On Sun, Oct 10, 2010 at 12:42 AM, Paul Wise
<address@hidden> wrote:
Apologies for the lateness of this reply.
On Sat, 2010-09-11 at 13:40 -0400, Robert Pyron wrote:
> I'm implementing a CHM reader in Python, based on Jed Wing's libchm,
> and guided by chmspec. So far I am able to read (and extract useful
> information from) #STRINGS, #SYSTEM, #TOCIDX, #TOPICS, #URLSTR, and
> #URLTBL. There is enough functionality already implemented to be
> useful to others, and I am almost ready to make an initial source
> release.
Is there any particular reason you are starting from scratch instead of
just using one of the many existing readers?
Oh, the usual reasons:
-- To scratch my own itch
-- None of the existing readers has everything I need, as far as I know.
I really want multiple tabs, plus the ability to resize text. Also, TOC,
index, and search.
-- An opportunity to learn a new technology (new to me, at least). My
current plan is to implement my viewer with Mozilla/XULRunner.
-- Kchmviewer comes very close to what I'm looking for. However, the
Windows version is built using the Qt HTML viewer, which in some
ways is seriously broken. For example, various versions of
Python##.chm are unreadable with kchmviewer.
-- I'm also aware of Ling Li's CHM Reader add-on for Firefox
(
http://sourceforge.net/projects/chmreader/ ). It does not function as
a stand-alone viewer. I will probably borrow some code from that project.
> I need to decide on a license for my code. Since I connect to libchm
> (which is LGPL), and since this code is not intended to be
> stand-alone, I would like to release it under LGPL. My problem is that
> comments in my code quote liberally from chmspec, which is GPL.
>
> Is it possible to get an exception to GPL in this case, as long as the
> quoted text remains as part of my source file (or files)?
Regardless of the licensing question I would prefer that you refer to
the spec itself rather than duplicating part of it in your project.
I've been thinking about that. I really like having pieces of the spec right
in front of me (i.e., in the source code) while I'm developing. However,
I can write a simple preprocessor to strip these pieces for a release
version.
On the licensing stuff, I'm not sure the GPL (or the LGPL) is the best
license for documentation anyway. In your case it would mean you would
have to distribute the DocBook source too. I need to think some more
about the right license.
Ouch. I just wanted to quote the documentation in my source code.
Now I'm definitely thinking the preprocessor idea is the way to go.