[Dovecot] Dovecot Version Numbers - Let's drop the Alpha?
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Here's my reasons. The 0.9x version are obsolete and people should be using the 1.0 Alpha versions which seem to me to do everything the previous versions did and more. It's as full featured and stable as all other IMAP servers and in my opinion is ready to be called 1.0.
The "Alpha" label scares people off and with the 0.9x version being obsolete I think it creates confusion for new users. It created confusion for me when I converted to dovecot, and distros are less ,ikely to include versions with the Alpha label.
Once you go to 1.0 you can start a 1.1 alpha series and finish it up. I know Timo has high standards but as a marketing issue I think that the Alpha label needs to go to get people to accept and use 1.0 in production.
Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
Marc Perkel wrote:
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Actually, what I would like to see is to start a beta phase for dovecot 1.0, declare 1.0 feature freeze and only commit bug fixes to this branch from now on.
Timo's work can continue with a 1.1alpha version, in CVS tip.
After some time of beta status (which may encourage more people to upgrade from 0.99 to 1.0) Timo could make an offical 1.0 release and distributions would be very likely to catch up.
Early adopters and those who want new features immediately (like dbox) should continue to follow the 1.1 version as they see fit.
Things that should be fixed before making a 1.0 beta, IMHO:
- included plugins should actually work
Of course this is all up to Timo, so, I am very interested in his point of view. Timo?
Best Regards, Michael Paesold
For political reasons, I would like to see alphaX promoted to "1.0". My boss had a cow the other day when I told her we are running alpha code in production. My defense was "you hadn't noticed because it works." Or at least call alpha6 something like "beta1" instead.
Jeff Earickson Colby College
On Fri, 16 Dec 2005, Marc Perkel wrote:
Date: Fri, 16 Dec 2005 06:42:09 -0800 From: Marc Perkel marc@perkel.com To: dovecot@dovecot.org Subject: [Dovecot] Dovecot Version Numbers - Let's drop the Alpha?
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Here's my reasons. The 0.9x version are obsolete and people should be using the 1.0 Alpha versions which seem to me to do everything the previous versions did and more. It's as full featured and stable as all other IMAP servers and in my opinion is ready to be called 1.0.
The "Alpha" label scares people off and with the 0.9x version being obsolete I think it creates confusion for new users. It created confusion for me when I converted to dovecot, and distros are less ,ikely to include versions with the Alpha label.
Once you go to 1.0 you can start a 1.1 alpha series and finish it up. I know Timo has high standards but as a marketing issue I think that the Alpha label needs to go to get people to accept and use 1.0 in production.
Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
My experience is that the daemon performs much better than "alpha"
would ordinarily indicate.
I very nearly implemented 0.9x, which would have, in retrospect, been
unfortunate. The only reasons I implemented dovecot at all initially
(what with the Alpha label and all) are twofold:
- I was ******very****** desperate for a imap server with indexing
- Redhat packaged it in RHEL
All the same, it's not quite done yet, and if those issues may
require an incompatible redesign of some section of the daemon, then
Alpha should remain, IMHO.
If the likelyhood of an incompatible rework is nil, then I vote to
move it on up to Beta. When the release time comes, we could have
some "Release Canidate" versions, and it would all be politically
easier to sell, and get new users on board. It would really, really
suck to be in position (like I am) where *I need this daemon*, but
had to weather major static over a label issue.
Anyway, that's my .02 for Timo and crew to duly consider :)
Also, my mail server & clients & I say 1024**4 thanks for the great
work!!
-CedricOn 16-Dec-05, at 10:12 AM, Jeff A. Earickson wrote:
For political reasons, I would like to see alphaX promoted to "1.0". My boss had a cow the other day when I told her we are running alpha code in production. My defense was "you hadn't noticed because it works." Or at least call alpha6 something like "beta1" instead.
