gmane.comp.lang.smalltalk.fonc

movies list (was: Figuring out what you all want to hear)

Jecel Assumpcao Jr — 2010-03-16T22:36:35

Thiago Silva wrote on Wed, 10 Mar 2010 05:06:53 -0300

Thanks for the tip! I have added these, and two more from Youtube, to

http://www.smalltalk.org.br/movies/

This really should be a wiki page so other people could add the stuff
they find, though keeping the spam out is always a problem.

Re: Code Bubbles -- Active Essay

Ted Kaehler — 2010-03-15T21:30:29

Turn to the proper page in the essay and see all the explanation there.
They are in "logical" order.
All of the methods in LWordWrapLayout come from rules.  If you
browse that class, you see all of the rules as Squeak code.
The active essay needs a table of contents that is a list of the rules.
Or, we could make a specialized rule browser to see them in When...Do...
format.


They are Squeak methods.  If we go far with this, we may want end-user-
friendly debugging tools.


Yes.


  --Ted.

Re: Code Bubbles

Julian Leviston — 2010-03-14T04:23:30

On 14/03/2010, at 1:10 PM, James McCartney wrote:


No, it's very good - definitely a step in the right direction. Introspection or breaking apart seems missing, though I'm sure we didn't see the entire system from this short demo - introspection is very useful as a tool for discovering more about how to code in a language.

I quite like Alice as a programming environment just in terms of the UI. It encourages discovery and experimentation.

Part of the trouble seems to be, though, that like smalltalk and self, programs seem to need to ensure they have the entire "world" at their disposal...

Julian.

Re: Code Bubbles

James McCartney — 2010-03-14T02:10:08

Also, I don't know if I need to say this, but my perception is that
people around here may hate this because of the strong functional
programming stance, but I think the key neat idea here is the idea of
fusing UI. So if possible, try to get past any other prejudices he or
you might have.

On Sat, Mar 13, 2010 at 5:49 PM, James McCartney <asynth-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

Re: Code Bubbles

James McCartney — 2010-03-14T01:49:29

http://www.youtube.com/watch?v=faJ8N0giqzw

Re: Code Bubbles

James McCartney — 2010-03-14T01:48:50

Conal Elliot's Tangible Functional Programming work seems like an
interesting way to shrink the divide between using and programming.


On Fri, Mar 12, 2010 at 7:59 PM, Julian Leviston <julian-62HJldmB3ajk1uMJSBkQmQ< at >public.gmane.org> wrote:

Re: Code Bubbles

Julian Leviston — 2010-03-13T03:59:12

Re: Flight Data Recorders, etc. (was: my two cents)

John Zabroski — 2010-03-12T23:53:17

Re: Code Bubbles

karl ramberg — 2010-03-12T18:56:03

This is very interesting.
How do you organize and manage these rule boxes when you have 37 or
more of them ? Do they collapse into a browser or some other other
structure?

And how do you debug the rules?
A value pane on the side of each of them with all the temp and
instance vars, and a step button would be nice.

Karl

Re: Code Bubbles

Andrey Fedorov — 2010-03-12T17:41:07

Re: Code Bubbles

Alan Kay — 2010-03-12T16:32:59

Re: Code Bubbles

Julian Leviston — 2010-03-12T16:27:05

Re: Code Bubbles

John Zabroski — 2010-03-12T16:26:12

Re: Code Bubbles

John Zabroski — 2010-03-12T16:17:54

Re: Code Bubbles

Julian Leviston — 2010-03-12T16:05:12

On 13/03/2010, at 2:45 AM, Dethe Elza wrote:


I disagree with this.

The difference seems fairly clear and obvious to me. A programming language must be turing complete... spreadsheets are not turing complete because they don't do program execution flow... looping and the like...

http://en.wikipedia.org/wiki/Turing_completeness

Julian.

Re: Code Bubbles

John Zabroski — 2010-03-12T15:53:19

Re: Code Bubbles

Dethe Elza — 2010-03-12T15:45:04

Hi folks,

I've been enjoying this discussion and happy to see this list come to life. Coming out of lurk mode to toss my $0.02 in.

On 2010-03-11, at 11:05 PM, Julian Leviston wrote:

Perhaps GMail is a bad example, but James Noble and Robert Biddle made a convincing (to me, at least) argument that Google's search bar is a programming language, in their "Notes on Postmodern Programming."

http://www.mcs.vuw.ac.nz/comp/Publications/archive/CS-TR-02/CS-TR-02-9.pdf

What I remember of their presentation at OOPSLA was their making a stronger case than in the paper above, so I will try to summarize it.

1. A programming language is defined by the types of problems we can solve with it
2. One common problem to start off CS students with is "write a program that returns the first thousand prime numbers"
3. Pulls up web browser, types "first 1000 primes" in search bar
4. Clicks on answer

Other have made the argument that Google is essentially the modern command-line interface, but I think this goes way beyond Google. Things like mash-ups are made possible by the View Source nature of the web, every web page becomes an API. The ubiquity of Javascript is what makes it powerful, not so much the actual syntax of Javascript (although I will go on record that I *like* Javascript).

And back to Andrey's example that Julian was responding to. We can write bookmarklets or GreaseMonkey scripts, or browser plugins that change every aspect of GMail. We can use GMail to make general Google Queries. We can embed spreadsheets and other executable code into GMail messages. How much goal direction and configuration is needed to become "a program?"

--Dethe

Composing Examples (was Re: Code Bubbles)

Markus Gälli — 2010-03-12T11:52:02

Am 12.03.2010 um 02:12 schrieb Julian Leviston:


I'd love to see an extension of the "Given When Then" approach of Cucumber towards "Given When Then Return" approach.
Then suddenly rules could be composed by plugging the "Return" values into the "Given".

More on example composition:

"Composing Tests from Examples" (Journal of Object Technology)
http://www.jot.fm/issues/issue_2007_10/paper4.pdf

JExample
http://scg.unibe.ch/research/jexample

PHPUnit
http://sebastian-bergmann.de/archives/848-Fixture-Reuse-in-PHPUnit-3.4.html

Cheers

Markus

Re: Code Bubbles

Ted Kaehler — 2010-03-12T07:45:17

Folks,
       The tasks are to explain the goal of a program clearly enough 
to be easily grasped by a person, and precise enough that it is a 
runnable specification.
       Here is an example of a specification that reads like an essay, 
but is also a computer program that runs.  It is the complete 
specification of text layout with word-wrap, and it includes the 
commands of an editor on that text.

http://www.vpri.org/pdf/m2010002_lobjects.pdf

       The PDF is best viewed with the zoom set to "actual size".
       In this example, we lean toward the "Literate Programming" 
approach.  This "Text Field Specification" active essay is one of 
many STEPS experiments at VPRI.
       Please forgive the shortcomings of this active essay.  As you 
well know, creating a good explanation and running code at the same 
time is not easy.

--Ted.

Re: Code Bubbles

Julian Leviston — 2010-03-12T07:05:50

Re: Code Bubbles

Andrey Fedorov — 2010-03-12T06:41:09