RE: [Caml-list] OCaml growing pains

RE: [Caml-list] OCaml growing pains

[Brandon J. Van Every]

Anil Madhavapeddy wrote:
>
> You are possibly the worst evangelist I have ever seen.  The good
> evangelists actually know quite a lot about the stuff they pontificate
> about.

Living a stone's throw from Microsoft as I do, and often wishing for a
stone, I am ROTFLMAO at what you say.  The book of Microsoft states that
knowing everything about a technology is clearly not necessary to market
it.  When OCaml has the popularity of C# or Windows, we'll talk about
what evangelist qualities led to that.  Meanwhile, you should realize
that the people who are best at excruciating technical detail are the
worst evangelists, because they aren't interested in being accessible to
anyone who doesn't meet their high standards of technical content.

Thus it is necessary to have more than 1 type of person to promote a
language, and more than 1 culture.    (The Python Software Foundation
still doesn't understand this, unfortunately).  I'm doing just fine
promoting OCaml to game developers right now, no matter how much Xavier
might shake his head.  That's because I understand game industry culture
far better than he does.  I know what their issues are and what I have
to say to address them.  I answer honestly about where OCaml is really
at, as best I can.  Even if someone thinks I answer incorrectly, the
more important truth is game developers go through the exact same futz /
misconception learning curve that I do.

Some things are not misconceptions.  OCaml does have baroque syntax, a
long learning curve for imperative programmers, technical flaws, a
relative but not absolute lack of libraries, a futzy toolchain on
Windows, only fledgling community organization, and no marketing
materials that would convince a suit to give it a whirl.

> Your posts about user groups would be tolerable if they were limited
> to one every few weeks, and not also followed up with reams of
> whines and moans about Bayesian filters conspiring against you.

Which is what actually happens.  You aren't getting ML S*attle announces
*and* whining, there'd be no reason for it if the announces actually
went through.  I tried again last night and still haven't managed it.
Maybe a very large code snippet would do the trick.


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...

// return an array of 100 packed tuples
temps
  int $[tvar0][2*100]; // what the c function needs
  value $[tvar1]; // one int
  value $[tvar2]; // one tuple
  int $[tvar3] // loop control var
oncePre
eachPre
  $[cvar0]=&($[tvar0][0]);
eachPost
  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
    $[tvar2] = alloc_tuple(2);
    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
    Store_field($[tvar2],0,$[tvar1]);
    $[tvar1] = Val_int($[cvar0][1]);
    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
    Array_store($[lvar0],$[tvar3],$[tvar0]);
  }
oncePost

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] OCaml growing pains

[Anil Madhavapeddy]

On Thu, Aug 12, 2004 at 01:48:38PM -0700, Brandon J. Van Every wrote:
> 
> Living a stone's throw from Microsoft as I do, and often wishing for a
> stone, I am ROTFLMAO at what you say.  The book of Microsoft states that
> knowing everything about a technology is clearly not necessary to market
> it.  When OCaml has the popularity of C# or Windows, we'll talk about
> what evangelist qualities led to that.  Meanwhile, you should realize
> that the people who are best at excruciating technical detail are the
> worst evangelists, because they aren't interested in being accessible to
> anyone who doesn't meet their high standards of technical content.

How terribly rude.  I sent you a private mail to avoid spamming the
list with yet more crap, and you forward it back here.  In fact, I'm 
starting to be convinced that you're some kind of surreal troll, given
your stunning lack of ability to shut up and write code.

I took out an amusing five minutes to google for 'brandon j. van every'
to see if I could find any MIT-licensed free code written by you, and
unfortunately failed to come up with a single hit.  It was drowned out
by your numerous posts to various lists asking the same repetetive 
questions so familiar to this list.

My personal highlights from the google results include:

http://mail.python.org/pipermail/python-list/2003-August/178976.html
http://coding.derkeiler.com/Archive/Lisp/comp.lang.lisp/2004-01/1329.html
http://lists.complete.org/freeciv-dev@freeciv.org/2003/12/msg00471.html.gz

I found reference to the 'free3d' library.  However, it was GPLed
(tsk tsk, how can I make money from it??!) and not available anywhere.

