This is true for the client.
Robert did work on fuzzy matches at the server side, later that month.
> Is the cdindex being developed at all?
Yes, July and August have been unproductive months concerning the
CD Index.
However this should not be overrated.
It is just that Robert and I had not much time for this project lately.
In my case it is a combined effect of heavy load at work, leaving
me mostly Friday-Monday and a surprisingly nice summer weather.
This meant my wife got a bigger share of the free weekends
(doing such horrible things like dragging me from my keyboard to
places like France and Holland) and thus programming got a smaller
share. Now that the lousy weather has started again, you can expect
to see more output from me.
My remaining free time from July on went into a project, of IMHO
very high priority to the free operating systems -accelerated 3d-
where development began to explode around that time.
While it is nice that open source has made it to the main stream
press, and that huge numbers of distributions are sold and the
stock markets trust gross amounts of money into the sellers, it
is very interesting to observe that these large numbers (and thus
large numbers of users) do correspond with only a relativley
small crowd of developers that are spread over lots of projects.
The 3d effort is not too large either: Mesa ~3-4, open glx ~4-5,
FreeBSD 3d integration ~2-3.
What I want to make clear with all this talk is that our little
CD Index project consisting of 2 gofers plus ~3-4 feedbackers is
not unusal.
> I was thinking of adding some support for it in Xmms, but I don't feel
> like writing an XML parser and everything to interpret the results.
This is something I could write for you. We just have to agree how
we interface. Send me a private email, and tell me how you want to
provide and receive data.
> Are there any ongoing work on the cdi_lib?
Let's say it this way. We have ideas and intentions but no
code yet. (see another mail, I'll post on that subject)
Easiest would be to connect your player to our server.
You know your player, we know the database.
Should be easy. And right now I rather have a crappy lib
as the result of this than no lib.
> BTW, I don't like the idea of providing a general cdi_lib interface in
> C++. A regular C interface is preferred if cdindex is going to be
> integrated into other applications (which probably is the whole idea
> of having cdindex in the first place).
See the mails from June 28th, 29th.
C is the way to go.
Regards,
Marc