes" either you must either feel that a malliable attitude is wasted on your workforce (which is still poor management) or that you do know enough about the topic to judge the AC's criticism of Torvalds' email.
My point about the "Mr. Authority" comment was that since you're so concerned with logical form that you'll identify an ad homonym why didn't you identify this other Informal Fallacy? It was the entire point of the AC's post!
I believe the rest is self explanatory.
... and maybe the ISDN ppl were ignorant of the deadline. It would suck to expect 6 months to finish a project and all of a sudden get a memo from your pointy-haired boss to get your ass in gear for a code/feature freeze next week. So someone picked a date, and noone can push it back now can they? Yes there were rumors of a fall release. Does that make the probably unexpected feature freeze any more devistating to an unready development team? Off to patchland the ISDN users go!
Not that Linux development should ever suffer from the curse of middle management, but it helps to keep the general on his toes.
so...
QUESTION AUTHORITY
Just on the off-chance anyone cares,
and for comparison purposes,
a lot of people in Ireland,
when speaking "Irish-Dialect" English,
use "yez" or "youse" as the 2nd person plural.
Y'all is sometimes heard too.
Choice of masters is not freedom.
As soon as they realize that the next release isn't going to be revolutionary enough to deserve the new major number
what constitutes a feature freeze?
When all the features are in place. They freeze it so that they'll be able to work on getting things stable instead of putting in new things all the time. A feature freeze does not mean that things are stable.
"Oppression and harassment is a small price to pay to live in the land of the free." -- Montgomery Burns.
I bet Linus has some corporate intentions behind this sudden hurry. And I bet it has something to do with their mystery processor. And possibly something about Amiga too.
Anyway I hope all this hurrying isn't going to hurt stability...
"Oppression and harassment is a small price to pay to live in the land of the free." -- Montgomery Burns.
So is XFS or ext3 or ext2+b*tree going to make it in to 2.4? Is any of this stuff in the 2.3 kernels yet, I didn't think it was.
This is my signature. There are many signatures like it but this one is mine..
Is the USB support in the 2.4 kernel going to be 'complete' enough to allow the use of USB modems? I've noticed that they're the same price as serial modems or even slightly cheaper these days and I need a new modem. :)
Is the TCP/IP stack going to be rewritten to be multithreaded (finally) in 2.3.x/2.4? (I haven't been following the 2.3 releases at all... I miss my Ethernet.... when does school start again? :-) This is a badly needed feature in the kernel, and I can't believe that it didn't have it already...
Anyway, if somebody has the answer to this question please email me at spong@wave.harvard.edu.
Thanks
Matt
"Software is like sex- the best is for free"
Gee a couple of big things come to mind, such as:
ISDN (already mentioned)
DEVfs
Linux-Raid Patches
Uniform-IDE/UDMA Driver
I personally dont see much value to a new kernel until these are in........
--John C
Red Hat 6.0 is based on kernel 2.2.5, although I wouldn't be surprised to see a 6.1 release (on a 2.2.x kernel) after 2.4 is released. Red Hat 7.0 is still a long ways away.
For more information, click here.
hahaha
That sounds like the problem I have with floppies. Once in a while I need them to transfer some files to and from a laptop.
My personal hope is that some of the changes in the FS and SCSI layers for 2.4 will help out the stability problems. Currently I can't scan with my SCSI scanner for more than a handful of scans before the sg device or machine locks.
No, y'all is still first person; all y'all is second person. ;)
Suse 6.2 ships Aug 9th
Also see:
http://www.linuxtv.org/
suse 6.1 gives you the option of installing a 2.2 or a 2.0 kernel.
Take a look at my logic again, I didn't make any value judgement on that point at all, because *I* don't know enough about the issue. I did assert that the management is "good enough for our purposes," which is more truly "good enough for my purposes". I don't think that "mr. authority said it" is a valid argument either, but so what?
Someone had to pick a date, someone did. It's not like there's a date that would please everybody, is there?
I believe that more work should be going into stabilising 2.2.x. AFAIK Alan is still unhappy with the number of unexplained failures.
Linux's "legendary stability" must be top priority.
And what's this stuff about "kernel 2.4 will be after all at this fall"? Autumn is still nine months away.
Think global.