> Which is what actually happens.  You aren't getting ML S*attle announces
> *and* whining, there'd be no reason for it if the announces actually
> went through.  I tried again last night and still haven't managed it.
> Maybe a very large code snippet would do the trick.

I wonder if this post will get through the filters given its "meta"
nature (some might describe it as "irrelevant").  I hope it doesn't ;-)

-- 
Anil Madhavapeddy                                 http://anil.recoil.org
University of Cambridge                          http://www.cl.cam.ac.uk

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

RE: [Caml-list] OCaml growing pains

[Brandon J. Van Every]

Anil Madhavapeddy wrote:
>
> How terribly rude.  I sent you a private mail to avoid spamming the
> list with yet more crap, and you forward it back here.  In fact, I'm
> starting to be convinced that you're some kind of surreal troll, given
> your stunning lack of ability to shut up and write code.

My apologies for responding publically when you sent it privately.  I
received your post in the wee hours of the morning and that aspect of it
got lost in the shuffle.  All of my [Caml-list] mail goes into one
folder, whether public or private.  I suppose I could play with the
Outlook 2000 "except if addressed directly to me" option.  Being MS
software I don't know that it'll work though.  :-)  If it doesn't,
sorry, these errors are going to happen sometimes.  Also I don't know
that I really want such e-mail separated out anyways.  On the balance, I
think I'll just try to be more careful, even early in the morning.

In any event, I didn't answer in a rude manner, call something crap,
accuse anyone of sp*m, etc.  I don't have anything to be ashamed of in
how I responded publically.  I am wondering why you choose to respond to
me, either privately or publically?  I am noticing now it's not the 1st
time you've sent me something privately that didn't serve any
demonstrable need.  I'm choosing to ignore a number of aspects of your
public post.  I don't mind giving an apology for things I should, and I
wouldn't want to taint the apology with other issues.

> I found reference to the 'free3d' library.  However, it was GPLed
> (tsk tsk, how can I make money from it??!)

It was never GPLed, it was LGPLed back in the day.

> and not available anywhere.

Try http://www.indiegamedesign.com/Free3d_2004.zip
I put it up there several months ago, but I do not announce it at all.
It is old, dead, probably useless code.  But, some student in
comp.graphics.algorithms was interested in it, so I put it up.  With a
MIT license.  If anybody is even mildly curious to see if it runs on a
X11 box, I'd be highly amused.  I gave that kid fair warning about what
it is and isn't.  Some day I suppose I'll have a Linux box again and
then I'll see if any of it runs.


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...

// return an array of 100 packed tuples
temps
  int $[tvar0][2*100]; // what the c function needs
  value $[tvar1]; // one int
  value $[tvar2]; // one tuple
  int $[tvar3] // loop control var
oncePre
eachPre
  $[cvar0]=&($[tvar0][0]);
eachPost
  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
    $[tvar2] = alloc_tuple(2);
    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
    Store_field($[tvar2],0,$[tvar1]);
    $[tvar1] = Val_int($[cvar0][1]);
    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
    Array_store($[lvar0],$[tvar3],$[tvar0]);
  }
oncePost

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] OCaml growing pains

[Mikhail Fedotov]

Brandon J. Van Every wrote:

>Mikhail Fedotov wrote:
>  
>
>>Good points below, but it seems you are loosing a major one:
>>you *never* can promote/advance
>>the langauge while going *against* its authors. (You can fork if of
>>course, but then you'll be on your own.)
>>    
>>
>
>I doubt the historical evidence supports your claim.
>
Conflicts are always forcing good/competent people to go away - go away 
from the conference,
go away from the site. Lowering quality of discussions. If only I'd get 
a buck for each such evidence...

> <>For instance, I seem to recall people who have disagreed with Guido 
> Van Rossum
> about the direction Python should take, who just went off and did 
> stuff, and whose
> work is now considered important to the Python community. 

