[Ann] Blog entry about using Pharo / Smalltalk to build a bot

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

[Ann] Blog entry about using Pharo / Smalltalk to build a bot

sergio_101
Hey, all..

I just released a blog entry entitled:

"Why I chose Pharo Smalltalk to build my Bot project"

which can be viewed here:


I would appreciate (and probably create new posts about) any input..

Thanks!



signature.asc (852 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

OmarAlmeida
Great article!
Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

EstebanLM
In reply to this post by sergio_101
nice! thanks :)

On 15 Aug 2017, at 20:23, sergio ruiz <[hidden email]> wrote:

Hey, all..

I just released a blog entry entitled:

"Why I chose Pharo Smalltalk to build my Bot project"

which can be viewed here:


I would appreciate (and probably create new posts about) any input..

Thanks!



Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

kilon.alios
In reply to this post by sergio_101
there was one claim you made in that post that got me thinking

""No one else knows Smalltalk." 

So I wonder how many developers amount to "no one".

Now my data is questionable to say the least but I dont think I am very far from the truth.

According to a quick estimate I found online there are 18.500.000 developers , around half of that are pros 
Now Tiobe Index has some numbers about language popularity, widely disputed but I will use that. Unfortunately Smalltalk does not even make top 50 but I think its close to that because it used to be in the top 50 of Tiobe Index once .... so .... 50th is around 0,4% which means that smalltalk is probably at least 0.1% Now if we multiply with the total amount of developers 

No |  Name of Language | Percentage of Popularity | Amount of total develpers
---------------------------------------------------------------------------------------------------------
1) Java  13% =  2.405.000
2) C 6,5% =  1.202.500
3) C++ 5,5% = 1.017.500
4) C# 4,2% = 777.000
5) Python 3,7% = 684.500
6) VB .NET 2,6%
7) PHP 2,3%
8) Javascript 2,1%
9) Pearl 2%
10) Ruby 2%
11) Swift 1,8%
12) Delphi 1,8%
13) Visual Basic 1,8%
14) Assembly 1,8%
15) R 1,8%
16) Go 1,7%
17) Matlab 1,6%
18) Objective C 1,5%
19) Scratch 1,5 %
20) Dart 1,3%
26) Cobol 1% = 185.000
28) Scala 0,9%
31) Erlang 0,7%
34) Lisp 0,7%
35) Fortan 0,7%
36) Lua 0,7% 
47) Haskell 0,5%
120) Smalltalk 0,1% = 18.500

I made up the position of smalltalk and its percentage but if you see how percentages are from language to language you will have to agree that I cannot be very far.
So the bad news is the sources are questionable to say the least the good news is however that we are close to the truth because the Python devs have also reported 1 million users. Now if this is so close to the truth that means two massively important things. I also know for a fact that on its popular days Delphi used to also have 1 million devs so its percentage is close to reality as well. The general picture is that the math at least seems to hold up. 

So what is the conclusion out of this ?

1) The vast majority of the popular languages , apart from Java, are not anywhere near as popular as they imply. Especially Javascript with only 300k developer , by the noise they generate you would believe they are like 300 million. Once you leave the top 5 , pretty much every language is very close.
2) 18.500 is nowhere near close to "no one" its actually closer to a small town or a very large village. So definitely a significant amount of Smalltalk devs. 
3) Coders use pretty much every language out there and they do not concentrate on few popular ones

On the other hand there are all sort of numbers online , one source gives Java over 6 million devs. But I do not think that the exact matter matters rather the general picture.

I have found the whole thing mind blowing to say the least and it completely reduces the meaning of "language popularity" showing that judging the popularity of the language by the noise it generates on the net is definitely a terrible metric. 

I am willing to bet that Pharo has around 3-6.000 users which is definitely not bad. 

Thank you for opening my eyes, I will from now on fight the illusion of language popularity.

Oh yes I loved your post, very well written, excellent work

On Tue, Aug 15, 2017 at 9:24 PM sergio ruiz <[hidden email]> wrote:
Hey, all..

I just released a blog entry entitled:

"Why I chose Pharo Smalltalk to build my Bot project"

which can be viewed here:


I would appreciate (and probably create new posts about) any input..

Thanks!


Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

John Pfersich
+1

Sent from my iPhone

On Aug 17, 2017, at 06:11, Dimitris Chloupis <[hidden email]> wrote:

there was one claim you made in that post that got me thinking

""No one else knows Smalltalk." 

So I wonder how many developers amount to "no one".

Now my data is questionable to say the least but I dont think I am very far from the truth.

