Smalltalk and Self

Andreas Raab andreas.raab at gmx.de
Thu Sep 1 04:29:26 UTC 2005


Marcus Denker wrote:
> Am 01.09.2005 um 00:12 schrieb Jecel Assumpcao Jr:
>> In short: the Self world lacked its Andreas and Ian which was the main
>> factor why Self and Squeak followed different paths.
>>
> But if you then look at the development of Tweak/Croquet
> and the latest discussions on the vm-list, it seems they learned their
> lesson and now are using a much more closed (hostile?)
> form of development...

Hey, what's your gripe, man? What exactly do you mean by "the latest 
discussions on the vm-list" (the latest discussions were about socket 
problems; before that we had discussions about how to use SVN/CVS, 
before that we were discussion problems with the 64bit changes - I fail 
to see how that relates in any way or shape with the above) and what 
exactly do you mean when you talk about a "hostile" form of development?

 From my point of view, Tweak and Croquet follow the same form of 
development that Squeak used to follow: A small, dedicated group of 
people moving the project forward, making sure it is focused on a goal 
and making sure that it gets done. There has been no change in our form 
of development that I can see.

> So something clearly did not went the right way with Squeak from their
> perspective.

Yes, that is true. There were some things that went "wrong" and that we 
felt could only be addressed by doing our own systems. Most importantly: 
We had lost control over our own destiny. That may be not too much of a 
problem if you're a hobbyist but if you are serious about doing 
something with Squeak on the scale of either Tweak or Croquet (or
Scratch or any of the Impara projects for that matter) you *must* be 
able to control your own future. It is not acceptable to stand 
helplessly on the sidelines when something very bad is about to happen 
and all you can do is scream and kick (and if you do, you get flamed 
etc). It is not acceptable if you need a fix in your system to go 
through the endless processes to get it approved. It is simply not 
doable. We have always recommended that people do exactly that even back 
in the SqC days and most of those groups who were serious took the 
advice. Such as Interval or Exobox or StableSqueak. And now, Scratch, 
Tweak and Croquet do the same (which goes to show that we actually take 
our own advice of the past).

So the lesson that we learned here is that we were right in the past, 
that you do have to take your future into your own hands and that you 
have to avoid error 33 like hell (we paid for it, too). The lesson is to 
stay focused, to keep the ball running. Discussions can be helpful but 
only up to a point - we all know (and you are just leaning it again) 
that people will generally resist *any* kind of change. At times, when 
it's important enough you just have to shove it down people's throats. 
And some will not like that one bit and you'll loose them. And others 
will join you *because* of the very changes. That's life.

> One thing that the Traits discussion again has shown that it is very  hard
> to innovate in Squeak. There is a huge hostility against doing new
> things, even if the alternative is complete standstill.

Heh. I've been saying this for years. Wanna know what a good alternative 
is? Roll your own. Then you've got control over your own destiny and can 
build the best system of all. Yiiihaaaa! ;-)

Cheers,
   - Andreas



More information about the Squeak-dev mailing list