__/ [ Mark Kent ] on Wednesday 28 June 2006 09:01 \__
> begin oe_protect.scr
> B Gruff <bbgruff@xxxxxxxxxxx> espoused:
>> On Wednesday 28 June 2006 00:07 Mark Kent wrote:
>>
>>> begin oe_protect.scr
>>> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> espoused:
>>>> A GPL requirement could have a chilling effect on derivative distros
>>>>
>>>> ,----[ Quote ]
>>>>| Many derivative distributions, then, seem to be on their own in a
>>>>| difficult situation where good intentions and creativity count for
>>>>| nothing beside the letter of the law.
>>>> `----
>>>>
>>>> http://software.newsforge.com/article.pl?sid=06/06/23/1728205&from=rss
>>>
>>> The FSF are quite correct about their position here, the licence is
>>> clear, and source code must be made available. If this provision is not
>>> upheld, the GPL would begin to look like the BSD licence, enabling those
>>> who wish to take code without returning their work to the pool to spread
>>> this limitation of others' freedom to GPLed works, which is clearly not
>>> acceptable (noting that at least one regular here doesn't seem to
>>> understand this, if not two or three).
>>
>> I agree of course, Mark, but I reckon it's even worse that that. This is
>> likely also a Configuration Control/Management issue. They are (I think)
>> taking the source from a "parent" distro (e.g. Debian), but not
>> re-distributing it, and possibly therefore not archiving it. If, after
>> they have taken that code, and put out their distro, Debian were to make a
>> change..... do you see what I mean?
>> I don't think it's a case of worrying that they are "holding something
>> back", but rather that by not archiving and distributing, they are in
>> danger of a configuration cock-up!
>>
>
> Hmmm, quite right... one wonders if there isn't a case for a shared
> respository for source-code? A bit like sourceforge, but wouldn't need
> to have all the bells & whistles, just lots and lots of storage,
> mirrored across the planet.
This kind of reminds me of ibiblio.
> In all honesty, one of the greatest crimes of all time is the number of
> chip & software designs which have just been "lost". Worse than that,
> this behaviour is actually /encouraged/ by most information retention
> policies in most companies, which tend to prefer their staff to not
> archive things - for generally good reason, I would add, which is to
> save embarassing information surfacing which shouldn't - but the result
> is that lots of info is lost.
>
> This is exacerbated by the reliance lots of organisations have had on
> proprietary tools over the years - how many documents have been lost
> because they're stored in a binary format which can no longer be read,
> or no longer be /easily/ read (which means they get junked - too much
> effort to do anything else...).
I still have some data, which I believe has lost its value. It's utterly
inaccessible, which makes it a ball of binary goo. It is unfortunate and
vexing to me, to say the last (Palm are among the culprits, by the way).
Much as in Pigrim's words, I want to keep my trail of data open and
accessible in the long run. I often need to refer back to old information,
which I sometimes keep indexed. But without some shallow standards, how do
you index in the first place? How do you impose hierarchy (levels) and
structure, for instance?
> The fantastic work which emulator writers have done over the years to
> keep the capability to run older binaries on new systems is amazing, but
> as chip designs are ever more bespoke, this kind of emulation will
> become increasingly difficult, if not impossible, so it could be that
> future generations of emulators will, in reality, rely on having source
> available.
>
> Hmm, wonder if I should start blogging my diamonds of wisdom :-))
I urge you too, as I already said before. It should take just minutes to get
started and you will have at least one reader.
Best wishes,
Roy
--
Roy S. Schestowitz | England - 1 Ecuador - 0
http://Schestowitz.com | GNU is Not UNIX ¦ PGP-Key: 0x74572E8E
9:50am up 61 days 14:53, 12 users, load average: 0.17, 0.98, 1.34
http://iuron.com - proposing a non-profit search engine
|
|