You are doing exactly that I've said before - express good point in the 
beginning but going
in the wrong direction. I doubt  there is an analogy to the case with 
python (can not say for sure until
checked it myself), except when you are talking about technical details. 
And when you are talking
about those details, you are talking about something that do not know 
well with people who wrote
it and know from the top to the bottom.

>Authority figures *could* provide organizational resources
>

They simply might not have them.

>>In addition, when all feedback from major players that you
>>are receiving is negative, it means
>>that you are going in the wrong direction and for some reason fail to
>>change it into the right one.
>>    
>>
>
>I don't think so.
>

Symptoms match. The key method to find the blind spot is to formulate 
verifiable criteria
to ensure the correctness of your own statements and if the choice of 
topics to discuss is right.
The same about the desision to start the discussion at all.

As I've said, I've lost a lot of time on discussions like yours. 
Patterns are very clear and common.

>biz-focused mailing lists are indeed better for the growth discussions.
>I'm just not ready to start such a list yet.
>
I can not imagine its audience. Web pages with success stories and 
remarks about problems
solved should do the trick, and success stories are already there.

>I was ready to start an OCaml Games mailing list, so I did so.
>
Is there much of professional game developers, and they really want to 
share their
*confidential* technology ? I doubt it will be ever anything other than 
newbies list.

>I was ready to start ML S*attle, so I did so.
>
But even with starting it you've managed to bump into spam filter. :)

That's the curse I've been talking about - without good criterias to
verify that you are going in the right direction,  you are loosing
ability not to bump into things when you are finally trying to start
something, things that most normal people do not bump into.
Game list has questionable audience,  meeeting announce
bumps into the filter, your concerns about language features
bump into the luck of your experience with the language, and
you've even managed to go into conflict with Xavier when
he is not in position to address your questions in the manner
they are asked (i.e. completely wrong person for conflict,
such conflict is as pointles as it can be).

>>The most obvious is http://www.ocaml.org - it does not seem to be
>>maintened anymore ( no mention
>>of 3.08 release), so *maybe* you have the chance to become the
>>maintainer if you ask the
>>right people.  Then you'll be able to show that you can do. :)
>>    
>>
>
>You put a smiley, so maybe you don't mean this so seriously.
>

I just believe that you are spending so much time in those duscissions 
that you've
become so *rusty* that a lot of *newbies* can compete with you in any 
good thing
you might want to do, even if you are not a newbie yourself. There is 
one hell of
symptoms in your messages.

It is like being on a drug.

>Praise Be to the caml-list Bayesian filter! It blesseth
>my postings, it is evil crap!  evil crap!  Bigarray!
>Unboxed overhead group!  Wondering!  chant chant chant...
>
>// return an array of 100 packed tuples
>temps
>  int $[tvar0][2*100]; // what the c function needs
>  value $[tvar1]; // one int
>  value $[tvar2]; // one tuple
>  int $[tvar3] // loop control var
>oncePre
>eachPre
>  $[cvar0]=&($[tvar0][0]);
>eachPost
>  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
>  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
>    $[tvar2] = alloc_tuple(2);
>    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
>    Store_field($[tvar2],0,$[tvar1]);
>    $[tvar1] = Val_int($[cvar0][1]);
>    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
>    Array_store($[lvar0],$[tvar3],$[tvar0]);
>  }
>oncePost
>  
>

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

RE: [Caml-list] OCaml growing pains

[Brandon J. Van Every]

John Skaller wrote:
>
> I think you need to consider that the INRIA team's goals
> are closer to 'showing it is possible for a well typed
> functional language developed using theory to actually
> also compile fast code fast' rather than 'taking over
> from other industrial languages'.

It is considered.  Of course, others have other considerations.  Only a
closed source development model would keep people away from something
useful.

> Ocaml is great for writing language translators.
> The *obvious* use for it for a game developer
> is to write a game scripting language in it --
> rather than try to write games directly in Ocaml.

