> From: Lassi Kortela <xxxxxx@lassi.io>
> Date: Sunday, October 06, 2019 12:48 PM
>
>> But I'm persuaded that we should have a Schemecomm mailing list, too....
>
>>> +1 for pre-SRFI topic list.
>
>> I'll start with Schemecomm.
>
> This would really be a good time to think about those subdomains and
> how to organize things....

For recommending "Action This Day" creation of new topics, I have used
the following requirements:

- The topic is cleanly separable from existing ones

- The topic is "big enough".

- There are "enough" people interested in it

- The topic has at least two different things people are interested in

- At least one of those things must have a commitment by a participant
  to produce concrete artifacts, SRFIs and/or libraries.

I moved from wild idea suggestion to "let's create Schemecomm" because
Amirouche has a specific, concrete idea about peer-to-peer
communications worthy of at least one SRFI, the base communications, and
applications he envisions on top of it beyond the existing file sharing
domain.  And then radio excited interest.

Schemecomm is technically a partial superset of Schemeweb, but I think
we can keep two distinct enough. A bit of research suggests web sockets
falls into it because they don't use HTTP or necessarily connect to a
web server per se.

Communications is *huge* in society, and breadth of topics.

We've got 3+ people each for peer-to-peer and radio, on a weekend.

As for the total set of lists today, looking at the basic building
blocks of a computer, CPU + memory are partly covered by the Scheme
language and its implementations, that's partly what I though pre-srfi
might include.

Storage is covered by Schemepersist, I/O beyond the most basic like DASD
by Schemecomm and Schemeweb.

Stepping back, include the user: he needs documentation, so Schemedoc.

He needs human interfacing, so there's Schemeweb.  But for more direct
interfaces I can see a Schemeui someday, also a partial superset of
Schemeweb.  But I emphasize "someday", for me this is firmly in the
category of "someone should do this, but I'm not going to"....

Schemesec for security is another I've thought we should eventually do
to satisfy what I gather are the ambitions of R7RS-large.

What else?  Here's digested lists of ACM Special Interest Groups, from
https://www.acm.org/special-interest-groups/sigs-by-knowledge-area,
minus some "obvious" exceptions, and areas already covered above:

SIGACCESS  - Accessibility and Computing
SIGMETRICS - Measurement and Evaluation
SIGMOBILE  - Mobility of Systems, Users, Data & Comp
SIGOPS     - Operating Systems

Specific application domains:

SIGAPP     - Applied Computing
SIGBED     - Embedded Systems
SIGBio     - Bioinformatics, Computational Biology
SIGDA      - Design Automation
SIGecom    - Economics and Computation
SIGGRAPH   - Computer Graphics
SIGHPC     - High Performance Computing
SIGMIS     - Management Information Systems
SIGMM      - Multimedia Systems
SIGSIM     - Simulation
SIGSPATIAL - Spatial Information
SIGUCCS    - University & College Computing Services

The following could be handled by expanding the remit of Schemedoc:

SIGDOC     - Design of Communication ("documentation")
SIGIR      - Information Retrieval
SIGKDD     - Knowledge Discovery and Data Mining

The computer science and the like topics:

SIGACT     - Algorithms & Computation Theory
SIGAI      - Artificial Intelligence
SIGARCH    - Computer Architecture
SIGCSE     - Computer Science Education
SIGEVO     - Genetic and Evolutionary Computation (e.g. genetic algorithms)
SIGITE     - Information Technology Education
SIGLOG     - Logic and Computation (e.g. proofs)
SIGMICRO   - Microarchitecture
SIGPLAN    - Programming Languages
SIGSOFT    - Software Engineering
SIGSAM     - Symbolic & Algebraic Manipulation

None seem ready for a topic list today or on a vaguely specific
date in the future.  I agree with Lassi that it's easy to get
overwhelmed, but Scheme is a survivor, it's got time to grow.

- Harold