Bugzilla vs. Gerrit

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Bugzilla vs. Gerrit

Stephan Herrmann-2
Hi Team,

I wondered about a batch of gerrit changes recently created by Eric Milles, some
of which gave no obvious reasons why a fix was needed in the first place. So on
one gerrit I asked him "why?" and he responded:
   "Andrey Lusktov and Lars Vogel have repeatedly asked me to submit gerrit
changes instead of opening bugs with proposed fixes/patches."

I haven't asked Andrey and Lars whether indeed they said "instead of", and I
haven't asked him if they spoke about Platform specifically or all of Eclipse,
but I think the JDT team should speak in one voice in such matters.

I personally prefer to use bugzilla first and work with gerrits only for working
out the details towards an agreed-upon goal. And in fact I believe the option to
use gerrit without bugzilla was intended only for trivial changes obviously not
needing a discussion, wasn't it?

I would even vote to keep the gerrit-only approach only for committers and route
all incoming JDT contributions through bugzilla.

Finally, I think it's important that every committer can easily see, whenever
communication already happened between a committer and a contributor, to the end
that we don't annoy contributors with contradictory requests.

What do others think?

best,
Stephan

PS: There's also the issue of separating real fixes from "style cleanup", but I
think we already agree that both kinds of changes should never be mixed in a
single contribution.
_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bugzilla vs. Gerrit

Andrey Loskutov
I *never* recommend Eric (or anyone else) to provide Gerrits *instead* of reporting bugs.
I prefer to have bugs reported first, not only in JDT project.
Independently to this, I also prefer that non trivial commits (with or without bug) have an appropriate commit message describing the fix.


Am 23. Februar 2020 12:58:44 MEZ schrieb Stephan Herrmann <[hidden email]>:

>Hi Team,
>
>I wondered about a batch of gerrit changes recently created by Eric
>Milles, some
>of which gave no obvious reasons why a fix was needed in the first
>place. So on
>one gerrit I asked him "why?" and he responded:
>"Andrey Lusktov and Lars Vogel have repeatedly asked me to submit
>gerrit
>changes instead of opening bugs with proposed fixes/patches."
>
>I haven't asked Andrey and Lars whether indeed they said "instead of",
>and I
>haven't asked him if they spoke about Platform specifically or all of
>Eclipse,
>but I think the JDT team should speak in one voice in such matters.
>
>I personally prefer to use bugzilla first and work with gerrits only
>for working
>out the details towards an agreed-upon goal. And in fact I believe the
>option to
>use gerrit without bugzilla was intended only for trivial changes
>obviously not
>needing a discussion, wasn't it?
>
>I would even vote to keep the gerrit-only approach only for committers
>and route
>all incoming JDT contributions through bugzilla.
>
>Finally, I think it's important that every committer can easily see,
>whenever
>communication already happened between a committer and a contributor,
>to the end
>that we don't annoy contributors with contradictory requests.
>
>What do others think?
>
>best,
>Stephan
>
>PS: There's also the issue of separating real fixes from "style
>cleanup", but I
>think we already agree that both kinds of changes should never be mixed
>in a
>single contribution.
>_______________________________________________
>jdt-dev mailing list
>[hidden email]
>To change your delivery options, retrieve your password, or unsubscribe
>from this list, visit
>https://www.eclipse.org/mailman/listinfo/jdt-dev

--
Kind regards,
Andrey Loskutov

https://www.eclipse.org/user/aloskutov
Спасение утопающих - дело рук самих утопающих
_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bugzilla vs. Gerrit

Lars Vogel-2
In reply to this post by Stephan Herrmann-2
Hi Stephan,

The PMC decision was that we do not necessary require bug reports for cleanup or code improvement patches to simply the contribution process.

Also the PMC decided to encourage such patches.

Contributions which changes something or fix something should have a bug report associated with it.

Also we dislike patches which mix cleanups and real changes 

Please share link to your discussion with Eric do that I can add that clarification there also.

Btw. it is IMHO awesome that Eric who IIRC works  on the Groovy tooling contributed to JDT and I definitely encourage this.

Best regards, Lars

Stephan Herrmann <[hidden email]> schrieb am So., 23. Feb. 2020, 12:59:
Hi Team,

I wondered about a batch of gerrit changes recently created by Eric Milles, some
of which gave no obvious reasons why a fix was needed in the first place. So on
one gerrit I asked him "why?" and he responded:
   "Andrey Lusktov and Lars Vogel have repeatedly asked me to submit gerrit
changes instead of opening bugs with proposed fixes/patches."

