Re: Summary of Debian Science BOF at DebConf
On 10/08/2010 08:37 AM, Andreas Tille wrote:
> On Thu, Oct 07, 2010 at 11:21:22PM +0200, Christian Kastner wrote:
>
> Thanks. I added a link to the DebianScience/ProblemsToWorkOn page.
I had already added a link in the "Make better use of DebTags [...]"
section on that page - did you perhaps miss it? Just asking; collecting
subtasks in another section would be fine, too.
> Here are my comments:
>
> - Paragraph "4. Enhance Task Pages": TBD (volunteers)?
> Volunteers are very welcome to join my Python coding. Some suggestions in
> how far the enhancement should be done are helpful as well.
It would be great if volunteers would also those suggestions to the wiki
page (replacing the TBDs) so that they could be debated; I assume this
would be helpful in future coding.
Should I point this out more explicitly in the wiki? I intended the page
to be a collaborative effort (edit at will, especially proposals), I
just fleshed out the initial parts.
> - Paragraph "5. Keeping Task Pages Up-to-date"
> The tasks pages are updated once a day - so what do you exactly mean?
With 4. and 5., I tried to mirror the two proposals in the "Make better
use of DebTags [...]" section (paragraphs 2 and 3) in the
ProblemsToWorkOn page.
I added a link back to that paragraph in the subpage for clarification.
> - Representation of Debian Science Tasks:
> Paragraph "Proposal #1"
> I agree from a DebTaggers point of view. It seems like a clean and and
> logical approach. However, matching to our current tasks seems a bit
> tricky because sometimes a task is mapped to a field (mapping with devel
> facet would be doable of course vor the devel packages). But sometimes
> we need to look into use:: and works-with:: fields. I currently do not
> have a good idea how to map this consistently. (This is no argument
> against your tagging, I just have no proper solution for the tasks.)
I added a clarification to the proposal. In general, I don't see this as
much of a problem: while we can't, as you point out, map tasks 1:1 to a
tag, we can certainly provide manual mappings. It becomes a coding problem.
> - I like the data:: facet suggestion
... and whilst looking into the mapping issue above, I found by chance
the role::data tag (I wasn't expecting that one there). With this tag,
the "alternative to proposal #1" could therefore be implemented without
a single hange to DebTags. I'll think this over again.
Reply to: