[ANN] Please use Gerrit from now on for Patch Review

classic Classic list List threaded Threaded
90 messages Options
Next » 12345 « Prev
Bjoern Michaelsen Bjoern Michaelsen
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

On Wed, Jun 20, 2012 at 02:55:53PM +0200, Lubos Lunak wrote:
>  If you don't want people to miss important information, you need to announce
> it properly and not as some offhand remark.

A mail with "ACTION REQUIRED" in the subject is not an offhand remark.
Having called it out with vigor in the ESC multiple times, for example here:

 http://nabble.documentfoundation.org/michael-meeks-suse-com-minutes-of-ESC-call-td3942653.html

is _not_ an offhand remark. The only thing less "offhand" I can do, is writing
on planet.documentfoundation.org that the gerrit installation is ready to go.
Incidentally when I did that, we got a >30 post flamewar on the list.

Best,

Bjoern
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Norbert Thiebaud Norbert Thiebaud
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Lionel Elie Mamane
On Wed, Jun 20, 2012 at 5:26 AM, Lionel Elie Mamane <[hidden email]> wrote:

> Ah, I understand. When the gerrit repos are the "true one source" and
> gerrit will do the "push" automatically once someone validates the
> patch in the web interface, what will "Committer" be? The one that
> uploaded the patch or the one that validated it in the web interface?

That will be dictated by the 'committer'/'author' fiels of the uploaded patch
it is not touched afaik by gerrit

>
>> and isn't the whole point of OpenID is 're-use'.
>
> And? That I might be able to 're-use' on another website in 5 years
> does not make it any easier or more attractive to go through a
> double-plus-more-complicated setup to get an OpenID identity
> (double-plus-more-complicated compared to clicking "register" and
> copy-pasting a random password generated by pwgen or "dd
> if=/dev/urandom count=1 bs=9 | uuencode -m foo". Plus the though
> choice of "OK, who (which OpenID provider) do I want to give the power
> to impersonate me"?)
>
> To add insult to injury, I cannot 're-use' *any* of my *existing*
> reusable authentication methods. I have an OpenPGP card, SSH keys in
> files, I have X.509 client certs (in files, on smartcard, ...),
> etc. Noooo, it has to be *yet* *another* method.

1/ I did not design nor wrote gerrit (or it would certainly not be in Java.
2/ It is a case of damned if you do, damned if you don;t.
The exact same complain about "*yet* another method" was raised
because our wiki _does_ not use OpenIDfor auth.
3/ I know that our sysadm are looking at setting up our 'own' OpenID
service, if/when that happen and is flexible enough
we may be able to patch/configure gerrit so that loggin is just one
click awya (just like google-based onpenid is today)

Norbert
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lubos Lunak Lubos Lunak
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Bjoern Michaelsen
On Wednesday 20 of June 2012, Bjoern Michaelsen wrote:
> On Wed, Jun 20, 2012 at 02:55:53PM +0200, Lubos Lunak wrote:
> >  If you don't want people to miss important information, you need to
> > announce it properly and not as some offhand remark.
>
> A mail with "ACTION REQUIRED" in the subject is not an offhand remark.

 It is, if it doesn't clearly say why the action is required.

> Having called it out with vigor in the ESC multiple times, for example
> here:
>
>
> http://nabble.documentfoundation.org/michael-meeks-suse-com-minutes-of-ESC-
>call-td3942653.html
>
> is _not_ an offhand remark.

"
* Use gerrit ! test it / give it a hard time (Bjoern)
        + the repo there will never get merged back,
          so play and try to do evil stuff.
        + https://gerrit.libreoffice.org/
        + Norbert to merge change-id tweak to git-hooks
"

 Besides, is every committer required to study ESC minutes?

> The only thing less "offhand" I can do, is
> writing on planet.documentfoundation.org that the gerrit installation is
> ready to go.

 No. Announcing something properly means announcing it in the right place
(this list), under an appropriate name and with clear description (such as
saying the important part in the first paragraph or at least saying it
clearly).

 Your blog or blog aggregator fail the first one, because not everybody reads
that. So until this thread, there has been only one announcement that was
sent to the right place that I can remember, where the name was so-so
("ACTION REQUIRED" is good, but "getting setup for gerrit" is not if you
don't tell people what gerrit is) and the description was fail ("we will move
to gerrit soon" doesn't mean much without futher info, and the remark about
losing commit access was easy to miss where it was). So in practice you have
just dropped a solution on us without saying much about it.

