[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kmail-devel
Subject:    Re: How to stop a maintainer from destroying KDChart? (was: kdenetwork/kmail)
From:       Don Sanders <sanders () kde ! org>
Date:       2003-01-15 10:56:46
[Download RAW message or body]

On Wednesday 15 January 2003 17:33, Karl-Heinz Zimmer wrote:
> On Wednesday 15 January 2003 05:53, Don Sanders wrote:
> (...)
>
> > > On Friday 10 January 2003 15:49, Don Sanders wrote:
>
> (...)
>
> > > > I am a maintainer of KMail, please accept this fact.
>
> (...)
>
> > I said I am a maintainer. That means I'm a maintainer right now
> > as I write these words, as agreed to in a discussion between all
> > core KMail developers on kde-core.
>
> Don, instead of discussing whether you are a maintainer of KDChart
> or not I would like to emphasize that your behavior is dangerous
> for KDChart.

What's KDChart? I guess you mean KMail.

> This is what happened:
>
>      1. Zack and Marc proposed to work on integration of
>         the make-it-cool branch features into HEAD branch.
>
>         They proposed on this list that they would like to
>         divide all changes and form a lot of small patches,
>         so they could be committed in the right order and
>         tested independendly from each other.
>
>         Nobody objected so they started with this big job.

In fact this is not accurate I objected.

>         After several days of hard work they had isolated
>         several dozens of little patches and were seeing
>         the end of the tunnel.
>
>
>      2. Waking up in the morning they got a shock when looking
>         at the screen and seeing that you had committed the
>         BIG patch - ignoring their efforts which you KNEW about.
>
>
>      3. Now Ingo Klöcker asked you:
>
>            "This example show that we will have to check for
>             every single patch since at least Nov 29th whether
>             it is in HEAD or whether it's not in HEAD.
>             And you want to tell us that your huge commit
>             saved time?"
>
>         you replied:
>
>            "I just committed 1MB of changes to a 3MB program.
>             I'm glad I'm primarlily relying on the users to
>             find bugs instead of peer review of code."

That's not all I said.

Because HEAD was reverted to the KDE_3_1_BRANCH any changes to HEAD 
since the KDE_3_1_BRANCH branch was created would have to be 
reapplied to HEAD whether or not I committed.

I also stated that I'm willing to go through make_it_cool and merge in 
outstanding changes. Now I've begun to do that, and have merged in 
all the small changes that I think didn't require discussion.

The remaining changes are larger or move involved and should be posted 
to the list. I've asked Zack to start with the folder jobs changes, 
please give him some time to prepare that.

>
> Dear Don,
>
> this is _not_ politics but me trying to make clear that your
> behavior is a big danger for KDChart: I am not against /you/ but
> against the way how you /acted/ during the last weeks.
>
> To show what I mean please let me cite another part of the dispute:
>
>     Replying to Ingos critics you wrote:
>
>         "Splitting everything into small patches was
>          introducing lots of bugs"
>
>     He replied correctly:
>
>         "Huh? The merge via small patches did never begin
>          because of your interaction. Remember? You made
>          your huge commit to HEAD (on Dec 26th) before Marc
>          and Zack started the "baby steps merge". So how
>          can you possibly claim that something that never
>          happened introduced lots of bugs?"
>
>
> To conclude IM(humble)O you are by far NOT behaving like
> a maintainer but like an egocentric person not seeing the
> risks coming from acting like this:  people stop being
> interested to contributing to KMail development when
> seeing your behavior.
>
> Like the KDChart maintainer wrote:
>
>      "Why couldn't you simply keep hacking on make_it_cool
>       and watch Marc and Zack do the work? Why did you have
>       to piss them off?"

I already answered this on kde-cvs here:

http://lists.kde.org/?l=kde-cvs&m=104191872427603&w=2

Basically splitting the changes into small patches would help other 
developers understand it, but it didn't and wouldn't provide any 
direct benefit to the users. 

This is due to the fact that I had a stable branch of all make_it_cool 
changes that were ready to go into HEAD. Hence I considered breaking 
make_it_cool into small diffs to be a waste of time; it could again 
introduce bugs, and make HEAD unstable and delay the integration of 
much requested improvements. Please note HEAD had been unstable for 
almost 3 weeks before my commit.

Also I offered to conduct a peer review of the make_it_cool 
integration, my committing code to HEAD in no way prevents that code 
from being peer reviewed, and it does not prevent other developers 
from understanding my work.

Furthermore after my commit HEAD was stable and it remains that way 
today.

> If that is 'politics' that forgive me for being a political
> person.
> in my opinion it is not politics but asking somebody to
> behave correctly - somebody who wants to be a maintainer!

I don't want to be a maintainer, I am a maintainer, this has been 
agreed to in a discussion on kde-core that involved all core KMail 
developers.

Technically I consider the make_it_cool integration I performed, a 
success, and I'm certainly not alone in feeling that way. I'm only 
aware of one remaining regression (In an IMAP folder, mail scrolls to 
the top when new mail arrives), I should fix that but have to 
reproduce it first.

Finally I fail to see how one can criticize the make_it_cool 
integration for not being discussed on the list and split into small 
patches, when the kroupware integration was also not discussed on the 
list (neither Zack nor I both core developers were consulted), and 
some of the kroupware changes were also not all split up into small 
changes before committing.

In the future I would like larger kroupware changes to be posted to 
the list before integration into HEAD, and I intend to do the same 
thing for make_it_cool changes.

Regarding the non-technical aspects of the make_it_cool integration. 
Yes feelings were hurt. However keeping contributors happy is not my 
primary goal, the welfare of the users has higher priority for me.

Don.

_______________________________________________
KMail Developers mailing list
kmail@mail.kde.org
http://mail.kde.org/mailman/listinfo/kmail
[prev in list] [next in list] [prev in thread] [next in thread]