cputney at wiresong.ca
Wed Jan 30 00:51:16 UTC 2008
On 29-Jan-08, at 3:42 PM, Joshua Gargus wrote:
> The benefits of popularity seem clear. There would be more smart
> people with more spare time to contribute good ideas and code.
> There would be more jobs and a better chance of making a living
> using the language. The second benefit would feed into the first,
> and vice-versa.
Well, I agree that smart people contributing to the community would be
a good thing. But popularity doesn't necessarily imply smart people,
it just means *more* people. I think the community we have today is
actually quite good. The "unpopularity" of Smalltalk acts as a filter.
To be a Smalltalker you've got to be smart enough to recognize the
benefits, confident enough to leave the mainstream, and resourceful
enough to overcome the obstacles that working in an "unpopular"
language entails. If Smalltalk were more popular, I doubt we would
actually get all that many more "smart people" than we have now.
Now, making a living using the language. Popularity would probably
bring more jobs, but it would also bring more programmers to compete
for those jobs. It would probably also lower the average salary of
Smalltalk jobs. That might or might not be a good thing.
> The question is whether the benefits outweigh the costs. Since the
> benefits seem obvious to me, I'll assume that you're really
> expressing skepticism about whether the benefits outweigh the
> costs. What do you think the costs are? (I can think of a few, but
> I'm curious about what others think)
I guess there are two costs. One is the effort and sacrifices required
to make Smalltalk popular. For example, we might try creating a Ruby-
on-Rails clone in Smalltalk, in order to take advantage of the current
vogue in web apps. That would be a fair amount of work, presumably
done by people who might otherwise be working on things that benefit
the existing community. Or perhaps Seaside could be "dumbed down" so
it could be marketed to the kind of developer that doesn't like the
"magic" of continuations. That makes Seaside worse for the rest of us.
The other cost is all the noise that would get introduced into the
community. Sure, Java has more libraries than Smalltalk, but most of
them are just crap. All they do is make it harder to find the good
stuff, and diffuse the energy of the community.
In general, I think we'd be better to focus not on popularity, but on
community. Yes, a certain size is required for the community to
function well, but beyond that there are diminishing returns from
further growth. As long as the VM gets maintained, libraries written,
bugs fixed, questions answered, newbies encouraged - as long as the
community is functioning - Smalltalk is sufficiently popular.
More information about the Squeak-dev