Re: [Pharo-dev] About method comments

[lb]

Hi, Stef
I have an idea for this problem,
1. method should have it's comment like class comment
2. method comment has two versions, one for developer another for reader
3. developder's comment will be saved in source or changes file
4. reader's comment will be saved in its own file
5. system browser can load  reader's comment of method when you want to read or 
write 


reader can write his comment using his native language, we can share all 
reader's comments on internet.


Best regards


Bing Liang

Re: [Pharo-dev] About method comments

[Stephane Ducasse]

A method comment should tell
- what it is doing, what are the preconditions
- then some concerns about the implementation

You can give a try to document methods of fileSystem or SUnit as Katas.

Stef


On Sat, Feb 10, 2018 at 5:54 AM, Sean P. DeNigris  wrote:
> Esteban A. Maringolo wrote
>> Commenting an accesor is redundant, IMO.
>
> Typically, yes, and there are other examples.
>
>
> Esteban A. Maringolo wrote
>> I think that having a public guideline on how to comment
>> methods would be positive for the community
>
> Interesting idea. Any suggestions?
>
>
>
> -
> Cheers,
> Sean
> --
> Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html
>

Re: [Pharo-dev] About method comments

[Sean P. DeNigris]

Esteban A. Maringolo wrote
> Commenting an accesor is redundant, IMO.

Typically, yes, and there are other examples.


Esteban A. Maringolo wrote
> I think that having a public guideline on how to comment
> methods would be positive for the community

Interesting idea. Any suggestions?



-
Cheers,
Sean
--
Sent from: http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html

Re: [Pharo-dev] About method comments

[Stephane Ducasse]

You see I take some core classes and I read the code and I comment
when I understand what a method does.
And I do that 15 min from time to time. And just doing this will improve Pharo.
When I learn something I take 2 min to add a comment.

Stef


On Fri, Feb 9, 2018 at 9:13 PM, Jan Vrany  wrote:
> Hi Esteban,
>
> On Fri, 2018-02-09 at 15:17 -0300, Esteban A. Maringolo wrote:
>>
>> And rename refactorings usually don't rename the comments
>> (hopefully).
>>
> Why hopefully? Java IDEs rename references to the method also
> in comments, I found it very nice. Why it should not? I'm just curious.
>
> Best, Jan
>

Re: [Pharo-dev] About method comments

[Jan Vrany]

Hi Esteban,

On Fri, 2018-02-09 at 15:17 -0300, Esteban A. Maringolo wrote:
> 
> And rename refactorings usually don't rename the comments
> (hopefully).
> 
Why hopefully? Java IDEs rename references to the method also
in comments, I found it very nice. Why it should not? I'm just curious.

Best, Jan

Re: [Pharo-dev] About method comments

[Esteban A. Maringolo]

2018-02-09 8:10 GMT-03:00 Sven Van Caekenberghe :
> name
>   "Returns the receiver's name"
>   ^ name
>
> Would also pass a simple test, while it is quite useless.
>
> I would say that any public, non-trivial methods need a comment.

Commenting an accesor is redundant, IMO.

And rename refactorings usually don't rename the comments (hopefully).

However I think that having a public guideline on how to comment
methods would be positive for the community (if these rules are
consensual).
After that some tools can use these guidelines to generate comments on
their own, or if you don't know how to comment, simply copy the
templates from the guide.

E.g. Private methods in Dolphin always start with "Private - ...", if
you remove the "private" category from the method (methods can belong
to multiple categories), then it removes the "Private - " prefix from
the comment as well.

Regards,

--
Esteban.

Re: [Pharo-dev] About method comments

[Norbert Hartl]

One more reason to do reviews. Now that it is super easy with iceberg there is 
less reason to do so.

Norbert

> Am 09.02.2018 um 12:10 schrieb Sven Van Caekenberghe :
> 
> 
> 
>> On 9 Feb 2018, at 12:04, Stephane Ducasse  wrote:
>> 
>> Hi guys
>> 
>> From now on I will VETO the introduction of code where methods are not
>> commented.
>> I'm FEDUP FEDUP FEDUP to spend my time commenting methods that I do
>> not know and understand, juts because the original author thought that
>> this is obvious.
>> 
>> NOTHING IS OBVIOUS
>> 
>> So I will check if we can have a little process that systematically
>> reject code without enough method comments.
>> 
>> Pharo should grow and pharoers too.
>> 
>> Stef
>> 
>> PS: now you can comment that x said that method should be
>> self-explanatory. This is bullshit. I'm not asking for stupid
>> comments! I'm asking for respect for maintainers and other
>> programmers.
> 
> +100
> 
> The same goes for class comments too, duh.
> 
> One problem is that it is hard to check the quality/meaning of the comment. 
> 
> "Comment later" would probably pass any check, as would "I am class XYZ".
> 
> Also, 
> 
> name
>  "Returns the receiver's name"
>  ^ name
> 
> Would also pass a simple test, while it is quite useless. 
> 
> I would say that any public, non-trivial methods need a comment.
> 
> So human judgement will be needed.
> 
> Sven

Re: [Pharo-dev] About method comments

[Sven Van Caekenberghe]

> On 9 Feb 2018, at 12:04, Stephane Ducasse  wrote:
> 
> Hi guys
> 
> From now on I will VETO the introduction of code where methods are not
> commented.
> I'm FEDUP FEDUP FEDUP to spend my time commenting methods that I do
> not know and understand, juts because the original author thought that
> this is obvious.
> 
> NOTHING IS OBVIOUS
> 
> So I will check if we can have a little process that systematically
> reject code without enough method comments.
> 
> Pharo should grow and pharoers too.
> 
> Stef
> 
> PS: now you can comment that x said that method should be
> self-explanatory. This is bullshit. I'm not asking for stupid
> comments! I'm asking for respect for maintainers and other
> programmers.

+100

The same goes for class comments too, duh.

One problem is that it is hard to check the quality/meaning of the comment. 

"Comment later" would probably pass any check, as would "I am class XYZ".

Also, 

name
  "Returns the receiver's name"
  ^ name

Would also pass a simple test, while it is quite useless. 

I would say that any public, non-trivial methods need a comment.

So human judgement will be needed.

Sven