> Incidentally when I did that, we got a >30 post flamewar on
> the list.

 If there weren't problems with communicating this, why the three biggest
subthreads start with Caolan's "why gerrit?", Kendy's "how the new workflow
is supposed to look like?" and Petr's "first need to conclude that gerrit is
in usable state"? This thread is here for a reason.

--
 Lubos Lunak
 [hidden email]
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Norbert Thiebaud Norbert Thiebaud
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

On Wed, Jun 20, 2012 at 8:52 AM, Lubos Lunak <[hidden email]> wrote:
>
>  Besides, is every committer required to study ESC minutes?

That is the _one_ message a week that, yes,  every committer should
read, or loose the right to complain about not being informed.

Norbert
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Bjoern Michaelsen Bjoern Michaelsen
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Lubos Lunak
On Wed, Jun 20, 2012 at 03:52:11PM +0200, Lubos Lunak wrote:
>  Besides, is every committer required to study ESC minutes?

Good question. IMHO not in general. But this is one of the rare occations,
where I think making a difference between volunteers and full-time paid
developers is sensible. I personally I think it can be expected from full-time
developers to follow at least the essential minutes posted to
[hidden email].

>  Announcing something properly means announcing it in the right place
> (this list)

... which is where the ESC minutes are posted.

>  If there weren't problems with communicating this, why the three biggest
> subthreads start with Caolan's "why gerrit?", Kendy's "how the new workflow
> is supposed to look like?" and Petr's "first need to conclude that gerrit is
> in usable state"? This thread is here for a reason.

I think there are multiple things to consider: First it seems, that some people
panicked and assumed this announcement suddenly meaning "every change *must* go
through changereview with gerrit" or any other implied policy change although
that was never stated.

Still thats a valid question, but I dont have to answer that alone. Frankly, I
expect members of the ESC to have a clear idea what we are spending resources
on and if there is any fundamental doubt to question that investment _early_ on.
So from the way the ESC works and as can be seen from the minutes the question
of "if we use gerrit" was already long decided, the open question was "how we
use gerrit".

I expect ESC members to have either an informed and qualified opinion on
ongoing topics or them completely trust those involved to do the right thing --
esp. when it is about the way we collaborate. That was unfortunately clearly
not the case here. So yes, there has been some leadership failure and
miscommunication here. How could anyone on the ESC, which is expect to lead and
make the final executive decision _not_ have at least looked at gerrit once
after months, unless he is completely and blindly trusting those implementing
it?

Anyway, we will clarify on the ESC call and should find some way to not make it
happen again.

Best,

Bjoern
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lionel Elie Mamane Lionel Elie Mamane
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Michael Meeks-2
On Tue, Jun 19, 2012 at 04:46:26PM +0100, Michael Meeks wrote:

> On Tue, 2012-06-19 at 17:04 +0200, Lubos Lunak wrote:

> My hope is that by encouraging the use of gerrit in parallel with the
> mailing list, the benefits will become sufficiently obvious over time
> that the old way of merging patches mailed to the list (while still an
> option) will seem to be the more annoying way to do things ;-)

> So - I think we should give gerrit a break; and play with it and see
> what happens - though I agree the absence of a 10 bullet TLDR; rational
> has been a bit of a frustration; I'm sure we'll get past that.

> Either way - we'll discuss this at the ESC on Thursday if you want to
> join in.

As I won't be able to join in this week, my views:

I have a very, very positive view of what gerrit is supposed to
do. As a person that asks for review, I'm constantly frustrated with
the process of sending a patch for review to the ML (git format-patch,
fire up email client, new email to list, attach patch, ...) and I look
forward to submitting with a simple "git push".

I dislike web interfaces, and from what I read, gerrit is good for
that, because one can do many (all?) things from the command line
(now, if our BTS could also have a command-line or email interface,
I'd be extra-plus happy).

However, our current setup *requires* an OpenID; is it an option to
make that optional (and allow people to e.g. use a "classic"
username+password for the web interface)?

My reasons for that is that signing up for an OpenID is quite an
involved process for people that don't already have one; at least for
people that think about the consequences (and don't like them); I
expect that would be at least the whole cryptogeek / cypherpunk
crowd, as well as the "privacy aware / control aware" crowd.