Jeff Earickson Colby College
On Fri, 16 Dec 2005, Marc Perkel wrote:
Date: Fri, 16 Dec 2005 06:42:09 -0800 From: Marc Perkel marc@perkel.com To: dovecot@dovecot.org Subject: [Dovecot] Dovecot Version Numbers - Let's drop the Alpha? I'd like to make a suggestion. Let's drop the Alpha and come out
with an official version 1.0.Here's my reasons. The 0.9x version are obsolete and people should
be using the 1.0 Alpha versions which seem to me to do everything
the previous versions did and more. It's as full featured and
stable as all other IMAP servers and in my opinion is ready to be
called 1.0.The "Alpha" label scares people off and with the 0.9x version
being obsolete I think it creates confusion for new users. It
created confusion for me when I converted to dovecot, and distros
are less ,ikely to include versions with the Alpha label.Once you go to 1.0 you can start a 1.1 alpha series and finish it
up. I know Timo has high standards but as a marketing issue I
think that the Alpha label needs to go to get people to accept and
use 1.0 in production.Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
| CCj/ClearLine - Unix/NT Administration and TCP/IP Network Services | 118 Louisa Street, Kitchener, Ontario, N2H 5M3, 519-489-0478 \________________________________________________________ Cedric Puddy, IS Director cedric@thinkers.org PGP Key Available at: http://www.thinkers.org/cedric
Cedric Puddy wrote:
My experience is that the daemon performs much better than "alpha" would ordinarily indicate.
and there's the nomenclature itself. "Alpha" means (or at least it used to mean) testing under control of the developers. "Beta" means testing under control of the users. It looks like we all agree that Dovecot is beyond both.
Jack
Jack Bailey wrote:
Cedric Puddy wrote:
My experience is that the daemon performs much better than "alpha" would ordinarily indicate.
Absolutely true.
and there's the nomenclature itself. "Alpha" means (or at least it used to mean) testing under control of the developers. "Beta" means testing under control of the users. It looks like we all agree that Dovecot is beyond both.
Well sorta ... In my experience with s/w development, I have used the alpha designation to mean that the code was fairly stable and usable, but not necessarily feature complete. Alpha code is only made available to selected and highly skilled users who can provide detailed bug reports including a test case that exposes the bug.
Beta code is deemed feature complete and ready for wider testing. Beta code should not need any great thrashing of internals to be ready for release, however if needed, do it. Beta s/w should also include at least a completed first-draft of the user documentation.
After an appropriate period of beta-testing and a final draft of the user documentation, a series of release candidates is started. When an appropriate time period has passed w/o any patches, the s/w is formally released.
On more occasions that I like to admit, I have been forced by scheduling requirements from senior management to fudge the above rules a bit. C'est guerre.
In the case of Dovecot, there seems to be aspects of the project that fit all of the above coding stages.
My (unsolicited) advice to Timo would be to create a branch tag to the project as soon as practical and apply the bug fixes to the branch as needed, as well as to HEAD. Feature development can continue on HEAD as planned. These new features would appear in later releases such as 1.(n+1), etc...
As soon as the bug rate to the branch tapers off, create the 1.0 release tarball from the branch tag. This method is roughly the same as the OpenBSD development team uses and it clearly works rather well. Nobody, in my experience, produces a tighter and cleaner software product than OpenBSD in both the free or commercial world.
The key here is to create a branch that will only have bug fixes applied which is separate from on-going development. A significant portion of the instability seen in Dovecot is to the continuing development. These two activities need to be separated to achieve the desired stability in the released code.
I have been using Dovecot for the last 3 years and have rarely had any trouble with it even though it has been been labeled -test and -alpha. My thanks to Timo and his merry elves for a filling need in the open source community. IMHO, Dovecot is already best-of-breed.
It is testimony to the competence of Timo and crew that Dovecot is rather stable. Separating bug-fixing from feature development can make Dovecot uncommonly so.
Just my $0.02 Ray
PS Long diatribes happen when one has free time on a cold rainy Saturday night.
Quoting Jack Bailey jjb@bcc.com:
Cedric Puddy wrote:
My experience is that the daemon performs much better than "alpha" would ordinarily indicate.
and there's the nomenclature itself. "Alpha" means (or at least it used to mean) testing under control of the developers. "Beta" means testing under control of the users. It looks like we all agree that Dovecot is beyond both.