I haven't asked Andrey and Lars whether indeed they said "instead of", and I
haven't asked him if they spoke about Platform specifically or all of Eclipse,
but I think the JDT team should speak in one voice in such matters.

I personally prefer to use bugzilla first and work with gerrits only for working
out the details towards an agreed-upon goal. And in fact I believe the option to
use gerrit without bugzilla was intended only for trivial changes obviously not
needing a discussion, wasn't it?

I would even vote to keep the gerrit-only approach only for committers and route
all incoming JDT contributions through bugzilla.

Finally, I think it's important that every committer can easily see, whenever
communication already happened between a committer and a contributor, to the end
that we don't annoy contributors with contradictory requests.

What do others think?

best,
Stephan

PS: There's also the issue of separating real fixes from "style cleanup", but I
think we already agree that both kinds of changes should never be mixed in a
single contribution.
_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev

_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bugzilla vs. Gerrit

Stephan Herrmann-2
Thanks Andrey, Lars, for clarification.

See in particular https://git.eclipse.org/r/158132

An example for a mixed change is https://git.eclipse.org/r/158128/

His other recent gerrits you can see under Related Changes (7)

Regarding groovy tooling: yes, contributions from s.o. who intimately knows the
code from maintaining a fork is a well-tried model :)
I'm even willing to review a few changes for the explicit benefit of the groovy
tooling, provided I'm able to understand what is the advantage, and that the
risk of changing JDT for all doesn't outweigh the benefit for groovy.

Stephan


On 23.02.20 13:09, Lars Vogel wrote:

> Hi Stephan,
>
> The PMC decision was that we do not necessary require bug reports for cleanup or
> code improvement patches to simply the contribution process.
>
> Also the PMC decided to encourage such patches.
>
> Contributions which changes something or fix something should have a bug report
> associated with it.
>
> Also we dislike patches which mix cleanups and real changes
>
> Please share link to your discussion with Eric do that I can add that
> clarification there also.
>
> Btw. it is IMHO awesome that Eric who IIRC works  on the Groovy tooling
> contributed to JDT and I definitely encourage this.
>
> Best regards, Lars
>
> Stephan Herrmann <[hidden email]
> <mailto:[hidden email]>> schrieb am So., 23. Feb. 2020, 12:59:
>
>     Hi Team,
>
>     I wondered about a batch of gerrit changes recently created by Eric Milles,
>     some
>     of which gave no obvious reasons why a fix was needed in the first place. So on
>     one gerrit I asked him "why?" and he responded:
>         "Andrey Lusktov and Lars Vogel have repeatedly asked me to submit gerrit
>     changes instead of opening bugs with proposed fixes/patches."
>
>     I haven't asked Andrey and Lars whether indeed they said "instead of", and I
>     haven't asked him if they spoke about Platform specifically or all of Eclipse,
>     but I think the JDT team should speak in one voice in such matters.
>
>     I personally prefer to use bugzilla first and work with gerrits only for
>     working
>     out the details towards an agreed-upon goal. And in fact I believe the
>     option to
>     use gerrit without bugzilla was intended only for trivial changes obviously not
>     needing a discussion, wasn't it?
>
>     I would even vote to keep the gerrit-only approach only for committers and
>     route
>     all incoming JDT contributions through bugzilla.
>
>     Finally, I think it's important that every committer can easily see, whenever
>     communication already happened between a committer and a contributor, to the
>     end
>     that we don't annoy contributors with contradictory requests.
>
>     What do others think?
>
>     best,
>     Stephan
>
>     PS: There's also the issue of separating real fixes from "style cleanup", but I
>     think we already agree that both kinds of changes should never be mixed in a
>     single contribution.
>     _______________________________________________
>     jdt-dev mailing list
>     [hidden email] <mailto:[hidden email]>
>     To change your delivery options, retrieve your password, or unsubscribe from
>     this list, visit
>     https://www.eclipse.org/mailman/listinfo/jdt-dev
>
>
> _______________________________________________
> jdt-dev mailing list
> [hidden email]
> To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/jdt-dev
>

_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bugzilla vs. Gerrit

Manoj Palat-2
Hi All,
 
 
+1 to Stephan's suggestion - we have discussed this in length and  agreed upon a best practice in JDT.CORE specifically because changes in jdt.core might have wide ramifications. 
 
