Posting to hackers and patches lists
Folks, can we avoid posting an email to both hackers and patches lists?
I understand why people do it, but it is best avoided, I think. If you
feel the need to keep patch discussion on hackers, please post just the
patch to patches and a summary to hackers.
Or better yet, have a URL to the patch in an email to hackers.
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Bruce Momjian <bruce@momjian.us> writes:
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.
Actually, I find that that is a truly awful habit and I wish that people
would *not* do it that way. There are two reasons why not:
* no permanent archive of the submitted patch
* reviewer won't know if the submitter changes the patch after he
downloads a copy, and in fact nobody will ever know unless the submitter
takes the time to compare the eventual commit to what he thinks the
patch is
regards, tom lane
Tom Lane wrote:
Bruce Momjian <bruce@momjian.us> writes:
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.Actually, I find that that is a truly awful habit and I wish that people
would *not* do it that way. There are two reasons why not:* no permanent archive of the submitted patch
* reviewer won't know if the submitter changes the patch after he
downloads a copy, and in fact nobody will ever know unless the submitter
takes the time to compare the eventual commit to what he thinks the
patch is
This requires the patch submitter to send an email every time they
update the URL. The problem with no archive is a problem though. It
works for me because I am around to supply versions but I see your
point --- perhaps we could make the system have a stable URL but allow
for versioning access. Maybe email is a fine interface, of course.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
On Thu, May 8, 2008 at 12:17 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Bruce Momjian <bruce@momjian.us> writes:
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.Actually, I find that that is a truly awful habit and I wish that people
would *not* do it that way. There are two reasons why not:* no permanent archive of the submitted patch
Yes. I can see how posting a URL to a patch would be convenient, but
having the permanent record of the patch as submitted is important.
What about uploading patches to the wiki? That way we have the
permanent record (change history), as well as the single authoritative
location for fetching the latest version.
* reviewer won't know if the submitter changes the patch after he
downloads a copy, and in fact nobody will ever know unless the submitter
takes the time to compare the eventual commit to what he thinks the
patch is
Well, as long as you send another message to the lists saying "I've
uploaded a new version of the patch, that URL again is <>". If you
just silently update the patch without telling anybody you're bound to
run into problems.
Cheers,
BJ
Brendan Jurd wrote:
On Thu, May 8, 2008 at 12:17 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Bruce Momjian <bruce@momjian.us> writes:
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.Actually, I find that that is a truly awful habit and I wish that people
would *not* do it that way. There are two reasons why not:* no permanent archive of the submitted patch
Yes. I can see how posting a URL to a patch would be convenient, but
having the permanent record of the patch as submitted is important.What about uploading patches to the wiki? That way we have the
permanent record (change history), as well as the single authoritative
location for fetching the latest version.
Right, I was assuming once the patch was uploaded it would be to our
infrastructure and would be permanent.
* reviewer won't know if the submitter changes the patch after he
downloads a copy, and in fact nobody will ever know unless the submitter
takes the time to compare the eventual commit to what he thinks the
patch isWell, as long as you send another message to the lists saying "I've
uploaded a new version of the patch, that URL again is <>". If you
just silently update the patch without telling anybody you're bound to
run into problems.
Yep.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Bruce Momjian wrote:
Folks, can we avoid posting an email to both hackers and patches
lists? I understand why people do it, but it is best avoided, I
think. If you feel the need to keep patch discussion on hackers,
please post just the patch to patches and a summary to hackers.Or better yet, have a URL to the patch in an email to hackers.
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.
What?! Did you just propose a patch tracker? Bruce? Hmm. I think I need
to get a new email client, because this one clearly corrupts the emails
I receive ;)
//Magnus
On Wed, May 7, 2008 at 8:28 AM, Bruce Momjian <bruce@momjian.us> wrote:
Brendan Jurd wrote:
On Thu, May 8, 2008 at 12:17 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Bruce Momjian <bruce@momjian.us> writes:
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.Actually, I find that that is a truly awful habit and I wish that people
would *not* do it that way. There are two reasons why not:* no permanent archive of the submitted patch
Yes. I can see how posting a URL to a patch would be convenient, but
having the permanent record of the patch as submitted is important.What about uploading patches to the wiki? That way we have the
permanent record (change history), as well as the single authoritative
location for fetching the latest version.Right, I was assuming once the patch was uploaded it would be to our
infrastructure and would be permanent.
