🚀 go-pugleaf

In article <RRJ2aB7w164w@mantis.co.uk>, mathew@mantis.co.uk (To the Ground) writes:
>Specifically, if we categorize people writing code as:
>
>a) People writing "free" software under the GPL,
>b) People writing free software, and
>c) People writing limited-distribution software,
>
>then if grant recipients' work is PD it can be used by all three categories
>of programmer, whereas if it's GPLed it can only be used by people in
>category a).

What about:

d) People who need to customize or enhance the derivative software.

If the derivative software is implemented by people in category (c) then
category (d) can't use it.

>> The usual example is the X window system, which is available free
>> from MIT.  But many manufacturers create specially-tuned versions
>> of X for their own machines, which they then sell for a significant
>> price, and distribute without source code.  I remember using X on
>> HP machines, when we found a serious bug in the server.  Since HP
>> didn't ship sources, we had no way to fix the bug short of going
>> back to the raw MIT distribution.
>
>So go back to the raw MIT distribution and create a specially-tuned version
>from that.  Distribute it under the GPL.  Compete by BEING BETTER, not by
>forcing other people to do what you want using legal threats.

He didn't say he wanted to implement a specially-tuned version; all he
wanted to do was fix a bug.  And he presumably wasn't planning on competing
with HP as an X server vendor; I expect this was just for his site's own
use (although I could imagine him passing along the patch to other sites).
If the source code were available he could just make the (perhaps) one-line
change and recompile; without it, he must do a full port just to fix a bug,
or beg HP to fix it for him and send him a new version.  And the result of
this is likely to have its own set of bugs, and probably won't be as well
tuned to the system (perhaps HP's version makes use of undocumented
hardware interfaces).

>If the GNU way is so great, how come HP can get away with selling an HP
>version of X? Why isn't there already an HP version of X available for
>free? Could it be that (*gasp*) the GNU way isn't as good at producing
>special limited-market versions of software as we have been led to 
>believe?

The fact that some specific programs haven't been distributed under the GPL
could hardly be a justification for condemning the entire philosophy.  HP
can get away with what they do because no one has yet had the inclination
to write a copylefted version.

>If free software works, let's see you develop a free software version of
>X for HPs and supply it with source.  Then I'll agree that the non-free
>HP version is unnecessary.  Until then, stop bitching and trying to force
>HP to release source.  Nobody's stopping you from out-competing them.

But he doesn't want to compete with them, he just wants his bug to be
fixed!

And what about my point about the possible undocumented hardware
interfaces?  If this were true, it would be virtually impossible for a
third party to write a good X server without licensing technology from the
system vendor.

-- 
Barry Margolin, Thinking Machines Corp.

barmar@think.com
{uunet,harvard}!think!barmar