That's not obvious to me.  It runs afoul of the Yet Another Scripting
Language problem.  The game industry already has scripting languages
that aren't running so well on current hardware, like Lua, Python, and
Ruby.  Customizing the heck out of a language is seen, rightly, as a
waste of time.  It may be that programming tastes change as the game
industry becomes more sophisticated in its methods.  But people who can
envision, "This is how and *WHY* I should write a custom scripting
language for my game," and who can lead that to commercial advantage +
success, are not a dime a dozen right now.  Rather, we read postmortems
about how some employee blew the company's techno-wad on some half-baked
compiler he thought would be Kewl to implement.

What's obvious to me, is C++ is a low level language.  Most of the game
industry still uses it for performance reasons.  Show us something
higher level, that also has performance, and also library and tools
support, and you definitely have the possibility of a sea change from
C++.

It doesn't really matter if INRIA isn't worried about industrialization.
To a large degree, other parties can do that.  What matters, is that
INRIA doesn't try to majorly get in the way of it.

> That way you bypass all the problems, and get to
> say how great Ocaml is for writing translators with :)
>
> Oh yeah, did I mention before I've spent 5 years developing
> such a tool already .. ?

Right, but you don't have critical mass and OCaml does.  What would it
take for you to develop critical mass?  That's not a rhetorical
question, I'm wondering what you envision as your strategy.  Feel free
to e-mail me privately about it.


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...

// return an array of 100 packed tuples
temps
  int $[tvar0][2*100]; // what the c function needs
  value $[tvar1]; // one int
  value $[tvar2]; // one tuple
  int $[tvar3] // loop control var
oncePre
eachPre
  $[cvar0]=&($[tvar0][0]);
eachPost
  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
    $[tvar2] = alloc_tuple(2);
    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
    Store_field($[tvar2],0,$[tvar1]);
    $[tvar1] = Val_int($[cvar0][1]);
    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
    Array_store($[lvar0],$[tvar3],$[tvar0]);
  }
oncePost

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

RE: [Caml-list] OCaml growing pains

[skaller]

On Fri, 2004-08-13 at 07:30, Brandon J. Van Every wrote:

> I don't think so.  I think it shows where the major players are at in
> their thinking and tastes.  OCaml is not this popular language, so why
> should we assume its major players know best about how to grow it?

I think you need to consider that the INRIA team's goals
are closer to 'showing it is possible for a well typed
functional language developed using theory to actually
also compile fast code fast' rather than 'taking over
from other industrial languages'.

Ocaml is great for writing language translators.
The *obvious* use for it for a game developer
is to write a game scripting language in it --
rather than try to write games directly in Ocaml.

That way you bypass all the problems, and get to
say how great Ocaml is for writing translators with :)

Oh yeah, did I mention before I've spent 5 years developing
such a tool already .. ?

-- 
John Skaller, mailto:skaller@users.sf.net
voice: 061-2-9660-0850, 
snail: PO BOX 401 Glebe NSW 2037 Australia
Checkout the Felix programming language http://felix.sf.net



-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

RE: [Caml-list] OCaml growing pains

[Brandon J. Van Every]

Mikhail Fedotov wrote:
>
> Good points below, but it seems you are loosing a major one:
> you *never* can promote/advance
> the langauge while going *against* its authors. (You can fork if of
> course, but then you'll be on your own.)

I doubt the historical evidence supports your claim.  For instance, I
seem to recall people who have disagreed with Guido Van Rossum about the
direction Python should take, who just went off and did stuff, and whose
work is now considered important to the Python community.  I don't
recall how adversarial those relationships were.  It is also certainly
possible to support a language in the absence of its canonical technical
list.  One just has to create a list that's more user friendly and that
has sufficient brains for the problems.  Languages with big audiences
certainly aren't restricted by 1 canonical technical list, for instance.
C, C++, Java, and C# are all quite beyond their authors.

> That's one bad side, but there is another one which is even
> worse. While spending all time on
> sorting out offences and stuff, you are not only loosing any
> chance to do anything good, but you
> even do not know if you are actually able to help (read:
> develop trivial ideas like "community
> should grow" into something implementable and implement them
> *without* going against tool
> authors and maintainers, choose proper style and attitude for
> messages in *tech* list etc).

