--- Jeroen Dekkers <[EMAIL PROTECTED]> wrote:

> It's not a done deal.

It _is_ a done deal.  If you want to change it,
rewrite all parts of plex86 from non-consenting parties
including me.  'nough said.



> The LGPL isn't a license for libraries, only often used for that.

Who cares?  Read the text again for the real message.

> I don't see what licensing has to do with architectural defects.

Architectural design, licensing, and many other factors
add together to form a complete "welcome mat" which
enables the sharing, contribution, interaction,
acceptance, draw etc for potential users/developers
both commercial and otherwise.

When a real architect puts together a plan, they
consider as many significant things as possible before
making the final blue print and certainly before
implementing.  What a lot of OSS/free/libre hackers
do mistakenly is considering only the technical and
philisophical parameters.  This is akin to designing
a sky-scraper without looking into building code.
Sounds good in class, sucks hairy ass in the
real world dude.

If you want a reasonably complex program to succeed
and you think it's gonna need the support of some
of the big boys, you better start thinking about
what makes them happy too.  This is where you can
take that GNU diatribe and flush it down the
toilet.

The GNU licenses in general were holding back
a lot of commercial support for quite some time.
I personally pinged RMS on a couple issues which
would have been simple to correct.  But the guy
_insists_ on maintaining his personal philisophical
spewing agenda.  When he could have simply corrected
problems at no real expense, and gained earlier
support.

So back to your comment.  GPL was _not_ acceptable
to previous relationships of mine, and LGPL was
the one people voted most for.


> There might be a reason for reinventing stuff. Maybe to do it better, or
> because you have another goal in mind.

Your point is well taken - sometimes you gotta reinvent
stuff.  What people are talking about is borrowing components
from other projects.  Which means, necessarily there
needs to be some commonality in an architecture to
facilitate that sharing.  Or one mother-fscking big
hack which is unmanageable.

If you want something new, persue something new.  If you
wanna borrow/share other code, persue a modular and common
interface.  Otherwise don't bother.

Also, if you want to borrow code, then that code
has to be innately "shareable" by license.  Therefore,
architecture interfaces and license should be part
of the original blue-print.  If you work backwards
from such potentialities as code sharing, perhaps
it's more clear that A) people who start projects
never really thought through the license decision
and B) the GNU people are steering people in the
wrong direction, posing the LGPL as "lesser".  In
fact it is more open, has less encumberances, more
flexible, and promotes the spirit and principal
of open/free/libre software more than GPL.  That
said, it is not perfect and needs a fix-up by somebody
who's mind is not on auto-repeat.


-Kevin

__________________________________________________
Do You Yahoo!?
Send your FREE holiday greetings online!
http://greetings.yahoo.com

Reply via email to