While the smalltalk alumni who lead Eclipse continue to replicate some of the things that made smalltalk so compelling in the first place, the original smalltalk codebases and communities are still around. The Long Tail of software development tools rules.
One thing to remember about Long Tail effects- is that something out in the tail can, in the right circumstances, reemerge as a fully fledged mainstream hit. The Touching the Void, Into Thin Air effect.
While I am certainly not about to make the argument that smalltalk is set to come back and displace .NET, Eclipse and whatever scripting framework du jours is oh so hot with the cool kids, there are some interesting green shoots around the platform worth noting.
Dabble DB has the blogosphere going brock wild (nice take here on Dabble and other situated online databases). Reviews for this database, targeted at the Read/Write, Programmable Web have been pretty much uniformly positive. Dabble DB is written in smalltalk running on the seaside web apps framework. Dabble, if it does break out into the mainstream, will surely drag some smalltalk developer adoption with it.
Meanwhile people are going ape dirt for online virtual collaboration. One of the emerging environments for building same: Croquet, which is…smalltalk-based. Anyone for open source peer to peer Second Life? It could be powered by smalltalk.
I guess some people would argue Dabble and Croquet are irrelevant to the because they are not "enterprisey". That’s OK- a great deal of the enterprise budget is spent on technologies that aren’t. Smalltalk fits the lesscode revolution in a way Java obviously doesn’t.
I argued recently that mass customisation, and or situated software is becoming a more important IT industry driver/or should I say productivity driver? Never mind the line-of-business decision maker, front line employees are the only truly active end points, and many of the new productivity tools require a very limited budget. [Just to be a bit overly contrarian and slightly mad and off topic for a second – why should lines-of-business define user tools choices, anyway? Maybe the LOB buyer will be squeezed in future, as the pendulum swings back, and we’ll just end up with IT-oriented service builders and business oriented service consumers… LOB people often criticise IT for being unresponsive, what about the LOB as a brake on getting things done? Corporations are decomposing, LOBs often make terrible platform decisions.]
Another evidence point on smalltalk, if somewhat tongue in cheek: wouldn’t it be just like HP to lay off Alan Kay, a guy about to help drive a new technical revolution? There are still some Xerox Parc concepts we haven’t seen instantiated yet. Mea culpa-this is a revised view on my part.
Why is Smalltalk still around? Talking to Cote the other day something struck me: "familarity breeds contempt"… but not in development or operational tools.
When it comes to tools, familarity breeds content. With the tools we know really well we can enter flow states. Lao Tzu‘s point about non-action, or wu-wei, expanded on by Chuang Tzu, was that the ability to get things done effortlessly is actually based on relentless, remorseless practice. A master butcher slices meat off the bone effortlessly and so never needs to sharpen his knife. He got there by practice. A hack, on the other hand, spends all their time sharpening the knife (or looking for a new one).
James Mc: Thanks for the pointer, I appreciate it.
James R: Good point by the other James: cincom could surely do more to "formally" brief the analyst community, people like me and Brenda Michelson would be good places to start. I have wondered before why you didn’t get into that. We’re not just bloggers.
disclaimer: Of the companies and technologies mentioned only Eclipse is a client.
Tags: smalltalk, cincom, dabble db, Alan Kay, Croquet, Second Life, Lao Tzu, Chuang Tzu, James McGovern, James Robertson, enterprisey, HP, Eclipse, situated software, Long Tail, Read/Write Web, Programmable Web