[Openmcl-devel] user submissions and open-sourceness

[p2.edoc at googlemail.com]

At 5:15 PM -0400 08/06/16, Jeremy Jones wrote:
>But contributing to a port in progress is much harder.  It requires a
>lot of communication.  Even that is possible if the people doing the
>port think that the contribution will outweigh the communication
>overhead.  I can't think of anyone who takes the time to clearly
>explain detailed technical issues better than Gary.  What should we be
>doing differently?

I am not inclined or qualified to tell Clozure to do anything. I'm 
sure we all have our own wish list priorities for CCL (mine are 
currently all in the UI), but Clozure are presumably so busy that 
nagging would seem counter productive.

Coming to mind is an analogy - driving a car on a feeder road up to a 
major route. It can seem a challenge to stand at the edge 
contemplating joining the juggernauts as they thunder by, so it might 
help to have appropriate ramps and sign posts where it would be 
fruitful to filter in.  It can be discouraging to put effort (no 
matter how ineffectual as far as others are concerned) into some 
project only to find it missed the bus (sorry for mixing metaphors). 
Just anticipation of this could be a disincentive to those thinking 
of contributing. By this I mean that it could help if there was 
guidance as to what, where and how contributions would be welcome. It 
may be disappointing to find that our contributions are not welcome 
by Clozure (inappropriate, sub standard, worthless, duplicating, side 
effecting, intrinsically obsolescent, stupid, ...), but if we are 
self motivated we can always pursue such ventures in parallel, and 
others users might possibly find them useful.

I'd like to see some sort of contributions facility. This would 
provide guidance on how to contribute, how to communicate in that 
regard, who is doing what, and a list of areas where contributions 
would be welcome.  Although CCL is free and Open Source (and how can 
anyone not be hugely grateful and motivated by that), it is the 
possession and in the control of Clozure, hence presumably any 
contributions for CCL need to be appropriate and welcomed by Clozure 
to have much chance of coming to fruition.  But that unilateral 
control also distances those that do not feel they are a part of the 
Clozure extended flock (are still on side roads).  Hence some voting 
system where everyone can make suggestions and indicate priorities 
might be useful.  Not structured so it could be any sort of style 
cramper on Clozure, rather enfranchising for potential contributors 
and users (and of course useful for Clozure).  Jeremy's survey (on 
info-mcl last year) allowed some folks to make some input, but it was 
a one shot.  I have in mind a rolling system that shows where work is 
in progress, where, when and how items are planned to be addressed, 
and the collective wish-list stretching ahead, where contribs are 
needed. The whole structured to be the least style cramping or energy 
sapping for Clozure, yet encouraging for contributors. And be a means 
by which the loud know-it-all types can be diluted by general common 
sense and goodwill.

The likes of Gary's recent clarification on the x86 32bit system 
thread: [not even at the point where external feedback/testing would 
be useful.  ("It's buggy as hell!" "We know!")] is useful.  >99% of 
what I have to say about CCL, I don't, as I assume it'd be time out 
of sych with requirements (the rest is probably anyway).

Just this minute I'd like such as a Who Calls facility in the UI, 
mouse-copy, curb the CCL Console window, m-. path translations that 
work out of the can, Search Files dialog, copy in the apropos and 
inspector dialogs, bracket flashing fix, process peek, non-modal 
multi-term apropos dialog, backtrace dialog inspect item mouse 
action, init file documentation, ... many more.
Do they exist somewhere already, should I wait till they appear, how 
long will that be, should I roll my own, if so how can I do this so 
it helps others (and doesn't overlap Clozure work or other 
simultaneous wheel inventors), what do others want of them, where do 
they belong within CCL if at all, ...

My top priority for CCL is robustness, and to that end it would seem 
that contributed code would be the better for being both embraced by 
Clozure and used by many. Contributions guidance would seem to help 
this cause in many ways.

Not to suggest that the equivalent hasn't been elegantly worked out 
in other projects, or that this is necessarily the right time for it, 
but I get the impression there's not yet much along these lines 
available now for CCL (other than 
<http://trac.clozure.com/openmcl/wiki/ToDo>).

The above a rambling 2c thought and not meant to be any demand or 
criticism of anyone (although any rolling contrib/wishlist facility 
would take effort, I think it could pay back for CCL and it's uptake).