Hi Cesar, El vie., 1 feb. 2019 a las 20:23, Cesar Rabak (<cesar.rabak@gmail.com>) escribió:
I'm having a good time reading to all those diatribes turned into admonitions to 'non initiated' into the arcane art of Smalltalk.
However, I see the need to inform you that as presented so far, the phrases bring too much emotional content instead of intelligent down to Earth updated to XXI century information!
I know, and I mostly agree with you. The emotional part is totally on purpose, at least from my side, because these kind of articles are meant to shock readers, and motion curiosity. I won't see them as "admonitions" but a compilation of false or dubious perceptions about Smalltalk. As this is WIP we could also add specific examples supporting the claims, although I didn't seen it too much in other articles of this type.
For example, all the mentions to adverbs like *slow*, l*ess* and so on, are subject to an analysis that only in a concrete context and specific case can be correctly assayed.
Again agreed. However I should say I need to see a backup yet to one of the greatest lies in software developement: "Python is easy" (which doesn't release smalltalkers from nothing, but is a nice reminder of rethoric power). One of my ideas behind asking in mailing-lists to well-known experts is to avoid ex silentio fallacy, but also is difficult to escape half-truth phrases.
Other class of objectionable arguments are about the implementations, they bring an enthymeme that the specific feature or approach is not desirable (e.g. fixed size bytecodes, etc.) without showing the reason for the need to "defend" the position it seems to me that we are at a bar counter discussing the tone of the color of the adversaries' sport teams uniforms...
The fixed size bytecodes is a bad example, I should remove it.
Even the "General Myths" compiled in github would need to be better addressed.
For each of them, in the proper realm they are more than acceptable as a reasonable proxy for the truth, notwithstanding disgust.
For example: for what "real world" "real time" application can an example be given that is not a convoluted "demonstration" of one vendor or an application developed for academic research that is abandonware or conversely was done in an earlier version of its Smalltalk environment?
Bonus point if an example is used in production (meaning people is collecting money due is use, and has been developed as recent need and not a decision of not redo/recode a legacy application.
Please note there are people which doesn't focus solely on industry goals, for example whatever you develop in academics today is to focus an extremely specific problem, and mostly impossible to measure in terms of money-making at least at scientific journal publication time. Factors of success in academic research software tools are completely different than industry, so yes it is very common to find "abandonware" there, but they weren't thought as a "adoptware" in the first place. Even if you do today a state-of-the-art genomic aligner, you can only measure real adoption after a few years (and citation factor is not a measure of adoption), and your users will be a few guys in laboratories and bioengineering niches.
My counter examples:
IBM dropped its Smalltalk business.
Interesting to see IBM mentioned as a referring player, here in Argentina is seen as one of the worst enterprises after the "IBM affair" [1] scandal. Tektronix used to employ Smalltalk in their intelligent instruments, no
longer: they follow another (industry standard) approach.
We ourselves migrated to non smalltalk solutions for the code repos...
Sad to read about your migration. I imagine wasn't something you did gladly. You know there are many articles and forums where Smalltalk is totally destroyed, without anyone to contend claims. And they aren't even "claims"... is just plain insulting without any experience to back up. Isn't any news anymore to read about weakness of Smalltalk implementations, lack of community support, quality, etc. That could be done with almost any developement tool, but strangely one cannot find so fierce criticism against bash, or R, when it's almost 2020 and debugging or testing with them is a complete nightmare. That doesn't mean we should support or reject any myths. I like to read both supports and constructive criticism. This is a complex subject and collecting experience samples/reports is something the community could use more.
*For every specific problem a specialized language would overperform all others. However certain languages are inherently broad scope enough to keep being mainstream.*
Smalltalk currently can only be considered as niche language/environment/whatever the purists would like to call it, but has not caught the World. . .
True, but I'm not sure that catching the World is a reasonable goal. Making life easier for those now working with Smalltalk as a means of living could be more reasonable, at least for this age. It doesn't mean current-generation smalltalkers hasn't learnt nothing or cannot share any knowledge. Cheers, Hernán [1] https://medium.com/worm-capital/the-ibm-hall-of-shame-a-semi-complete-list-o...
My 0.019999
-- Cesar Rabak
On Fri, Feb 1, 2019 at 5:15 PM Hernán Morales Durand < hernan.morales@gmail.com> wrote:
Added to the list. Nice catch!
Cheers,
Hernán
El vie., 1 feb. 2019 a las 2:33, stephen@heaveneverywhere.com (< stephen@heaveneverywhere.com>) escribió:
How about:
Smalltalk is less efficient than "low-level" languages.
Smalltalk can't do "real time."
Garbage collection is less efficient than manual memory management.
Garbage collection is really slow if you use complex data structures.
stp
--
Stephen Travis Pope Santa Barbara, California, USA
http://HeavenEverywhere.com http://FASTLabInc.com https://vimeo.com/user19434036/videos *http://heaveneverywhere.com/Reflections <http://heaveneverywhere.com/Reflections>*
--
On Jan 28, 2019, at 5:22 AM, Reinout Heeck <vwnclist@soops.nl> wrote:
Smalltalk is a programming language.
R
-
On 20-1-2019 23:31, Hernán Morales Durand wrote:
Hi there,
I just created a GitHub repo to collect myths around Smalltalk-based technologies: Pharo, Squeak, VW, VAST, Smalltalk/X, GNU/ST, etc. in the spirit of the Falsehoods lists [1-4].
This is just a draft now but please feel free to add falsehoods based on your own experiences. Examples are greatly appreciated.
https://github.com/hernanmd/falsehoods_smalltalk
Cheers,
Hernán
[1] https://infiniteundo.com/post/25326999628/falsehoods-programmers-believe-abo... [2] https://chiselapp.com/user/ttmrichter/repository/gng/doc/trunk/output/falseh... [3] https://github.com/googlei18n/libphonenumber/blob/master/FALSEHOODS.md [4] https://meta-package-manager.readthedocs.io/en/develop/falsehoods.html