I already learned from the Python Software Foundation crowd not to rely
upon the authority figures for anything.  Authority figures *could*
provide organizational resources, but they may choose not to do so, for
benign or malicious reasons.  In that event, one simply has to do it
oneself.  That's acutally preferrable if an authority figure isn't any
good at addressing a particular problem.

Open source is primarily about having a route around obstructions.  I
wouldn't be interested in OCaml if there weren't solutions to the worst
case scenarios.

> In addition, when all feedback from major players that you
> are receiving is negative, it means
> that you are going in the wrong direction and for some reason fail to
> change it into the right one.

I don't think so.  I think it shows where the major players are at in
their thinking and tastes.  OCaml is not this popular language, so why
should we assume its major players know best about how to grow it?

I've encountered tons of crabby Python developers who don't want to hear
about business or language growth.  In fact, many of them explicitly say
they want the language to stay small so that they won't have to deal
with... idiots, or whatever other horrible thing they think would happen
if something pierced their personal techno-bubble.  It's a pattern of
introversion among techies.  I don't easily fathom that mentality
myself.  I see money and jobs working on what people actually want to
work on.  Part of the problem may be that the most introverted have
already solved this problem for themselves, and don't really want
anything interfering with their pleasurable status quo.

biz-focused mailing lists are indeed better for the growth discussions.
I'm just not ready to start such a list yet.  I was ready to start an
OCaml Games mailing list, so I did so.  I was ready to start ML S*attle,
so I did so.  If OCaml proves to be commercially viable for me, then I
will start a biz list about it.

> >This is called getting things done.  Where's your index of local user
> >groups?  Where are the announces?  There is nothing at
> >http://caml.inria.fr at all.  What transmission vehicle if not
> >caml-list?
>
> The most obvious is http://www.ocaml.org - it does not seem to be
> maintened anymore ( no mention
> of 3.08 release), so *maybe* you have the chance to become the
> maintainer if you ask the
> right people.  Then you'll be able to show that you can do. :)

You put a smiley, so maybe you don't mean this so seriously.  But I have
to observe: I'm not here to spend buckets of man hours to gratuitously
prove my credibility.  I look for the simple solution and implement it.
It's not rational to take on the burdens of acquiring and maintaining
someone else's website when all I really want, right now, is a
transmission vector for ML S*attle announces.  If I thought caml-list
couldn't serve that role, I'd start another mailing list.  As it is, at
least I do have comp.lang.ml and comp.lang.functional (without
controversy).

I'm certainly not ready to become OCaml's official webadmin.  People
around here don't even like me, so I hardly feel obligated to do some
Herculean, unappreciated task to benefit them.  Besides, if you look at
my website you'll realize how weak my webadmin skills are.  If I have
success with OCaml in the arenas I most care about, I'll organize
*other* people to take on such burdens.

There is one community organizational question I can address
immediately, however.  So, I will do that in a new thread, leaving
behind this more incendiary conversation.  I don't personally have a
problem with incendiary conversations, as sometimes the pointed needs to
be voiced.  But, they do lead people to flames, as people often don't
take the pointed very well.


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...

// return an array of 100 packed tuples
temps
  int $[tvar0][2*100]; // what the c function needs
  value $[tvar1]; // one int
  value $[tvar2]; // one tuple
  int $[tvar3] // loop control var
oncePre
eachPre
  $[cvar0]=&($[tvar0][0]);
eachPost
  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
    $[tvar2] = alloc_tuple(2);
    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
    Store_field($[tvar2],0,$[tvar1]);
    $[tvar1] = Val_int($[cvar0][1]);
    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
    Array_store($[lvar0],$[tvar3],$[tvar0]);
  }
oncePost




-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

Re: [Caml-list] OCaml growing pains

[Mikhail Fedotov]

Hi,

Brandon J. Van Every wrote:

>Xavier Leroy wrote:
>  
>
>>Some of your recent Usenet
>>postings left me shaking my head in disbelief, not knowing whether to
>>laugh or cry.
>>    
>>
>
>I was going to reply privately, taking this comment of yours in stride.
>I was composing my reply inline, dealing with some industrial growth
>issues.  When I got farther down in your post, I realized how nasty your
>response actually was, and how disinterested you are in some things I'm
>interested in.  I don't take public nastiness sitting down, so here's my
>reply.
>
>  
>
Good points below, but it seems you are loosing a major one: you *never* 
can promote/advance
the langauge while going *against* its authors. (You can fork if of 
course, but then you'll be on
your own.)

