[V3dot10] Re: loading scripts via package universes

[Keith Hodges]

Sorry Edgar but that is a load of rubbish, and is EXTREMELY unfair.
> I warn several times to Keith to play in teams.
>   
Teams need leaders, vision, direction, management, discussion, action,  
and integration. When I see a team I will let you know. When I was a lad 
I used to sail 18 berth yachts , in my teens I lead small flotillas of 
yachts. I know what team work is, I know what it looks like when it 
works and when it doesn't. The latter ends with a broken boats! 

What I see here is a few people, including myself bumbling along in a 
rough direction, trying to make a disparate bunch of pieces work 
together. And we all know its not easy to get everything working at once 
in all versions. Worse than this, while we are trying to put pieces 
together at the same time we have a goal of taking things apart!

For me, one day I am working on gjallar in squeak 3.8, the next on 
something else in 3.9, and then more recently on some seaside stuff in 
3.10. Its enough to keep anyones head spinning.
> Seems he end re-writing his own Squeak, being incompatible with existent
> code and blame us.
>   
The truth is, that I myself listened to the users requests. Lets think 
back.... requests to be able to categorize tests. So did I do my own 
project or write my own squeak. No, I did something for the benefit of 
the community. I extended SUnit to have a test categorization facility, 
not just one but three different methods, and at the same time making 
SUnit potentially compatible with SSpec as well.

So, the users requested it, it was discussed on the mailing list, and 
where is it now? In 3.10 I am afraid not, hmm wasted effort eh. Wouldn't 
be the first time. Team work hmm.

Thinking back, emails flew back and forth about split-join, I write some 
code I wrote the tests, in 3.10, fraid not, never mind eh. Ok so my code 
isnt perfect, none of my code is I regret, but did it get looked at?  
Hmm, it was in the 3.9 to 3.9.1 proposed prototype process scripts page 
for months as one of the first fixes in there.

Next request, a build and test system. Hang on, wasnt this the MAIN GOAL 
of 3.10? Yes I remember, improved tools. So I build one, ok so the 
script trigger code was in ruby, (ralphs is a shell script, I would take 
ruby every time, and fully tested using rspec, I spent a month on 
testing alone) So I build and demonstrate a build system with automatic 
running of tests and reporting of results. What happened to it? Is it 
used by 3.10 no its not. More wasted effort? Probably.

Ok, final fling. What is the biggest pain in the neck in squeak 
development, what does the release team(s) complain about the most? 
Answer: Monticello. So what might just happen to be the most important 
thing to fix, for the sake of the team? For the coherence of the 
community? Could it possibly be monticello?

So do I spend 3 month working on improving monticello for my health? I 
dont think so it is a nightmare with lots of different factors all 
working in opposition to each other. (Not to mention that the monticello 
in 3.10 is the odd one out, it has no fixes from the other versions and 
it is guaranteed incompatible, i.e. it can't load any of the other 
monticellos)

In doing this other projects stand to suffer, you know those projects 
which might be in danger of paying my bills, you know, putting food on 
the table that kind of thing, just so I an finally give up this crappy 
washing up job that gets me in such a bad mood all the time, like now!

And finally, I suggest releasing Universes from restrictions to be able 
to run arbitrary scripts. At least two other people seem to think that 
it is a useful idea, but one person doesnt, on the grounds that scripts 
are not "undoable" when neither is monticello undoable without my weeks 
worth of work I put into making it undoable. Further more there is a 
really cool undo mechanism potentially available via the System Editor 
(which needs testing and commissioning). Who cares if not absolutely 
everything is undoable? It never will be, it's completely a pie in the 
sky idea. Its best to have visibility as to what is undoable and what isnt.

Adding scripts to universes, using the packages mailing list for 
universe issues, and opening universes to be editable by everyone are 
three suggestions of mine, all of which seek to open up the 
possibilities for collaboration teamwork and inclusiveness. The first 
enables a wider variety of items to be managed in universes, including 
setups, window tidying scripts, mantis bug fix loads and anything else 
you can think of. The second and third enables the universes to be kept 
up to date more easily.

> He sure have good intentions, but  the road to hell is made with good
> intentions stones.
>   
If you want a fork then fork your own. I for one haven't got time anymore.

Little did I know it at the time, my personal life has been in limbo for 
a while leaving me with a lot of spare time. So I had a whole year to 
spend, on squeak, with over 60 hours a week to invest in moving squeak 
forward. I am happy to follow some direction, some leadership, I am 
happy to take the initiative on occasion. I have done so. Where has been 
the willingness to take on board any of these initiatives, or 
alternatively to provide some leadership for that matter?

Now I find I am snowed under with too much to do and not enough time to 
do any of it, and what am I STILL doing. I am still trying to make a 
positive contribution to the community through getting monticello to 
work better.

What does it get me, It gets me emails like that, which is totally and 
utterly insulting.

Keith