Anthony Towns dijo [Wed, Jun 14, 2006 at 10:21:41AM +1000]: > > > 1) travel sponsorship seems very obscure; [...] > > Branden did not get any money because he specified that he could > > do with anything between 0 and 500USD. The goal is usually to not > > pay people who can actually afford the trip on their own. > > Right, the problem is that in this case (aiui) it resulted in him not > being able to attend. That might mean you're not communicating with > potential attendees very well: maybe it needs to be made clearer that > the question isn't "how much would you like?" (lots!!) but "what's the > minimum amount of support you need in order to get to debconf" Please keep in mind that Branden, as the then-DPL, was part of the sponsorship committee, and he had as much a say as anybody else. He got exactly the amount of money he requested - He requested for no money, I suppose, because at some point we announced we were pretty tight on budget, and asked everybody to check their request for the minimum they could pay in order to attend. > Maybe an alternative would be to have fields like: > > What's your expected flight cost [________] > The orga team estimate accommodation costs to be $y / day > The orga team estimate food costs to be $x / day > > [x] I won't be able to attend debconf unless I get sponsorship for: > _____ % of my travel costs > [ ] accommodation > [ ] food Well, it was not worded that way, but we _did_ request for that information. On one hand, you could select your food preferences between: No food requested Regular Vegetarian Vegan (strict vegetarian) Other (contact committee) You could select your desired accomodation from: Sponsored accomodation in a 4-6 people room Not sponsored accomodation in the executive hotel And for travel sponsorship, if you selected "I want travel sponsorship", you were requested for the total travel cost and for the amount you could pay by yourself, meaning youexpected only the remaining to be sponsored. Yes, we took some time to provide the estimates on lodging/food prices, but in the end they were all there - and we gave by mail rough estimates to whoever requested them. > > > 3) the distance between the hacklab and the lecture rooms we've had > > > the past couple of years has been really bad to the point where it's > > > discouraged people from going to talks > > can you be more specific at which location it was very bad, very > > good or just great? > > Oslo was fine, Helsinki and Oaxtepec were difficult. Well... I have to take a good portion of the blame here - In Helsinki, I barely went to the talks as it was too far away. Keeping that in mind, I still decided (yes, largely on my own, as there was not many people to check this with) to go for the tower, as 400m didn't look like _that_ much to me - In the end, it turned out to be a point many people complained about. And in retrospective, we could have even not used the tower except for DebianDay and hold the talks in the second hacklab, which was alwaays very sparsely populated. Anyway, point taken. Next time we go to Oaxtepec... Oh, sorry :) > > > 4) it's not entirely clear what the point of the conference is, there > > > are more than a few possibilities: > > > -- educate people with talks > > > -- provide an opportunity for DDs to get together and hack on stuff > > > -- give people a chance to meet face to face, socialise, > > > reduce aggression and brainstorm new ideas > > > -- provide an excuse to have some fun rather than just hack and flame > > > each other > > > -- prove Debian is just as cool as Ubuntu because we can > > > have expensive conferences too > > The point of the conference is to inspire people about debian. > > Which people? Do you mean inspiring people to use Debian? Inspiring > people who haven't contributed before to start? Inspiring people who > already contribute to do more? Inspiring people who've stopped to start > again? Inspiring people to look at different aspects of Debian and work > on them instead? > > How about the people who are already inspired to work on Debian and > are doing as much as they can -- eg the release managers or the orga > team itself? > > Do the other orga team people share that view of the point of the > conference? It is very hard to define what is Debconf's specific goal - Maybe because all of that points are important. Maybe thy will be ranked in different order by each of us, but all of those points are important. For me, the most important part is the social one. Linking faces to names, and linking $spent_good_time and reduce $beers_owed_to with them as well. Which people? All of them. We didn't think much of outsiders on the first Debconfs; last and this year, we held a very successful DebianDay, and we invited everybody who is local and is just interested even in getting to know what/who Debian really is. Of course, the main target is the people already inside. No, the orga team is not the target. We didn't run the conference so that the orga team had a good time - That would be much easier to do if we hired a second orga team :) Personally, I know that I'm better running conferences than coding infrastructure or fixing bugs, and I see Debconf as my main contribution to Debian being better - at least for 2005 and 2006, and hopefully 2007. > > They should be thrilled about the breadth and width of it, > > Not everyone thinks that way; other people get annoyed at the focus on > derivatives, or legal issues, or social issues, when what inspires them > about Debian is Debian itself, or the technical issues. Again, there are hundreds of possible definitions on what Debian is. I still think that "Debian itself" is a social project, and that it has a technical product. For me, social issues are as important (not more. In the end, we are a social project, but with a clear technical goal) as technical ones. And part of the bottom line is that we might have over 900 opinions on what Debian is and what is important for it. > The idea probably isn't to focus on what went wrong, but on what cool > new things can be done next year -- the day trip is a great idea, and > the debcamp preceeding debconf is a great idea, but they've been done, > so what's the next great idea? Putting paper selection in the hands of > the attendees? Getting lots more people to go? Something else? Debcamp is Oslo's innovation. Daytrip is Brazil's, IIRC. DebianDay is Helsinki's. Mexico's is... A waterfall? ;-) No, really... I think -and that's something we discussed a lot since last July- that we have converged to our ideal size. About paper selection... I am not a fan of the "everybody vote!" idea, although I am also not strongly against it. > > > I'm also not sure why debconf remains free for all comers; > > we dont try to encourage everyone and her dog to come. > > Really? I thought that was at least the theory behind "Debian Day"? DebianDay, yes. But DebianDay just happens very near in time and space to Debconf. It is not Debconf. > [charging for attendence] > > yes, that would be a totally differnet conference and i think > > also a pretty differnet set of attendees. Here i am interested in > > whom we would loose and who would come, additionally. > > Well, every year should be a different conference, really; the important > thing is what sort of conference it is. If you want to keep it a small, > focussed, technical conference that's great -- it's then just a matter > of doing that as well as possible. I don't like the idea of charging, as it would undermine one of the most important -again, to me- factors of the conference. We all work for Debian, and maybe this conference is a bit of what Debian can pay us back. We generate enough benefit to the world so that the world (i.e. our sponsors) give enough resources to spend a great time together, in one of the longest conferences I know of. Of course, whoever wants to pay for his stay is requested to do so (and many people actually do), and people who want to help with more than their expenses are recognized as sponsors. > One of the most inspiring things at LCA this year for me was going to > see a talk by Van Jacobson (as in -the- Van Jacobson) about optimising > the kernel level implementation of networking from the ethernet driver > to cope with 100Gbps networking -- it was /highly/ technical, approached > it with a broad analysis (ending up going from a specific kernel driver > all the way to userspace), had both low level details (why doubly linked > lists are bad) and high level principles (why end-to-end processing is > fundamentally more efficient), covered lots of historical detail, and was > even immediately relevant to the software I use every day. > > It also didn't have anything to do with me personally -- I'm not a kernel > guy, or even a glibc guy, so it wasn't remotely "this is something you > should do better in the future, aj", it was just "here's some cool stuff > I did, and look how fun it was, doesn't it make you want to go do fun > stuff too?" As a counter-example: In Mexico, I started the CONSOL conference. It is quite big for our country's standards (we have had 500-800 attendees, for four days of work, with up to five simultaneous tracks). Every year, we invited many internationally known important Free Software personalities, to get more attention and more people. The last year I was directly involved, I pushed and managed to bring Larry Wall - For the Perl fan I am, it was a great achievement. His talk was obviously greeat - I enjoyed it immensely even though I was doing the simultaneous translation (not an easy task, believe me!) Some other 20 or 30 people were very interested, following him - But I was shocked to see some people _sleeping_ here and there at the auditorium! Why? How? Well... They were not up to the level. Tell me, in LCA, how many people you think were able to really make sense out of Van Jacobson's talk, and follow it with the interest you did? I'd guess that not too many (or I don't really know what LCA is about). Good thing you could hear a Teacher as him - But do you honestly believe LCA was the right forum for him? Greetings, -- Gunnar Wolf - gwolf@gwolf.org - (+52-55)5623-0154 / 1451-2244 PGP key 1024D/8BB527AF 2001-10-23 Fingerprint: 0C79 D2D1 2C4E 9CE4 5973 F800 D80E F35A 8BB5 27AF
Attachment:
signature.asc
Description: Digital signature