Current state of the code (was: Re: [Demexp-dev] Issues on classification)

[David MENTRE]

Hello,

I've not spammed you in the recent weeks, it does not mean that nothing
is done. :)

If a take a previous point:

David MENTRE <address@hidden> writes:

> The remaning modules to write are thus:
>
>  - classification (only simple keywords and path matching)

done. More exactly, a sort of.

>  - server (taking commands from network and executing it). To accelerate
>    the development, the server will be single-threaded (poor performance
>    but easy to design, no locking issues)

first big step to do. Unfortunalty, at this step I would have to
integrate previously written module and I would probably discover that
some code is missing, is not well designed, etc. So probably more work
for the last percents of job. There is no such a thing as a linear
job. ;)

>  - network messaging (the minimal required set)

done. Or I think so. :) I've decided to use RPC method for segmentation
of XDR encoded messages onto TCP. It basically means that I follow
section 10 of RFC 1831. "I" is not a exact denomination neither. I fact,
all the dirty job has already be done by Gerd Stolpmann in his RPC
module so I have integrated his code (with few minor modifications). The
net effect is that part of the code should be bug free[1]. :)

Otherwise, I have defined a set of network messages that I think is
minimal for a functional prototype. Some of them would have to be
modified to match latest code integration.

>  - integrate Isabelle's work and make it emit network messages

second big step to do. I need to integrate nework code and probably
modify her code to match latest modification of networking.

BTW, d2, have you a patch to propose on isa's code?

Any way, the first prototype is not quite there but progress is made. :)

Yours,
d.

[1] Side note on software engineering and free software:

 - free software allows easy code reuse, and more importantly bug free
   code reuse. An interesting thing from a software engineering point of
   view. Of course, there is no such thing as a simple code copy/paste.

 - however, I had to modify slightly the original code, both at
   interface and implementation level. Yet another example of why a
   Software Component would had never been useful. Yet another good
   point for free software: the ultimate software component! :)
-- 
 David Mentré <address@hidden>
   http://www.linux-france.org/~dmentre/david-mentre-public-key.asc
 GnuPG key fingerprint: A7CD 7357 3EC4 1163 745B  7FD3 FB3E AD7C 2A18 BE9E