[Pharo-project] feedback on CLFramework

Stéphane Ducasse stephane.ducasse at inria.fr
Sun Sep 27 04:15:58 EDT 2009


>>
> Sorry, but i don´t like that format of code. For me is not readable.  
> I don´t
> know when starts a block or if is a loop or condition block.That  
> problem is
> more big when exists various identations. I´m coming of VB and C#  
> and that
> way seems anarchist for me :)  my cerebellum is limited for that way  
> of
> write code.

smalltalk allows you to write nice and compact code so take it as a  
opportunity.
In python you would be forced to use tab at the right place.


	[:each | 		....			]


	[:each|
			...
				
			]


> Why my format not is "Smalltalkish" way?

read all the code in the image and you will see nowhere

	do:
	[
	:ecah |
	



	]

or

	(expression) ifTrue:
	
	[



	]
		ifFalse:

	[



	]


	


> Perhaps I´m wrong (i have short
> time with that language) but I believe indeed than Smalltalk don´t  
> have a
> style defined (Thanks to God)

Read Kent beck book (best Smalltalk pattern)
have a look at Smalltalk with Style on my web page.

I can tell you I cannot read your code well.
So if you want to really have fun with Smalltalk, change.

>>
> Yes. I´m not add comments basically for my poor english.
> Today i could try it. Is a relax work for sunday :)

just write
subject verb complement
and code.

I'm responsible for XXXX
I collaborate with YY and ZZ

Here are three important methods:
	-
	-
	-

Here is a way to invoke me.
	
>> Keep working.
>> I still believe that a windowSpec based on a literal array is better
>> than a byte array.
>>
>>
>
> Of course, but I don´t know do it of another way. Literal array is  
> very
> complicated for me.
> Xml serialization and Moose serialization don´t be possible. The code
> generated is HUGE, and fails
> with complex composite morphs.

if you store them as literal arrays probably but inside a string I do  
not see why.

> The one way for me is a mechanism where each
> morph write his state in a stream. All states togethers will build  
> the final
> morph. I believe that mechanism works, is elegant and readable. And  
> is easy
> detect problems too, but i need a little more of time for do it. I  
> want
> implement various widgets ( a grid is MUCH needed ) and last, before  
> of 1.0
> version, the serialization mechanism.
>
> Thanks for the feedback :)
>
> PS. I added recently the CLSplitterPanel widget ->
> http://www.youtube.com/watch?v=qbd9bbfxNB4

So far I could not build an interface with two buttons.
>
>
> Regards
>
>
>
>
> -- 
> View this message in context: http://n2.nabble.com/Re-feedback-on-CLFramework-tp3718474p3718991.html
> Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
>
> _______________________________________________
> Pharo-project mailing list
> Pharo-project at lists.gforge.inria.fr
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project





More information about the Pharo-dev mailing list