Ian,
This is a partial reply as I am struggling to finish an urgent paper, and
we also have problems with part of our file-store containing pop-forum
informtion.
You wrote:
I've had a quick glance at what the bham install script produces, and it's
a little different to what's in your repo, so now I don't know which source
to cohere around.
The situation is much more complex than your wording suggests, and no quick
glance can provide the information you require, though the later part of
your message indicates that you've understood that!
Waldek aimed to produced, and succeeded in producing, a port that (to a
first approximation) replaced the *core* 32 bit Poplog linux sources with
64 bit sources. He also produced 'build' scripts to produce an executable,
basepop11 and saved images, and cleaned up some of the structure of the
poplog sources system.
But that left out a considerable amount of documentation and also a lot of
pop11 libraries including teaching materials -- which replaced much of the
mixed up contents of the old auto/lib/teach/help/doc/ref files with a new
'packages' system that allowed required packages to be made accessible as
needed by different users. Each package had its own auto/lib/teach.... set
of files and possibly also demo files.
That extension to poplog proved essential for our use of it in teaching and
research in Birmingham, though I don't know if anyone else ever used it. It
was all included in the 32-bit tar files previously used.
I think it is fair to say that even fairly experienced Poplog users found
it difficult to do anything with Waldek's port, until I made it easy to
fetch a version that combined his portion with the remaining directories
(including missing core documentation) though there were still some messy
features of the interface to the packages directory.
I believe David Young hopes to produce a revised (64bit and otherwise
extended) version of his vision and neural net packages.
There are also libraries provided by Steve Leach that should be added, with
a view to replacing/rationalising the sections system (unless I've
misremembered).
I think if you read my download documentation
https://www.cs.bham.ac.uk/research/projects/poplog/V16/AREADME.html
and the current fetch and install script described here:
https://www.cs.bham.ac.uk/research/projects/poplog/V16/AREADME.html#installation-process
you'll get a much better understanding than any answer to a short question
can provide.
My feeling is that before taking any decisions about what to put in a new
github package you should fetch and install the current version using my
scripts (which use all of Waldek's port -- after unpacking it and
re-packing with missing documentation and user libraries) and perhaps look
at what needs to be redesigned to integrate Waldek's core into a more
usable and extendable system.
By playing with it and finding out what works well and what doesn't you'll
get a much better idea than asking any questions with yes/no answers. I am
sure you understand this!
But it will take time, and I don't know how much time you can spare.
At present I have two problems: desperately trying to finish some overdue
papers and temporary loss of access to some of our file-store which is on
an old machine that has died and is currently being (slowly) brought back
to life, not helped by the campus lock-down which limits what our computer
officers can do.
The download script references your work which is great,
but the readme at https://github.com/hebisch/poplog says "...currently only
core part". So is that complete? What's missing?
I have tried to outline an answer to that above but the full answer needs
more detail.
Would there be a consensus that your version is the one to start from?
Waldek's version is an absolutely essential core part of a new rationalised
64 bit Poplog. But for most people it is an unusable skeletal core. What is
needed for 'flesh' in the long term should be decided somehow before you
produce any new github installation. There may be a better solution than
my hasty package.
The rest of your message outlined some reasonable steps towards a new
combination. But it will need a lot of thought if it is to produce a design
for the next decade!!
A minor point:
There is a problem with motif that has convinced me there's a part of the
system source that needs to handle CtrlC interrupts in poplog graphic
windows instead of killing poplog as happens now.
That doesn't happen in an xved window with motif. So I suspect there's
something in the xved sources that handles CtrlC with motif that needs to
be copied to the main pop11 graphic window sources for the case when motif
is included.
I may have got some detail wrong, but there's definitely something not yet
right and I suspect it is fixable. But it requires a level of programming
knowledge that I lack and will not have time to acquire in the foreseeable
future.
Of course this will not affect people who don't use motif, but there are
some nice things that can be done only with motif, including use of some of
ISL's tools, I think.
Written in haste, apologies for any errors/omissions, etc.
Aaron
