Hi Michael,

michael meeks wrote:

>       I understand the process is not your fault, (my frustration is not
> primarily with you :-) - and that clearly some simple level of
> functional spec is necessary for QA to be done etc. But please don't
> defend this over-complicated, painful and substantially pointless
> process.

The process is not a fault at all. I don't want to start a discussion
about the spec process on this list, but I also didn't want to ignore
your (IMHO) unqualified rant.

A specification is necessary to get the approval of all stake holders (I
hope you don't deny the necessity of getting these approvals), so
nothing that might be relevant to *any* of them should be omitted.

You might argue that the specification does contain too much things that
*currently* appear useless, but you should consider that none of us has
a working crystal ball, so we don't know wich information about a
particular feature might be important in one or two years. So we try to
have everything in the spec that IOHO *might* be needed later. This is
the same motivation that lets us write our code in a way that it deals
even with strange conditions that might never happen (OK, at least we
try to ;-)), something I assume you will not see as substantially
pointless (though I agree that this also is painful at times).

A specification shows its value over time. If you need to change
something later it's always good to know why this something was done in
the way it was done. As an example, it helps you to avoid damaging
something unintentionally later on. A bad (or missing) specification
very easily leads to a poor implementation quality with a large
potential of regressions in later changes. I can show you enough "good"
examples where exactly this happened because people provided only "some
simple level of functional spec" as you suggested.

I understand that you obviously are not interested in the stuff any
longer once it is integrated, but you should respect that other people
have to deal with this code a little bit longer. Of course it might be a
little bit painful for you to do some extra work that *you* don't need
at that time (the "core" developers feel the pain as you do!), but
please try to understand that a product like OpenOffice.org is the work
of a team (creating a software product is more than just code hacking)
and that we need an organized way to keep the team fit for work.

Nevertheless I hope to see you at the OOoCon and I'm looking forward to
having some interesting and inspiring conversations with you, of course
also about the spec process. :-)

Best regards,
Mathias

-- 
Mathias Bauer - OpenOffice.org Application Framework Project Lead
Please reply to the list only, [EMAIL PROTECTED] is a spam sink.


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to