When you are going against authors and mainters of some tool, it takes 
time. Your time. Their time.
The net result is the flame; good ideas are usually expressed in the 
beginning of discussion but
rarely developed any further.

That's one bad side, but there is another one which is even worse. While 
spending all time on
sorting out offences and stuff, you are not only loosing any chance to 
do anything good, but you
even do not know if you are actually able to help (read: develop trivial 
ideas like "community
should grow" into something implementable and implement them *without* 
going against tool
authors and maintainers, choose proper style and attitude for messages 
in *tech* list etc).

In addition, when all feedback from major players that you are receiving 
is negative, it means
that you are going in the wrong direction and for some reason fail to 
change it into the right one.

The net result may be that you become an expert at offences (and 
excuses) but can not do
anything when you have noone to battle with (i.e. when the real work 
starts). Been there myself,
lost a lot of time.

>- people visit Seattle from other cities and move there
>- people need motives to come to meetings, i.e. location, parking, beer
>- establishing critical mass in tech hubs is important to language
>growth
>- when other cities finally want to do it, they know who to contact
>- repetition is the key to all learning
>- announces every 3 weeks aren't anything out of anyone's life
>- those that don't care can skip it upon reading the subject line
>
>This is called getting things done.  Where's your index of local user
>groups?  Where are the announces?  There is nothing at
>http://caml.inria.fr at all.  What transmission vehicle if not
>caml-list?
>
>  
>

The most obvious is http://www.ocaml.org - it does not seem to be 
maintened anymore ( no mention
of 3.08 release), so *maybe* you have the chance to become the 
maintainer if you ask the
right people. Then you'll be able to show that you can do. :)

>// return an array of 100 packed tuples, just in case
>temps
>  int $[tvar0][2*100]; // what the c function needs
>  value $[tvar1]; // one int
>  value $[tvar2]; // one tuple
>  int $[tvar3] // loop control var
>oncePre
>eachPre
>  $[cvar0]=&($[tvar0][0]);
>eachPost
>  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
>  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
>    $[tvar2] = alloc_tuple(2);
>    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
>    Store_field($[tvar2],0,$[tvar1]);
>    $[tvar1] = Val_int($[cvar0][1]);
>    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
>    Array_store($[lvar0],$[tvar3],$[tvar0]);
>  }
>oncePost
>  
>

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners

[Caml-list] OCaml growing pains

[Brandon J. Van Every]

Xavier Leroy wrote:
> Some of your recent Usenet
> postings left me shaking my head in disbelief, not knowing whether to
> laugh or cry.

I was going to reply privately, taking this comment of yours in stride.
I was composing my reply inline, dealing with some industrial growth
issues.  When I got farther down in your post, I realized how nasty your
response actually was, and how disinterested you are in some things I'm
interested in.  I don't take public nastiness sitting down, so here's my
reply.

> Finally, my parents taught me not to use "I want" in polite company,
> so I find your demands somewhat rude.  Posting to caml-list isn't a
> right, it's a privilege.

Having posts blocked for stupid reasons by machines is rude.  I hope you
don't start entertaining the notion that getting exasperated at stupid
machines is uncalled for.

> Why don't you just put the details of your meeting on a web page and
> post a short message "Next meeting on <such date>, see http://URL for
> practical details"?

- people don't click on URLs when they're busy.

> You do realize that > 95% of the subscribers don't
> leave in Seattle and couldn't care less about the delicacies and car
> park available there, right?