According to a quick estimate I found online there are 18.500.000 developers , around half of that are pros 
Now Tiobe Index has some numbers about language popularity, widely disputed but I will use that. Unfortunately Smalltalk does not even make top 50 but I think its close to that because it used to be in the top 50 of Tiobe Index once .... so .... 50th is around 0,4% which means that smalltalk is probably at least 0.1% Now if we multiply with the total amount of developers 

No |  Name of Language | Percentage of Popularity | Amount of total develpers
---------------------------------------------------------------------------------------------------------
1) Java  13% =  2.405.000
2) C 6,5% =  1.202.500
3) C++ 5,5% = 1.017.500
4) C# 4,2% = 777.000
5) Python 3,7% = 684.500
6) VB .NET 2,6%
7) PHP 2,3%
8) Javascript 2,1%
9) Pearl 2%
10) Ruby 2%
11) Swift 1,8%
12) Delphi 1,8%
13) Visual Basic 1,8%
14) Assembly 1,8%
15) R 1,8%
16) Go 1,7%
17) Matlab 1,6%
18) Objective C 1,5%
19) Scratch 1,5 %
20) Dart 1,3%
26) Cobol 1% = 185.000
28) Scala 0,9%
31) Erlang 0,7%
34) Lisp 0,7%
35) Fortan 0,7%
36) Lua 0,7% 
47) Haskell 0,5%
120) Smalltalk 0,1% = 18.500

I made up the position of smalltalk and its percentage but if you see how percentages are from language to language you will have to agree that I cannot be very far.
So the bad news is the sources are questionable to say the least the good news is however that we are close to the truth because the Python devs have also reported 1 million users. Now if this is so close to the truth that means two massively important things. I also know for a fact that on its popular days Delphi used to also have 1 million devs so its percentage is close to reality as well. The general picture is that the math at least seems to hold up. 

So what is the conclusion out of this ?

1) The vast majority of the popular languages , apart from Java, are not anywhere near as popular as they imply. Especially Javascript with only 300k developer , by the noise they generate you would believe they are like 300 million. Once you leave the top 5 , pretty much every language is very close.
2) 18.500 is nowhere near close to "no one" its actually closer to a small town or a very large village. So definitely a significant amount of Smalltalk devs. 
3) Coders use pretty much every language out there and they do not concentrate on few popular ones

On the other hand there are all sort of numbers online , one source gives Java over 6 million devs. But I do not think that the exact matter matters rather the general picture.

I have found the whole thing mind blowing to say the least and it completely reduces the meaning of "language popularity" showing that judging the popularity of the language by the noise it generates on the net is definitely a terrible metric. 

I am willing to bet that Pharo has around 3-6.000 users which is definitely not bad. 

Thank you for opening my eyes, I will from now on fight the illusion of language popularity.

Oh yes I loved your post, very well written, excellent work

On Tue, Aug 15, 2017 at 9:24 PM sergio ruiz <[hidden email]> wrote:
Hey, all..

I just released a blog entry entitled:

"Why I chose Pharo Smalltalk to build my Bot project"

which can be viewed here:


I would appreciate (and probably create new posts about) any input..

Thanks!


Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

sergio_101
In reply to this post by kilon.alios
Oh! I like your approach..

Let me quickly clarify my situation. In that section of the article, I was referring to my workplace. In my case, no one else in the dev department knows smalltalk. In a great many workplaces, devs don’t really want to learn a new language / framework without going off to a developer summer camp or something like that.

I have found that once a developer gets to a certain level of competence, they can quickly (in a weekend?) get up to speed on the use of a new language / framework to get their project done.

For a long time, I felt the sting of the “most popular language lists” as companies wanted to start all projects in PHP so that if the senior devs moved on, they could just hire some kids “cheap, just out of school” to work on the project.

i do think there are lots of things about to change Pharo’s place in the world, especially the Pharo MOOC..

there was one claim you made in that post that got me thinking

""No one else knows Smalltalk." 

So I wonder how many developers amount to "no one".





signature.asc (852 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

kilon.alios
Yeah all that sounds familiar to me

Its great you made this post because as you see its all about the illusion of popularity than actual popularity. Exposing Pharo like this is a great way to gain "popularity". 

My numbers by the way can be widely disputed however there is a common pattern here that pretty much every other "popular language top 20" agrees on , that "popular" languages , according to Tiobe Index top 20, dont even make 50% of the actual languages people use. I have seen several language ranks, like PYPI (which of course is highly biased in favour of Python) , poplang etc.

They all agree on this. The disagree on how much popular each language is , or how many developers are out there, but what they do not disagree on is that popular languages take only half of the pie. 

50% is far lower than expected, I expected top 10 to concentrate at least 70%.  The inability of the top 20 to do that just blows my mind.

The reason why there are not many people using Smalltalk is not because PHP is popular. Quite the opposite , its not that popular in reality . The reason is that even "unpopular" languages take only a pathetic 0-1% user base there are still thousands upon thousands of them out there which is what helps them to make that other 50%. 

So the question that arises, and this is what surprised me , is not "why people use popular languages" ?

But rather "why people use unpopular languages". Because if 1 in 2 coders use languages that are less than 1% thats is kinda of big deal.

