Re: [datatracker-rqmts] Slides for the face-to-face BoF next week

[Fred Baker <[e-mail]>]

On Nov 1, 2010, at 3:28 PM, Paul Hoffman wrote:

> Greetings again. I have uploaded the proposed slides for the BoF to the IETF site; see <http://www.ietf.org/proceedings/79/slides/iddtspec-0.pdf>.
> 
> Of course, it would be swell if people wanted to talk about any of the topics in the draft or proposed in the slides before the meeting. :-)


Thanks for pointing out the slides.

One of the slides wonders what the rule is for identifying drafts associated with a working group. As I'm sure you know, there has been a convention for a long time in naming: draft-ietf-<wg>-*-nn.txt is a working group draft, and draft-<author>-<wg>-*-nn.txt is an individual submission to a working group. draft monikers that mention no working group are general submissions going nowhere in particular. I'm pretty sure that Henrik's existing tool looks for "draft-ietf-<wg>" and "anything containing <wg>" for the two areas.

I like the convention; I find it useful. I do go a little crazy with the set of people that don't choose to use it, though; I get people posting drafts and then wanting to discuss them for several meetings in my WG but not follow the convention, which means that I have to manually track things. Since a "convention" is not a "rule" (eg, I don't care to be a hard-ass about it), I put in the extra work, but there is a part of me that would like the tool to be able to remember for me in some sense. "documents that follow the convention plus those I add".


Something else the slides note - there could be some pain in various forms regarding personal lists. There might be; I'm not sure why anyone would want to know what drafts I was tracking (it's not like we have drafts about specialized delivery of pornography or etc), but I could imagine someone trying to do corporate intelligence by tracking who is tracking what drafts. A simple solution would be to keep our private lists on our private computers. That might, for example, be a reasonable use of a cookie or some kind of local preferences file. If we're keeping the list on the machine, I would think that per-user lists would not be used in daily processing; they would be compiled into a set of rules of the form "if a draft whose name matches /<regex>/ changes, notify {list}". Individual draft names are an obvious special case, as are /draft-*-<wg>-*.txt/.