[Apologies for reposting this, but I'm interested in other's feedback
and the old thread's topic has definately wandered. Plus on this copy
I took the time to M-x ispell-buffer... ;-) ]
········@hex.net writes:
[...]
> A _weakness_ of the "Open Source" movement [moreso, perhaps, than the
> "Free Software" movement] has been that it has tended to be pretty
> derivative.
>
> To a great extent, it is _not_ innovative.
"Innovation" is a marketing buzzword. There is no such thing.
[...]
> "Free Software" projects have certainly built some useful things, but
> it is not clear that there is a useful model for creating truly _NEW_
> things.
>
> That being said, I'm not sure that there is _any_ open model out there
> for that, at present. Universities used to be nexuses for this sort
> of thing, from whence came such _truly innovative_ things as:
>
> - BSD Unix, Sockets, much of TCP/IP [BSD]
> - Mach [CMU] (love it or hate it, it _was_ innovative)
> - X11 [MIT]
> - Common Lisp [MIT, CMU]
> - Emacs [MIT]
Hmmm, this would be the BSD extensions to the Unix software released
to colleges by AT&T (essentially as open-source) -- after initially
being developed by a bunch of righteous hackers killing cycles after
their Multics project was axed -- Multics being an attempt to combine
the best features of all the OS's-de-jour into one uber-OS... (At this
point my pitiful grasp of the pre-history of Unix is exhausted, but
I'll bet anyone who was there could recite several more orders of
derivation.)
There were API's for networking before BSD sockets. I just know it in
my soul. Why did this one survive? Ahhh, the *source* was widely
available.
[Mach I don't know enough about to comment on. But I know the HURD
uses it -- which means it too is Free Software.]
Ahhh, and X11 -- such a dismal plummet from the functionality and
capability of its predecessors. Ill-conceived, written by students and
rife with bugs. Released as open source by MIT, when far superior
systems were available from several established vendors -- today
universally known as *the* Unix GUI. Who today remembers NeWS?
Common Lisp -- surely a divine bolt of innovation! Completely without
precedent? Uhhh, what was it that was "Common" about it? Preceded by
how many dialects? The collective effort of how many disparate,
competing commercial concerns collaborating under the X3J13 banner?
The Emacs we know today -- descended from TECO, sired by the Father of
Free Software, nurtured by a community of devoted users who in all
probability have sunk a man-millennium into refining and improving it.
[We may criticize it from time to time, but I dare say most of us
compose those critiques within an Emacs buffer.]
> Much of the base stuff that the Free Software community depends
> _heavily_ on came out of megaprojects sponsored through universities.
>
> There is _no_ "free software" project of comparable complexity in
> design or intent to, quite frankly, _any_ of these systems.
Arguably with the exception of Common Lisp[1], *all* of the examples
you've cited are either Open or (research) Community Source
projects. And a couple of them *are* Free Software. This leaves me
particularly baffled by your last assertion.
[1] IMO CL is very much the product of an open research community, but
my knowledge is entirely anecdotal.
> There may be maintenance work, rewrites, or modifications taking
> place. Many of the early designers of X are fruitfully moving onwards
> to provide substantive new enhancements to X11 today.
>
> The point is that to design something completely new of similar scope
> to X11 represents a project requiring years of _design_ effort even
> before there is any working code.
>
> There's not a model in place to support that kind of effort. Not in
> the realm of "Free Software." Nor, it seems, much of anywhere else.
Well, we agree on a couple of points: (Okay, I'm stretching that last
statement to cover my second assertion below.)
-- Commercial, cathedral-style development doesn't support this kind
of effort[2].
-- Much of the contemporary academic "research community" is busily
trying to re-style itself as a cathedral[3]. A really shoddily run
cathedral with absolutely no experienced engineers...
[2] This is not to imply there isn't cool basic research being funded
from some corporate coffers. But we're talking about mega-man-year
software development projects here.
[3] And we as a society have ourselves to thank for that -- but that's
a rant for another day.
This is the crux of my assertion:
The Open Source and Free Software communities are the havens for
the free exchange of knowledge/ideas/code today that the DARPA
community, academic think-tanks, etc. were in past decades.
And as to "innovation," well, what you cite as examples of innovation,
I see as classic examples of the kind of incremental, inch-by-inch,
often-backsliding and entirely *derivative* progress that typifies all
meaningful human endeavor.
-- Kaelin
"Kaelin Colclasure" <······@everest.com> wrote in message
······················@soyuz.arslogica.com...
> [Apologies for reposting this, but I'm interested in other's feedback
> and the old thread's topic has definately wandered. Plus on this copy
> I took the time to M-x ispell-buffer... ;-) ]
M-x ispell-buffer *definitely* missed something ;-)
Kaelin Colclasure <······@everest.com> writes:
...
