Re: Imploding pkg.m

[c.]

Il giorno 13/lug/2012, alle ore 00.36, Juan Pablo Carbajal ha scritto:

> On Thu, Jul 12, 2012 at 11:40 PM, c. <address@hidden> wrote:
>> Jordi,
>> 
>> Il giorno 12/lug/2012, alle ore 18.57, address@hidden ha scritto:
>>> 
>>> Can we please move the private functions of pkg.m back into pkg.m as
>>> subfunctions? I don't find one-function-per-file to be easier to work
>>> with than multiple functions per file.
>> 
>> The idea was to make it more easy to find subfunctions that can be removed,
>> once all simplifications have been made I see no problem with going back to
>> subfunctions.
>> 
>> Unfortunately I did not have enough time to complete my plan to remove 
>> redundant subfunctions
>> and it seems no one else took up the job ...
>> 
>> But I still at list a few that make very little sense, e.g. 
>> "absolute_pathname".
>> Maybe this should a task for one of the codesprints at OctConf? Can you wait 
>> until then before you put everything together again?
>> 
>> 
>>> You don't do
>>> one-function-per-file in any other language, do you?
>> 
>> actually I sort of do ;)
>> 
>> 
>>> - Jordi G. H.
>> c.
>> 
> 
> We should have a talk about pkg.m in the meeting that is for sure.
> Also is probably one of the only functions where I can do significatn
> contributions, I might be able to make a code sprint on it, maybe
> carandraug will join.
> 
> I also prefer to work with multiple files, but I guess we shouold put
> pkg.m back together anyway. Once we could see what parts can be
> removed/improved.


Here's an example changeset for removing absolute_pathname.
It's not tested yet, if you like it I'll clean it up before pushing. 

c.

open_5VxTOKCf.txt
Description: Text document