So you’ve read Kent Beck’s Smalltalk: Welcome to the Balkans:
I don’t think the real issue is standardization.
It’s true that each dialect of Smalltalk has quirks, but version differences are not unique to Smalltalk. It can be challenging to move Ruby programs between different versions of Ruby and it can be challenging to move C code between different operating systems and compilers.
Seaside is proof that a significant body of code can be written to be portable across at least 7 different Smalltalk dialects. It takes a hefty glob of Grease (a Smalltalk portability layer), but it can be done.
Aida and Swazoo are additional examples of significant applications that have been ported to multiple dialects of Smalltalk. They are built on top of Sport (another Smalltalk portability layer).
Writing portable Smalltalk is a solvable problem.
Kent’s proposed solution is a good idea in and of itself:
Have the tiniest possible core defined in terms of test cases. Build a shared library on top of that, implemented in terms of the core. Include numbers, collections, meta-objects, code structure, and code loading. None of this parcel/bundle/package/pundle/category nonsense. Compete on VMs, graphics libraries, and enterprise-y tools.
But expecting a set of vendors to agree on the the “tiniest possible core”, when they can’t agree on a set of common methods is unrealistic.
Here’s what sent Kent off on his diatribe:
All the tests pass in Squeak finally. I’ve been making the changes in place…I load my code back into Monticello … I go to load into VW. Kaboom. The meta information is snarled. The monkey patches are gone. Merging does nothing sensible.
and it’s sent me off on my own diatribe:)
I’ve been working in Smalltalk for 25 years and this problem is just as bad now as it was 25 years ago! It is virtually impossible to do cross-dialect development in Smalltalk. It was unfortunate for Kent that Monticello was partially ported, because it gave him the illusion that cross-dialect development was possible:
Right on, Kent! Existing customers may not demand cross-dialect development, but prospective Smalltalkers are being turned away and turning your back on cross-dialect development is turning your back on the Smalltalk community.
You know the story of the Smalltalk balloon:
The cover of that issue depicted the “land of Smalltalk” as a remote island, and that triggered a connection for me out of which was born the fanstasy of liberating Smalltalk from the ivory tower by balloon ascent.
Smalltalk isn’t just balkanized, Smalltalk is an archipelago where every dialect is stranded on its own island! It would be a luxury to have a hot air balloon to transport code modifications back and forth between dialects!
It is technically feasible to port Monticello to every dialect of Smalltalk.
It is an idea whose time has come and gone about 50 times! You can claim that a file-based Smalltalk wouldn’t have this problem, but Monticello is file-based development for Smalltalk.
Monticello makes cross-dialect development possible.
The Monticello code base has not had the attention to portability that Seaside and Aida have, but seriously, Monticello isn’t under active development. A one off port is perfectly adequate. I did a one off port of Monticello to GemStone in December of 2006. It’s not incredibly difficult and it has not required a lot of maintenance.
Monticello has withstood the test of time in the Squeak community and it does a pretty good job:
Envy and Store are not viable candidates. Monticello is a light-weight solution compared to Store and Envy.
If you can afford to port Seaside, then you can afford to port Monticello.
So why don’t we just do it?
It isn’t realistic to expect all Smalltalk dialects to live on the same island, but to continue to perpetuate the current isolationism between dialects is just pathetic:
Thank you, Kent for the kick in the butt!
And Happy Valentines Day!