httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Stein <>
Subject LGPL (was: faster malloc for multithreaded programs)
Date Thu, 04 Nov 1999 03:05:40 GMT
On Wed, 3 Nov 1999, Ben Laurie wrote:
> Speaking as chair of the license committee I think it is a useful
> discussion. I'm somewhat upset that even LGPL is too dangerous (but if
> that's really true, where are we with standard C headers and so forth?).

We don't ship the C headers/libs as part of our distribution, so there
isn't a problem there.

> I'd be particularly interested to hear if IBM legal are changing their
> opinion.

I hope you mean this in the sense that "IBM legal is helping us out in
understanding the ramifications," rather than "we can ship/use the LGPL
and it is the Right Thing To do, but we won't if IBM can't use it."  In
other words, I hope that we aren't held back from doing the Right Thing.

[ and don't be silly... this isn't truly my hard-line position... I might
even shy away from my ideal here if it precludes IBM's assistance and
participation with Apache :-) ]

> However, isn't the crux that the _object code_ may be derivative? Do we
> care? Anyone wishing to avoid that problem can avoid the LGPLed stuff
> (so long as we make it optional, of course).

I thought the issue was putting LGPL code into our distribution. Sure, if
we make changes, we'll send them back to the author and include those
changes in our source release. However... will the LGPL affect our Apache
license at all?

IMO, it shouldn't. We aren't deriving from the library -- we're just using
it as part of a complete work.

But hey... I haven't looked over the LGPL in a long while...

> Also, it has just occurred to me, as the owners of Apache, we are
> actually permitted to publish under multiple licences. Hmm. Perhaps I
> don't really want to go there.

Good idea :-)


Greg Stein,

View raw message