I wrote a little while ago about developing a Windows conduit for Plucker,
and this is an update on where I am.
I originally wrote the spider/parser from scratch and was able to get the
web pages, parse them, and create the database on the Palm without
problems. This works well for our company now since we control the content
that is sync'd. However, there are a number of shortcomings to this
approach, so I've decided to switch tactics.
I've now started (and am a good ways through) converting the python code to
c++ and using it in the conduit. Right now, it will read text files and
put them on the Palm. I've kept the same class structure as the original
(1.1?) Plucker code so updates will be easier. The structure of the code
also allows me to easily create a standalone (non-conduit) version of the
spider/parser/writer. I intend it to work exactly the same way as the
current Plucker does except it works quicker and on demand when a hotsync
is done.
Since I am essentially copying the GPL'd code, I am assuming that my code
would be GPL'd as well, but I don't really understand the GPL. This is
fine and I will share with whomever wants it when I get something decent.
I am using VC++ 6.0 on Windows, but from reading the conduit docs, the code
& classes are supposed to work on the Mac as well.
I also created a web based channel selection program in java & mysql
running on Linux (or windows) to allow users to select channels and then
sync their individual channels when hotsyncing (a la Avantgo). It's okay,
but I don't think our company will end up using it (except for me).
Any comments?
Bill
P.S. again thanks for such a great product.