People like that (yes, I'm one of them) will balk at the requirement
of giving a third party (and anybody able to twist their arm... like
the surveillance agencies of governments) unlimited power to
impersonate them (to websites that use OpenID). So they'll want to run
their own OpenID end points; on the surface that's easy, but it
actually took me *days* and poring over standards to find a nice one
that will work with Gerrit, and I had to patch it myself.

 - gracie looks like it would do the job (just run it on your
   desktop), but it is bitrotten and does not work *at* *all* with
   recent python modules (see e.g. http://bugs.debian.org/src:gracie)

 - local-openid looks like a godsend, but gerrit won't interoperate
   with it; I now patched it (days and hours of efforts...), so
   hopefully it will become a good solution soon :)

 - prairie: alpha version last updated in 2008. ugh.

 - SimpleID seems to be nice too, and a good alternative to
   local-openid; just don't try (as I did) to use the same URL for
   your identity and for running SimpleID, it won't work.


My point is basically that it is too much of an investment for a
casual contributor... If we could make that easier by allowing plain
username+password (or exporting bugzilla accounts over OpenID? I guess
that would be *more* work), I feel it would lower the barrier to entry
to gerrit.

--
Lionel
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
David Ostrovsky David Ostrovsky
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by sberg
On 20.06.2012 14:11, Stephan Bergmann wrote:

> On 06/19/2012 09:32 PM, David Ostrovsky wrote:
>> On 19.06.2012 19:24, Petr Mladek wrote:
>>> Sounds good but how many people would know about the comments? How hard
>>> would be to find them?
>> https://gerrit.libreoffice.org/#/c/179/4/
>> (may be you need to login into gerrit with your openId)
>> You can see it immediatelly: if and how much and for wich file exactly.
>>
>> For me this is one of the most valuable features of gerrit: inline
>> comments.
>> On comment column from 17 files Michael has commented in 5 files.
>> This is already really good, isnt't? But it going to be even better:
>> the submitter can respond (and he surely will, if he doesn't understand
>> what the reviewer meant):
>> in the context of this file/line.
>
> Still, this removes the comments from many people's (potential) sight.
> The IMO big advantage of the "everything on a single mailing list"
> approach is that everybody is forced ;) to see everything (modulo
> information overload), so that e.g. a comment given on one
> contributor's patch is picked up "by osmosis" by other contributors
> too (so one would hope).
>
> I know there is no golden road to spreading information most
> effectively, but I personally tend to prefer spreading/consuming too
> widely over too narrowly.
>
>> I got one question with gerrit so far:
>> how can other people contribute code snippet into foreign gerrit patch
>> (so called extend it)?
>> During my work on gbuildi'fication of pyuno module Stefan helped me with
>> some scp2, Windows and Mac OS X specific stuff.
>> But he can not put a change set into my gerrit patch.
>> So he created a couple of patches and sent it to ML, I applied the
>> patches and pushed the next iteration to gerrit.
>
> To be honest, the main reason I just dumped my changes onto the ML is
> that I couldn't get comfortable with the gerrit web UI.  But hopefully
> the command line (which I haven't started to use yet) will suite me
> better...

May be I'm missing something obvious here, but how would it change the
things if you would use command line instead of web UI?
AFAIKs it can not be solved with gerrit: only i can change my gerrit
patch/change.
The only way i can think of: you would have to create your own gerrit
patch and make it depends on my.
But then tinderboxes must know, that these two patches *must* be chained
together to be successfully verified.

This flow can be trivially simulated with native git command with
patches: i will send you an almoust ready patch and would ask:
Could you please take care of scp2 and Mac OS X specific stuff in my patch?

You would just do:
git am # my patch
fix some stuff:
git -am commit ...
make && make dev-install
and then you would push my and your changes *together*.

Note: master is green all the time (!), you commited your changes under
your own user (!) and i do not have to mess around with scp2 and friends ;-)
As a casual contributor with small free time window i'm looking for a
solution for this flow with gerrit.

Regards
David

_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Bjoern Michaelsen Bjoern Michaelsen
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Lionel Elie Mamane
Hi,

On Wed, Jun 20, 2012 at 09:47:48PM +0200, Lionel Elie Mamane wrote:
> However, our current setup *requires* an OpenID; is it an option to
> make that optional (and allow people to e.g. use a "classic"
> username+password for the web interface)?

no.