Jack
While traditionally alpha testing was an internal process, it is often used now for external testing when there is no feature freeze. Beta testing would then mean that there is at least a partial feature freeze if not a total feature freeze.
I would not argue with the idea that the current code is beta quality, but it is the author's perogative to name it alpha or beta as the author chooses. The author's reputation is at stake, and he or she has to be the one to decide when and how to release the code because it is their reputation at stake.
What many want -- releasing it now as release quality code, when many like myself and apparently the author don't think it is release quality now -- is what is humorously called gamma testing, and should be avoided IMHO.
-- Eric Rostetter The Department of Physics The University of Texas at Austin
Go Longhorns!
Am 17.12.2005 um 16:17 Uhr -0800 schrieb Jack Bailey:
and there's the nomenclature itself. "Alpha" means (or at least it used to mean) testing under control of the developers. "Beta" means testing under control of the users.
Beta means "feature freeze, bug fixes only". And apparently Timo doesn't consider 1.0alpha feature-complete, yet - think of the dbox code that went in recently.
hauke-- /~\ The ASCII Ribbon Campaign Hauke Fath \ / No HTML/RTF in email Institut für Nachrichtentechnik X No Word docs in email TU Darmstadt / \ Respect for open standards Ruf +49-6151-16-3281
I'm not sure I agree with the straight jump to "1.0 release". If Timo feels it still needs testing/developpement/whatever I'll believe him.
That being said dovecot is much more stable then alot of beta code I've seen out there. So bumping it up to "beta 1" would be a good idea I think. It would show those people that didn't have the brain capacity to get anything better then an MBA (managers I think the're called :) ) that the product is moving forward.
Jd
Jeff A. Earickson wrote:
For political reasons, I would like to see alphaX promoted to "1.0". My boss had a cow the other day when I told her we are running alpha code in production. My defense was "you hadn't noticed because it works." Or at least call alpha6 something like "beta1" instead.
Jeff Earickson Colby College
On Fri, 16 Dec 2005, Marc Perkel wrote:
Date: Fri, 16 Dec 2005 06:42:09 -0800 From: Marc Perkel marc@perkel.com To: dovecot@dovecot.org Subject: [Dovecot] Dovecot Version Numbers - Let's drop the Alpha?
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Here's my reasons. The 0.9x version are obsolete and people should be using the 1.0 Alpha versions which seem to me to do everything the previous versions did and more. It's as full featured and stable as all other IMAP servers and in my opinion is ready to be called 1.0.
The "Alpha" label scares people off and with the 0.9x version being obsolete I think it creates confusion for new users. It created confusion for me when I converted to dovecot, and distros are less ,ikely to include versions with the Alpha label.
Once you go to 1.0 you can start a 1.1 alpha series and finish it up. I know Timo has high standards but as a marketing issue I think that the Alpha label needs to go to get people to accept and use 1.0 in production.
Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
I'm sure some of my customers would also freak if they knew I was using "Alpha" code. Besides, who expects a 1.0 version to be perfect?
Jeff A. Earickson wrote:
For political reasons, I would like to see alphaX promoted to "1.0". My boss had a cow the other day when I told her we are running alpha code in production. My defense was "you hadn't noticed because it works." Or at least call alpha6 something like "beta1" instead.
Jeff Earickson Colby College
On Fri, 16 Dec 2005, Marc Perkel wrote:
Date: Fri, 16 Dec 2005 06:42:09 -0800 From: Marc Perkel marc@perkel.com To: dovecot@dovecot.org Subject: [Dovecot] Dovecot Version Numbers - Let's drop the Alpha?
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Here's my reasons. The 0.9x version are obsolete and people should be using the 1.0 Alpha versions which seem to me to do everything the previous versions did and more. It's as full featured and stable as all other IMAP servers and in my opinion is ready to be called 1.0.
The "Alpha" label scares people off and with the 0.9x version being obsolete I think it creates confusion for new users. It created confusion for me when I converted to dovecot, and distros are less ,ikely to include versions with the Alpha label.
Once you go to 1.0 you can start a 1.1 alpha series and finish it up. I know Timo has high standards but as a marketing issue I think that the Alpha label needs to go to get people to accept and use 1.0 in production.
Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
Just 30 minutes ago I converted my boss from our old UW IMAP server to alpha 5 running on a new Sun box. I'm not going to tell him we're running on alpha code because he'd have a whole herd of cows. I sure hope I don't regret this.
Marc Perkel wrote:
I'm sure some of my customers would also freak if they knew I was using "Alpha" code. Besides, who expects a 1.0 version to be perfect?
Jeff A. Earickson wrote:
For political reasons, I would like to see alphaX promoted to "1.0". My boss had a cow the other day when I told her we are running alpha code in production. My defense was "you hadn't noticed because it works." Or at least call alpha6 something like "beta1" instead.
Jeff Earickson Colby College
On Fri, 16 Dec 2005, Marc Perkel wrote:
Date: Fri, 16 Dec 2005 06:42:09 -0800 From: Marc Perkel marc@perkel.com To: dovecot@dovecot.org Subject: [Dovecot] Dovecot Version Numbers - Let's drop the Alpha?
I'd like to make a suggestion. Let's drop the Alpha and come out with an official version 1.0.
Here's my reasons. The 0.9x version are obsolete and people should be using the 1.0 Alpha versions which seem to me to do everything the previous versions did and more. It's as full featured and stable as all other IMAP servers and in my opinion is ready to be called 1.0.
The "Alpha" label scares people off and with the 0.9x version being obsolete I think it creates confusion for new users. It created confusion for me when I converted to dovecot, and distros are less ,ikely to include versions with the Alpha label.
Once you go to 1.0 you can start a 1.1 alpha series and finish it up. I know Timo has high standards but as a marketing issue I think that the Alpha label needs to go to get people to accept and use 1.0 in production.
Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
I 100% agree with you. All of this version numbering needs to start
making sense. Call it 1.0.0 already. Nobody is going to take it
seriously if it's never released.
On Dec 16, 2005, at 6:42 AM, Marc Perkel wrote:
I'd like to make a suggestion. Let's drop the Alpha and come out
with an official version 1.0.Here's my reasons. The 0.9x version are obsolete and people should
be using the 1.0 Alpha versions which seem to me to do everything
the previous versions did and more. It's as full featured and
stable as all other IMAP servers and in my opinion is ready to be
called 1.0.The "Alpha" label scares people off and with the 0.9x version being
obsolete I think it creates confusion for new users. It created
confusion for me when I converted to dovecot, and distros are
less ,ikely to include versions with the Alpha label.Once you go to 1.0 you can start a 1.1 alpha series and finish it
up. I know Timo has high standards but as a marketing issue I think
that the Alpha label needs to go to get people to accept and use
1.0 in production.Thoughts?
-- Marc Perkel - marc@perkel.com
Spam Filter: http://www.junkemailfilter.com My Blog: http://marc.perkel.com
On Fri, Dec 16, 2005 at 10:43:21PM -0800, Erik Petersen may have written:
[snip whinging]
Why don't you let the project manager (i.e. Timo) decide what to do with the software he has spent so much time working on. If you don't like the name or the version number, use something else.
It really is that simple.
Cheers,
Brian T Glenn delink.net Internet Services
Brian T Glenn wrote:
On Fri, Dec 16, 2005 at 10:43:21PM -0800, Erik Petersen may have written: [snip whinging]
Why don't you let the project manager (i.e. Timo) decide what to do with the software he has spent so much time working on. If you don't like the name or the version number, use something else.
It really is that simple.
Cheers,
if it's for cosmetic reasons only anyway ... get the source, change version number and compile ^^
Florian
--On Tuesday, December 20, 2005 2:54 PM +0100 Florian Schnabel florian.schnabel@docufy.de wrote:
if it's for cosmetic reasons only anyway ... get the source, change version number and compile ^^
One could even fork it under a new name and call it version 1.0. The person creating the fork would be responsible for maintaining fixes for what's essentially a 3rd party release branch. (I suspect Timo doesn't want that headache while continuing development on the trunk.)
I'll call the next release 1.0.beta1.
