[Sugar-devel] coping with the traffic in sugar-devel

[Gary C Martin]

On 5 Oct 2010, at 17:05, Walter Bender wrote:

> On Tue, Oct 5, 2010 at 12:01 PM, Gary Martin <garycmartin at googlemail.com> wrote:
>> On 5 Oct 2010, at 12:05, Bernie Innocenti wrote:
>> 
>>> On Tue, 2010-10-05 at 11:19 +0200, Tomeu Vizoso wrote:
>>>> What about marking in some way those incoming emails that belong to a
>>>> thread that has included the word 'bernie'?
>>> 
>>> That's really hard... Procmail, my email sorting software, is stateless.
>>> 
>>> 
>>>> There's also the option of just going through the subjects, read the
>>>> threads that look interesting then marking all of them as read when
>>>> finished.
>>> 
>>> That's better, but it adds delay: I open my mailing list folders to
>>> check for new mail only once in a while, when I'm done with everything
>>> else.
>>> 
>>> 
>>>> I'm starting a new thread about this because I often hear from people
>>>> who have missed important developments because cannot cope with the
>>>> traffic.
>>>> 
>>>> I'm also on way more lists that I can read, but with a bit of
>>>> automatic labelling and organization I think I manage to miss little
>>>> of the relevant info.
>>>> 
>>>> Any other tips?
>> 
>> Just mucking around with a quick hack, how about this? It's a SOM for the month of Sept of every email from iaep, dextrose, and sugar-devel that mentioned "bernie" (including emails from bernie). That's about ~1171 emails filtered down to ~171 that mentioned "bernie" and mapped. I dialled up the number of top features from the usual 200 terms to 400 for a little more textual depth:
>> 
>>        http://people.sugarlabs.org/garycmartin/2010-Sept-dev-dex-iaep-som-bernie.png
>> 
>> And just for comparison and so I'm not picking on Bernie, here's the same thing for "gary":
>> 
>>        http://people.sugarlabs.org/garycmartin/2010-Sept-dev-dex-iaep-som-gary.png
>> 
>> I could do this for any keyword (the more unique the better), so it would be easy to perhaps map "etoys" or some other term a summary map might be of interest, also easy to include more mail-lists into the data set.

Here's one for the Etoys keyword:

	http://people.sugarlabs.org/garycmartin/2010-Sept-dev-dex-iaep-som-etoys.png

>>> Cc'ing people conservatively is standard practice on many high-traffic
>>> lists. Today the old arguments of wasted bandwidth are just ridiculous.
>>> 
>>> Occasionally someone complains for receiving dupes, but all modern MUAs
>>> provide ways to suppress them. Sascha elegantly solved the issue by
>>> setting the "Mail-followup-to" header to exclude himself from replies
>>> (if the remote MUA supports it).
>>> 
>>> --
>>>   // Bernie Innocenti - http://codewiz.org/
>>> \X/  Sugar Labs       - http://sugarlabs.org/
>>> 
>>> _______________________________________________
>>> Sugar-devel mailing list
>>> Sugar-devel at lists.sugarlabs.org
>>> http://lists.sugarlabs.org/listinfo/sugar-devel
>> 
>> _______________________________________________
>> Sugar-devel mailing list
>> Sugar-devel at lists.sugarlabs.org
>> http://lists.sugarlabs.org/listinfo/sugar-devel
>> 
> 
> fun!! when will be be able to click on the words and go to the email
> message in the archive :)

I'm procrastinating on the implementation :) but now that the wiki supports image maps, that is likely the easy, if not most graceful/elegant, route**. Clicking on a word would link to the N emails where the term was used.

** really this should be a nice dynamic/interactive map web app, where the text labels are added dynamically over the contour map images, with panning/zooming/search/filter type features... And the ability for admins to add in new data resources to be mapped :)

Regards,
--Gary

> -walter
> 
> -- 
> Walter Bender
> Sugar Labs
> http://www.sugarlabs.org