> People like that (yes, I'm one of them) will balk at the requirement
> of giving a third party (and anybody able to twist their arm... like
> the surveillance agencies of governments) unlimited power to
> impersonate them (to websites that use OpenID). So they'll want to run
> their own OpenID end points;

But having a third party that is as trustworthy as TDF shouldnt be too hard as
there are already lots and lots of OpenID providers. And if you are paranoid
you would use your OpenID account just for one purpose -- that will give you
enough plausible deniability.

>  - local-openid looks like a godsend, but gerrit won't interoperate
>    with it; I now patched it (days and hours of efforts...), so
>    hopefully it will become a good solution soon :)

That sounds great for those who care about this ;)

> My point is basically that it is too much of an investment for a
> casual contributor... If we could make that easier by allowing plain
> username+password (or exporting bugzilla accounts over OpenID? I guess
> that would be *more* work), I feel it would lower the barrier to entry
> to gerrit.

I think you are part of a very, very rare demographic there (no wordpress, no
google, no launchpad, no yahoo, no blogger, no myspace, no flickr) -- we
vaguely considered running a TDF OpenID provider in the distant future, but so
shied away from that for the nontrivial cost (security is hard to get right) --
your investigation of local-openid made that somewhat more of an option, but I
would wait until a second guy (in addition to you) considers this vital -- then
we can think about using your experience with local-openid (and maybe
explicitly limit that openid to only TDF services on request). But first lets
see if there is actually somebody else caring for this.

Best,

Bjoern
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Norbert Thiebaud Norbert Thiebaud
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by David Ostrovsky
foreword:  please trim the quotation when you reply...

On Wed, Jun 20, 2012 at 3:27 PM, David Ostrovsky <[hidden email]> wrote:
>
> May be I'm missing something obvious here, but how would it change the
> things if you would use command line instead of web UI?

Choice is a great thing.

> AFAIKs it can not be solved with gerrit: only i can change my gerrit
> patch/change.

really ?
I just did exactly that on https://gerrit.libreoffice.org/#/c/229/

> The only way i can think of: you would have to create your own gerrit patch
> and make it depends on my.
> But then tinderboxes must know, that these two patches *must* be chained
> together to be successfully verified.

gerrit has the notion of 'depend' on and the tinderbox _can_ be taught
about that, but really there is no reason to get unbuildable commit in
On of the main advantage of gerrit 'review' workflow is that stuff can
be rewritten even after having been 'published' for review...
if a patch is borked, then it need to be amended. pushing a patch on
top of it to fix it is ugly and render bisection pretty hard.

Norbert
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
David Ostrovsky David Ostrovsky
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

On 20.06.2012 22:47, Norbert Thiebaud wrote:
> On Wed, Jun 20, 2012 at 3:27 PM, David Ostrovsky<[hidden email]>  wrote:
>> AFAIKs it can not be solved with gerrit: only i can change my gerrit
>> patch/change.
> really ?
> I just did exactly that onhttps://gerrit.libreoffice.org/#/c/229/
Wow! no bad, not bad, but now this patch is not pushable any more (in
this state) to repo,
because by doing so it would preserve your identity (because the last
change set always win).

You have simple hijacked it ;-)

While claiming other people's work to be your own may be not a problem
in other contries,
here in gemany it is: in fact minister of defence and other politicians
stepped down for doing exactly that (copy/paste parts of their
dissertation in that case).
And i would really like to see an commit author's face, if reviewer
would say:
Hey dude, i just entered some more comments to clarify what you have
exactly done in your 10.000 changed line patch and
have promoted it to repo ... with my user as author!
This is obviously a no go.

To make the things right and preserve the gerrit patch to be
repo-pushable all the time,
you have to conduct even more severe civil crime: to forge other
people's identity ;-)
In your sample you have to push a new change set under Björn's (!) user
(or Stephan under mine in my question):

https://gerrit.libreoffice.org/#/c/229/

but then you would definitelly be put in prison for that ... and you
said that you wanted visit LO congress this year ...
;-)

Regards
David

_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lionel Elie Mamane Lionel Elie Mamane
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Bjoern Michaelsen
On Wed, Jun 20, 2012 at 10:46:24PM +0200, Bjoern Michaelsen wrote:

> On Wed, Jun 20, 2012 at 09:47:48PM +0200, Lionel Elie Mamane wrote:

>> However, our current setup *requires* an OpenID; is it an option to
>> make that optional (and allow people to e.g. use a "classic"
>> username+password for the web interface)?

> no.

Ah. Too bad.

>> People like that (yes, I'm one of them) will balk at the
>> requirement of giving a third party (and anybody able to twist
>> their arm... like the surveillance agencies of governments)
>> unlimited power to impersonate them (to websites that use
>> OpenID). So they'll want to run their own OpenID end points;

> But having a third party that is as trustworthy as TDF shouldnt be
> too hard as there are already lots and lots of OpenID providers.

Using a TDF OpenID provider to login at gerrit.libreoffice.org would
be OK since I'm authenticating with the TDF... So the TDF being able
to "impersonate me" on its own system... err... its sysadmins can do
that whatever I do. Freedesktop too, since our repos are at
freedesktop, so the freedesktop admins can meddle with our repos, and
we decided this is OK.

But frankly, why should Google, AOL, Wordpress or another person be
able to impersonate me at the TDF systems?

> And if you are paranoid you would use your OpenID account just for
> one purpose -- that will give you enough plausible deniability.

OK, say I open a Google account for each website that wants an OpenID
login for me. How does that improve the situation with regards of
Google being able to impersonate me at these websites?

>> My point is basically that it is too much of an investment for a
>> casual contributor... If we could make that easier by allowing plain
>> username+password (or exporting bugzilla accounts over OpenID? I guess
>> that would be *more* work), I feel it would lower the barrier to entry
>> to gerrit.

> I think you are part of a very, very rare demographic there (no
> wordpress, no google, no launchpad, no yahoo, no blogger, no
> myspace, no flickr)

It is not about *having* an account there, it is about deciding one of
these people should be able to use my TDF account. Else I'd have just
opened an account at one of these websites.

--
Lionel
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lionel Elie Mamane Lionel Elie Mamane
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Bjoern Michaelsen
On Wed, Jun 20, 2012 at 01:43:26PM +0200, Bjoern Michaelsen wrote:
> On Wed, Jun 20, 2012 at 12:26:23PM +0200, Lionel Elie Mamane wrote:

>> When the gerrit repos are the "true one source" and gerrit will do
>> the "push" automatically once someone validates the patch in the
>> web interface, what will "Committer" be? The one that uploaded the
>> patch or the one that validated it in the web interface?  IMHO it
>> would be nice if it would be the one that validated in the web
>> interface.

> Author will be the whoever was the original author. Commiter will be
> be the guy/gal pressing the submit button. Everyone giving a
> codereview+1 will be a signoff in the commit.

So you are saying "Committer" will be the one giving codereview+2?
That's nice!

I also understand our policies will *not* be automated? By policies, I
mean:

 - libreoffice-3-5 needs *one* review
 - libreoffice-3-5-5 needs *three* reviews

In a dream world, gerrit would be configured for these policies, and
would automatically push a patch that has *one* codereview+2-for-3-5
to libreoffice-3-5, but automatically push a patch that has *three*
codereview+2-for-3-5-5 to libreoffice-3-5-5.

I understand gerrit is not able to "understand" such policies and we
will continue to enforce them "manually" by giving only codereview+1
unless there are already two other codereview+1.

--
Lionel
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lionel Elie Mamane Lionel Elie Mamane
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Bjoern Michaelsen
On Wed, Jun 20, 2012 at 02:34:13PM +0200, Bjoern Michaelsen wrote:
> On Wed, Jun 20, 2012 at 02:11:31PM +0200, Stephan Bergmann wrote:

>> Still, this removes the comments from many people's (potential)
>> sight. The IMO big advantage of the "everything on a single mailing
>> list" approach is that everybody is forced ;) to see everything
>> (modulo information overload)

> I can assure you that I am not forced to see your comments on the mailing
> list. Indeed, unless I am on CC or the subject sounds very thrilling the mail
> body never passes my eye.

Exactly: this is not about *forcing* people to see, but about a "push"
model rather than a "pull" model. With publish-to-ML, it is "push" and
we can "easily" (for some value of "easily") sort out what to look at
from the subject line.

Gerrit seemingly gets us to a "pull" model; we have to go look for
patches to review, be it on the web interface or by command-line
query. It becomes a separate action from reading the ML.

What I fear the most in that is that I have no way to mark a patch as
"I won't review it, not my area / I don't know / don't understand /
...". With publish-to-ML, I just mark the post / whole thread as
"read". With gerrit, I fear I will see the same patch ever and ever
again in my queries...

And the "gerrit daily digest" idea floating around is not really,
really helpful there, exactly because we lose the capacity of scanning
for interesting patches by subject and the capacity to individually
mark patches as "not to be looked at again".

