[htdig-dev] Nearing first complete version of HtDig 4.0

[Neal R. <ne...@ri...>]

Hey all,
   Anthony Arnone and I are nearing the completion of the first version of 
htDig 4.0.

   At the moment we're fixing up the native Win32 build of it.  The next 
few steps are as follows.

   1) Polish new htdig.exe
   2) Final code/directory cleanup
 	(perhaps a reason to move htdig to subversion??)
   3) Polish build process
   4) Formal QA process

When #4 is begun we'll formally announce the new version as beta and 
update the htdig website.

   We'll also be tracking & including the current changes to CLucene 0.9 
until we reach a point of pain to be determined later.  After that we'll 
lock the version of CLucene used in htDig 4.0.

   The tarball used to build HtDig 4.0 will include versions of CLucene, 
HTML Tidy and Berkeley DB (already included in htDig 3.2).

-------

I would like some feedback on htsearch.exe.

I actually think we should cease providing a compiled cgi for searching. 
We should ship a polished PHP script for searching and possibly Perl & 
Python scripts as well.  There are various bindings for CLucene, so all 
of these can be made into htdig bindings by adding a bit of glue code that 
looks at the htdig.conf file etc.

Why?  Nearly every system this will be installed on will have PHP, Perl or 
Python.  Any of these 3 scripts would allow much better customization of the
output by users.  Continuing to maintain cgi code in C/C++ is not very 
attractive when these scripting languages do all the work for you.

What do you think?  Any strong objections?

Note that one of the most used contributions to htDig 3.1.6 & 3.2 is a PHP 
script which calls the htsearch cgi/exe and parses the output to allow 
customization...

Thanks

-- 
Neal Richter
Sr. Researcher and Machine Learning Lead
Software Development
RightNow Technologies, Inc.
Customer Service for Every Web Site
Office: 406-522-1485