directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alex Karasulu" <akaras...@apache.org>
Subject Re: [CONSTANTS] Changing AT and OC
Date Thu, 19 Apr 2007 17:39:28 GMT
Hi all,

On 4/19/07, Emmanuel Lecharny <elecharny@gmail.com> wrote:
>
>
>
> On 4/19/07, Ole Ersoy <ole.ersoy@gmail.com> wrote:
> >
> >
> > SNIP
> >
> > > PS. There are two unwritten rules you should always think about before
> > > proposing a change :
> > > - if it's not broken, don't fix it
> > > - respect existing conventions even if you don't like them
> >
> > Emmanuel, initially you liked the suggested change.
> > David seems to like it as well.  The reason
> > I'm mentioning this  is above you are saying
>
>
> I said, 'do it', then alex said don't. And he explained why. I respect
> Alex rationals, and I may have said 'do' a little bit too quickly. My bad.
>

We no you did ok Emmanuel.  This is why we have a community.  We all may not
consider all the ramifications.

Stefan Seelman I just noted responded with a very valid concern.  Why change
a public interface out of a whim when other's may depend on it.  He's right
and his justification for not making the changes were more compelling than
mine.  I just don't want to type more or have to break up lines of code with
big constants rather than logic in the code.  LAPTOP screen realestate is
limiting for me.

But back to Emmanuel's 'do it'.  This is fine.  We all need to speak our
minds and there is nothing wrong with that ... things need to be put on the
table right?  And objectors must stand up and object supplying their
rational.  So don't worry we took care of it.

> - respect existing conventions even if you don't like them
> >
> > What does that mean exactly (In terms of the current context)?
>
>
> This is a general rule, and it's not targeting you. It's for everyone, me
> included. The perfect example is that I should not have told you 'go'
> without asking Alex before.
>

Again I think you did what you thought was right and that's ok.

Also, if it's not broken don't fix is VERY ambiguous.
> > Lets just stick to facts and concepts ok?  Communicate clearly
> > and simply with examples if necessary.
>
>
> Just a general sentence. I try to say 'focus on serious issues'. We are
> still waiting the RPM packager you promized month ago, and DAS is serious
> business too. Don't get lost in small details. Keep the good work on, keep
> focus on DAS, you are close to a good solution, I think.
>

Yes I agree with this completely.  This not a critical issue but after
Stefan Seelmann's commentary I will have to agree that we must not change
these public constants without the deprecation process.

BTW Ole the DAS stuff you're doing is very cool.  I hope this little
disagreement does not take focus away from that.  That is a serious matter
for us all and the convention for constant naming is not.

You said do it.  Alex said don't do it.
> >
> > This is a democracy.  We all have to agree before
> > something is done.  That means we have to discuss
> > it and make sure everyone has clearly understood each other's
> > rationale.
>
>
> Sometime it's a litle bit more complicated. This is a meritocracy. Alex
> merit is enormous, and we should respect that. Even if he is wrong, he
> deserves respect. We all owe him a lot, because he started this project.
> When he says "don't", I back him, just because Alex is 99% right, when i'm
> only 50%. I won't fight for the 1%, if it's not important.
>

Well I do want people to leverage my experience in this stuff to do the
right thing.  Not just to listen to me.  I want people to challenge my ideas
and I am wrong several times.  So without that I can mislead this project.
Please keep questioning but please read and try to absorb what I am saying
instead of overlooking my comments.

Alex

Mime
View raw message