Heck, I dont think patch submitters really care. And Ill do whatever
is in the dev faq.
But Its a heck of a lot easier (for me) just to send them in email.
Plus it seems awkward to move a discussion thats taking place on
-hackers over to patches... Granted I could post to patches first,
wait an hour then send an email to hackers/reviewer and say hey!
updated patch here! But it hardly seems worth it to me... In fact I
would argue -patches should go away so we dont have that split.
Magnus Hagander wrote:
Bruce Momjian wrote:
Folks, can we avoid posting an email to both hackers and patches
lists? I understand why people do it, but it is best avoided, I
think. If you feel the need to keep patch discussion on hackers,
please post just the patch to patches and a summary to hackers.Or better yet, have a URL to the patch in an email to hackers.
I think it would be helpful for us to provide an infrastructure where
people who don't run their own servers to store their patches at a
stable URL where they can keep updating the content. I did that with
the psql wrap patch and it helped me.What?! Did you just propose a patch tracker? Bruce? Hmm. I think I need
to get a new email client, because this one clearly corrupts the emails
I receive ;)
I have suggested a patch tracker as optional for people before on this
list:
http://archives.postgresql.org/pgsql-hackers/2008-04/msg00626.php
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
* Magnus Hagander (magnus@hagander.net) wrote:
What?! Did you just propose a patch tracker? Bruce? Hmm. I think I need
to get a new email client, because this one clearly corrupts the emails
I receive ;)
If you want an email and web-based tracking system, RT is wonderful
(http://bestpractical.com/rt/)...
Enjoy,
Stephen
Alex Hunsaker wrote:
In fact I
would argue -patches should go away so we dont have that split.
+1 I think the main argument for the split is to keep the "large"
patch emails off the hackers list, but I don't think that limit is so
high that it's a problem. People have to gzip their patches to the
patches list fairly often anyway.
Alex Hunsaker wrote:
Right, I was assuming once the patch was uploaded it would be to our
infrastructure and would be permanent.Heck, I dont think patch submitters really care. And Ill do whatever
is in the dev faq.
But Its a heck of a lot easier (for me) just to send them in email.
Sure, then just keep sending them via email. I often go through several
revisions a day as I get feedback and having all that email volume seems
wasteful.
Plus it seems awkward to move a discussion thats taking place on
-hackers over to patches... Granted I could post to patches first,
wait an hour then send an email to hackers/reviewer and say hey!
updated patch here! But it hardly seems worth it to me... In fact I
would argue -patches should go away so we dont have that split.
The goal is for the patches list to just discuss patches, but often
there are user API issues that come up after the patch is submitted, and
people often want that discussion on hackers. The current email split
can certainly be awkward.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Stephen Frost wrote:
* Magnus Hagander (magnus@hagander.net) wrote:
What?! Did you just propose a patch tracker? Bruce? Hmm. I think I need
to get a new email client, because this one clearly corrupts the emails
I receive ;)If you want an email and web-based tracking system, RT is wonderful
(http://bestpractical.com/rt/)...
STOP!
We really really do NOT need to have this discussion every month of the
calendar.
cheers
andrew
On Wed, May 7, 2008 at 9:03 AM, Bruce Momjian <bruce@momjian.us> wrote:
Alex Hunsaker wrote:
Plus it seems awkward to move a discussion thats taking place on
-hackers over to patches... Granted I could post to patches first,
wait an hour then send an email to hackers/reviewer and say hey!
updated patch here! But it hardly seems worth it to me... In fact I
would argue -patches should go away so we dont have that split.The goal is for the patches list to just discuss patches, but often
there are user API issues that come up after the patch is submitted, and
people often want that discussion on hackers. The current email split
can certainly be awkward.
A big part of my problem with the split is if there is a discussion
taking place on -hackers I want to be able to reply to the discussion
and say "well, here is what I was thinking". Sending it to -patches
first waiting for it to hit the archive so I can link to it in my
reply on -hackers seems pointless and convoluted.
But if thats what you want, thats what ill try to do from now on :)
For instance the patch Tom reviewed of mine yesterday only -hackers
was copied, so I maintained that but also added -patches because I was
sending in a patch...
I think It will be an ongoing problem though especially for new people
as they probably wont understand the "logical" split...
