> Hi POI folks,
>
Hi.
> after a rather long time-out I am back to the project and willing
> to start working again. However, before I continue with HPSF
> development I'd like your votes on what to do first and what to
> do later.
The prioritization, if you like our opinion is fine, but its not binding. Its
the real pupose of this "whether to back out Andy's code changes" that requires
a vote.
> The current situation is intolerable because HPSF is simply
> dysfunctional. In February, I released a working and properly
> functional HPSF package and documentation. In April, Andy
> noticed that HPSF uses its own set of little-endian classes and
> submitted bug 8354.
No we did mention it before it was committed.
> I agreed that it would be nice if HPSF used the general
> little-endian classes and also agreed to switch over to them.
> However, I also made clear that this issue is not an urgent one,
> because HPSF has only very little impact on POI's performance
> and
> memory usage. I gave priority to completing the documentation
> and
> writing junit tests over exchanging the little-endian classes.
> Since then until recently I could not spend any time on the POI
> project and watched it only remotely.
Note for 5 months.
> In May, Andy tried to do the exchange of the little-endian
> classes
> himself. Unfortunately he broke the code and checked the broken
> code back into the CVS repository. This is the current
> situation.
> Run the poibrowser to see that it is no longer able to recognize
> property sets.
You left out the part where I asked for your help to fix it but you
didn't and still don't have time to do so. I want to move this forward not back.
By my estimation there is about 1 class that needs fixing that I don't understand
how to fix. If you would help me fix that, it would probably start working, if there
are other changes they're probably minor.
This isn't really about that though is it, its your wounded pride. If you don't want
to have to deal with others messing with your code, by all means don't opensource it.
Personally, I hope you do create unit tests and documentation and etc, but I'm honestly
not counting on it. You've shown very little inclination to work on POI and have never
shown any inclination to collaborate with any of us and "when you have time" hasn't
happened
in the last 5 months
> What I'd like from you is to vote on the development priorities
> for HPSF, since Andy and I could not get to an agreement. I
> outlined my priorities in <http://jakarta.apache.org/poi/hpsf/tod
/> o.html> /
> and repeat the top four of them here:
>
> 1. Rollback Andy's changes in order to have HPSF working again.
-1 - WORK WITH ANDY to help get HPSF working again with his changes.
2. Complete writing the HPSF documentation.
+100
3. Write up some junit tests.
+100
4. Use the general POI little-endian classes (see bug 8354).
-1 - redundant about #1
> Andy wants the switch to the general POI little-endian classes
> done first. This is unacceptable for me because I regard a
> correct
> functionality of HPSF plus documentation plus junit test much
> higher
> than an improved code-reuse which gives no benefit to the user
> and currently does not even work.
This is childish. You don't like that I touched your code and don't want to
help me fix one little class, you'd rather roll back all the changes and throw
away the only substantial work done on HPSF in the last 5 months. So is HPSF
your own private little project?
I'll only +1 rolling back my changes and freezing it until rainer "has time" if we
can move HPSF into the scratchpad until it matures. This will step it out of the
way and let others who "have time" donate something else that others can work on
without
rainers help.
> So please cast your votes on the priorities outlined above!
> Thanks!
-1 - a more reasonable alternative: quit sulking and help me understand
http://jakarta.apache.org/poi/javadocs/javasrc/org/apache/poi/hpsf/ClassID_java.html#ClassID
(scroll to the method clearly marked REWRITE ME REWRITE ME)
Read this method before you vote! and if someone other than rainer (who donen't have
time)
wants to give me a hand in figuring that method out, that would be appreciated and
probably
make this whole argument moot!
Buenos Dias,
Andy
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