> Ahhh, and X11 -- such a dismal plummet from the functionality and
> capability of its predecessors. Ill-conceived, written by students and
> rife with bugs. Released as open source by MIT, when far superior
> systems were available from several established vendors -- today
> universally known as *the* Unix GUI. Who today remembers NeWS?
I do. It shows what happens when you choose (as a company) to
"transfer" (quotation marks essential) a technology and when you
choose not. Who today remembers NFS?
Cheers
--
Marco Antoniotti =============================================================
NYU Bioinformatics Group tel. +1 - 212 - 998 3488
719 Broadway 12th Floor fax +1 - 212 - 995 4122
New York, NY 10003, USA http://galt.mrl.nyu.edu/valis
Like DNA, such a language [Lisp] does not go out of style.
Paul Graham, ANSI Common Lisp
Marco Antoniotti <·······@cs.nyu.edu> writes:
> Kaelin Colclasure <······@everest.com> writes:
>
> ...
>
> > Ahhh, and X11 -- such a dismal plummet from the functionality and
> > capability of its predecessors. Ill-conceived, written by students and
> > rife with bugs. Released as open source by MIT, when far superior
> > systems were available from several established vendors -- today
> > universally known as *the* Unix GUI. Who today remembers NeWS?
>
> I do. It shows what happens when you choose (as a company) to
> "transfer" (quotation marks essential) a technology and when you
> choose not. Who today remembers NFS?
Exactly! :-)
And thanks to that transfer many of NFS's successors have avoided some
of the more gratuitous flaws in NFS's design... ;-)
[Daintily need to turn on flyspell-mode... Definitively. :-p ]
-- Kaelin
From: Johan Kullstam
Subject: Re: The myth of "innovation" [repost, new subject]
Date:
Message-ID: <m2r93u1k8p.fsf@euler.axel.nom>
Raffael Cavallaro <·······@mediaone.net> writes:
> In article <············@nnrp1.deja.com>, ········@my-deja.com wrote:
>
> > Any rational person who has observed your responses would clearly see
> >that it is you who have "snapped", and have very little control over
> >your emotional responses. Furthermore, most unlike you, I refrain from
> >making viscious sub-human wishes about your bodily harm.
>
> I see another unsupecting victim has stepped on the naggum-mine and had
> a leg blown off. Oh well.
>
> I've said it before, and in all seriousness:
>
> Erik Naggum is the single largest force impeding the more widespread use
> of common lisp.
ironically, it was erik naggum who stirred me to look at common-lisp
in the first place.
> People come to c.l.l,
ah, he was perhaps kinder and gentler over in comp.emacs.
> see the sort of treatment that
> others receive at his hands, and quickly conclude that something is not
> quite right in common lisp land.
here's a free hint for avoiding naggum flamage. erik usually doesn't
turn on the blowtorch until a couple of messages have gone back and
forth. resist the urge to respond. this will end the pain. i have
*never* seen erik pile on a second flame without a response to the
first.
--
J o h a n K u l l s t a m
[········@ne.mediaone.net]
Don't Fear the Penguin!
From: Erik Naggum
Subject: Re: The myth of "innovation" [repost, new subject]
Date:
Message-ID: <3184585482010427@naggum.net>
* Johan Kullstam <········@ne.mediaone.net>
| ironically, it was erik naggum who stirred me to look at common-lisp
| in the first place.
I'm glad to hear that. I hear it quite often. As in "if Erik Naggum
thinks something is great, it must have some _significant_ merit."
| ah, he was perhaps kinder and gentler over in comp.emacs.
Probably not. Just a lot fewer arrogant ignorants with an attitude.
| here's a free hint for avoiding naggum flamage. erik usually doesn't
| turn on the blowtorch until a couple of messages have gone back and
| forth. resist the urge to respond. this will end the pain. i have
| *never* seen erik pile on a second flame without a response to the
| first.
Thanks. And this is indeed good advice, as is "if shot at, look for
the "trespassers will be shot" sign, don't just return the fire".
#:Erik
--
Solution to U.S. Presidential Election Crisis 2000:
Let Texas secede from the Union and elect George W. Bush their
very first President. All parties, states would rejoice.
>>>>> "Kaelin" == Kaelin Colclasure <······@everest.com> writes:
Kaelin> Marco Antoniotti <·······@cs.nyu.edu> writes:
>> Kaelin Colclasure <······@everest.com> writes:
>> > Ahhh, and X11 -- such a dismal plummet from the functionality
>> > and capability of its predecessors. Ill-conceived, written by
>> > students and rife with bugs. Released as open source by MIT,
>> > when far superior systems were available from several
>> > established vendors -- today universally known as *the* Unix
>> > GUI. Who today remembers NeWS?