Specifically for jdt.core:
 
(a) Please have a bug for *all* the changes including cleanups, except for very minor changes like test case, routine maint. etc. and provide the justification in the bug for the change 
(b) In the gerrit, have bug <number> title in the comment so that we have a map 
(c) If its a cleanup only bug, please use M1 as the milestone to commit
 
Also, I believe the committers in jdt.core can ask for a bug - in jdt.core we request to have a bug for all the contributions instead of we asking on every gerrit.
 
Last but not the least, thanks to the contributions of Eric Milles. Please treat the above as an input about the general practice so that we have lot more quality patches coming in.
 
Thanks,
Manoj
----- Original message -----
From: Stephan Herrmann <[hidden email]>
Sent by: [hidden email]
To: [hidden email]
Cc:
Subject: [EXTERNAL] Re: [jdt-dev] Bugzilla vs. Gerrit
Date: Sun, Feb 23, 2020 7:14 PM
 
Thanks Andrey, Lars, for clarification.

See in particular https://git.eclipse.org/r/158132

An example for a mixed change is https://git.eclipse.org/r/158128/

His other recent gerrits you can see under Related Changes (7)

Regarding groovy tooling: yes, contributions from s.o. who intimately knows the
code from maintaining a fork is a well-tried model :)
I'm even willing to review a few changes for the explicit benefit of the groovy
tooling, provided I'm able to understand what is the advantage, and that the
risk of changing JDT for all doesn't outweigh the benefit for groovy.

Stephan


On 23.02.20 13:09, Lars Vogel wrote:

> Hi Stephan,
>
> The PMC decision was that we do not necessary require bug reports for cleanup or
> code improvement patches to simply the contribution process.
>
> Also the PMC decided to encourage such patches.
>
> Contributions which changes something or fix something should have a bug report
> associated with it.
>
> Also we dislike patches which mix cleanups and real changes
>
> Please share link to your discussion with Eric do that I can add that
> clarification there also.
>
> Btw. it is IMHO awesome that Eric who IIRC works  on the Groovy tooling
> contributed to JDT and I definitely encourage this.
>
> Best regards, Lars
>
> Stephan Herrmann <[hidden email]
> <[hidden email]>> schrieb am So., 23. Feb. 2020, 12:59:
>
>     Hi Team,
>
>     I wondered about a batch of gerrit changes recently created by Eric Milles,
>     some
>     of which gave no obvious reasons why a fix was needed in the first place. So on
>     one gerrit I asked him "why?" and he responded:
>         "Andrey Lusktov and Lars Vogel have repeatedly asked me to submit gerrit
>     changes instead of opening bugs with proposed fixes/patches."
>
>     I haven't asked Andrey and Lars whether indeed they said "instead of", and I
>     haven't asked him if they spoke about Platform specifically or all of Eclipse,
>     but I think the JDT team should speak in one voice in such matters.
>
>     I personally prefer to use bugzilla first and work with gerrits only for
>     working
>     out the details towards an agreed-upon goal. And in fact I believe the
>     option to
>     use gerrit without bugzilla was intended only for trivial changes obviously not
>     needing a discussion, wasn't it?
>
>     I would even vote to keep the gerrit-only approach only for committers and
>     route
>     all incoming JDT contributions through bugzilla.
>
>     Finally, I think it's important that every committer can easily see, whenever
>     communication already happened between a committer and a contributor, to the
>     end
>     that we don't annoy contributors with contradictory requests.
>
>     What do others think?
>
>     best,
>     Stephan
>
>     PS: There's also the issue of separating real fixes from "style cleanup", but I
>     think we already agree that both kinds of changes should never be mixed in a
>     single contribution.
>     _______________________________________________
>     jdt-dev mailing list
>     [hidden email] <[hidden email]>
>     To change your delivery options, retrieve your password, or unsubscribe from
>     this list, visit
>     https://www.eclipse.org/mailman/listinfo/jdt-dev
>
>
> _______________________________________________
> jdt-dev mailing list
> [hidden email]
> To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/jdt-dev
>

_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
 


_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev
Reply | Threaded
Open this post in threaded view
|

Re: Bugzilla vs. Gerrit

Lars Vogel-2
In reply to this post by Stephan Herrmann-2
Thanks Stephan, I also answered in the Gerrit to Eric. Thanks for the
notification!

