Lars Gullik Bjønnes wrote:
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| Lars Gullik Bjønnes wrote:
| > Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| > | | You've just described my very plan ;-). I just thought it was
| > obvious
| > | so I didn't care to write it so explicitly. The problem is just that
| > | Lars won't allow step 1.
| > Very right. If the bus thing happens after step 1, we are left with
| > no
| > spellchecker.
|
| Man, we are not talking about the linux kernel. This is not rocket
| science we are doing here. If I am hit by a bus, another competent
| developer will be able to continue the work. We have quite a few
| here...
|
| So Angus, is this "NO!" clear enough for you? ;-)
I still do not understand why you have to do step 1 first, and why
this cannot wait until last.
If you did your homework, you would have understood ;-)
Fact 1: The SpellBase interface doesn't encapsulate well the fact that
the engine is a library or a process.
Fact 2: The controller has to test for the cases where the inner engine
is a process
Fact 3: The controller is also doing some views (in the MVC sense) which
is bad.
Those three facts make the spell thing pretty complicated. So the only
solution (IMO) is to cleanup the API (the base class and the controller
algorithm).
Cleaning one API is much easier done with one engine. Once the API is
set and the logic flow is good, we can add any other engine we want
(ispell, pspell, enchant, etc).
I don't want to care with three engines in the interim period. This is
just a waste of time and I have better things to do with my time. So for
me, step 1 is a must. If it is not for you, then please do the work
yourself.
Abdel.
