0) TL;DR: What's the proposal about: ------------------------------------ We had a BoF session at DebConf12; based on this and several other discussions among the current travel sponsorship team, Gaudenz and I propose to make some changes to the information gathered for travel sponsorship and to the evaluation process. We don't claim this is the perfect system, but we hope it is an improvement on the current system, within the limitations of wanting to have something we can get consensus for in a fairly short time. 1) Asking for one more number. ------------------------------ There seems to be a lack of common understanding about what "the amount unable to pay" field in penta means. We propose to ask for 3 numbers. a) total travel cost b) amount requested c) minimum useful amount (with less than this, I cannot attend) There currently seem to be roughly two main understandings of the "amount unable to pay" field corresponding to (b) and (c); by asking both we can fairly handle both views. We also have a reasonable way to do partial sponsorship. 3) Asking more questions ------------------------ We want to ask 3 questions, and rate the answers. Each answer need only be a few sentences, or longer if the applicant feels it is necessary/helpful. i) What are your current contributions to Debian? ii) How will your attending this DebConf benefit Debian? Examples include collaboration with other attendees, giving talks, and volunteering at DebConf. iii) Why do you request help paying for your travel costs? 4) Using references ------------------- For both 3(i) and 3(ii), the applicant could include the names of references, and probably have those references send email to some alias. The idea is to extend the set of people "known" to the sponsorship committee. 5) Evaluation Process --------------------- The rough idea is to use an n point scale for each question, and to (budget permitting) offer sponsorship first to every applicant with median rating above k in all 3 categories, and then to all applicants with median rating at least k in 2 of 3 categories, where k < n. As proposed, we do not explicitly consider amounts here, so we don't envision skipping over expensive, but otherwise deserving applications. This will possibly be compensated by being more explicit about need being a criterion. I guess in practice we have to have some sanity check on the amount applied for, and in general we should probably leave some discretion to the committee to deal with unusual circumstances. 6) About DebConf Newbies ------------------------ So far we think the same process could work for a separate DebConf Newbies budget, at least if the idea is to bring people already contributing to Debian, but not yet (or not recently) attending DebConf. It could still be worthwhile to have such separate budget in order to encourage applications from new people. 7) What this proposal does not cover ------------------------------------ There are quite a few issues "in the air" that this proposal does not cover. - There has been talk about how we can and should get the travel sponsorship done earlier, see e.g. http://lists.debconf.org/lurker/message/20120620.073520.90c3fab4.en.html - Related to that, we need to have some actual estimates for travel costs, based presumably on guesses about geographic distribution. - Also related, it has been suggested by a few people to generalize the travel sponsorship team/process to handle all related issues (e.g. travel for sprints). - The technical details of how to handle requests and evaluations are not decided yet (other than probably not penta. - There has also been discussion about the composition of the "herb" travel sponsorship team. - Feedback for the applicant about his ratings and transparency of the money spent. These are all interesting and important topics, but hopefully we can discuss them separately from what is proposed here.
Attachment:
pgpUrj6IG698d.pgp
Description: PGP signature