[Python-Dev] Wave module support for floating point data

Terry Reedy tjreedy at udel.edu
Sun Mar 2 00:04:09 CET 2014


On 3/1/2014 3:25 PM, Antoine Pitrou wrote:
> On Sat, 01 Mar 2014 15:08:00 -0500
> Terry Reedy <tjreedy at udel.edu> wrote:
>> On 3/1/2014 2:57 PM, Sebastian Kraft wrote:
>>> Hi everybody,
>>>
>>> more than a year ago I have submitted a patch to enhance the Wave module
>>> with read/write support for floating point data.
>>>
>>> http://bugs.python.org/issue16525
>>>
>>> Up till now this patch has not been applied nor did I get feedback if
>>> anything needs to be enhanced or changed before it can be committed.
>>> I have never been in contact with your development process and Python
>>> core developers, so please tell me what I should improve...
>>
>> Please subscribe to core-mentorship list and post your question there.
>
> I don't understand this response. You seem to be assuming that
> Sebastian is asking for guidance,

I am reading what he posted, which ended with "please tell me what I 
should improve...".

This sort of question-request is routinely posted, in much the same 
words, on core-mentorship, and routinely gets a response there.

 > but he's simply telling us about a  patch that hasn't received
 > any review yet, despite having been posted one year ago.

If that were all he said (and it is not) it would not be too useful. 
There are about 4000 open issues. About half have the 'patch' keyword 
set, and probably more have patches without the keyword.  About 700 with 
the keywork have seen no activity for a year.  Some fraction of those 
have never received a review. I would guess at least 100. In this sense, 
there is, unfortunately in my opinion, nothing too special about this 
issue or patch.

If you are really interested in this subset of issues, someone should do 
a custom search against the database. Issues that are open, have a patch 
(a file name ending in .diff or .patch) posted more than a year ago, and 
have no subsequent responses from a core developer, would be a start.

Perhaps we should add a 'reviewed' field to the table of uploads, which 
would be automatically marked True when is a completed Rietveld 
response. Perhaps there should be a way to connect review messages, as 
opposed to mere comment messages, to the patch they review. We should 
think about this in the process re-design Nick has planned.

> It's not obvious he has been doing something wrong that
> he needs to be taught about.

The same could initially be said of all the similar posts on 
core-mentorship. But please change 'wrong' to 'incomplete or something. 
Most patches get revised and augmented without being 'wrong'. This is 
especialy true of patches from people 'have never been in contact with 
[our] development  process'.

What makes Sebastian's request worth extra attention is the extra 
information that he, the author of the neglected patch, is still around 
and desirous of discussing, editing, and augmenting the patch as necessary.

-- 
Terry Jan Reedy



More information about the Python-Dev mailing list