[Debtags-devel] Re: Recent progress

[Benjamin Mesing]

Hello, 

I wanted to add my opinion towards the time stamping issue. I think that
timestamping is a good thing to do to support the maintainers. It should
not be too hard to give the tags a date when they entered the vocabulary
would it? Another option would be to regulary release a new version of
the vocabulary and give it a version number 0.0.0.1 :-) A new field
could be added to the package information which carries the version of
tagging information used to tag the package (perhaps like the standards
version). 
This information is sufficient to allow the maintainers to check which
tags were added to the database, they could check them add the
neccessary ones and increment the tag-version field in their control
file. Of course we have to offer a tool to print the tags which where
added between to versions of the vocabulary. In my opinion this is a
quite unintrusive way that will integrate smoothly in the packaging
process.
 
> > More tags => more work => less adoption
> 
> Haha. So let's reduce tags to the very minimum. User utility reduced to
> almost nothing, but developer adoption garanteed!
> Sorry for being sarcastic.
I agree with Erich here. If the number of tags get too bloated tags
itself might become unusable. No one will know the whole vocabulary than
and this will make using it a time consuming way, because you would have
to browse hundreds or even thousands of tags. This is nothing I want to
do! If I want complete freedom I use a full text search entering
whatever I like. In fact this is the way I do most of my searches now.
Additionally you will scare of the package managers if they have to scan
through a huge vocabulary to tag their packages.

My opinion is to make the vocabulary not too detailed. If you have
something very specific in mind, for example a tool to manage your
bookmarks choose e.g. "web::browser" combine this search with a full
text search for "bookmark" oder "manage" or even both and most likely
you will have what you are searching for (I just tried it, but currently
this example does not work well, because neither gnobog,
egroupware-bookmarks, or sitebar was tagged with web::browser -- I found
those packages by performing a full text search for bookmark && manage).
I know that a lot of you will disagree with a vocabulary being not to
much detailed and combining it with a fulltext search. The whole point
of a vocabulary is to speak in common terms so to avoid inconsitent
data. And I think you are right -- as long as professionals are involved
this would be the way to go -- we need something with the focus on "easy
to learn" in opposite to "easy to use". A search for tags might help to
cope with a complex vocabulary (as I offer it in my packagesearch
application) but I think it is improtant to keep it not too bloated.



Btw.: Enrico, you promised me to rename devel::environment to devel::ide
some time ago :-)

Greetings Ben