Question about Universes

> It doesn't depend upon that compiler hack. That hack provides the
> ability to put textual data in as an appendix to a method. It is an
> entirely independent facility that could be perhaps better included as a
> trait for those that want it.

The version I found in the latest Pharo-Dev images has the compiler
hack loaded and depends on it (or the package definitions).

> I don't think it is fair to use code critics as an argument.

You can't deny that references to variables and classes that do not
exist in the system are not a serious bug.

> To me it
> appears that you have developed all of these tools for your own use.

Code Critics is part of the refactoring browser by Don Roberts and John Brant.

The refactoring browser is one of the most amazing pieces of Smalltalk
code I have ever seen. I am maintaining the  code-base in Pharo (and
previously in Squeak), because it makes me much more productive in my
daily work. Code critics provides a huge set of high quality tests for
free that can quickly point out flaws in code. It is not always
correct, but it is usually very good to find the problematic parts of
a system.

>> not implemented anywhere. Of course there is not a single test.
> What are you talking about: Sake-Tests.

I am sorry. I was not aware of that package. I should have checked online.

Cheers,
Lukas

--
Lukas Renggli
http://www.lukas-renggli.ch

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Ramon Leon wrote:
> > I don't think it is fair to use code critics as an argument.
>
> I think it's perfectly fair, be it code critic or lint, if it finds
> bad code then it finds bad code.  It's not like you don't know about
> code critics and Lint is already in Squeak,
No it is not.

I have seen one passing comment about something called SwaLint, on the
mailing list that is about all.
> so it's your choice if you choose not to avail yourself of tools that
> point out badly written code.  He writes about the tools he's working
> on at his blog http://www.lukas-renggli.ch/blog which is easily
> subscribed to.
Thanks, I didnt know
> There's no reason anyone can say they're unaware of what he's doing
> unless they're choosing to be unaware.
Well clearly that is not true.

Keith


_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Lukas Renggli wrote:
>> It doesn't depend upon that compiler hack. That hack provides the
>> ability to put textual data in as an appendix to a method. It is an
>> entirely independent facility that could be perhaps better included as a
>> trait for those that want it.
>>    
>
> The version I found in the latest Pharo-Dev images has the compiler
> hack loaded and depends on it (or the package definitions).
>  
The package definitions do not depend on it, it allows the compiler to
ignore code after 6 quote marks.

myMethod

^ self

""""""
Some text here as documentation.

Bob will also use this facility, because it provides a place to put
scripts verbatim without having to escape quotes.
Again you dont have to use it.
>> I don't think it is fair to use code critics as an argument.
>>    
>
> You can't deny that references to variables and classes that do not
> exist in the system are not a serious bug.
>  
Yes I can, because plainly it hasnt effected use for while. If the tests
run then its not a serious bug. These sot of artifacts are unwanted side
effects in a rapidly evolving code base, they are not a big deal in the
grand scheme of things.
>> To me it
>> appears that you have developed all of these tools for your own use.
>>    
>
> Code Critics is part of the refactoring browser by Don Roberts and John Brant.
>
> The refactoring browser is one of the most amazing pieces of Smalltalk
> code I have ever seen. I am maintaining the  code-base in Pharo (and
>  
Well I never liked it, when I used it last I only had a 700Mhz machine
to work on, it was slow, buggy, menu items I used frequently were
missing. In short it was impossible to actually use.
> previously in Squeak), because it makes me much more productive in my
> daily work. Code critics provides a huge set of high quality tests for
> free that can quickly point out flaws in code. It is not always
> correct, but it is usually very good to find the problematic parts of
> a system
ok

Keith

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

But we do not want hacks. :) and hidden dependencies.
Frankly I do not understand why this is not possible to use a method  
(since a method is a script with a name).

>>>
>>
> Well I never liked it, when I used it last I only had a 700Mhz machine
> to work on, it was slow, buggy, menu items I used frequently were
> missing. In short it was impossible to actually use.

But it is working since years now.
I never understood why some Squeakers did not the value of it.
For Pharo I would like to have deeper integration of RB engine.

Stef

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

Stéphane Ducasse wrote:
> But we do not want hacks. :) and hidden dependencies.
> Frankly I do not understand why this is not possible to use a method  
> (since a method is a script with a name).
>  
Because methods have to compile, or they are not accepted.

In bob's case these are scripts that will run in a different image, so
do not necessarily compile in Bob's image.

In mantis's case these are the bug reports downloaded for searching and
reference purposes.

I dont regard this as a hack since Smalltalk is designed to have this
flexibility of supporting differnnt languages, and syntaxes. If you dont
like it then why not remove #compilerClass altogether.

If you call this a hack then what is WAFileDirectory? You would think
that using methods to store data in the image has never been done before.

Keith

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

> Where is the documentation?

  http://www.refactory.com/RefactoringBrowser/BrowserPagesLint.html

For an overview of RB:

  http://www.refactory.com/RefactoringBrowser/index.html

Lukas

--
Lukas Renggli
http://www.lukas-renggli.ch

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

On Feb 20, 2009, at 8:44 AM, Keith Hodges wrote:

but these images do not require to extend the compiler.

What I do not understand is how the browse also shows you this extra  
code
at the end of methods.
My gut feeling is that there is certainly a clean way to do that.

>
>
> Keith
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>


_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

 
> but these images do not require to extend the compiler.
>
>  
There is no extension to the compiler. All it does is give the compiler
the source up to the marker. The compiler is not changed at all. Matthew
just felt that it was cleaner to subclass the compiler, than to put the
methods in the class itself, as I used to do.

Keith

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project