On Wed, 28 Feb 2001, Michael Haggerty wrote:

> I'm a newcomer to the ht://Dig community, but I'm about to go
> elbows-deep into the source code--specifically for the htdig
> component--to add some features required by a client.

Splendid!

> replacements for the standard components.  Has anybody considered the
> need for a big-time refactoring of the classes, to try to tease apart
> separate elements of the functionality?  Ideally I'd like to package
> some of the more useful components as separate libraries that could
> also be used outside of ht://Dig.

A lot of the work in the 3.2 tree has been to do exactly this. The new
htword/mifluz project is an entirely separate library and I suspect
someone could take htnet and make it into a new library too. One of the
main stumbling blocks on the search side is that htsearch needs to be
refactored/rewritten. We have a new parser/query system contributed by
Quim Sanmarti, but I have yet to start replacing the current htsearch.
(When I do, I'll let people know because things will suddenly break.)

At this point, I'm not going to personally do too much more in terms of
infrastructure for 3.2 because I want to see the thing get out the
door. However, I always consider contributions!

> to get temporary read/write CVS access to the source code on
> SourceForge?

Done. It will take some time before things sync, though. BTW, the current
development on 3.2 is on the htdig-3-2-x branch and the mainline is pretty
broken. You should either repopulate the mainline from the htdig-3-2-0-b3
release or branch off of htdig-3-2-x.

> If that is not acceptable, what is the feeling of the existing
> developers getting lots of patches from me?

I don't mind patches. :-)

My suggestion is to take a look at the 3.2 code if you haven't already and
think about what sorts of "components" you need besides what's there
now.

Cheers, (and welcome!)
-Geoff


_______________________________________________
htdig-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/htdig-dev

Reply via email to