>>
>> I do. It shows what happens when you choose (as a company) to
>> "transfer" (quotation marks essential) a technology and when
>> you choose not. Who today remembers NFS?
Kaelin> Exactly! :-)
Kaelin> And thanks to that transfer many of NFS's successors have
Kaelin> avoided some of the more gratuitous flaws in NFS's
Kaelin> design... ;-)
Ah, but what do people _actually use_?
- X is flawed, but ubiquitous.
- NFS is flawed, but ubiquitous.
NeWS may not have been _technically_ flawed, but it had the _license_
flaw that using it meant forever shackling yourself to Sun
Microsystems. IBM and HP and Digital were understandably reluctant to
do that, from whence came the rebellion that put X11 in place.
[Obviously vastly larger groups of people have felt comfortable enough
shackling their "computing futures" to Wherever Microsoft Is Taking
You Today... That basically results from the business world having
been _potently_ shackled to IBM for many years, and being comfortable
with that sort of thing...]
--
(concatenate 'string "cbbrowne" ·@hex.net")
<http://www.ntlug.org/~cbbrowne/>
The *Worst* Things to Say to a Police Officer: Hey, is that a 9 mm?
That's nothing compared to this .44 magnum.
········@hex.net writes:
> Kaelin> And thanks to that transfer many of NFS's successors have
> Kaelin> avoided some of the more gratuitous flaws in NFS's
> Kaelin> design... ;-)
>
> Ah, but what do people _actually use_?
>
> - X is flawed, but ubiquitous.
> - NFS is flawed, but ubiquitous.
My comment above was an aside intended as an example of how source
code released to the community was enhancing the current state of the
art. AFS, CODA, etc. all benefitted significantly from the fact that
their implementors had NFS to draw from -- if only as a source of
ideas and lessons learned.
> NeWS may not have been _technically_ flawed, but it had the _license_
> flaw that using it meant forever shackling yourself to Sun
> Microsystems. IBM and HP and Digital were understandably reluctant to
> do that, from whence came the rebellion that put X11 in place.
Oh, I'm sure NeWS had its flaws too. But my point is, *at the time* it
embodied something near the state of the art, while X was a throwback
on the interface front coupled with some really muddled ideas about
what network bandwidth was good for. (I understand MIT had fairly
stupid bit-mapped frame buffers -- but a really spiffy and largely
un-utilized Ethernet to play with.)
And yet, as so often happens, the packaged, polished and supported-by-
a-real-company software -- which appeared to have *all* the near-term
advantages -- ultimately was selected for extinction. And the
community-supported "research quality" code base won big over the long
haul.
This is certainly not to imply that Open Source guarantees
longevity. Yes I'm sure dozens of OSS projects fizzle and disappear
every day -- just like hundreds (thousands) of other software
projects. The difference is, it takes place where people who have
similar needs or concerns have (relatively) full view of what
happened. "Hmmm, so the world's *not* ready for Y." Or "Z was a really
great idea, but they just couldn't pull the right contributors
together." And in the latter case, there are artifacts of "Z" left to
comb through next time it's tried.
When your project dies in a cathedral, in all likelihood all your hard
work really was for nothing.
> [Obviously vastly larger groups of people have felt comfortable enough
> shackling their "computing futures" to Wherever Microsoft Is Taking
> You Today... That basically results from the business world having
> been _potently_ shackled to IBM for many years, and being comfortable
> with that sort of thing...]
This brings up a related point -- commercial support simply *cannot*
be cost-effectively maintained for a "legacy" user base. People who
deride MS, or any software vendor, for "forcing" them to continually
upgrade their hardware and software just don't grasp the economics of
the situation. And yet those *user's* gripes are also economically
sound -- they're paying for hardware and features they don't really
need.
But again -- this is a long-term cost, which means for modern business
practices that it's invisible.
-- Kaelin
From: Tim Bradshaw
Subject: Re: The myth of "innovation" [repost, new subject]
Date:
Message-ID: <nkj8zq1u6rg.fsf@tfeb.org>
Kaelin Colclasure <······@everest.com> writes:
> Oh, I'm sure NeWS had its flaws too. But my point is, *at the time* it
> embodied something near the state of the art, while X was a throwback
> on the interface front coupled with some really muddled ideas about
> what network bandwidth was good for. (I understand MIT had fairly
> stupid bit-mapped frame buffers -- but a really spiffy and largely
> un-utilized Ethernet to play with.)
>
Did you use NeWS? It was *achingly* slow: far slower than even X11R3
(I think R4 was the first one to perform reasonably).
--tim