For interest: Magma in 3.9a-7008

[Chris Muller]

> twenty-five (!) minutes. The installation otherwise progresses fine,
> except for a warning towards the end of the installation:
> "CompiledMethod>>who has been deprecated. use #methodClass and
> #selector
> directly", but you can still proceed in this case. The same
> installation
> in 3.8-6665 doesn't even take a minute.

I see the same slowness but not sure why.  The relatively idle CPU
seems very odd, as if there were network delay or delay on the other
end.

There are several stages to the load process.  The initial stages
before "Loading..." (i.e., "Writing Definitions...") appear to be the
most-significantly affected between 3.9/294 and 3.8/231.  This is the
stage where it identifies all the (prereq) packages and creates them,
you can see them being added in the Monticello browser in the
background.  The "Loading..." stage seems to be about the same between
the two (although I did not measure it with a stopwatch).

The Monticello-md.294 version appears to have less support for
"required packages" in that they no longer show up in the bottom-pane
of the Repository explorer when a package is selected as they do in
Monticello-avi.231.

I managed to shoehorn Monticello-avi.231 into 3.9 to try to equalise
the conditions.  I had to hack in a couple of methods which may be
skewing but it appeared not to help.  I'm not sure what reliable
conclusion I can draw from this, probably none..

I'm sorry I don't have any clear answers!  I would go ahead and ask on
squeak-dev or the monticello list.

I am glad to see adventurers blazing the trail to 3.9 ahead of me,
thanks!  I will try to catch up to you soon but it might still be a
while.  Just keep in mind that I have no idea how well Magma works in
3.9 right now, you should at least run the test cases successfully in
3.9 before doing anything beyond playing around.

Regards,
  Chris