Base

Hi Tom, *

The problem with the Base guide is not that people aren't keen. There are at
least 2 people that joined Documentation specifically to work on it rather than

Indeed.

on other guides. The problem seems to be that Base is a bit quirky and keeps
suffering odd regressions. What works and what doesn't is very unpredictable in
each release and even what 'should' work is not always clear.

Unpredictable...is an understatement. Many things that used to work in
OOo 3.2.x, and for which the pre-existing OOo Base documentation was
written, now either no longer work at all, or only partially.

Paradoxically, the built-in help was actually much more detailed in OOo
before the switch to HC2/wiki-help, and since then it has been quite
simply unhelpful. Try carrying out a meaningful search in the Base help
and you will soon find out why, it is enough to drive anyone nuts.

Add to that the new features that were added to OOo during 3.x codebase
development, many of which were either not fully functional (charts in
reports), buggy (SQL parser engine), or quite simply got screwed up when
the LibreOffice project took over (see current Base bug list), and you
have a documenter's nightmare in the making. The bug fixes to Base that
the OOo developers had in their child workspaces whilst Oracle was
pulling the plug on OOo have not been integrated into the LibreOffice
code because there is a lack of clarity as to whether these are indeed
freely available (since by their very nature, they were CWSes by Oracle
employees, not trunk code, and thus not yet approved for submission into
the main source tree).

We need to attract devs to work on Base. At the moment there are none
apparently. Perhaps now that IBM contributed all that code to Apache the
OpenOffice Base might take a large step forwards but again that makes LO's Base
difficult to write for unless we can get that code base into LO too.

The devs would not only need to work on Base to patch it up, but also
provide documentation for the functions that get developed / completed /
corrected.

As for the Apache project, I can't see it being able to produce anything
of that ilk in any short timeframe, as the people there are still trying
to work out which bits of code they are actually allowed to integrate
into their source tree. They are still faced with many licensing issues
about various parts of their project, including how, or whether to, host
the documentation that has accrued over the years within the OOo
project, because a lot of it is under an Apache-incompatible licence.
Things might get sorted out, but it will be later, rather than sooner.

To put it bluntly, Base is currently in some kind of interdimensional
space where it is being left to degenerate as other parts of the code
progress.

Alex

Hi,

To put it bluntly, Base is currently in some kind of interdimensional
space where it is being left to degenerate as other parts of the code
progress.

I would *very* much regret if that were to happen... I wonder what we
might possibly do, maybe acting together, to agitate for some remedial
work on the component. I see Base as being an asset with great
potential value.

Hi David,

I would *very* much regret if that were to happen... I wonder what we
might possibly do, maybe acting together, to agitate for some remedial
work on the component. I see Base as being an asset with great
potential value.

I fear that developer ressources are too thin on the ground for that to
make much difference, or else, as I have been told in the past, get a
support contract and pressure that support company to fix things within
Base.

I loved working with Base in the previous versions of OOo, and if I had
the time, I would certainly delve into the current code and fiddle
around, but alas I do not earn my living from writing/repairing code and
it would take me a very long time to learn how to do so :slight_smile:

Alex

Hi :slight_smile:
I think the first step would be trying to get Apache to give TDF a copy of the
code that IBM gave them. I think Base in Symphony was more advanced.

I think a more time-consuming task would be to find out who the devs are that
worked on Base when it was under Sun. Then perhaps we could write to them and
ask if their work could be released under the LGPL.

Regards from
Tom :slight_smile:

Hi Tom,

I think the first step would be trying to get Apache to give TDF a copy of the
code that IBM gave them. I think Base in Symphony was more advanced.

I get the impression that you haven't used Symphony and its DB
capacities, or I doubt you would have stated that. It has a fairly nice
UI overall, but it is crap for DB work (and I'm putting that mildly).
Unless I have really missed the point, all Symphony offers is a JDBC
connector setup tool, which is as about as much use as a hessian bag
with weights in and tied around your feet to swim across the Channel in.

Once you have the db connector source set up, all you can then do is
copy a whole table into a Calc sheet - wow !! Fairly blew me away - not.

When I think back to how good Lotus SmartSuite was, I really feel sorry
for those who followed IBM down the path of give up SS for Symphony.

IBM's view is very server centric, i.e. you should have all of your
views, queries, procedural language, etc, running on the server backend,
and basically the client-side gets the crumbs. This might be fine for
large organisations that have structured IT departments, but for
everyone else it is far and away below expectations.

You can not even open or create a database document with Lotus Symphony.
Queries, forget it. Reports, forget them. Table management, forget it.
Sorry, for me, it does not even qualify as "second grade", more like
"third world".

I think a more time-consuming task would be to find out who the devs are that
worked on Base when it was under Sun. Then perhaps we could write to them and
ask if their work could be released under the LGPL.

Ah, Tom, ever the eternal optimist :wink:

I know who they are, but most of them, if they are still in employment
at Oracle are muzzled, or else they will perhaps, just maybe, be working
on the Apache project. However, of those former Sun/Oracle developers
that have showed up at the Apache project, I have yet to see a single
Base developer chip in.

Alex

Hi :slight_smile:
Well that's a bit disappointing but also a relief. So, we need to somehow
attract new people into a quagmire.

Regards from
Tom :slight_smile: