Hi, guys
Intro
First, please, don't take my words too hard to offend anyone. Just want
to say my (user's) impression of the trends in OOo developement.
I attended community since the beta of 1.0.0 and started reporting bugs
as an ordinary user. Later I became an ordinary admin supporting OOo on
150 PCs. I was excited by the project and couldn't wait enough for the
next stable release.
Core
Since some 1.1.3, it seems that quality progress is degrading. BETA
should give space to catch important bugs BEFORE they could dig into
final product, however I know serious bugs that remain unfixed for
years, survive even major versions and many minor "stable" releases.
Now, beta testing makes little sense for me.
Few of these bugs impose data loss of some degree (i.e. 54567, 35178,
32785, 21116, 35094, 58602) and by means of the guidelines, should block
final release until got fixed. Simple solution -lowering the priority
flag (dosen't look too good to have dataloss-grade bugs in the release,
yeah?). I encountered even worse scenarios, when bug was simply marked
as "workform"; dosen't matter that the user did nothing explicitely bad,
just used the program in the intuitive way that could differ from
programmer's "the holy and only" one, or from reasons, why it is done
the way it is.
After very bad start of OOo 2.0 (unable to save to network share on
Linux, bug 54567 and related), I need now to test every _stable_ verion,
whether is it worth deploying, or not. From my point of view, the last
"stable" release has been 1.1.4. Since then, OOo BECAME CONSTANT BETA,
whatever officially marked, slowly getting to RC with 2.0.3
I must say that not all bugs suffer this way. Import from WW8 is
constantly getting better (everyone wants to see it happening even
faster ;o) and bugs are even fixed. It seems to be of high priority;
anyhow, it's great. And for sure there are another teams that keep that
level of flexibility.
Conclusion
I think that developers deserve respect, and users who report problems
correctly deserve it too. We all invest time and effort. Reporting the
problem in understandable and reproducible way is said to be "half way
of solving it" and it's not easy. Not many users are able doing this and
I think that good reports AND GOOD REPORTERS (not that I consider
myself) are valuable property.
If nobody cares to solve them, let's face it. Be loud but fair telling
that. Keep the appropriate priority and say "noone will fix this for 10
years" and user would decise, whether he wouldn't use OOo for that time,
or start learning programming, or hire someone to fix it. Let there be
"stable" releases with known critical bugs. Politically unbearable?
Denying, hiding or bypassing the reported problems wouldn't make anyone
happy -the affected users get confused and would eventually turn away,
and new ones would face and (hopefully) report them again, or even turn
away immediately. Someone would have to read the reports again. Etc.
I know that exciting new functions gain much more fun, however getting
own code to solid, nearly-bugless state should also be cause of honor
for every programmer, shouldn't?
Epilogue
Reporting bugs and watching them solved has excited me with projects
such as OOo, and also Mozilla, Mozilla Firefox or Thunderbird. However,
two of them have just passed the "baby" stadium some year ago, and
became quite solid and serious, so that it's harder to find severe bugs
in there. Yes, there are glitches, but it is real pleasure to use them.
E.q. with Firefox, I have had no objection from any user ever! In few
cases, I've faced crappy webs.
Not so with the OOo. It's pitty but my users consider it to be
low-buck-low-quality product because they face code or usability
problems sometimes, either mere bugs, or annoying behaviour, or useless
interface (namely MAIL MERGE one).
The OOo is PAINFULLY NEAR to "JUST WORKS" for PAINFULLY LONG TIME.
Every admin knows that people like to shuffle, tend to make the computer
responsible for their failures. It is even harder to advocate the OOo,
because sometimes they are just right. E.q. the mail merge is bad enough
so that it forced us to buy several M$Office licenses.
There should be some clear strategic point of view: whether do we want
product with exciting functions but too much bugs to be considered
useful and nice by the end Joe user, or to have less shiny product that
just works, very stable and reliable. Or some kind of best balance
between these.
Thank You very much for Your time You've invested in OOo
Best regards
Peter Tuharsky
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
