[squeak-dev] I Want

keith keith_hodges at yahoo.co.uk
Fri Feb 19 02:11:37 UTC 2010


On second thoughts I don't think I should take this sitting down. I  
will reply to you because I think you are at heart a reasonable person.

>   It also included some pathetic arse licking and gang behaviour  
> from others

The others you refer to are ones who actually bothered to look at my  
stuff, and actually contributed their input to it, as opposed to  
Randal who was just talking about something he had never even  
investigated. They were arguing from knowledge not from ignorance.

> and reminded me of nothing as much as my 9 year old son's own  
> battles at school.

I agree entirely with that, I recognise a bully when I see one, and  
there are several around here.

>  I think that Keith is essentially a mediocre talent

I agree with that too. I only have 34 years experience of coding, but  
I never claimed to be as good as anyone else. I only do high level  
coding if I can get away with it, but I have been doing social process  
engineering stuff for 20 years.

My social process analysis of trunk is that it is effectively  
bottlenecked behind one person, and it locks out many contributors  
like myself. the proof is in the number of effective contributors,  
that can refactor anything without breaking the process, and I think  
that value is arguably the same or less than we had. I suspect it is  
actually zero.

It is precisely because us mediocre contributors such as myself want a  
process that we can use without some "guru" type telling us our  
mediocre code is crap, that we wanted a peoples-process in the first  
place. We want an actual process NOT a person.

For example, if I am refactoring the compiler, and I break it, how can  
I publish that to trunk and ask anyone to help me? It will not load in  
their image via trunk will it? You see the amount of actual  
collaborating faciliated by trunk is minimal.

Andreas did not give us a process he gave us a repository manned by a  
person. Opening a shared repository does not a process make. We  
already had one 3 years ago, and you were welcome to contribute there  
and have your contributions added to the auto-build or LPF. It also  
accepted patches and changesets packaged up as mcz files.

We want a process we can use to control the image we want to build,  
hence the goal for a kernel image. Sonce the goal for the community is  
and has been a kernel image for as long as I can remember then a  
process needs to be deigned that is actually capable of producing a  
kernel image. We mediocre types want a process where we can put our  
mediocre ideas, and get help from non-mediocre people like you to make  
them work and shine.

I benefited greatly from contributing fixes and ideas to mantis and  
having excellent feedback there, however we wanted more than that we  
wanted the ability to have multiple unstable and experimental builds.  
The world has moved on from CVS, we are now in the territory of  
multiple branching.

So all power to mediocre coders, that is what smalltalk is all about.  
If we wanted to be coding superstars we would have learned C++ in the  
first place. I did real work in Hypercard and I am proud of it.

> who has some good ideas but who doesn't execute very well.

Where where you when the board cancelled 3.11 altogether? I did not  
see you volunteering. I waited until no one else wanted the job. The  
future we were told was spoon, or squeak 5.0. Which apparently was 1  
month away at the time!!!! I don't see Craig getting the hassle I was  
given for being later on his deadlines than I ever was.

I see it is all right to support and endorse a mediocre, perhaps even  
slightly out of his depth person to do a job you don't want to  
volunteer for, when it suits you. I see it is also alright to drop  
them when it no longer suits you. The mediocre person is fair game to  
be walked over.

I also see that it is alright for anyone to walk all over someone else  
who is not so good as them, just because you think you are gods gift  
to coding, and you know better than everyone else and cant be bothered  
to invest the time and energy to understand what has happened in the  
community in the three years or more that you have been absent from it.

Having been blessed with non-mediocre status is apparently a licence  
to treat everyone else as below being worth talking to. Even people  
that had previously invited you to invest time and money, and other  
peoples money, on a project.

If Andreas had been involved in the release team at all himself, he  
would have known that Metthew as 3.11 team leader had been sequestered  
by the board (aka Randall) (again without discussion) to work on  
Squeak 4.0 the relicencing deal instead, so no only did the board make  
me effectively a man down, they also told me to SLOW down on 3.11  
because it would have to wait for the re-licencing.

> Keith himself said in the IRC log three years of work on continuous  
> integration was wasted by the new trunk.  Three years is simply too  
> long in this space.

Well it was about 1 and a bit years from when we officially took over  
3.11, I was incorporating 3 years of work, in the final project. SUnit  
test runner was written in Sept 2006ish Bob version one was written in  
ruby in January 2007. Bob2 was finished in Feb 2009.

This community and the board needs to watch that video on open source  
projects and poisoned people several times. The FIRST principles in  
that video were


When you lack any of those things in your leadership, you will  
inevitably end up poisoning people. Tell me that has not happened here?

In real life, I am a carer for a person with 30 multiple personalities  
who is extremely sensitive, and I ran an open house for 8 years. I  
have spent 15 years helping victims of abuse, incest, rape and  
torture. I am, I am told a fairly nice guy. Seriously the squeak  
community needs to look at what it has become if it can have this  
effect on me.

Trust is an important one. I trusted the board, and my employer  
trusted both the board and me. I simply trust that if I am asked to do  
a job, and I am doing it. That the board will actually talk to me  
about things that effect my work, and they will provide a layer of  
protection from the wider public or anyone else who does not know the  
details of progress.

Humility, is also important, one might ask if you have someone who has  
been thinking about the problem for 3 or 4 years they might be worth  
getting together in a round the table discussion about it. To have the  
arrogance to come up with the solution on the back of an envelope and  
then impose it on the rest of us without discussion is not humility.

Politeness, well I reserve politeness for those who deserve it. I know  
I am wrong on this one, and I am working on it.



p.s. I might vote after all. Chris Muller is the ONLY person running  
for the board who actually produces anything in the form of loadable  
packages for other Squeakers to use across the board. He knows the  
difficulties of maintaining a package for several forks.

More information about the Squeak-dev mailing list