In article <cgl2vu$g45$1@digitaldaemon.com>, Sean Kelly says...

>To me, it's because the end result has to follow an established set of interface standards and such if it's to feel like a standard library.  I have some trepidation as well, but I think some degree of organization is necessary if this will have any hope of supplanting Phobos.

Yeah. I got it now. I'm agreeing with you now.

>But I think it should be clear that no one should try and tell another *how* to do something.  Especially in their area of expertise.  I think we all feel pretty much the same way as you--if this is going to work we'll need to cooperate, and if it turns into a power struggle then I suspect we'll all leave the project.

Sounds good. I think the biggest difficulty is that committees tend to consist of people who want to be on committees. I don't have a better answer though.

Arcane Jill

"J C Calvarese" <jcc7@cox.net> wrote in message news:cgktp2$d5r$1@digitaldaemon.com...
> Don't think of it as an authoritarian committee. Think of it as a group of concerned D citizens. We'll only have the power of our coding skill. If
someone
> doesn't want to join, we won't try to make them. If we start laying down a
bunch
> of stuffy rules for our members and no one does anything constructive, we
won't
> accomplish anything. But if we develop common-sense solutions and good
code
> rather than pontificate, we could do well.

How does the Boost committee work? It seems to be rather successful.

"Arcane Jill" <Arcane_member@pathlink.com> wrote in message news:cgl53j$h6u$1@digitaldaemon.com...
> Sounds good. I think the biggest difficulty is that committees tend to
consist
> of people who want to be on committees. I don't have a better answer
though.

I think it would be best if I wasn't on the Diemos committee, as I think it would be good for D to establish some organizational momentum that is independent. One of my big goals for D is that it reach that "tipping point" where it is self-sustaining regardless of whether I am pushing or not.

In article <cgl4pn$gs4$1@digitaldaemon.com>, Arcane Jill says...
>
>In article <cgl27o$fok$1@digitaldaemon.com>, Sean Kelly says...
..
>So you're quite right to say there are two different purposes here. And with that understanding, I support the proposal. "Phoenix" seems to be the preferred name for the new arena, so I'll add my support to that. /Presumably/ (although I may have misunderstood this) the top level namespace within Phoenix will just be "std", to take over from Phobos. Is that right? If that's so then perhaps people could develop stuff in Deimos without restraint, and then propose it to the DSLG for possible movement to Phoenix (instead of Phobos) when it reaches a certain level of maturity?
>
>Jill


My belief is that "std" is one of the things that it can't be. If we use "std",
it won't be obvious that it's something different than Phobos. When someone sees
import somethingdifferent.stream;
he/she can tell right away, somethingdifferent is needed.

I think it should be short and self-explaining (and catchy would be nice, too).

We have a forum called "Phobos Rising", but "phobosrising" is too much typing. I was thinking more along the lines of "phoenix", "newt", "newstd", etc.

I started a forum topic to talk about choosing a name: http://www.dsource.org/forums/viewforum.php?f=31

jcc7

In article <cgl4pn$gs4$1@digitaldaemon.com>, Arcane Jill says...
>
>/Presumably/ (although I
>may have misunderstood this) the top level namespace within Phoenix will just be
>"std", to take over from Phobos. Is that right? If that's so then perhaps people
>could develop stuff in Deimos without restraint, and then propose it to the DSLG
>for possible movement to Phoenix (instead of Phobos) when it reaches a certain
>level of maturity?

Exactly.


Sean

In article <cgl53j$h6u$1@digitaldaemon.com>, Arcane Jill says...
>
>Sounds good. I think the biggest difficulty is that committees tend to consist of people who want to be on committees. I don't have a better answer though.

True enough.  Though I'd like to believe that none of us are doing this because we want to be on a committe :)


Sean

In article <cgl5mo$hf2$1@digitaldaemon.com>, Walter says...
>
>
>"J C Calvarese" <jcc7@cox.net> wrote in message news:cgktp2$d5r$1@digitaldaemon.com...
>> Don't think of it as an authoritarian committee. Think of it as a group of concerned D citizens. We'll only have the power of our coding skill. If
>someone
>> doesn't want to join, we won't try to make them. If we start laying down a
>bunch
>> of stuffy rules for our members and no one does anything constructive, we
>won't
>> accomplish anything. But if we develop common-sense solutions and good
>code
>> rather than pontificate, we could do well.
>
>How does the Boost committee work? It seems to be rather successful.

Boost works in a similar way to what I was envisioning here.  The committee consists of all mailing list members.  Formal submissions require a list sponsor who volunteers to be the review manager (so there's no one person with more authority than any other).  All list members are welcome to submit formal reviews and it is the review manager's job to sort through all the information and reject or accept the submission.  Pre-submission evaluation is kind of unstructured and happens both on the list and in other forums.  In our case, I had proposed that Deimos be the venue for pre-review discussion, since that's pretty much what it was intended for in the first place.  The full description of the Boost process is here: http://boost.org/more/formal_review_process.htm


Sean

"Sean Kelly" <sean@f4.ca> wrote in message news:cglale$js5$1@digitaldaemon.com...
> In article <cgl5mo$hf2$1@digitaldaemon.com>, Walter says...
> >
> >
> >"J C Calvarese" <jcc7@cox.net> wrote in message news:cgktp2$d5r$1@digitaldaemon.com...
> >> Don't think of it as an authoritarian committee. Think of it as a group
of
> >> concerned D citizens. We'll only have the power of our coding skill. If
> >someone
> >> doesn't want to join, we won't try to make them. If we start laying
down a
> >bunch
> >> of stuffy rules for our members and no one does anything constructive,
we
> >won't
> >> accomplish anything. But if we develop common-sense solutions and good
> >code
> >> rather than pontificate, we could do well.
> >
> >How does the Boost committee work? It seems to be rather successful.
>
> Boost works in a similar way to what I was envisioning here.  The
committee
> consists of all mailing list members.  Formal submissions require a list
sponsor
> who volunteers to be the review manager (so there's no one person with
more
> authority than any other).  All list members are welcome to submit formal reviews and it is the review manager's job to sort through all the
information
> and reject or accept the submission.  Pre-submission evaluation is kind of unstructured and happens both on the list and in other forums.  In our
case, I
> had proposed that Deimos be the venue for pre-review discussion, since
that's
> pretty much what it was intended for in the first place.  The full
description
> of the Boost process is here:
http://boost.org/more/formal_review_process.htm
>
>
> Sean


That sounds quite good!

How about taking this over to dsource? Brad has kindly set up a forum here: http://www.dsource.org/forums/viewforum.php?f=31&sid=1ddb9ad9d82f7ab9d1cfbe4 85cd722a1

FWIW, I think it's in our best interests to get this part sorted out before most anything else.

In article <cglblu$kdv$1@digitaldaemon.com>, antiAlias says...
>
>How about taking this over to dsource? Brad has kindly set up a forum here: http://www.dsource.org/forums/viewforum.php?f=31&sid=1ddb9ad9d82f7ab9d1cfbe4 85cd722a1
>
>FWIW, I think it's in our best interests to get this part sorted out before most anything else.

Agreed.  And done :)


Sean

On Thu, 26 Aug 2004 17:02:43 +0000, Arcane Jill wrote:

> In article <cgl2vu$g45$1@digitaldaemon.com>, Sean Kelly says...
> 
>>To me, it's because the end result has to follow an established set of interface standards and such if it's to feel like a standard library.  I have some trepidation as well, but I think some degree of organization is necessary if this will have any hope of supplanting Phobos.
> 
> Yeah. I got it now. I'm agreeing with you now.

Agreed. One big problem with C and C++ always goes like: "where is my library" or "oops, which implementation are you using then?... and which one is broken?... ok, here is a shell script so we support both now."

Part of the reason for the spreading of languages like python or php
etc. IMO is that there is one central place where you will always find
the reference standard library (in working code). This makes it quite
clear that every other implementation that is not compatible is to be
considered wrong. When people talk about posix on the other hand, you
cannot simply point someone to "http://this-is-posix.org/source/download/"
and the problem is solved. So even if the reference implementation is
horribly broken, you at least know whom to address.

>>But I think it should be clear that no one should try and tell another *how* to do something.  Especially in their area of expertise.  I think we all feel pretty much the same way as you--if this is going to work we'll need to cooperate, and if it turns into a power struggle then I suspect we'll all leave the project.
> 
> Sounds good. I think the biggest difficulty is that committees tend to consist of people who want to be on committees. I don't have a better answer though.
> 
>

Well... there does not necessarily need to be _one_ big committee that decides about everything. The concept of python's special interest groups feels very useful and seems to work. I don't know about the details though.

cheers,
  peter.