On 8 June 2018 at 20:49, Sean P. DeNigris <sean@clipperadams.com> wrote:

Ben Coman wrote
> the most important are...
> performance, portability and development speed.

I wonder about this because Ruby and Python are no speed demons and it's
hard to imagine anyone (except maybe lisp) beating Smalltalk in development
speed.

So its not all three at once, but different choices for different domains.

And maybe Smalltalk is trumped by the next two most important...��

�� Personal familiarity (E)

�� Team familiarity (E)

where such "familiarity" encompasses text-based versus image-based systems.

And we can't do much about these...

�� Extending existing code��(E)

�� Already used in group (E)

So we just promote what we can influene.

��

I didn't read the article, but it seems to me the main factor in popularity
is random chance because the industry is a pop culture where trends are more
about being cool (or getting a job) than any underlying principles.

And this reinforces "Personal familiarity" and perhaps a bit like Coca Cola adverts.

Their aim is not to make you rush from the TV to go and get a Coke,

but just the next time you are standing in the corner store deciding��

which you might choose, its��Coke that comes to mind.

That said, IMHO mainstream popularity is a curse. The sweet spot is a
critical mass where the core and desired libraries are available and
maintained.

Agreed.

cheers -ben