Alex Hunsaker wrote:
A big part of my problem with the split is if there is a discussion
taking place on -hackers I want to be able to reply to the discussion
and say "well, here is what I was thinking". Sending it to -patches
first waiting for it to hit the archive so I can link to it in my
reply on -hackers seems pointless and convoluted.
Yea, that is a problem. Adding a new patch to patches while discussing
on hackers is a receipe for confusion.
But if thats what you want, thats what ill try to do from now on :)
For instance the patch Tom reviewed of mine yesterday only -hackers
was copied, so I maintained that but also added -patches because I was
sending in a patch...
Yea, sending to both is probably the worst. I would just post to hackers
and mention you sent a new version of the patch to patches --- they
usually show up the same time.
I think It will be an ongoing problem though especially for new people
as they probably wont understand the "logical" split...
Yep, I can hardly explain it.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Alex Hunsaker wrote:
A big part of my problem with the split is if there is a discussion
taking place on -hackers I want to be able to reply to the discussion
and say "well, here is what I was thinking". Sending it to -patches
first waiting for it to hit the archive so I can link to it in my
reply on -hackers seems pointless and convoluted.But if thats what you want, thats what ill try to do from now on :)
For instance the patch Tom reviewed of mine yesterday only -hackers
was copied, so I maintained that but also added -patches because I was
sending in a patch...I think It will be an ongoing problem though especially for new people
as they probably wont understand the "logical" split...
Patches are an integral part of the conversation about development, I
think trying to split them up is awkward at best. Do people really
still think that the potential for larger messages is really a problem?
By the way, what is the actual size limit on hackers vs patches. I
would imagine that most patches would already fit in the current hackers
limit, especially since you can gzip.
"Matthew T. O'connor" <matthew@zeut.net> writes:
Patches are an integral part of the conversation about development, I
think trying to split them up is awkward at best. Do people really
still think that the potential for larger messages is really a problem?
Personally I'd be fine with abandoning -patches and just using -hackers.
We could try it for awhile, anyway, and go back if it seems worse.
By the way, what is the actual size limit on hackers vs patches.
They do have different size limits; we'd have to raise the limit on
-hackers if we do this. Marc would know exactly what the limits are.
regards, tom lane
On Thu, May 8, 2008 at 1:54 AM, Matthew T. O'connor <matthew@zeut.net> wrote:
Patches are an integral part of the conversation about development,
I'd go further than that. Patches ARE conversation about development,
they are just in C rather than English.
Having one list for the parts of the conversation that are written in
C and another for the parts that are in English is bizarre, in my
opinion. Especially since you almost always want to accompany your C
code with some English commentary.
Cheers,
BJ
* Alex Hunsaker <badalex@gmail.com> [080507 11:38]:
A big part of my problem with the split is if there is a discussion
taking place on -hackers I want to be able to reply to the discussion
and say "well, here is what I was thinking". Sending it to -patches
first waiting for it to hit the archive so I can link to it in my
reply on -hackers seems pointless and convoluted.
Note that even though I'm not a fan of the split, the "wait to hit the
archive" problem is not really a problem.
If you sent it, and you know it's message-id, and you can link directly
to it: such as:
http://archives.postgresql.org/message-id/34d269d40805070837q19f1144eu8c316fa1cf6d8780@mail.gmail.com
a.
--
Aidan Van Dyk Create like a god,
aidan@highrise.ca command like a king,
http://www.highrise.ca/ work like a slave.
Tom Lane wrote:
Personally I'd be fine with abandoning -patches and just using -hackers.
We could try it for awhile, anyway, and go back if it seems worse.
I'd be good with that. The split never made much sense for me.
On Wed, May 07, 2008 at 12:20:04PM -0400, Tom Lane wrote:
"Matthew T. O'connor" <matthew@zeut.net> writes:
Patches are an integral part of the conversation about
development, I think trying to split them up is awkward at best.
Do people really still think that the potential for larger
messages is really a problem?Personally I'd be fine with abandoning -patches and just using
-hackers. We could try it for awhile, anyway, and go back if it
seems worse.
This would make it a little tougher on me as far as maintaining the
patches section of the PostgreSQL Weekly News, but I'll deal with it
if I need to :)
Cheers,
David.
--
David Fetter <david@fetter.org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david.fetter@gmail.com
Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate
