[Squeakfoundation]Concrete SqF projects

[Paul Fernhout]

Andreas.Raab@gmx.de wrote:
> [Snip]
> Starting some of those projects might also help finding
> (at least interim) project officers for some of the issues 
> that will stay in the long-term focus of SqF.

Agreed. Defining projects is also a good way to produce feedback on the
principles and the purpose of a Chaordic organization as an iterative
process.
  http://www.chaordic.org/what_des.html

> So what could we start with in practice?!  [snip]
> * Establish a web-presence of Squeak image releases
> Meaning a central list pointing out what versions exists, 
> when they were released, what VM to use with it, 
> an FAQ per release, etc. etc. etc.
> 
> * Establish a web-presence of Squeak VMs
> Meaning a central list of compiled VMs that are available, 
> references to the source code for each version, 
> telling people that "If the VM for version XYZ
> hasn't been compiled on ABC, please drop us a copy", 
> FAQs for VMs/platforms,
> etc. etc. etc.
> 
> * Establish a web-presence of Squeak packages [GOODIEs] [ADDONs] etc.
> Meaning a central list of available packages, goodies, 
> add-ons, telling people what this thing does, the Squeak version 
> it has been designed for, the versions it has been successfully 
> tested on, perhaps a swiki for posting bugs etc.
[snip]
* Review and coordinate a migration of Stable Squeak 
> into the main release
> [snip]
* Write and post a monthly "Mini Squeak FAQ" to the mailing list
> [more snipped]

One could also add to the organizing tasks determining who wrote the
package, the licensing status for each contribution (Squeak-like, public
domain, BSD, GPL, etc.), and contact info for the authors. Another
related non-programming task could be going through the Squeak mailing
list and mining it for summaries, ideas, discussions, and so on to build
a Squeak and Smalltalk related knowledgebase.

These are all great projects. They all would certainly help address the
issues Glyph Lefkowitz raised  on the list recently (April 19, 2001). He
linked to a page describing his experience where he downloaded
approximately 50 changesets and found they would not run at all on the
setup he used, or that if they ran they often conflicted with each other
when he installed more than one at a time.
  http://twistedmatrix.com/users/glyph/rant/sqcr.html
 
With the exception of the mini-faq, one question to think about is why
haven't these consolidation and organization projects happened already?

I certainly wouldn't say most people in the Squeak community are lazy or
without resources. Time and time again people on the Squeak list
demonstrate perseverance, generosity, and lots of hard work. There are
also several Wikis where this sort of effort could be happening at no
cost. As you point out, there are also many files already out there at
easily referenced URLS, so part of the job is done already. 

Perhaps this hasn't happened yet because managing this level of
complexity is very hard work, and exceeds some magic threshold of what
can be easily done by people currently attracted to Squeak. These
projects if comprehensive would take an enormous amount of (often boring
yet difficult) work sustained over a long period of time (think RedHat
for Squeak). Some of that work is unavoidable. But, as a fundamentally
bright but lazy programmer (which is not as incompatible with sustained
hard work as one might think, paradoxically), I wouldn't want to see a
few creative people doing all this work if it can be avoided or easily
distributed across the community as part of daily practice.

I think Tim Rowledge is, as the Brits say, "spot on" when he wrote in
another reply: "a really useful thing to do whilst attacking these would
be to make an effort to make tools (code, conceptual, procedural) to
help do it next time around."

Alan Kay presents a vision of the Dynabook where children learn math and
science by collaboratively making a space war game. Squeak now seems to
be (especially with eToy on Morphic) the premier environment for making
such a game as an individual (just fork and go). However, I feel, it
does not seem to yet be the premier collaborative programming
environment for other than at best a small tightly-knit mostly
face-to-face group (contrasted with, say, Doug Engelbart's 1968 Augment
demo).
  http://www.bootstrap.org/

I think the lack of these well organized indexes and presentations and
module details is not the problem itself. I see it more as a symptom of
the actual problem. That problem is the lack of more such tools for
collaborative community software development in Squeak and a lack of
related infrastructure for complexity management of community developed
Smalltalk software. To date the Squeak community process has been mainly
about making Squeak work better. I say, we need to move on in a big way
to using Squeak to make the community process work better. Yes,
something like Nebraska is a step in the right direction, as would be,
in my opinion, something like a peer-to-peer ENVY. 

Obviously a doctor would immediately treats the injury if a patient
comes in, say, with a broken leg from jumping on a trampoline at a party
(and so too these projects you outline could and probably should get
started now by hand to at least some extent). However, it is is likely
the doctor would also point out the risks to the patient should he or
she jump on a trampoline again in the future as part of a group. Due to
wave dynamics, seemingly innocent backyard trampolines can cause
terrible injuries seemingly at random -- literally shattering bones into
thousands of fragments -- if more than one person jumps on a trampoline
at a time. 
  http://www.cpsc.gov/cpscpub/pubs/085.html
So, how can we make the Squeak "trampoline" safe for a community to jump
on all at once (and still keep it fun and worth using)?

Tim's suggestion is the path between and does seem the best of both
worlds, and again shows why he would be a good choice for "benevolent
dictator for life". 

And reiterating the comment Cees made to me about duplication of effort
(posted earlier) there is no reason we can not have many people
organizing lists, many people making lists and tools, and many people
making tools, all without much coordination. The marketplace of the web
will decide among the successes by whose work gets used and linked to.
In a large community, duplication of such effort is not necessarily a
bad thing.

I might add, how people work together (or not) within the Squeak
Foundation is a completely different issue from perhaps having the
Squeak Foundation provide a united front or one voice to the world from
a Squeak marketing standpoint. 

-Paul Fernhout
Kurtz-Fernhout Software 
=========================================================
Developers of custom software and educational simulations
Creators of the Garden with Insight(TM) garden simulator
http://www.kurtz-fernhout.com