We already have separate libreoffice-commits and libreoffice-bugs
mailing lists so as not to flood the "main" development mailing list
with those; maybe we could use a "libreoffice-review" mailing list to
get *one* mail for each gerrit action? Gerrit action = +1/-1 validate,
+1/+2/-1/-2 codereview, gerrit automatic push, comment added, patch
changed, new patch, ... And all emails concerning one patch/change in
one thread, just like the bugzilla mails for one bug are all in one
thread.

I'd subscribe to such a mailing list (and obviously quickly scan by
subject and not read most threads).

As I review very few patches, keeping me happy in this respect is
probably not high priority, except maybe as a long tail argument (if
we have 100 committers reviewing one patch per two months, that's
still 600 reviews per year... Worth having for the project).

--
Lionel
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Norbert Thiebaud Norbert Thiebaud
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by David Ostrovsky
On Wed, Jun 20, 2012 at 11:57 PM, David Ostrovsky
<[hidden email]> wrote:
> While claiming other people's work to be your own may be not a problem in
> other contries,
> here in gemany it is: in fact minister of defence and other politicians
> stepped down for doing exactly that (copy/paste parts of their dissertation
> in that case).

This is software development... not Academic papers.

> And i would really like to see an commit author's face, if reviewer would
> say:
> Hey dude, i just entered some more comments to clarify what you have exactly
> done in your 10.000 changed line patch and
> have promoted it to repo ... with my user as author!
> This is obviously a no go.

Then fix your own patch...
As I explained on IRC: someone that _is_ a Committer can do some
modification and still push the patch with you as author and him as
commiter (git allow that, if we used svn like some other Indians, your
scenario - author does not appear in the log unless he is the commiter
- would be the norm.

a given patch can have only one author... you don't get shared credit
on a single patch...
the alternative is to push a broken patch and then another patch to
correct it... that is pushing breakage that render bisection very
painful only to be pedant about 'authorship'. no thanks

You said: "it can not be solved with gerrit: only i can change my gerrit
patch/change."

I illustrated that this assertion was false. I did not suggest that it
was the preferred way to do it, and as a matter of fact 'Committer'
have a way to do it more nicely by preserving the 'author' information
while correcting the patch for them.

>
> To make the things right and preserve the gerrit patch to be repo-pushable
> all the time,
> you have to conduct even more severe civil crime: to forge other people's
> identity ;-)
We do that all the time with patch we collect from the ML... very
often the commtiter 'polish' the patch and still credit the original
author for the whole thing..
the alternative is to reject the patch and tell the author to come
back with a fixed version... not exactly the kind of dev-friedlyness
we are aiming at.

> but then you would definitelly be put in prison for that ... and you said
> that you wanted visit LO congress this year ...

go ahead, sue me. Me, and every copy-editor of every german newspaper...

Norbert
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Winfried Donkers Winfried Donkers
Reply | Threaded
Open this post in threaded view
|

RE: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Bjoern Michaelsen
> > Still, this removes the comments from many people's (potential) sight.
> > The IMO big advantage of the "everything on a single mailing list"
> > approach is that everybody is forced ;) to see everything (modulo
> > information overload)
>
> So, IMHO that advantage not only has its drawbacks (information overload),
> it is also mostly an illusion. Peoples forced _potential_ sight isnt really helping
> us here at all.
> And the drawback (information overload) is hitting those hardest, who have
> not finetuned their procmail for month/years -- read: newcomers and
> volunteers.
 
Not wanting to interfere, just to provide some feedback:
being a volunteer and being on the brink of newcomer and not-quite newcomer, the mailing list gives me a lot of information. Comments on submitted patches can be very informative for me. And you may know that information overload is something that I suffer from more than others.

I also learn a lot about gerrit and change-processes in groups, by the way :)

All the best,

Winfried
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
sberg sberg
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by David Ostrovsky
On 06/20/2012 10:27 PM, David Ostrovsky wrote:

> On 20.06.2012 14:11, Stephan Bergmann wrote:
>> On 06/19/2012 09:32 PM, David Ostrovsky wrote:
>>> I got one question with gerrit so far:
>>> how can other people contribute code snippet into foreign gerrit patch
>>> (so called extend it)?
>>> During my work on gbuildi'fication of pyuno module Stefan helped me with
>>> some scp2, Windows and Mac OS X specific stuff.
>>> But he can not put a change set into my gerrit patch.
>>> So he created a couple of patches and sent it to ML, I applied the
>>> patches and pushed the next iteration to gerrit.
>>
>> To be honest, the main reason I just dumped my changes onto the ML is
>> that I couldn't get comfortable with the gerrit web UI. But hopefully
>> the command line (which I haven't started to use yet) will suite me
>> better...
>
> May be I'm missing something obvious here, but how would it change the
> things if you would use command line instead of web UI?

