|
Boost : |
From: Gary Powell (powellg_at_[hidden])
Date: 2001-11-15 12:34:13
>>
Please remember that in a corporate setting individuals wishing to
use boost in their product or its SDK must solve a number of
problems, not all of them technical. Reuse is often a driven by a
bottom up, grass roots effort. Engineers must solve technical
problems, educate peers and management, and address the "not invented
here" syndrome. Reuse efforts can easily be derailed; obtaining
approval from the legal department can be a major hurdle. It may seem
like large corporations would have sufficient legal resources to
examine multiple licenses. This might be true if reuse was a
corporate wide initiative. But, for bottom up efforts, legal
resources can be scarce. Besides the problem concerning the number of
licenses, some legal departments also have problems with the license
in some of the libraries that requires boost related copyright
notices to appear in 'user documentation'. As I see it the goal is to
encourage as many programmers as possible to use the boost libraries.
At least some programmers are only seeking more uniformity from the
licenses. This is not a radical change or some kind of watering down.
<<
Ok,
After having slept on this idea/problem for a few nights, I now agree that
we should revisit it. Both for the problems with foreign language
translation, general acceptance etc. It is an indication that boost is
"growing up" to be more formal but this is a good thing here. (It's like
unit testing, a pain but you know you need it.)
The problem with the current set of copyright/copyleft's, is that
corporations and institutions insist on their own copyrights and merging all
of these into one totally acceptable copyright is a large hassle. (Witness
the current mess we have.)
For instance, I would prefer that "boost" get mentioned in user
documentation, of course to a non code slinging user, its meaningless, but
to other engineers it might promote more uses of boost. And what does
"mentioned" and "user documentation" really mean? Are online links
sufficient? Do the authors and contributors need to be listed? Do you have
to mention libraries, that are in boost, but that you didn't use? Simpler is
probably better, but those members who are consultants would most likely
like to have their name listed. (Might encourage them to submit more work as
a method of self advertisement.) But I'd rather not prevent a library from
being accepted because it had this requirement, if we decided that was too
much.
The next step of course it to get all the current authors of boost libraries
to agree to modify the current copyrights, and then distill the
"basic/essential text" and the poll them on this new copyright and edit all
the files. (I suppose this should be done sooner rather than later as the
editing job will only get worse.)
So Mat, if you are volunteering, this job needs a point person.
Yours,
-gary-
PS
Just to make the task it more manageable, I would suggest that only the
current accepted library authors need be considered on the precise wording.
(Ok, you can clutter the listserv with "final" drafts, but the initial
drafts need only go to those most affected.)
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk