Houston, we have a problem.

[Marc Mutz]

On Thursday 05 September 2002 23:47, Dirk Mueller wrote:

> > It's not easy for me to do this, but failing any positive effect, I
> > ask to revoke Don's CVS rights at least until 3.1.0 is out of the
> > door.
>
> The short answer: No.
<snip>

I guess you misinterpreted the past paragraph. By no means I meant to 
revoke his rights _now_. I just pointed out that _if_ this wider 
discussion does not yield any result and the situation stays the same, 
only then this has to be talked about.

Sorry.

> I can't decide if Don's patch is a feature or a bugfix and I don't
> think its my job to decide about it, because I hardly contributed a
> single line of code to KMail. Those who wrote most parts of KMail
> will be able to judge much better, and I encourage them to explain
> their reasoning for their opinion.

Ok, here come my reasons for not wanting this patch in at this time (I 
do agree with the basic design, though!):

- Don earlier on said that he would postpone the patch until 3.1 is out. 
  (It's in one of the mails I sent the msg-id of) This lead me to 
  postpone reviewing the patch, and focusing on my own todo list. Only 
  after the workload of resolving conflicts in his tree became too big 
  for him he decided to commit earlier.

- The patch does not fix a single bug. It's just the preparation of new 
  features (context menus in the separate reader and the search window).

- Along with the undoubted cleanup by introduction of KMCommands, there 
  came other changes, most notably the introduction of KMReaderMainWin, 
  which is basically a cut'n'paste of parts of KMMainWin. By all 
  accounts, I do not consider this code a _cleanup_. Now, we have yet 
  another code point to check for when e.g. the context menu for 
  messages is to be changed. This is a _hack_ preparing to make the 
  popup work in the separate reader window for 3.2. That there is no 
  (working) popup is no bug, but a missing feature.

- We have some showstoppers (mostly in the encryption/signed mails 
  department) that need attention much more urgently. Reviewing and 
  testing such a large, intrusive and unneeded (for 3.1!) patch results 
  in manhours "wasted" that could work on more urgent things.

> So, the issue that remains is the question about wether or not his
> patch should be committed for KDE 3.1 or not at this time of the
> release cycle (The fact that it is already in CVS doesn't matter for
> me, it can be reverted at any time once there is a consensus).
>
> This, however, is a conflict that can be solved on a pure technical
> base. If the patch causes more problems than it is worth, it should
> be dropped for 3.1 release.

The KMCommands class is a long-to-mid term investment in code 
maintainability. It is completely unneeded for 3.1. The patch as a 
whole (which contains much more that that) has already caused compile 
errors and a runtime bug.

What I want to say is that IMO, the patch is worth nothing for 3.1 
(albeit much for the future), but has already proven to cause many 
problems (one should think a patch at least compiles or it compiles 
following a "oops" commit by the author after a few minutes....)

> If you think that reviewing and a little
> cooperation could make it work in time for the release, then by all
> means go for it.

I do think that reviewing and cooperation could make it work in time. 
There's only the problem that doing so will likely usurp _much_ time 
that would be better spent fixing _real_ bugs and annoyances instead.

> I believe that Don is reasonable enough to revert the changes if you
> actually point out the technical problems with it and explain why
> there is not enough time to fix them before the release. If you
> can't.. hmm.
>
> I would like to point out that I'm really trying to be impartial in
> this case, but seeing a bunch of developers who are angry (maybe for
> a reason, I don't want to deny that) about the commit from
> Don and are now trying to find a sort of "revenge",

I do not think I was out for "revenge". I wanted a problem solved that 
discussion alone did not solve up until now. Maybe your mail is a 
starting point for a more emotionless debate. I do hope so. That's 
basically what I wanted to achieve. That's why you call on mediators 
when discussions get stuck, isn't it?

Marc

-- 
They [RIAA,MPAA] are trying to invent a new crime:
interference with a business model.
                           --Bruce Schneier, Crypto-Gram 08/2002
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: signature
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20020906/98585dc6/attachment.sig>