historical note

Jecel Assumpcao Jr jecel at merlintec.com
Sat Jun 23 22:20:27 UTC 2001

On Saturday 23 June 2001 13:52, Lex Spoon wrote:
> You're citing an article where the guy says he can't make up his
> mind....  Which track do you consider the dead end?  :)

While he does say that each paper contradicts the previous one, it 
didn't seem that way to me as I read them.

The dead end path is Unix and C. And I already thought that as I urged 
my university department to adopt them back in 1982. I felt it would 
get them much further than the VersaDOS and Pascal that they were using 
and I was right. By 1989 I was trying to get people to replace 
Wordstar, DOS and daisy wheel printers with Write for Windows (1.0, 
which wasn't the latest version back then, but it was what they had) 
and dot matrix printers. I was fully aware that this was also a dead 
end solution.

I am not against following a dead end if where we are now is really bad 
and the construction crew is just starting work on "the right way". But 
it is silly to hope to keep on going in that direction forever.

> IMHO, the original Worse is Better presents a false dichotomy.  There
> is never a 100% solution to a problem, and instead you can spend more
> or less time designing before you start building.  No matter how much
> time you spend designing, you are doomed to eventually have a design
> assumption broken.

The second paper, "Worse Is Better Is Worse", says exactly the same 
thing. But I didn't see this as really contradicting the conclusion.

> This is not doom and gloom!  Software in many environments is
> flexible, so the problems can usually be fixed.  What I'm really
> arguing is: Software should stay alive, and should never be
> completely frozen.

Which is why it was very nice to see GNU Smalltalk come back from the 
dead after having suffered just such a fate for many years. We are back 
to the community/artifact discussion, here.

> Ah well.  I'm not sure what impact these sweeping discussions have on
> anything, but it's fun to muse about.

I hope these off topic considerations haven't been annoying too many 
people on this list. But they have a vital impact on me - I have bet 
everything I have that RPG is wrong and sometimes "the right thing" 
does win. Sometimes software problems can't be fixed, no matter how 
many billions of dollars you throw at them. If people care enough about 
them (which very rarely happens) then they are willing to start over on 
a different path.

People used to really hate me when they asked "This is ridiculous! I 
have a whooping 4 MB of RAM on my machine but my programs can only use 
640KB. When will they fix this?" and I would reply "They already have. 
But you need to give up DOS and those applications. Nobody will even 
remember SideKick in a few years anyway..."

I was obviously totally insane in those days. Still am ;-)

-- Jecel

More information about the Squeak-dev mailing list