Re: [greenstone-users] Displaying multiple metadata values individually?

From jens wille
DateWed, 17 Aug 2005 18:32:39 +0200
Subject Re: [greenstone-users] Displaying multiple metadata values individually?
In-Reply-To (43035AC0-8010809-atp-rub-de)
hi axel!

schild wrote:
> thanks for the comment. You are right, the outhtml only works for
> IExplorer such that in my latest versions of the javascripts I
> changed them to also work with mozilla. I think I already posted
> this solution a while ago on the list ("Re: [greenstone-users]
> Referencing & querying the search engine", 27.06.2005).
sure, just forgot about that posting ;-) i will try it some time.

> However Jens I must say I disagree with your comment that this is
> only an "undesirable hack". Sure there are probably better
> solutions from a computer scientist point of view but that is
> what I can claim for my solution: It works fine for me (never
> caused me any trouble at least), and leaves a lot of flexibility
> for specific (personal) needs.
well, i didn't mean to disrespect your effort, it was a great idea.
it's just that it doesn't appeal too much to me: apart from the
difficulties writing browser independent (let alone /working/
standards compliant) javascript, it requires quite a bit of effort
on the user's side (at least at the moment).
besides, it utterly requires javascript to be enabled in the
user's/visitor's browser (usually, this only results in not having
these flashy image effects, but in this case it's a bit more spoiling).

but you're right, it's flexible as hell :-)

> I think that you could implement this feature in C++ and coping
> for the requirements of all users (different search engines,
> different index names, etc, some people would want to search for
> a combination of metadata values, etc) but I think it won't come
> for without a larger effort (and will be most likely not
> generic).
yeah, it might be _very_ difficult to implement - nonetheless, i
consider it the better way.
imho, adding functionality this way cannot be more than a temporary
solution - it just belongs to the core of the application!

> But if somebody feels up to adapting the C++ code I will be
> grateful as well...