Buncha slackers, those volunteer developers... :-)
Eh? 1.2.15? Was that a typo, or should I be upgrading my 1.2.13 laptop?
Honestly! The aliens have been trying to teach humanity Transmeta-ese for generations, and yet there are some humans that still can't speak the language correctly. Motherbase, indeed! Where would the processor go, eh?
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
It's in Shakespere, so it must be ok. :)
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
I have never, ever heard anyone talk about the "autumn" season, in ANY context, in or around the United States, by ANYONE, other than visiting Britons.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
Although I generally say 'fall' more than 'autumn,' I do use both.
But then I use two different pronunciations of 'either' interchangably, and I strongly support the use of 'y'all' as the new standard 2nd person plural pronoun. Maybe I'm just a freak.
-- This and all my posts are in the public domain. I am a lawyer. I am not your lawyer, and this is not legal advice.
Don't forget, the last day of fall is Dec 20th. I'm guessing that this is mainly a call to get people serious about closing on 2.4, and it's easier to be serious if you're at least somewhat serious about a deadline. In this case, Linus is trying to give us 2.4 as an early Christmas present, it would appear (even though he hasn't come out and said so).
I'm sure that the kernel will go out when it's ready, and no sooner. Nonetheless, it's good to state some goals up-front, and a "limited development cycle" goal for 2.4 is probably a good idea. Saying "feature freeze" and "2.4 in the fall" at least gets people starting to think in that manner, rather than leaving it open-ended and sprawling, as it happened with Linux 2.2.
(Personally, I feel 2.0 should've been 1.4, and 2.2 should've been 2.0, in retrospect. At least the numbering, relative amounts of features and release dates are more consistent that way. Ahhh well....)
--Joe--
Program Intellivision!
Please pardon my bout of mental flatulence.
LATEST-IS-1.2.13
--Joe
--
Program Intellivision!
I still don't think that 2.4 will be out by fall.
Althought its only been a few months since 2.2 came out and 2.3 is already about to be frozen.
There was a lot of new features being worked
on and not merged with the main kernel months
before 2.2 came out when 2.1 was frozen, and were
shot into 2.3 as soon as it was open season.
-Omar
Omar El-Domeiri
I was planning my next machine to be on 2.2... It now looks like Linus, et al, may be finished with 2.4 before I even order any hardware.
Does anyone know if SuSE or Red Hat is planning another release based on 2.2?
Geeky modern art T-shirts
Feature freeze != code freeze. Even the best programmers occasionally have bugs in their code, and that's why a feature freeze is a Good Thing, as it lets the coders focus on fixing the bugs.
---
"'Is not a quine' is not a quine" is a quine.
"'Is not a quine' is not a quine" is a quine.
Quine "quine?
Running a fresh new kernel is like putting on a warm pair of boxers strait out of the drier.. Mmm....
Corndog
Thousands of Linux advocates and Transmeta watchers are debating the significance of a recent email sent from Linus Torvalds to Alan Cox on 8-3-99. The email read in part:
/. thousands have suggested that he is implying that Transmeta has created a magic invisibility device which Torvalds will use to battle the fiendish crime lord Erik "The Red" Blowhard. /. "I think that Linus didn't mean to cc. the kernel list on this one. He's spilled the beans on the whole Transmeta operation. Think about it. Invisibility suits. This rules. FIRST POST."
/.
Probably. I really think it's a matter of "if this device really doesn't
have any ordering constraints, then we can use the new nifty feature to
make it invisible to most users".
The big question is what did Linus mean by those fateful words. On
"Wow man!" Shouted flak in a post to
Anthropologists have postulated that Linus' mails receive so much attention because he is seen as a sort of divine "priest king" by his followers. "Unfortunately, this means that if the pace of development on the kernel ever slows, the crazed worshippers may sacrifice Linus to regain the favor of their gods. History shows this pattern occurring again and again." commented Dr. Rajeev Papshigali of the University of Utah. "If the sacrifice is unsuccessful, we may see Linux users losing their faith and joining other strange sects. Possibly they may even convert to one of the daemonology cults that originated at UC Berkeley."
An AC suggested that this was an out-of-context quote from a mail about standard pc bus architecture. This was quickly moderated down as "flame bait" as was another post wondering if Linus could scratch his nose without having the event posted on
--Shoeboy