Linux kernel getting buggierhttp://news.zdnet.co.uk/0,39020330,39267255,00.htm
http://kernelslacker.livejournal.com/37851.html
Pazite devojčice i dečaci...ovo su mainteneri produkcijskog kernela i Fedore!
Da vidimo što tu ima:
Citat:
"I believe the 2.6 kernel is slowly getting buggier. It seems we're adding bugs at a higher rate than we're fixing them," Morton said, in a talk at the LinuxTag conference in Wiesbaden, Germany, on Friday.
Što bi s onim "sve će to narod pozlatiti":
Citat:
One problem is that few developers are motivated to work on bugs, according to Morton. This is particularly a problem for bugs that affect old computers or peripherals, as kernel developers working for corporations don't tend to care about out-of-date hardware, he said. Nowadays, many kernel developers are employed by IT companies, such as hardware manufacturers, which can cause problems as they can mainly be motivated by self-interest.
"If you're a company that employs a kernel maintainer, you don't have an interest in working on a five-year-old peripheral that no one is selling any more. I can understand that, but it is a problem as people are still using that hardware. The presence of that bug affects the whole kernel process, and can hold up the kernel — as there are bugs, but no one is fixing them," said Morton.
"If you're a company that employs a kernel maintainer, you don't have an interest in working on a five-year-old peripheral that no one is selling any more. I can understand that, but it is a problem as people are still using that hardware. The presence of that bug affects the whole kernel process, and can hold up the kernel — as there are bugs, but no one is fixing them," said Morton.
Ha? Pokvarenu korporativnu developersku gamad boli neka stvar za tamo neki opskurni hardver na kojem ne mlate lovu...gdje je su ti "milijuni" dobrovoljaca da uskoče i otklone bugove? :)
O kvaliteti koda ovih par naivaca i koji šalju svoje ponosne "uratke" i nadaju se da će postati slashdot heroji ako im, ne daj bože, kod završi u glavnom stablu:
Citat:
The mm tree is what Linus' tree is going to look like in three months time. A lot of stupid bugs get in. I wish people would send me code that compiles — probably about 75 percent do," he said.
LOL!
Da vidimo što Dave kaže, kako je to u enterparadajz RHAL okruženju:
Citat:
The rebase to 2.6.14 took us from 356 open bugs to 376 in the space of a week. Another week later, up to 388. At the time, focus was on getting FC5 in a releasable state, so FC4 bugs were somewhat neglected at this time. By the time we got around to rebasing to 2.6.15, FC4 was at 488 bugs. A week or so later after the rebase, we're up into the 500's!
Whooo, koje cifre! Lakše malo Dave, zna se da open-source programeri ispravljaju bugove u "roku nekoliko minuta"!
Malo o (ne)kvaliteti bug reporting alatića:
Citat:
Like all poor workmen though, I blame the tools in this case. Bugzilla *really* sucks for forwarding information around. The best approach I've found is to forward the mail I get from bugzilla to the relevant maintainers, however in the cases where it only really becomes clear whats going on after multiple comments, it's easy to lose context, and not have all the right bits in place to forward, putting you back to asking people "hey, look at http://...". The real pain point appears when the upstream maintainer needs more info. Sometimes they'll be real rock stars and create a bugzilla account and just ask them themselves. Other times, I get to play middle-man, which goes something like..
maintainer -> me -> bugzilla -> user -> bugzilla -> me -> maintainer.
This quickly falls apart (see the note about losing context above) unless you're paying attention to bugzilla 24/7, have a huge memory for details, and aren't swamped with 1000 open bugs all vying for your attention at the same time.
maintainer -> me -> bugzilla -> user -> bugzilla -> me -> maintainer.
This quickly falls apart (see the note about losing context above) unless you're paying attention to bugzilla 24/7, have a huge memory for details, and aren't swamped with 1000 open bugs all vying for your attention at the same time.
O fragmentaciji developmenta i njenom utjecanju na overall kvalitetu distribucije upstream/downstream patcheva (čitaj: ne zna se tko pije tko plaća):
Citat:
On top of all of this, this is just one source of bug reports. Being a sucker for punishment, I read bugmail from several other distro's bugtrackers. It has value for a number of reasons.
(pusta pravdanja...)
All this adds up to a ton of bugmail, and I could easily spend a whole day doing nothing but triaging bug reports. During all this time, I rarely get a chance to look through upstream kernel bugzilla. When that first came about, I jumped in with a few others and stuff was working out pretty well, bugs got squished, dupes got closed, more info was asked for etc. I get a mail for every single comment that gets entered in that bugzilla, yet after spending all day fighting distro bugzillas, I never seem to have the motivation to fight one more, especially as it's a) gotten out of hand somewhat (1228 open bugs right now, some dating all the way back to 2.6.1 -- by comparison, Fedora bugzilla has 681 open bugs right now, against a base-level of 2.6.16 (a bunch are earlier bugs that have been requested to retest on.
(pusta pravdanja...)
All this adds up to a ton of bugmail, and I could easily spend a whole day doing nothing but triaging bug reports. During all this time, I rarely get a chance to look through upstream kernel bugzilla. When that first came about, I jumped in with a few others and stuff was working out pretty well, bugs got squished, dupes got closed, more info was asked for etc. I get a mail for every single comment that gets entered in that bugzilla, yet after spending all day fighting distro bugzillas, I never seem to have the motivation to fight one more, especially as it's a) gotten out of hand somewhat (1228 open bugs right now, some dating all the way back to 2.6.1 -- by comparison, Fedora bugzilla has 681 open bugs right now, against a base-level of 2.6.16 (a bunch are earlier bugs that have been requested to retest on.
I konačno...zaključak:
Citat:
There is an endless supply of bugs, and not a huge number of people jumping on them.I'm not convinced that the reason for this is because everyone is too interested in adding new features. I think a more likely explanation is just the fact that we're getting better at finding bugs, and getting better (and more) reports from end-users. The problem is one of being completely overwhelmed. When I look at Fedora bugzilla at the start of a day, I honestly don't know where to begin.
Low hanging fruit such as "you forgot to enable this config option doofus" is a quick fix, so that kind of thing gets picked off quickly. The nastier bug reports such as slab corruption bugs seem to linger for a long time, eventually being closed out with "haven't seen this in a while, I'll close it", which gives me the creeps. We're changing *thousands* of lines of code each week, a lot of the nasty bugs get dismissed (by their reporters in most cases) without explanation as to what actually fixed the problem.
Low hanging fruit such as "you forgot to enable this config option doofus" is a quick fix, so that kind of thing gets picked off quickly. The nastier bug reports such as slab corruption bugs seem to linger for a long time, eventually being closed out with "haven't seen this in a while, I'll close it", which gives me the creeps. We're changing *thousands* of lines of code each week, a lot of the nasty bugs get dismissed (by their reporters in most cases) without explanation as to what actually fixed the problem.
Brr...čini se da je situacija poprilično horidna.
A kad se tek sjetim Linusovih bisera o intefejsima i "nepotrebnosti" stabilnog kernel ABI-ja... :) To bi toliko trivijalnih bugova riješilo, a sve što treba jest pravilan softver inženjering...
[Ovu poruku je menjao cynique dana 08.05.2006. u 20:38 GMT+1]
Re: Linux kernel getting buggier