The problem you brought up ("contribute code snippet into foreign gerrit
patch") would likely not change at all with web vs cl.  My remark was
merely meant as a general rant against "yet another idiosyncratic web
UI."  ;)

Stephan
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
sberg sberg
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Norbert Thiebaud
On 06/21/2012 08:08 AM, Norbert Thiebaud wrote:
> On Wed, Jun 20, 2012 at 11:57 PM, David Ostrovsky
> <[hidden email]>  wrote:
[...]

> As I explained on IRC: someone that _is_ a Committer can do some
> modification and still push the patch with you as author and him as
> commiter (git allow that, if we used svn like some other Indians, your
> scenario - author does not appear in the log unless he is the commiter
> - would be the norm.
>
> a given patch can have only one author... you don't get shared credit
> on a single patch...
> the alternative is to push a broken patch and then another patch to
> correct it... that is pushing breakage that render bisection very
> painful only to be pedant about 'authorship'. no thanks
>
> You said: "it can not be solved with gerrit: only i can change my gerrit
> patch/change."

So it's probably more of an "it cannot be solved with git" (evolving a
patch in discrete steps and having information about the discrete steps
recorded in the final git history, yet hiding the steps in a single unit
for purposes of bisecting; one longs for such a feature occasionally).

As long as the standard gerrit workflow is to automatically preserve the
initial patch's authorship when pushing it (whereas having to manually
add comments about additional person's polishing activities into the git
commit message is acceptable, as with the current patch-on-ML workflow),
this is IMO OK.

Stephan
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Norbert Thiebaud Norbert Thiebaud
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Lionel Elie Mamane
On Thu, Jun 21, 2012 at 12:53 AM, Lionel Elie Mamane <[hidden email]> wrote:
>
> What I fear the most in that is that I have no way to mark a patch as
> "I won't review it, not my area / I don't know / don't understand /
> ...".

No but...
1/ you can 'star' patch
2/ patch are presented in reverse chronological order

so you can scan the list of open... start the one you thing are 'interesting'

the next day the top of the pile will be new patche.. you scan until
you get to stuff you already saw and again 'star' stuff you care

then, to concentrated on the one you stared you add 'is:stared' in the
search filed (top right).

Norbert
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Lionel Elie Mamane Lionel Elie Mamane
Reply | Threaded
Open this post in threaded view
|

Re: [libreoffice-projects] [ANN] Please use Gerrit from now on for Patch Review

On Thu, Jun 21, 2012 at 02:10:05AM -0500, Norbert Thiebaud wrote:
> On Thu, Jun 21, 2012 at 12:53 AM, Lionel Elie Mamane <[hidden email]> wrote:

>> What I fear the most in that is that I have no way to mark a patch as
>> "I won't review it, not my area / I don't know / don't understand /
>> ...".

> No but...
> 1/ you can 'star' patch
> 2/ patch are presented in reverse chronological order

> so you can scan the list of open... (...)

> the next day the top of the pile will be new patche.. you scan until
> you get to stuff you already saw (...)

Ah yes, that's workable.

--
Lionel
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Bjoern Michaelsen Bjoern Michaelsen
Reply | Threaded
Open this post in threaded view
|

Re: [ANN] Please use Gerrit from now on for Patch Review

In reply to this post by Lionel Elie Mamane
On Thu, Jun 21, 2012 at 07:09:15AM +0200, Lionel Elie Mamane wrote:
> But frankly, why should Google, AOL, Wordpress or another person be
> able to impersonate me at the TDF systems?

If you created an account at one of those, you are trusting them. The trust
issue is with account creation, not with usage. Once you have a google account
it is automatically OpenID enabled. Even if you never used it yourself, google
is perfectly able to impersonate you. The same is true for an
email/password-login and any external mail provider.

Best,

Bjoern
_______________________________________________
LibreOffice mailing list
[hidden email]
http://lists.freedesktop.org/mailman/listinfo/libreoffice
Next » 12345 « Prev