I don't think there's a need yet for a new branch, since most of the changes in Dovecot core are bugfixes, and the rest are simple enough changes that they can't really break. And I'm not planning on adding any new larger features to Dovecot core either.
dbox format lives just fine in its own little world. If it's buggy, it doesn't really matter as it doesn't affect mbox/maildir at all.
The added plugins also can't break things since they are plugins and aren't used by default. Of course, there's that one compilation problem that I'll still need to fix..
As for when 1.0 comes, well, there are still too many unsolved bugs. It's good that people report backtraces from crashes here, but too often I just can't figure out how to reproduce them and can't see anything wrong with the code either. It would really help if people tried reproducing the bugs themselves and told me how it can be done..
On Thu, 22 Dec 2005, Timo Sirainen wrote:
It's good that people report backtraces from crashes here, but too often I just can't figure out how to reproduce them and can't see anything wrong with the code either. It would really help if people tried reproducing the bugs themselves and told me how it can be done..
One problem in reproducing an assert/crash is that the person's mailbox has changed by the time I find the dovecot core file and run gdb against it. It would be really cool if i_assert would save a copy of the INBOX (and any other open files) into the user's home directory at the time it generated the core. That way maybe I could have a data file to work with.
Another idea... I know there is the rawlog feature, but I don't want to turn it on for everybody, since asserts are very rare now. Any chance some kind of memory buffer for recent user actions could be kept in memory, then dumped to a file in the person's homedir if an assert occurs?
This would answer the two unknown questions: 1) What was the user doing?, 2) What was the state of the file he was doing it to?
I am open to suggestions for better procedures at my end too.
Jeff Earickson Colby College
On 22.12.2005, at 21:02, Jeff A. Earickson wrote:
Another idea... I know there is the rawlog feature, but I don't want to turn it on for everybody, since asserts are very rare now. Any chance some kind of memory buffer for recent user actions could be kept in memory, then dumped to a file in the person's homedir if an assert occurs?
This would answer the two unknown questions: 1) What was the user doing?, 2) What was the state of the file he was doing it to?
Hmm. I think I could do that as a plugin, but I'm not sure how useful it would be. The commands might tell why the mailbox got into such a state that corruption occurred, but might as well not..
Timo Sirainen tss@iki.fi writes:
On 22.12.2005, at 21:02, Jeff A. Earickson wrote:
Another idea... I know there is the rawlog feature, but I don't want to turn it on for everybody, since asserts are very rare now. Any chance some kind of memory buffer for recent user actions could be kept in memory, then dumped to a file in the person's homedir if an assert occurs?
This would answer the two unknown questions: 1) What was the user doing?, 2) What was the state of the file he was doing it to?
Hmm. I think I could do that as a plugin, but I'm not sure how useful it would be. The commands might tell why the mailbox got into such a state that corruption occurred, but might as well not..
*Something* along these lines would be very helpful. Ever since I started using the alpha versions I have had Dovecot hang on me at various times. Try as I might, I can't reliably reproduce the hang.
Looking at truss output, it seems that that Dovecot thinks it has finished sending a response to the client and has re-entered the poll loop. But the rawlogs and the log on my client say that Dovecot has *not* finished sending the response.
I've asked about this before and you said to try a patch which added some debug output to the logs (http://dovecot.org/patches/hang-debug.diff). I posted the output of that but didn't hear back--I was hoping you found the cause and fixed it anyhow :)
As you say it would be best if I could show you exactly how to reproduce the problem. I'll keep trying to find a recipe for reproducing the problem but I'm not sure what else to do. Today it hung while I tried to read two different large messages (~4k lines of text). Immediately after the second hang I copied the problematic folder, inserted a new message just like the ones that caused the hangs, etc. But it never hung again.
participants (16)
-
Andy Cravens
-
Brian T Glenn
-
Cedric Puddy
-
Eric Rostetter
-
Erik Petersen
-
Florian Schnabel
-
Hauke Fath
-
Jack Bailey
-
Jean-Daniel Beaubien
-
Jeff A. Earickson
-
Kenneth Porter
-
Marc Perkel
-
Mark Plaksin
-
Michael Paesold
-
Raymond Lillard
-
Timo Sirainen