[Kexi] KOffice status survey

[Jaroslaw Staniek]

Inge Wallin said the following, On 2007-12-04 14:06:

> We were discussing the tagging of the next alpha together with KDE 4.0RC2 on 
> IRC today, and the topic of the current status of KOffice came up.  It turns 
> out that nobody really knows the full status, so we decided to do a little 
> survey.
> 
> Would the maintainers of each component please reply to this mail in the 
> mailing list and state here:
>  * which features that are still not implemented, but which you want for 2.0?

* Turning KexiDB as Predicate library _out_ of KOffice
* putting CSV import/export back for reusing in kofficelibs

>  * which features that exist, but are still far from usable?

Not ported in Kexi (and not using Qt4 improvements): Forms, simple printing
Those alone can be treat as whole applications.

I imagine very long period before there's chance for any freeze in Kexi 2.0.
The reason is as obvious as KDE 4 - we and particulary I, need to split my 
time and give some chance for KDE 4 internals, especially Windows port for the 
KDE framework and some strategic KDE applications other than KOffice.

> You do *not* have to name every feature in the program, and especially not 
> those that only have polishing left.

We know the main showstopper for any app would be the level of support of 
underlying file format, to avoid data loss and users frustration. I believe a 
failure here would be worse than delays in our situation.
For everyone it's ODF, except for Kexi which maintains its own file format 
consisted of XML subformats (QT Desingner's), some custom ones. Chunks of ODF 
are the matter of the future beyond 2.0 (e.g. embedding whole ODT templates) - 
ODF provides nothing for database apps except connectivity information for 
apps like word processors, and this is not especially bad thing.

On top of that Kexi's storage model would improve by migrating to the newest 
stable SQLite. This is not started and isn't a matter of replacing a library - 
semi-automatic import/export features and/or conditions backward compatibility 
have to developed.

Scripts, what is another whole big layer, can take extensibility to the next 
level - we do not maintain the object model yet - one that can stay solid for 
years allowing for backward compatibility. Scripts are central part of Kexi, 
as it's more development tool than any other app.

Adam Pigg is implementing introductory stages for object model in his much 
appreciated reporting module [1]. There is not much reuse, largely because of 
we're short on manpower. Again, the reporting module is something like a whole 
application.

To put more vivid thing at the end, we are going to have one more driver in 
2.0, providing Sybase and MSSQL db support. This one comes from a new, 
vigorous developer Sharan Rao, which I met at Akademy this year.

[1] http://kde-apps.org/content/show.php/Kexi+Report+Part?content=58208

-- 
regards / pozdrawiam, Jaroslaw Staniek
  Sponsored by OpenOffice Polska (http://www.openoffice.com.pl/en) to work on
  Kexi & KOffice: http://www.kexi.pl/en, http://www.koffice.org
  KDE3 & KDE4 Libraries for MS Windows: http://kdelibs.com, http://www.kde.org