Best regards, Lars

On Sun, Feb 23, 2020 at 2:44 PM Stephan Herrmann
<[hidden email]> wrote:

>
> Thanks Andrey, Lars, for clarification.
>
> See in particular https://git.eclipse.org/r/158132
>
> An example for a mixed change is https://git.eclipse.org/r/158128/
>
> His other recent gerrits you can see under Related Changes (7)
>
> Regarding groovy tooling: yes, contributions from s.o. who intimately knows the
> code from maintaining a fork is a well-tried model :)
> I'm even willing to review a few changes for the explicit benefit of the groovy
> tooling, provided I'm able to understand what is the advantage, and that the
> risk of changing JDT for all doesn't outweigh the benefit for groovy.
>
> Stephan
>
>
> On 23.02.20 13:09, Lars Vogel wrote:
> > Hi Stephan,
> >
> > The PMC decision was that we do not necessary require bug reports for cleanup or
> > code improvement patches to simply the contribution process.
> >
> > Also the PMC decided to encourage such patches.
> >
> > Contributions which changes something or fix something should have a bug report
> > associated with it.
> >
> > Also we dislike patches which mix cleanups and real changes
> >
> > Please share link to your discussion with Eric do that I can add that
> > clarification there also.
> >
> > Btw. it is IMHO awesome that Eric who IIRC works  on the Groovy tooling
> > contributed to JDT and I definitely encourage this.
> >
> > Best regards, Lars
> >
> > Stephan Herrmann <[hidden email]
> > <mailto:[hidden email]>> schrieb am So., 23. Feb. 2020, 12:59:
> >
> >     Hi Team,
> >
> >     I wondered about a batch of gerrit changes recently created by Eric Milles,
> >     some
> >     of which gave no obvious reasons why a fix was needed in the first place. So on
> >     one gerrit I asked him "why?" and he responded:
> >         "Andrey Lusktov and Lars Vogel have repeatedly asked me to submit gerrit
> >     changes instead of opening bugs with proposed fixes/patches."
> >
> >     I haven't asked Andrey and Lars whether indeed they said "instead of", and I
> >     haven't asked him if they spoke about Platform specifically or all of Eclipse,
> >     but I think the JDT team should speak in one voice in such matters.
> >
> >     I personally prefer to use bugzilla first and work with gerrits only for
> >     working
> >     out the details towards an agreed-upon goal. And in fact I believe the
> >     option to
> >     use gerrit without bugzilla was intended only for trivial changes obviously not
> >     needing a discussion, wasn't it?
> >
> >     I would even vote to keep the gerrit-only approach only for committers and
> >     route
> >     all incoming JDT contributions through bugzilla.
> >
> >     Finally, I think it's important that every committer can easily see, whenever
> >     communication already happened between a committer and a contributor, to the
> >     end
> >     that we don't annoy contributors with contradictory requests.
> >
> >     What do others think?
> >
> >     best,
> >     Stephan
> >
> >     PS: There's also the issue of separating real fixes from "style cleanup", but I
> >     think we already agree that both kinds of changes should never be mixed in a
> >     single contribution.
> >     _______________________________________________
> >     jdt-dev mailing list
> >     [hidden email] <mailto:[hidden email]>
> >     To change your delivery options, retrieve your password, or unsubscribe from
> >     this list, visit
> >     https://www.eclipse.org/mailman/listinfo/jdt-dev
> >
> >
> > _______________________________________________
> > jdt-dev mailing list
> > [hidden email]
> > To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> > https://www.eclipse.org/mailman/listinfo/jdt-dev
> >
>
> _______________________________________________
> jdt-dev mailing list
> [hidden email]
> To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> https://www.eclipse.org/mailman/listinfo/jdt-dev



--
Eclipse Platform project co-lead
CEO vogella GmbH

Haindaalwisch 17a, 22395 Hamburg
Amtsgericht Hamburg: HRB 127058
Geschäftsführer: Lars Vogel, Jennifer Nerlich de Vogel
USt-IdNr.: DE284122352
Fax (040) 5247 6322, Email: [hidden email], Web: http://www.vogella.com
_______________________________________________
jdt-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/jdt-dev