[Bioc-devel] All Package Maintainers Please README!

[Kasper Daniel Hansen]

On Oct 28, 2007, at 10:02 PM, Diego Diez wrote:

> Hello All,
>
> just come to this thread and I would like to share my opinion. I
> usually don't log any changes to the svn log for several reason,
> mainly have to do with bad practices (I've been doing it into a
> separate file in inst/doc). I was planing to change to the svn way
> for the next devel/release cycle though. I would like to have some
> flexibility for the log so I prefer adding a specific character at
> the beginning of each line in order to identify them. I don't like
> too much putting something as long as INTERNAL but if a suitable
> character like # can be safely used I'll definitely adhere to this
> policy. As a side comment it may be interesting to add the character
> to the lines that are usually going to be less common. What do you
> think use to take more space/lines in your commit logs? User focused
> comments or detailed developer focused changes? If developer focused
> changes use to be longer in description it might make more sense to
> add the character just to the blogged lines. That will have the
> additional effect of no blogging by default that may be a good thing
> (as Marc pointed out).

I think this post raises the interesting question of who this blog- 
style news listing is intended for. If it is developers (which I  
assume), I would vote for keeping it super easy and not worry too  
much about formatting etc. If the intention is more user-oriented I  
think it would be better to try to start a changeLog style thing  
which should be written separately from the svn log. Such an approach  
(which I would like anyway) is already supported by some packages and  
we could make a standard function like changeLog() that displays it.  
That way a user can keep track of changes, something which I  
sometimes have found very useful indeed, especially if algorithms are  
changed. The svn log should - in my opinion - be a way of  
communicating with your co-developers and should not necessarily be  
understandable for anyone who does not have a decent insight into the  
package.

Kasper



> Best,
>
> Diego.
>
>
>
> On Oct 27, 2007, at 5:10 AM, James W. MacDonald wrote:
>
>> Marc Carlson wrote:
>>> Not a problem Jim,
>>>
>>> I was really just trying to get out of hitting shift-3 a few  
>>> thousand
>>> more times.  We obviously have to go with what most people want.
>>> I just
>>> wanted to put my two cents in.
>>
>> Maybe I haven't been very clear. Consider the following as a commit
>> message:
>>
>> This part of the commit message will be posted. Regardless of the
>> number
>> of sentences or apparent number of lines.
>>
>> #This part of the commit message will not be posted. Even though  
>> there
>> appears to be several lines here, the Python code only sees one.
>> Therefore a single # at the beginning of the paragraph will cause  
>> this
>> whole paragraph to be scrubbed, regardless of the length. Thus, until
>> you have made a few thousand commits you will really not be hitting
>> shift-3 that often...
>>
>> But if you add a newline, the next bit will be posted.
>>
>> #Unless you comment that part out as well.
>>
>> Jim
>>
>>
>>>
>>>     Marc
>>
>> -- 
>> James W. MacDonald, M.S.
>> Biostatistician
>> Affymetrix and cDNA Microarray Core
>> University of Michigan Cancer Center
>> 1500 E. Medical Center Drive
>> 7410 CCGC
>> Ann Arbor MI 48109
>> 734-647-5623
>>
>> _______________________________________________
>> Bioc-devel at stat.math.ethz.ch mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>
> --
> Dr. Diego Diez
> Bioinformatics center,
> Institute for Chemical Research,
> Kyoto University.
> Gokasho, Uji, Kyoto 611-0011 JAPAN
> diez at kuicr.kyoto-u.ac.jp
>
> _______________________________________________
> Bioc-devel at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/bioc-devel