Web interface for gnuCash

[Conrad Canterford]

On Tue, 2006-08-08 at 11:56 +0200, Graham Leggett wrote:
> I have reached the point where Gnucash in it's current form is no longer
> useful, and I have to make a decision to either a) develop the software
> further and submit patches for inclusion into the trunk, or b) drop
> Gnucash and invest the effort in something else.

Or c) take a fork, and develop it in the direction you want it to go.
Personally, I think that would be bad for all concerned, but if you feel
strongly enough about it, this is an option for you.

> So far, evidence suggests that even if I coded up patches, they would be
> rejected, because, as you state very clearly, "this is not out goal for
> Gnucash, period".
> This is very unfortunate, as Gnucash has an excellent user interface, and
> would make a very good platform for a larger accounting system.

The point that Christian and Derek are making is that it is not the goal
of the developers, and accordingly they are not going to support any
change that requires a lot of work from them but gets them no nearer to
what they want to achieve with the program. If you can come up with a
plan and process that will achieve your goals, and which does not
require them to completely rewrite the rest of the program, your plan
will be considered on its merits (primarily for stability, future growth
options and long term maintainability).

> > Not helpful. It rather seems to me you haven't understood how such a
> > volunteer-based project works at all.
> Having been involved for many years with one of the largest (and oldest)
> volunteer software projects, I would disagree, but then you are free to
> believe what you will.

Surely you must understand that the current developers devote staggering
amounts of their own personal time to develop and maintain gnucash. They
each of them have their own bugbears that they work on to bring the
program closer to what they want from the software. If I may dare to put
words into their mouths, that and a desire to preserve probably the best
free software accounting program are their primary motivation for the
considerable amounts of work they do for our benefit.

You cannot fault them for rejecting any suggestion that they should
devote the next several years worth of their precious development time
to your suggestion instead of working on issues and features which are
affecting them or which they believe are important. This is all the more
the case given that we have just come from a period of development which
has been very short on new features because of the significant amounts
of rewriting that were necessary to implement the gnome2 conversion. You
can hardly be surprised that a suggestion that you want them to do the
same all over again (only bigger) is not leaped (is that correct
American spelling?) on with great joy.

Now, if you are actually suggesting that you want to undertake this
project, and that you are prepared to make it work alongside the current
gnome2 interface, then I dare say the developers will be very happy to
hear your thoughts on how you might achieve this and will offer
constructive suggestions on what is and is not workable.

Personally, I can see some value to a web interface. But there are some
very fundamental issues that need to be addressed before I think such an
interface is a viable option. The first and foremost of those would be
the ability to support multiple simultaneous users, and the second would
be to complete the separation of the gui from the logic. These both
introduce further dependencies (not the least of which is addressing the
issue of the data storage backend).
In short, you are not suggesting a small project. If you want to see
gnucash move in that direction, by all means start planning and
discussing it. If you were to undertake one of the outstanding issues I
pointed out above (or others I'm not aware of), I suspect you will be
quite actively encouraged and helped. But please do understand that we
do not have a pool of developers to throw at the idea. You will be
supported and encouraged as long as you are not breaking things, but
unless you can attract other people to assist you, the current
developers have their hands more then full enough with what they want to
achieve.

Conrad.