People tend to think that Smalltalk is not widely used as it is not in the (stupid) TIOBE index. But Smalltalk is not only a language but a fully dynamic object system (the language is just built in using objects and messages and fully adoptable which makes it hard to compare to regular programming languages). Compared to any other it is also the most pure object oriented systems.
Mostly other languages started to be "object oriented" for fashion reason without following good OO practice. There is also a very simple reason why Smalltalk is not highly ranked in TIOBE index. You do not need a webbrowser or compiler manual. You can concentrate on the problem solving and lively development in the Smalltalk system instead of googling for keywords, syntax and API's.
Nonetheless there are so many projects and applications that are built in Smalltalk - unfortunately they are not that visible in the first place even when some of them influence us daily like financial applications, container management, railway planning and ERP systems.
I was surprised today to see that someone started a #thingsPeopleBuiltWithSmalltalk campaign on Twitter. Nice initiative!
So if you know something built in Smalltalk then just twitter it:
https://twitter.com/hashtag/thingsPeopleBuiltWithSmalltalk?src=hash
Lets make 2015 the "There is no excuse not to know about Smalltalk" year ;)
There's another – much bigger – reason why Smalltalk isn't on the TIOBE index: the index is predicated on nonsensical data, making it totally meaningless rubbish. Just look at the January index. Languages such as Erlang, Go, Groovy, and Haskell (which were on the top 50 list just a couple of months ago) have dropped off like a ship's anchor. What explains this??
ReplyDeleteAnd why are languages such as Ada, OpenEdge ABL, PostScript, and Max/MSP ranked so highly on the top 50?? These languages are hardly relevant in today's IT world.
The TIOBE index is an insult to our collective intelligence. It cannot be trusted to guide our decision-making. It hardly qualifies as a valid data point, even, for any kind of reasonable analysis.
I use langpop.corger.nl as a substitute for TIOBE. It's not perfect, either, but I trust the data a whole lot more.
Berke Durak expressed similar sentiments 7 years ago: http://lambda-diode.com/programming/the-tiobe-index-is-meaningless