Re: htdig: Windows??


James Mcpherson (j.mcpherson@mailbox.uq.edu.au)
Mon, 27 Jul 1998 22:40:28 +1000 (GMT+1000)


On Mon, 27 Jul 1998, Torsten Neuer wrote:
[snip]
>
> Why not then use Ada? This would provide you with even more machines
> where htDig can be run on, since there are still a lot of machines with no
> Java byte-code interpreter available for them.
> It would also cut down the bugs, since the language is known to be able
> to verify most of the programs during compile-time.
> Despite the use of interpreted byte-code, htDig would stay a native binary
> executable, thus not slowing down the processes of indexing or searching.
> With regards to database interfaces, Ada comes with interfaces to most of
> all commonly used programming languages.. even COBOL, so there should
> be no problem at all to create bindings to whatever database there is on the
> market. I've worked quite a lot with Ada programs in the past that did this
> interfacing without causing any trouble.

[Please don't anybody take this personally - it's my opinion only]

I don't think that is such a good idea. how many people out there have
access to a free ada compiler? Is the gnu version up to scratch compared
with commercial versions? (I'm not sure that it is). And one other thing,
it is my firm belief that if you write a program in a language such as
c, c++, java, perl, lisp or pascal then it is much easier to (a) maintain,
(b) hack on and (c) acquire the good will of sysadmins who have to install
it than if it's written in a language such as modula-2/3, ada, eiffel or
other languages which do not have such a large base of competent
programmers. I know that this is a rather flameable comment, but if you
are in the position (as I am) of having to maintain the search engine for
a corporate website, and want to be able to go to the source to fix a
bug (if necessary), then you don't really want to have to pick up more than
one or two major languages in order to get the job done. Hence, I guess,
the popularity of perl.

Personally, I would like to see a new version of ht://Dig written such
that there was a core engine with a well-defined api and sample
reference implementation of UI functions which people could hack on
in other languages if they so desired. I think that this would give
local ht://Dig maintainers the capability to be more flexible in their
delivery of their service, which is something that we see more and more.

If people can wait a few weeks, I might be able to produce something
like this myself (I'm about to change jobs).

How does this sound?

cheers,
James C. McPherson
Unix Systems Administrator
IT Service
University of Queensland Library

-- 
 I do not speak for the University of Queensland or the University Library
J.McPherson@mailbox.uq.edu..au   | this space _not_ for rent. 
j.mcpherson@library.uq.edu..au   | remove erroneous char to reply properly

---------------------------------------------------------------------- To unsubscribe from the htdig mailing list, send a message to htdig-request@sdsu.edu containing the single word "unsubscribe" in the body of the message.



This archive was generated by hypermail 2.0b3 on Sat Jan 02 1999 - 16:26:55 PST