> I think it would be useful to have a stand-alone gui program which would
> have as inputs a URL to be converted and checkboxes for many of the
> command-line parameters of "plucker-build". I guess what I'm thinking
> of would be an extention of the plucker-build gui program.
(tapes Mike's mouth quiet =)
I started this a long time ago (November of 1999), in GTK, before I
had my life stripped from me and sold to the dot-com corporation I currently
work for. I could probably pick it back up and complete it at some point,
but I've been pretty-much shopping it around for a maintainer (someone more
familiar with gtk and toolkit programming than I am). An old screenshot can
be seen here:
http://www.gnu-designs.com/pd-buddy.png
I started thinking about a better GUI for this, a new interface
design, which would be more portable to Windows as well as the Linux and
OS/2 users. A web-based gui would work too, but it would require a plugin,
not ideal.
I'll put something together this weekend, I have a few
Plucker-related projects on my plate, and part of my weekend is dedicated to
addressing them.
(Incidentally, I fixed the snapshot breakage. It was related to when
I renamed the module in the cvs, and worked after that, but I restored that
script from a backup which was prior to that date, and screwed it up
recently. OOPS!)
> Then if I'm browsing in Netscape or Konqueror, I could highlight a URL
> of a document that I want to convert, and paste it into this program,
> and quickly select the depth, stayonhost, image-depth and other
> parameters without having to enter the exact syntax of plucker-build.
Actually, I'm going to be also adding the ability to gather the
content from the Build-It page soon, after I add the new options from the
parser into it. That will work for now. I've been tinkering with a side
project I started back in January or so called PODS, the Palm Open Directory
Syndicate (yes, it needs a new name) which currently holds 543 separate
Palm-formatted URLs in a mysql database. All of these were hand-scraped from
the web, search engines, websites, online forums, etc. and are all
publically accessible.
Once I wrap a system around that which allows people to "log in" and
store a "profile", I can then begin to automate this process. In the
meantime, I'll be putting a few new tools on the website which will allow
you to select from newspapers (dropdown + submit), sports (dropdown +
submit) and so on, and people can gather their own pdbs runtime. It's
getting XML'ized also, because with PODS, that's the best way to deal with
these URI objects.
Until I figure out the legalities of storing off-site content
(transcoding, meaning offsite from the original publisher of that content,
cnn.com for example), I'll have to be careful about how I approach it.
Thanks for the tip and motivator. The top Plucker things on my plate
are (in no particular order)
1. Fix the Plucker Bookmark Assistant (PBA), and add a region to
deal with a toolbar folder which Plucker bookmarks can be
"plucked" from. Also grok IE, Opera, links, lynx, and Mozilla
bookmarks. If someone has Konqueror's format, send me an example,
so I can add that also.
2. Update the Build-It section to handle the new parser options. Add
the ability to actually GATHER the content online as well.
3. Bug tracking goes live. A few things have to be squeezed out of
it before it goes live, and Mike had a formatting request for one
of the bug pages, so I'll fix that and put it up for everyone to
work their bugs into.
4. Plucker Desktop Buddy (PDB) fixed and xml'ized with the new GTK
class structure (gnome-xml and friends) which allows us to change
the GUI to be anything we want, abstracted from the actual source
code of the application logic.
5. Perl parser needs a revisit, and a rewrite to make it current
with the Python parser. It too will handle HTML and XML content
as well as the .site files from Sitescooper and the .subs files
from AvantGo.
Keep the ideas and suggestions coming!
/d
