[libre-riscv-dev] pyobjc, gnu objective-c and gnustep

[Jean-Paul Chaput]

On Tue, 2020-03-31 at 11:42 +0000, lkcl . wrote:
> *click* what the hell am i saying??  jean-paul, we've got a clear need
> for improvements to coriolis2, it is *really* easy to justify putting
> in an application for NLNet funding, here.

Hello Luke,

This is a complicated matter. As a team of one, I have limited resources,
so I must prioritize (and I know myself, I'm not very good at that).
Anyway, the short term goal is to have the 180nm circuit done and
working by October. In order to do that I/we must not "finish"
Coriolis, but put it to a point it can fulfil that objective.
I'm convinced it is doable, but we should go straight to the point,
because I'm sure that lots of unexpected issues will pop-up and
I will have to tackle them quickly.

My "development cycle" are alternate steps of "adding features" then
"improve code at constant features".

So, having more hands will really be a plus (as you already proposed in
an early mail). The problem is to have someone(s) with that triple skills
(ASIC, algorithmic, software) and it has been proven difficult on my
end. To have an example of the kind of code I get, you could take a
look to the whole capacitor stuff in coriolis/oroshi/python/
(and you will easily see where I modified the code with my unique
 coding style). This was made by a good electrical engineer, but
with embryonic software knowledge. The code is to draw the layout
of one or a matrix of analog capacitors. The comments are made for
doxygen.

And besides, all of our students have already chosen their internship
(in the industry). I think we may have a good one in June.
And lastly, due to the Covid-19 and University shutdown we cannot
hire anybody, should we have a candidate.

I absolutely do not want to give you the impression that I do not
want more people working on Coriolis, quite the contrary.
But I want to expose the context as rightfully as it is.
The question being how to circumvent those initial conditions
to get more efficient people working on Coriolis?

I've taken a look at GNUstep, it seems a bit quite by there...
(maybe I was on the wrong site). For the record, WidowMaker is
one of the more memory efficient window manager I have ever seen
(still can work on my UltraSparc 1 / NetBSD).

However I will remove Boost as soon as possible, it has proven to
be a pebble in my shoes for too long. And migrating towards Python3.

I'm currently making my Python code a bit more PEP8 compliant then
backporting some of the work of Jock in soclayout. *But* I will keep
some style that you may find bad. Paradox is that the "comma" at the
start of line was showed to me by an early Python adopter back around
2000...


Best regards,
-- 

      .-.     J e a n - P a u l   C h a p u t  /  Administrateur Systeme
      /v\     Jean-Paul.Chaput at lip6.fr
    /(___)\   work: (33) 01.44.27.53.99              
     ^^ ^^    cell:      06.66.25.35.55   home: 09.65.29.83.38

    U P M C   Universite Pierre & Marie Curie
    L I P 6   Laboratoire d'Informatique de Paris VI
    S o C     System On Chip