Which also collapses the stereotype of hype addicted coder. Instead we see a coder that likes to try a variety of things. 

Which in turn that Pharo is as "threatened" by popular language as much as it is by unpopular languages. 

Unpopular languages are not used so much on big project as main languages but they are used for much more minor tasks. 

We should not forget how Python started. It did not became popular because it was used as a programming language. Actually its creator has been quoted in the early history of Python stating that he designed the language only to be a scripting language and to be used for very small task of only a few lines of code. NOT , I repeat , NOT as a programming language. This is his actual word.

Which is how he explained the minimalism of the syntax, he also targeted Python to education to help people started coding.

World domination was never in his plans at all. Even later he still insisted that Python should never been used as a replacement for C. Yet scientists have done this very thing, as Python is extremely popular with scientists. 

This shows the power of small beginnings and I think this is also can teach a very important lesson for Pharo that filling the small gaps , targeting the minor tasks can be a massive boost to the popularity of the language.

Because more and more people used python for just a replacement to bash or just for doing something simple, then started slowly and steadily using it for more and more complex tasks. 

This sneaky tactic can be a great way to promote pharo to your coworkers. No need to convert your entire project to PHP. Just start using Pharo for minor tasks, maybe clean up the code, analyse and visualise the code base with Roassal , parse some text files etc.

This way your coworkers will stat seeing you using this obscure language with no need for you to start ranting how amazing Pharo is and no need for them to learn Pharo. Instead of seeing something that is too good to be true they will see something that "just works". 

Of course you did something much better and built an entire project on Pharo which is why I think your blog posts is such a great advertisement for Pharo. Well done :)

On Fri, Aug 18, 2017 at 3:55 PM sergio ruiz <[hidden email]> wrote:
Oh! I like your approach..

Let me quickly clarify my situation. In that section of the article, I was referring to my workplace. In my case, no one else in the dev department knows smalltalk. In a great many workplaces, devs don’t really want to learn a new language / framework without going off to a developer summer camp or something like that.

I have found that once a developer gets to a certain level of competence, they can quickly (in a weekend?) get up to speed on the use of a new language / framework to get their project done.

For a long time, I felt the sting of the “most popular language lists” as companies wanted to start all projects in PHP so that if the senior devs moved on, they could just hire some kids “cheap, just out of school” to work on the project.

i do think there are lots of things about to change Pharo’s place in the world, especially the Pharo MOOC..

there was one claim you made in that post that got me thinking

""No one else knows Smalltalk." 

So I wonder how many developers amount to "no one".




Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

Stephane Ducasse-3
In reply to this post by sergio_101
Hi sergio

Tx for your nice article.
I hope you got success with Pharo.
Could we write a small success story?
Stef

On Fri, Aug 18, 2017 at 2:54 PM, sergio ruiz <[hidden email]> wrote:

> Oh! I like your approach..
>
> Let me quickly clarify my situation. In that section of the article, I was
> referring to my workplace. In my case, no one else in the dev department
> knows smalltalk. In a great many workplaces, devs don’t really want to learn
> a new language / framework without going off to a developer summer camp or
> something like that.
>
> I have found that once a developer gets to a certain level of competence,
> they can quickly (in a weekend?) get up to speed on the use of a new
> language / framework to get their project done.
>
> For a long time, I felt the sting of the “most popular language lists” as
> companies wanted to start all projects in PHP so that if the senior devs
> moved on, they could just hire some kids “cheap, just out of school” to work
> on the project.
>
> i do think there are lots of things about to change Pharo’s place in the
> world, especially the Pharo MOOC..
>
> there was one claim you made in that post that got me thinking
>
> ""No one else knows Smalltalk."
>
> So I wonder how many developers amount to "no one".
>
>
>
>
> ----
> peace,
> sergio
> photographer, journalist, visionary
>
> Public Key: http://bit.ly/29z9fG0
> #BitMessage BM-NBaswViL21xqgg9STRJjaJaUoyiNe2dV
> http://www.Village-Buzz.com
> http://www.ThoseOptimizeGuys.com
> http://www.coffee-black.com
> http://www.painlessfrugality.com
> http://www.twitter.com/sergio_101
> http://www.facebook.com/sergio101

Reply | Threaded
Open this post in threaded view
|

Re: [Ann] Blog entry about using Pharo / Smalltalk to build a bot

Sean P. DeNigris
Administrator
In reply to this post by kilon.alios
kilon.alios wrote
This sneaky tactic can be a great way to promote pharo to your coworkers.
No need to convert your entire project to PHP. Just start using Pharo for
minor tasks, maybe clean up the code, analyse and visualise the code base
with Roassal , parse some text files etc.
I love this idea! Take a task no one is particularly invested in and show off Pharo's advantages by adding real value to the team. Very sneaky... I mean effective... indeed!
Cheers,
Sean