I will not argue whether or not Smalltalk and Lisp were right all along. However, the fact that Smalltalk and Lisp are where they are is not simply a matter of technical merit. There were many, many, political decisions made by the owners of the technology which participated in its successes and failures. Proprietary, commercial, expensive, and the varying corporate entities one had to deal with in order to use this software was a huge factor. Us small guys who would want to use this fantastic enabling technology had our limitations in acquiring this technology. You had to have money and really want to spend it in this manner. This is not to make derogatory statements about their decisions. But they did have an impact. Had the best of Smalltalk and Lisp been open sourced 20 to 30 years ago. Would this be a very different world? I think so. However, the income and profits of those corporations would have been different also. It wasn't our decision to make and we were not living in their shoes. We would have to become knowledgeable about the economics, the state of the competition, the state of their own business, etc. to become even reasonably qualified to judge. Even then different people make different decisions because they think differently or even possibly are looking for different results. Open source was a different and very foreign world back then. It was a very foreign thought to open source your intellectual property and believe that you could earn a living from it. You do have to think different to be a creator of open source software and to be successful at making a living from it. What if Sun put the effort into Strongtalk that they put into Java? Where would we be? I personally believe Smalltalk and Lisp had it right. Own the entire environment and machine. Turtles all the way down. Unfortunately the world isn't so clean. However, we do have a very nice open source Smalltalk inspired tool in Pharo. While the past was not in our control. We do have a say about the future. :) Lets make the future great! Shalom Aleichem Jimmie On 01/20/2016 10:13 AM, Dimitris Chloupis wrote:
indeed python and C++ are not just diffirent but two languages going the exact opposite directions, Python emphasizing speed of development , C++ emphasizing speed of execution.
In C/C++ every decision must be based upon performance, this not how any other language works , even Java is not that worried about performance, sure its important but it is not the main goal.
Hence its not surprising that C++ is the undisputed choice for very performance orientated fields, like Gaming, audio , science that involves heavy calculation. Look at python how popular it has become for scientists but AFTER the fact that so many C++ libraries have been made for python like numpy/scipy and Blaze.
So you are absolute to say that some people migrate to dynamic languages not just for dynamic types and is also to say that people stick with C++ and C not because of static types. If that was the case python would not implement optional static types and C++ would not implement generics.
The important thing to note here is that we no longer at 80 and 90s , nowdays you can combine any language with anything else and the vast majority of projects are written in multiple languages.
I dont agree with the remark "Smalltalk and Lisp were right all along" , obviously they were not because the entire world would have been using smalltalk and lisp which we dont. To claim so is to try to simplify coding and with every simplification it loses a part of the truth.
On Wed, Jan 20, 2016 at 5:57 PM Martin Bähr <mbaehr@email.archlab.tuwien.ac.at <mailto:mbaehr@email.archlab.tuwien.ac.at>> wrote:
Excerpts from Dimitris Chloupis's message of 2016-01-20 12:36:37 +0100: > we have witnessed 3 great migrations of coders > 1) The migration from Assembly to C/C++ and other high level languages > 2) The migration from C++ as the dominant force of coding to Java > 3) The migration from static types languages to dynamic typed languages
that's a great observation! and you are right, dynamic typed languages have won and smalltalk and lisp had it right all along.
> On the matter of python getting optional static typing I can say this and > predict this, static type will never become anywhere as big for python as > generic types are in static type languages and I say that because I have a > good understanding of the python culture.
that's a good point too. what i am interested in is the fact that having types available in these languages, research can finally look for conclusive evidence of how much advantage types really give, because all other differences between eg python and c++ are eliminated.
i simply do not believe that any findings about types, by comparing python with c++ is valid because they are such different beasts of languages.
greetings, martin.
-- eKita - the online platform for your entire academic life -- chief engineer eKita.co pike programmer pike.lysator.liu.se <http://pike.lysator.liu.se> caudium.net <http://caudium.net> societyserver.org <http://societyserver.org> secretary beijinglug.org <http://beijinglug.org> mentor fossasia.org <http://fossasia.org> foresight developer foresightlinux.org <http://foresightlinux.org> realss.com <http://realss.com> unix sysadmin Martin Bähr working in china http://societyserver.org/mbaehr/