- people visit Seattle from other cities and move there
- people need motives to come to meetings, i.e. location, parking, beer
- establishing critical mass in tech hubs is important to language
growth
- when other cities finally want to do it, they know who to contact
- repetition is the key to all learning
- announces every 3 weeks aren't anything out of anyone's life
- those that don't care can skip it upon reading the subject line

This is called getting things done.  Where's your index of local user
groups?  Where are the announces?  There is nothing at
http://caml.inria.fr at all.  What transmission vehicle if not
caml-list?

To grow, OCaml needs more than 1 mailing list devoted to the
uber-technical.  Two attitudes you could take here.  (1) "Fine, Brand*n.
Go start all your own lists.  Knock yourself out."  (2) "Yes actually
INRIA would like to facilitate these efforts to grow OCaml."

> - I and many other caml-list regulars don't wish to discuss business
>   issues with you.  I don't discuss business on open mailing lists.

That's definitely not an open source attitude.  I would say that Python
and Perl have more powerful models of business promotion than the
'closed doors' model you say you prefer.

> - The growing pains you mention weren't apparent to us before you
>   started making such a noise on this list.

If you do not see the growing pains in OCaml, it is because you're not
much interested in issues of industrialization and evangelism.  On this
list I have heard people discussing the standard library and what
INRIA's role should be with it.  Someone tried to volunteer a paid
compiler guy, so that they could get some business insurance on what's
happening with OCaml.  People still think http://www.ocaml.org/ is the
proper website, and I've seen no movement on that issue since I've been
on this list.

You are at a pre-Python level of industrialization.  Python is on every
programmer's lips, but commands a measly 2% market share (to be
generous).  It has a gazillion libraries that OCaml doesn't have, yet
still lacks glaringly in large scale desktop applications development
and 3D graphics.  It can't mount a business-friendly marketing campaign
because of the techies currently in charge of it.  Whereas OCaml is in
the "What's that?" stage.  So if Python has growing pains, you have
growing pains.  Unless your attitude is similar to Matthias Blume's,
that SML/NJ is only good for publishing papers for other academics.

> - You're most welcome to create your caml-biz list and
>   discuss whatever you want there.  Actually, I feel you
>   aren't interested in discussions as much as in
>   asserting your preconceptions, which makes you prime
>   material for blogging.

A gratuitous piece of managerial theory for you today:
http://www.teams.org.uk/shaper.htm

> - As I explained above, posting to this list isn't a right, so we
>   are not at all "stuck with you here".

But what responsibility do you feel, Xavier, for building communities?
At the most basic level, are you only interested in people who play the
game your way, on your terms?  Or do you want OCaml to grow into
something big and really useful to tons of people?  If you want the
latter, you will have to cut people some slack.

> Thanks for your attention.

Do I receive yours in fair exchange, regarding communities?


Cheers,                         www.indiegamedesign.com
Brand*n Van Every               S*attle, WA

Praise Be to the caml-list Bayesian filter! It blesseth
my postings, it is evil crap!  evil crap!  Bigarray!
Unboxed overhead group!  Wondering!  chant chant chant...


// return an array of 100 packed tuples
temps
  int $[tvar0][2*100]; // what the c function needs
  value $[tvar1]; // one int
  value $[tvar2]; // one tuple
  int $[tvar3] // loop control var
oncePre
eachPre
  $[cvar0]=&($[tvar0][0]);
eachPost
  $[lvar0] = alloc(2*100, 0 /*NB: zero-tagged block*/ );
  for(int $[tvar3]=0;$[tvar3]<100;$[tvar3]++) {
    $[tvar2] = alloc_tuple(2);
    $[tvar1] = Val_int($[cvar0][0+2*$[tvar3]]);
    Store_field($[tvar2],0,$[tvar1]);
    $[tvar1] = Val_int($[cvar0][1]);
    Store_field($[tvar2],1,$[tvar1+2*$[tvar3]]);
    Array_store($[lvar0],$[tvar3],$[tvar0]);
  }
oncePost

-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners