1 00:00:00,000 --> 00:00:01,760 Hello everybody! 2 00:00:01,760 --> 00:00:04,030 Is there anybody who has seen 3 00:00:04,030 --> 00:00:08,460 my talk on Monday, about Blends or SoB? 4 00:00:08,460 --> 00:00:12,390 Ok, some people. Well, I'll repeat something 5 00:00:12,690 --> 00:00:14,850 only few things, because I think 6 00:00:14,850 --> 00:00:18,480 it became important even more after you've heard the other talk 7 00:00:19,040 --> 00:00:21,660 I'm a little bit astonished because this is 8 00:00:21,660 --> 00:00:26,630 kind of a by-product of my main topic 9 00:00:26,630 --> 00:00:28,630 and this ended up in the main talk room. 10 00:00:30,820 --> 00:00:34,370 What about MoM? 11 00:00:35,200 --> 00:00:37,060 It's called Mentoring of the Month 12 00:00:37,060 --> 00:00:39,860 and Mentoring of the Month was one of my ideas 13 00:00:40,250 --> 00:00:43,620 which came exactly at the same time, 14 00:00:43,620 --> 00:00:50,350 January last year (my own mail was sent one day later 15 00:00:50,350 --> 00:00:51,770 but was just prepared) 16 00:00:52,110 --> 00:00:55,430 and we had one person on our mailing list 17 00:00:55,430 --> 00:00:58,710 who wrote "Dear Debian Packagers, we are looking for one 18 00:00:58,710 --> 00:01:01,800 of you to help us change the history of Healthcare" 19 00:01:01,800 --> 00:01:07,490 actually he was quite engaged and wanted 20 00:01:07,490 --> 00:01:14,260 to have a really large, heavy hospital management system packaged 21 00:01:15,050 --> 00:01:18,110 and we had it on our todo list 22 00:01:18,110 --> 00:01:20,070 but it was so hard that 23 00:01:20,070 --> 00:01:24,560 we would not be able to do it without any help from upstream 24 00:01:24,560 --> 00:01:26,560 because it was quite complex and 25 00:01:26,560 --> 00:01:31,880 I'm personally a physicist and I've no idea about all those stuff, so 26 00:01:31,880 --> 00:01:34,670 I can have the knowledge about packaging 27 00:01:34,670 --> 00:01:37,870 and upstream has the knowledge about upstream program, so 28 00:01:37,870 --> 00:01:39,870 in my opinion this is quite a good fit. 29 00:01:39,870 --> 00:01:42,590 And at the same time 30 00:01:42,590 --> 00:01:46,910 I had written the announcement of the Mentoring of the Month 31 00:01:50,900 --> 00:01:56,900 It was by chance connected to the tenth birthday of Debian Med 32 00:01:56,900 --> 00:01:59,810 and I learnt in these ten years of Debian Med that 33 00:01:59,810 --> 00:02:05,060 there are people showing up and say "Hey, I want to support you a little bit, 34 00:02:05,060 --> 00:02:08,090 I would like to do something" 35 00:02:08,090 --> 00:02:11,020 and then you heard never again something from this person. 36 00:02:11,020 --> 00:02:16,070 Anybody else had the same experience? 37 00:02:16,070 --> 00:02:19,620 In the non-free world 38 00:02:19,620 --> 00:02:23,490 there is a boss and gives money to somebody and 39 00:02:23,810 --> 00:02:27,430 if this somebody doesn't do what he's supposed to do 40 00:02:27,430 --> 00:02:29,430 he will be fired: that's easy. 41 00:02:29,430 --> 00:02:32,250 In the Free Software world is not that easy, 42 00:02:32,250 --> 00:02:35,420 so what to do to get at least some 43 00:02:35,420 --> 00:02:38,990 slight handle on this person? 44 00:02:39,750 --> 00:02:43,900 I came up with the idea to 45 00:02:44,500 --> 00:02:49,450 say "I will spend my time to teach you, 46 00:02:49,450 --> 00:02:54,570 and you are supposed to do some work and do some reports, and so on" 47 00:02:54,870 --> 00:02:59,410 I will tell you how it works in 48 00:02:59,410 --> 00:03:04,420 detail, but first I want to show you another interesting thing 49 00:03:04,420 --> 00:03:06,420 in the Debian Med team 50 00:03:06,420 --> 00:03:10,960 we have about 23 Debian Developers and Maintainers who are 51 00:03:10,960 --> 00:03:13,650 not all very active, but 52 00:03:13,650 --> 00:03:17,470 because we're doing some mentoring 53 00:03:17,470 --> 00:03:19,470 we are advertising this subproject 54 00:03:19,470 --> 00:03:23,850 we have won 10 Debian Developers who confirmed 55 00:03:23,850 --> 00:03:25,850 that they are only in Debian because 56 00:03:25,850 --> 00:03:27,850 Debian Med project exists. 57 00:03:27,860 --> 00:03:29,330 Debian Med is a leaf project 58 00:03:29,330 --> 00:03:31,440 nobody really cares much about it 59 00:03:31,440 --> 00:03:33,040 in the main Debian universe. 60 00:03:33,040 --> 00:03:36,310 It's not important for several reasons, but 61 00:03:37,060 --> 00:03:40,860 even if this small unimportant project can get 62 00:03:40,860 --> 00:03:42,680 ten developers who otherwise 63 00:03:42,680 --> 00:03:43,790 would not be in Debian, 64 00:03:43,790 --> 00:03:46,440 we could do a lot of stuff more. 65 00:03:46,440 --> 00:03:48,440 One point of this is 66 00:03:48,470 --> 00:03:51,580 because we really care for newcomers. 67 00:03:51,580 --> 00:03:53,580 Seven of this people 68 00:03:53,800 --> 00:03:55,880 even dived into other fields 69 00:03:55,880 --> 00:03:57,880 inside Debian, 70 00:03:57,880 --> 00:04:02,240 so, they do not even necessarily be active in Debian Med anymore 71 00:04:02,240 --> 00:04:04,240 and eight of them are working 72 00:04:04,240 --> 00:04:06,240 actively in Debian Med. 73 00:04:07,110 --> 00:04:09,420 What is Mentoring of the Month? 74 00:04:10,040 --> 00:04:14,920 The mentor, currently only me, 75 00:04:14,920 --> 00:04:16,920 but it's not necessarily the case, 76 00:04:17,510 --> 00:04:20,850 just dedicates a part of his spare time, 77 00:04:20,850 --> 00:04:24,270 it's anyway my spare time what I'm doing for Debian 78 00:04:24,270 --> 00:04:28,670 I cut a little bit from packaging myself, 79 00:04:28,670 --> 00:04:32,120 and spends the time for his student. 80 00:04:32,760 --> 00:04:37,830 He should try to guide his student kindly into all secrets of Debian packaging 81 00:04:38,010 --> 00:04:42,090 and I think that learning by example is the best way you can do something 82 00:04:42,090 --> 00:04:50,380 Do some really help to reproduce "hello" packages, which is fine but 83 00:04:50,380 --> 00:04:52,380 it's not full of pitfalls. 84 00:04:52,380 --> 00:04:56,600 real life examples have lots of pitfalls we need to work around 85 00:04:56,960 --> 00:05:00,050 and in this process I've also 86 00:05:00,050 --> 00:05:02,050 faced something which 87 00:05:02,050 --> 00:05:04,050 make some trouble for me and 88 00:05:04,050 --> 00:05:06,790 that it's even more trouble for the newcomer. 89 00:05:07,160 --> 00:05:13,060 If the newcomer has a package he's interested in, 90 00:05:13,060 --> 00:05:17,510 and basically we had lot of these students 91 00:05:17,510 --> 00:05:18,960 that are upstream developers 92 00:05:18,960 --> 00:05:21,720 so have a real interest to get it packaged, 93 00:05:21,720 --> 00:05:24,950 they learn it on the real example. 94 00:05:24,950 --> 00:05:30,010 And it's not necessarily upstream developers, 95 00:05:30,010 --> 00:05:33,980 so the student is free to pick a package 96 00:05:33,980 --> 00:05:37,410 and I, as mentor, have some vetoing rights 97 00:05:37,410 --> 00:05:40,210 just because I have sometimes experienced 98 00:05:40,210 --> 00:05:42,680 that it's too hard to do this in one month 99 00:05:44,410 --> 00:05:46,050 You can say "well, it's nice if you 100 00:05:46,050 --> 00:05:48,110 want to package this, but this doesn't work" 101 00:05:48,110 --> 00:05:50,290 For instance, we have in Debian Med 102 00:05:50,290 --> 00:05:52,870 some kind of todo list 103 00:05:52,870 --> 00:05:54,870 we have this 104 00:05:54,870 --> 00:05:58,380 task pages and we have here some packages 105 00:05:58,380 --> 00:06:02,510 where some work is done in SVN or Git 106 00:06:02,510 --> 00:06:06,830 and you can just continue something already started 107 00:06:06,830 --> 00:06:08,830 or we have 108 00:06:09,530 --> 00:06:12,410 some other things which are not really started 109 00:06:12,410 --> 00:06:16,800 so we have this long todo list 110 00:06:16,800 --> 00:06:20,590 that students could pick from or can do something else. 111 00:06:21,970 --> 00:06:25,590 So, the goals of Mentoring of the Month: 112 00:06:25,590 --> 00:06:29,930 the package in question is finished and uploaded, after one month. 113 00:06:30,450 --> 00:06:34,040 And the student is able to do advanced packaging tasks 114 00:06:34,040 --> 00:06:38,110 and he's also introduced into the communication channels of Debian Med 115 00:06:39,970 --> 00:06:44,450 I always want to make a student communicate 116 00:06:44,450 --> 00:06:49,340 via the mailing list: I personally don't like 117 00:06:49,830 --> 00:06:52,290 mails popping up in my private mailbox 118 00:06:52,290 --> 00:06:56,710 there should be always some public record which you can seek for 119 00:06:57,150 --> 00:06:59,860 and so, communication and the package. 120 00:07:01,580 --> 00:07:06,480 The communication, as I said, on mailing list. 121 00:07:06,480 --> 00:07:10,990 I don't know why, but I always face people who 122 00:07:11,360 --> 00:07:14,710 are too shy to post to mailing list or whatever, 123 00:07:14,710 --> 00:07:17,220 but this is something I really want the student to do. 124 00:07:18,590 --> 00:07:23,360 And other readers might be bothered a little bit by this education 125 00:07:23,750 --> 00:07:26,570 it should be tagged by the 126 00:07:26,570 --> 00:07:28,570 [MoM] tag 127 00:07:28,570 --> 00:07:30,570 so people could filter out or 128 00:07:30,570 --> 00:07:35,990 perhaps there are even people who are only interested in this mentoring and they could read it as well. 129 00:07:38,500 --> 00:07:44,140 I have set five rules: one of these is responsiveness, 130 00:07:44,730 --> 00:07:49,490 so if the mentor is posting something on the mailing list, with a [MoM] subject 131 00:07:49,490 --> 00:07:52,940 the student should really try hard to 132 00:07:52,940 --> 00:07:56,400 to comment on this. 133 00:07:56,400 --> 00:08:01,480 "I have no idea what you mean, can you help me or explain better?" 134 00:08:01,550 --> 00:08:05,460 The student should just respond with something. 135 00:08:05,460 --> 00:08:08,260 It's quite frequent that you post something and 136 00:08:08,260 --> 00:08:11,200 to think that message is received but 137 00:08:11,200 --> 00:08:14,900 it is not really, so this is handshake principle you know from electronics. 138 00:08:14,900 --> 00:08:18,250 You send and the receiver says "yes, I got it". 139 00:08:21,250 --> 00:08:24,250 The student should also confirm understanding: 140 00:08:24,700 --> 00:08:30,570 He should say: "this is ok" or "what does it mean?" or "please, explain" 141 00:08:30,570 --> 00:08:36,150 or he can also say: "I would prefer this or that" over mine, because 142 00:08:36,150 --> 00:08:39,810 I'm not that strict on my opinion, 143 00:08:39,910 --> 00:08:46,420 because newcomers maybe have more ideas about upstream program than me. 144 00:08:46,420 --> 00:08:51,230 So they are way more right than I, or 145 00:08:51,730 --> 00:08:57,130 maybe there's an alternative and I do not want to put my strong opinion 146 00:08:57,130 --> 00:08:59,130 onto this student. 147 00:08:59,790 --> 00:09:05,020 The student should also follow changes, he should subscribe to the commit list 148 00:09:05,020 --> 00:09:13,250 or alternatively should pull from the repository and read the log files. 149 00:09:14,260 --> 00:09:16,760 And the student should also try to find 150 00:09:16,760 --> 00:09:18,760 a useful comment to confirm that 151 00:09:18,760 --> 00:09:20,760 he understand what was commited, because 152 00:09:20,760 --> 00:09:26,740 sometimes I can explain to the student "please do this or that change", 153 00:09:26,740 --> 00:09:31,310 but sometimes it's way longer to explain what to do 154 00:09:31,690 --> 00:09:35,910 than just commit a small change, put a comment on it 155 00:09:35,910 --> 00:09:41,880 and then the student he get also the message, and it's more straightforward. 156 00:09:41,880 --> 00:09:47,890 It's not that I want to do all the coding to put on the student. 157 00:09:49,120 --> 00:09:53,330 If there's something available in the repository, 158 00:09:53,330 --> 00:09:57,210 the student should confirm that he was able to build a package. 159 00:09:57,210 --> 00:10:03,040 You cannot always be sure that students just check out everything 160 00:10:03,040 --> 00:10:05,440 and know the processes of building 161 00:10:05,520 --> 00:10:11,990 I want him to say "I was able to build, and everything's fine, I'll continue working on this" 162 00:10:16,850 --> 00:10:20,890 In case of problems the student should ask on the Debian Med mailing list 163 00:10:20,890 --> 00:10:23,540 also with a [MoM] subject, 164 00:10:24,130 --> 00:10:27,500 and he should show to us that he's working on a problem 165 00:10:27,920 --> 00:10:30,040 so others might join, or not 166 00:10:30,780 --> 00:10:34,200 things should be sorted out via open discussion, 167 00:10:35,070 --> 00:10:40,850 and there are possibly a lot of silent lurkers on the mailing list 168 00:10:40,850 --> 00:10:44,400 we do not even know who is subscribed and who's reading 169 00:10:44,730 --> 00:10:47,610 and perhaps other people could learn. 170 00:10:48,230 --> 00:10:53,100 Alternatively, you could do the MoM with the 171 00:10:53,100 --> 00:10:55,100 IRC channel 172 00:10:55,100 --> 00:10:59,860 I admit I'm not so frequently on IRC, basically only on DebConfs, but 173 00:10:59,860 --> 00:11:03,620 if the student prefers it, we could do some meetings 174 00:11:03,620 --> 00:11:06,260 also to get some faster communication. 175 00:11:07,260 --> 00:11:12,010 And the student should post reports: what he has done, 176 00:11:12,010 --> 00:11:14,010 what he's doing everyday 177 00:11:14,010 --> 00:11:17,370 how was the progress working, and I think these reports are 178 00:11:18,150 --> 00:11:22,210 the way to prove that he's really working 179 00:11:22,210 --> 00:11:25,810 on what we agreed upon, because 180 00:11:25,810 --> 00:11:29,850 I agreed upon teaching this student and the student 181 00:11:30,260 --> 00:11:35,310 just reports back and shows that he understand something. 182 00:11:37,520 --> 00:11:44,300 In summary I can say that Mentoring of the Month is work, for both 183 00:11:45,030 --> 00:11:52,130 but I really think that if you work on it, it also shows that it's fun to work 184 00:11:52,130 --> 00:11:56,080 because both people can learn something 185 00:11:56,080 --> 00:11:59,790 the mentor trades his spare time for the work of the student, 186 00:12:02,220 --> 00:12:06,000 he also tries to train the "silent observers" 187 00:12:06,000 --> 00:12:10,780 and the focus is not only on the technical details but also on learning 188 00:12:10,780 --> 00:12:14,600 to know relevant communication channels 189 00:12:14,600 --> 00:12:18,600 this is also quite an interesting thing because 190 00:12:18,600 --> 00:12:21,970 we have a lot of people working silently on their own and 191 00:12:21,970 --> 00:12:24,980 they do not really know how to find friends. 192 00:12:27,370 --> 00:12:31,970 After hearing the talk this morning, about 193 00:12:31,970 --> 00:12:36,230 the women in Debian, I've seen 194 00:12:36,230 --> 00:12:44,260 that paultag has some statistics about Google Summer of Code's 0% women applicants 195 00:12:44,260 --> 00:12:50,580 out of 181, but in Gnome Outreach Program for Women they had 100 applicants, 196 00:12:50,580 --> 00:12:53,520 I wondered if I should I explicitly dedicate 197 00:12:53,520 --> 00:12:55,520 every second month for women only, 198 00:12:55,520 --> 00:13:01,880 I don't know. I don't know if it helps because we are quite a niche project 199 00:13:01,880 --> 00:13:06,920 if there are really so much women but I could consider to try this 200 00:13:06,920 --> 00:13:11,300 if other projects might adopt this idea of MoM 201 00:13:12,010 --> 00:13:15,580 it could be some way to help also here. 202 00:13:16,550 --> 00:13:20,270 As a result we had in the first year 203 00:13:20,270 --> 00:13:27,260 four MoM periods and in this year one, 204 00:13:27,260 --> 00:13:30,720 I've not students every month, so well.. 205 00:13:30,720 --> 00:13:36,820 And the most interesting thing is actually the first one 206 00:13:36,820 --> 00:13:38,820 it's not yet uploaded 207 00:13:38,820 --> 00:13:43,090 but in this MoM work, upstream learnt that they need 208 00:13:43,090 --> 00:13:45,410 to change their build system and changed a lot 209 00:13:45,410 --> 00:13:47,520 and so we need to wait for the next upstream 210 00:13:47,520 --> 00:13:50,990 which is supposed to happen soon. 211 00:13:50,990 --> 00:13:57,760 And finally we got something into Debian, or will get something really soon 212 00:13:58,430 --> 00:14:02,500 which we never managed over the ten years of Debian Med existence. 213 00:14:02,500 --> 00:14:06,080 This is a really valuable thing, 214 00:14:06,080 --> 00:14:08,080 to train people 215 00:14:08,080 --> 00:14:12,110 because otherwise we wouldn't manage that. 216 00:14:13,440 --> 00:14:16,200 Finally I think this is 217 00:14:16,200 --> 00:14:17,460 kind of a good result 218 00:14:17,970 --> 00:14:21,290 it also have that people put their name in the queue, 219 00:14:21,290 --> 00:14:25,350 we are using a wikipage, I'll show the link at the end 220 00:14:25,350 --> 00:14:29,370 there's a table and you put your name on the table 221 00:14:29,370 --> 00:14:32,040 it's quite informal. 222 00:14:32,040 --> 00:14:40,760 Some people consider doing it, and say "So, what can I do?" as I said I cannot fire anybody 223 00:14:40,760 --> 00:14:42,760 As a conclusion, 224 00:14:42,760 --> 00:14:48,270 I think the time spent into mentoring is really worth the effort. 225 00:14:49,250 --> 00:14:55,820 Because you have not student for every month, the workload is bearable 226 00:14:55,820 --> 00:14:59,560 somehow, so I think it's not very much more than 227 00:14:59,560 --> 00:15:02,360 what I'm doing anyway, because 228 00:15:02,360 --> 00:15:06,260 I do not restrict my teaching work on the mailing list to this project: 229 00:15:06,260 --> 00:15:08,260 I'm teaching on the mailing list anyway. 230 00:15:08,260 --> 00:15:13,170 And so it's a little bit more formalised and a bit more strict 231 00:15:13,170 --> 00:15:15,170 and hopefully inviting. 232 00:15:16,830 --> 00:15:19,800 And actually I'm also learning myself, because 233 00:15:20,240 --> 00:15:23,010 I've been a Debian Developer for fifteen years 234 00:15:23,700 --> 00:15:28,050 and not really sure if I have read every document which should be read 235 00:15:28,050 --> 00:15:31,880 and specifically I have not read them in the last ten years. 236 00:15:31,880 --> 00:15:37,370 So, the student is new and reads the documents and I'm learning 237 00:15:37,370 --> 00:15:39,120 I'm really learning a lot of features 238 00:15:39,120 --> 00:15:42,310 I would never have learnt if I wouldn't have done this teaching 239 00:15:44,700 --> 00:15:49,240 The major advantage for me it's that by the training 240 00:15:49,240 --> 00:15:52,940 we could pool the knowledge of upstream 241 00:15:52,940 --> 00:15:56,250 and ours of packaging stuff and this is a very efficient way 242 00:15:56,250 --> 00:16:00,780 to get packages which are really hard for us to do. 243 00:16:02,500 --> 00:16:06,320 So, Sponsoring of Blends: what does it mean? 244 00:16:06,320 --> 00:16:10,600 Well I learnt that the Blends concept is 245 00:16:10,600 --> 00:16:13,010 widely unknown amongst newcomers 246 00:16:13,010 --> 00:16:15,700 and also amongst the Debian Developers. 247 00:16:16,070 --> 00:16:18,560 I'm talking about Blends since 248 00:16:18,560 --> 00:16:21,000 2003, DebConf3 249 00:16:21,390 --> 00:16:24,090 and with different names, 250 00:16:24,090 --> 00:16:28,910 but people are widely ignoring this topic 251 00:16:29,030 --> 00:16:34,760 or think "that's a cool idea but we have no time to do it". 252 00:16:35,270 --> 00:16:39,680 And, well, this is fine: everybody can spend his time as he wants. 253 00:16:39,680 --> 00:16:45,660 But the thing is that newcomers might desperately seek for sponsors 254 00:16:46,300 --> 00:16:50,230 and simply don't know how to find one. 255 00:16:51,380 --> 00:16:53,500 But in principle every Blend should 256 00:16:53,500 --> 00:16:55,850 have a team around and should be 257 00:16:55,850 --> 00:16:59,110 really involved in sponsoring the packages. 258 00:17:00,360 --> 00:17:03,780 But some Blend doesn't fly in the sense that 259 00:17:03,850 --> 00:17:07,310 there are some maintainers having a package 260 00:17:07,310 --> 00:17:10,080 and asking for sponsoring on debian-mentors list 261 00:17:10,470 --> 00:17:12,710 So, what can you do? 262 00:17:13,640 --> 00:17:17,050 I think we could kill two birds with one stone. 263 00:17:17,520 --> 00:17:19,300 We can get the packages sponsored 264 00:17:19,300 --> 00:17:21,980 after proving that you understood 265 00:17:21,980 --> 00:17:23,980 the Blends techniques 266 00:17:25,920 --> 00:17:27,700 Think about this. 267 00:17:28,050 --> 00:17:30,230 I personally want to 268 00:17:30,840 --> 00:17:34,750 make some Blends concept way more popular than it currently is. 269 00:17:34,750 --> 00:17:38,490 And the maintainer of the package which could be in a Blend 270 00:17:38,490 --> 00:17:41,610 is seeking for sponsor, so we can find a common solution. 271 00:17:43,600 --> 00:17:47,320 Also here, there are some rules: Sponsoring of Blends means... 272 00:17:47,320 --> 00:17:53,020 The first rule is the maintainer should confirm 273 00:17:53,020 --> 00:17:58,070 that he understood the Blends principles, he has read Blends documentation 274 00:17:58,440 --> 00:18:02,800 is a member of some team on Alioth or reading the development mailing list. 275 00:18:02,800 --> 00:18:07,170 I've frequently observed that there are ITPs 276 00:18:07,170 --> 00:18:09,170 or an ITP list 277 00:18:09,170 --> 00:18:12,420 which are relevant for - say - 278 00:18:12,860 --> 00:18:16,340 Geographic Information Systems, Multimedia, Games 279 00:18:16,340 --> 00:18:19,330 and these people are single maintainers 280 00:18:19,330 --> 00:18:21,020 not really connected to a team, and 281 00:18:21,020 --> 00:18:23,670 I would say "Would you consider 282 00:18:23,670 --> 00:18:28,150 maintaining this on GIS team, or in Debian Science team 283 00:18:28,150 --> 00:18:30,150 to get some connection to people?" 284 00:18:31,320 --> 00:18:35,910 I really hope that if 285 00:18:35,910 --> 00:18:37,910 this seek for sponsor 286 00:18:40,750 --> 00:18:43,800 makes obvious that the person has no contacts 287 00:18:43,800 --> 00:18:47,100 I try to get this contact via the Blends. 288 00:18:47,900 --> 00:18:52,100 Second rule is that your package is maintained in a VCS 289 00:18:52,100 --> 00:18:54,100 used by the Blends team 290 00:18:54,650 --> 00:18:58,880 the technical background behind this is that we are 291 00:18:59,400 --> 00:19:05,390 checking the VCS with Blends tools and it shows up here 292 00:19:05,390 --> 00:19:07,390 so if your package is in SVN 293 00:19:08,010 --> 00:19:11,760 all the informations are updated from SVN 294 00:19:11,760 --> 00:19:15,480 if you really put your package there 295 00:19:15,580 --> 00:19:21,310 you have to do nearly nothing for it showing up in that page. 296 00:19:22,170 --> 00:19:26,340 One thing is to make the 297 00:19:26,980 --> 00:19:32,100 sponsee aware of this, and the other thing is that we get the information straight 298 00:19:32,990 --> 00:19:35,680 Your package is listed on the Blends tasks pages, 299 00:19:35,680 --> 00:19:38,100 which I've shown you, which is really really easy: 300 00:19:38,100 --> 00:19:43,360 you just put an additional line in the Control file. 301 00:19:43,770 --> 00:19:47,770 It's very easy and then after one day 302 00:19:48,030 --> 00:19:50,880 when the cron job is running, it is there. 303 00:19:50,880 --> 00:19:53,290 They will know about your package 304 00:19:53,290 --> 00:19:59,630 and you get some extra chance that somebody might pick it and sponsor it. 305 00:19:59,970 --> 00:20:03,150 And the fourth rule is that if you are unable to find a sponsor 306 00:20:03,150 --> 00:20:06,590 on the specific Blend list, even after posting there at least twice 307 00:20:06,590 --> 00:20:08,590 I will sponsor it for you. 308 00:20:08,590 --> 00:20:11,940 You should seek in your own team, 309 00:20:11,940 --> 00:20:15,260 and if the team doesn't go with it, I will do it. 310 00:20:15,830 --> 00:20:17,260 So these are the rules. 311 00:20:19,060 --> 00:20:24,560 This concept is quite new and we have only two 312 00:20:24,560 --> 00:20:27,590 cases for the moment: 313 00:20:27,880 --> 00:20:30,210 one was sponsored and it was fine, 314 00:20:30,210 --> 00:20:32,860 the other one... I sent him some change requests 315 00:20:32,860 --> 00:20:34,860 saying "please change this and that" and 316 00:20:34,860 --> 00:20:36,860 got no response. 317 00:20:39,030 --> 00:20:44,050 I think two or three months. 318 00:20:44,920 --> 00:20:46,740 The question was "How new is new?": 319 00:20:46,740 --> 00:20:51,340 I do not really remember when I announced it but two or three months. 320 00:20:53,340 --> 00:20:57,480 And I'm also lurking on relevant mailing lists and 321 00:20:57,480 --> 00:21:01,590 I'm reading debian-mentors and 322 00:21:01,590 --> 00:21:04,300 I'm trying to catch maintainers in need of a sponsor. 323 00:21:04,300 --> 00:21:09,050 I'm also lurking, for instance, on the lists of derivatives 324 00:21:09,050 --> 00:21:12,960 OsGeo distribution which 325 00:21:12,960 --> 00:21:17,950 is a Ubuntu derivative 326 00:21:18,460 --> 00:21:21,050 and in principle competing with the Debian GIS team. 327 00:21:21,680 --> 00:21:25,000 They don't know each other, they also want to make 328 00:21:25,000 --> 00:21:27,440 a GIS distribution of Free Software, 329 00:21:27,440 --> 00:21:29,440 they don't know of each other, 330 00:21:29,960 --> 00:21:34,370 they also don't know that there's a Debian Java team but 331 00:21:34,910 --> 00:21:36,970 there would be some need, 332 00:21:36,980 --> 00:21:38,470 and they're asking questions like "how can I package 333 00:21:38,470 --> 00:21:40,470 Java applications?" then 334 00:21:40,470 --> 00:21:42,470 I've written to their mailing list "well, there's the Debian 335 00:21:42,470 --> 00:21:45,760 Java list" and then nothing happened. 336 00:21:48,280 --> 00:21:51,200 If I see that there's some package near to be ready 337 00:21:51,200 --> 00:21:52,900 I post to them 338 00:21:52,900 --> 00:21:56,880 "you could get into Debian straight if you follow these four rules". 339 00:21:58,430 --> 00:22:02,060 I think this thing is new, so I'm always 340 00:22:02,060 --> 00:22:04,060 sending a link to the wikipage 341 00:22:04,060 --> 00:22:06,290 and hope that something will happen. 342 00:22:07,320 --> 00:22:11,620 If you might notice that there's some candidate for this 343 00:22:11,620 --> 00:22:15,370 Sponsoring of Blends rules, please redirect them to me 344 00:22:15,370 --> 00:22:23,150 I hope I can bear the rush you put on me now. 345 00:22:25,320 --> 00:22:30,800 This is what I just put on the slides when I was sitting in your talk. 346 00:22:30,800 --> 00:22:37,640 Mentoring of the Month, Sponsoring of Blends, and also face-to-face meetings. 347 00:22:37,640 --> 00:22:42,440 I'm coming to the repetition of what I've said on Monday. 348 00:22:43,570 --> 00:22:46,870 I want you to have a very quick look to 349 00:22:46,870 --> 00:22:51,480 what happened since 2011 when we started doing yearly sprints in our team 350 00:22:52,060 --> 00:22:55,940 and in these sprints we are not only Debian Developers 351 00:22:55,940 --> 00:22:58,550 there are upstream developers and users. 352 00:22:58,810 --> 00:23:01,480 About 20-25 people 353 00:23:01,480 --> 00:23:04,910 sitting together, we had in 2011 in January 354 00:23:04,910 --> 00:23:07,740 last year in February, and this year also in February 355 00:23:08,690 --> 00:23:11,690 and the effect is really amazing. 356 00:23:11,690 --> 00:23:14,930 By the way, thanks to Debian for sponsoring those sprints 357 00:23:15,760 --> 00:23:19,740 You see here the usual statistics, you might ignore that 358 00:23:19,740 --> 00:23:22,170 I'm always this large red line 359 00:23:22,170 --> 00:23:23,820 but what you see here 360 00:23:23,990 --> 00:23:27,340 in 2011 you have a solid base compared to this one 361 00:23:27,340 --> 00:23:29,340 these are newcomers 362 00:23:29,340 --> 00:23:31,170 and these newcomers are constant 363 00:23:31,460 --> 00:23:36,070 and actually these people who are longstanding (these are longstanding ones) 364 00:23:36,550 --> 00:23:39,420 may drop their activity. 365 00:23:39,420 --> 00:23:42,780 In 2011 you have some obvious 366 00:23:42,780 --> 00:23:47,050 increase in the upload of Debian Med packages 367 00:23:47,050 --> 00:23:50,560 next matrix is the discussions on the Debian Med users list. 368 00:23:51,020 --> 00:23:54,470 You see quite the same effect, right? 369 00:23:54,470 --> 00:23:58,600 We have another thing: the Debian Developers mailing list 370 00:23:58,600 --> 00:24:01,130 also 2011 371 00:24:01,350 --> 00:24:04,500 it is an increase, a measurable increase. 372 00:24:05,890 --> 00:24:08,810 Who is fixing bugs in the Debian Med team? 373 00:24:08,810 --> 00:24:10,810 You see since 2011 374 00:24:10,810 --> 00:24:12,810 a lot more people fixing bugs. 375 00:24:14,740 --> 00:24:19,320 And commits to the Debian Med VCS: same effect. 376 00:24:19,320 --> 00:24:21,320 It's quite consistent. 377 00:24:21,320 --> 00:24:27,150 And finally, in the packages we have in Debian Med we have a lot of biology packages 378 00:24:27,150 --> 00:24:29,150 and at the sprints we had actually 379 00:24:29,150 --> 00:24:31,150 biology developers. 380 00:24:31,370 --> 00:24:35,270 So the increase is in this biology section, you see here? 381 00:24:35,270 --> 00:24:37,740 The other are quite constant increasing, but 382 00:24:37,740 --> 00:24:40,060 you don't see this increase. 383 00:24:40,060 --> 00:24:44,620 If you meet face-to-face with interested people 384 00:24:44,620 --> 00:24:47,100 you get some drastic effort for your project 385 00:24:47,530 --> 00:24:50,370 even if it such a side project, 386 00:24:50,370 --> 00:24:52,040 a leaf project like Debian Med. 387 00:24:53,080 --> 00:24:56,120 I can't stop saying that if we would do all this stuff 388 00:24:56,440 --> 00:25:01,440 in Multimedia, GIS, Games or so, we could have lot of more people 389 00:25:03,080 --> 00:25:06,730 We always have the discussion here, and 390 00:25:06,730 --> 00:25:08,730 in the two talks, or three talks I've heard 391 00:25:08,970 --> 00:25:12,120 that people have trouble finding friends in Debian on 392 00:25:12,120 --> 00:25:14,120 IRC or on the mailing lists. 393 00:25:14,120 --> 00:25:16,120 Well, it's not that hard 394 00:25:16,120 --> 00:25:19,950 the answer is create inviting teams around certain topics. 395 00:25:19,950 --> 00:25:22,620 If somebody comes to Debian 396 00:25:22,620 --> 00:25:24,190 with a certain interest 397 00:25:24,190 --> 00:25:26,610 and everybody has some interest to 398 00:25:26,610 --> 00:25:28,010 install Debian 399 00:25:28,010 --> 00:25:30,010 they shouldn't do it otherwise. 400 00:25:30,010 --> 00:25:32,230 It should be natural 401 00:25:32,230 --> 00:25:34,500 for me, if I am a biologist 402 00:25:34,500 --> 00:25:37,440 I'm looking around at what distribution has the most 403 00:25:37,440 --> 00:25:38,890 biology packages, 404 00:25:38,890 --> 00:25:40,890 "ah! that's Debian! Let's try it!" 405 00:25:40,890 --> 00:25:43,560 and then you come and see there's a team behind it 406 00:25:43,940 --> 00:25:45,380 and then you will contact the team 407 00:25:45,380 --> 00:25:49,640 this is kind of a natural thing since 2011 or so. 408 00:25:49,990 --> 00:25:51,500 Formerly there wasn't so much, so 409 00:25:51,500 --> 00:25:53,500 create strong teams and 410 00:25:53,500 --> 00:25:56,510 you will make it very easy 411 00:25:56,510 --> 00:26:01,470 for others to find new members for a team. 412 00:26:01,600 --> 00:26:04,070 This is my definition for a team: 413 00:26:04,070 --> 00:26:06,060 waking up in the morning and realizing 414 00:26:06,060 --> 00:26:08,680 that somebody else has solved your problem from yesterday. 415 00:26:09,080 --> 00:26:12,270 I really have realized that 416 00:26:12,450 --> 00:26:15,510 we have some Debian Developer in the team 417 00:26:16,270 --> 00:26:20,000 who just joined the Debian Med team because it's fun in this team. 418 00:26:20,000 --> 00:26:24,670 He's Thorsten Alteholz, he's not connected to biology 419 00:26:24,670 --> 00:26:26,670 or medicine anyway, but he said 420 00:26:26,960 --> 00:26:28,670 "Well, that's fun, let's join it" 421 00:26:28,670 --> 00:26:33,160 and he's doing QA work, and he made some 422 00:26:33,500 --> 00:26:37,390 in December, some "Advent Calendar" 423 00:26:37,390 --> 00:26:40,090 to push people to fix a bug everyday 424 00:26:40,090 --> 00:26:42,780 to close the windows. 425 00:26:42,780 --> 00:26:44,780 This is very fun and it works 426 00:26:44,780 --> 00:26:47,440 because we try to invite people. 427 00:26:48,370 --> 00:26:50,300 So, these are some links: 428 00:26:50,300 --> 00:26:53,570 our policy document, which is also helpful for newcomers 429 00:26:53,570 --> 00:26:58,150 maybe this is even a fourth point that a newcomer can read, 430 00:26:58,150 --> 00:27:01,240 the main point is how to behave inside a team. 431 00:27:01,240 --> 00:27:04,230 Mentoring of the Month, Sponsoring of Blends, 432 00:27:04,230 --> 00:27:06,560 and well, that's basically all. 433 00:27:06,560 --> 00:27:08,560 I will upload this PDF document 434 00:27:08,560 --> 00:27:09,970 you can click on the links. 435 00:27:09,970 --> 00:27:11,970 And now, well, I talked 436 00:27:11,970 --> 00:27:13,970 two thirds of this BoF 437 00:27:13,970 --> 00:27:16,770 I just want to hear new ideas 438 00:27:16,770 --> 00:27:18,770 and any comments. 439 00:27:22,560 --> 00:27:24,010 Raise your hands, please. 440 00:27:34,450 --> 00:27:40,980 [Geert] Can you tell us more about organizing sprints? 441 00:27:41,310 --> 00:27:49,090 [Andreas] Yes, I think there's a mailing list debian-sprints@lists.debian.org 442 00:27:49,090 --> 00:27:55,580 There's a mailing list, debian-sprints 443 00:27:55,580 --> 00:27:58,570 and there you can announce that you want to do a sprint 444 00:27:58,570 --> 00:28:01,330 and then you should create the wikipage 445 00:28:01,330 --> 00:28:04,950 with some informations given, 446 00:28:04,950 --> 00:28:14,600 and you fill in location...There's a template you just fill... Lucas? 447 00:28:16,790 --> 00:28:20,680 [Lucas] Yes, about sprints: please organize sprints, we have 448 00:28:20,680 --> 00:28:25,340 the budget for that, if you think it's useful for your team, 449 00:28:25,340 --> 00:28:34,160 just contact me at the debian-sprints mailing list, and it's really easy to organize it. 450 00:28:35,910 --> 00:28:38,500 [Asheesh] If I understood the question correctly, it was about 451 00:28:38,500 --> 00:28:41,260 what are the steps required to organize a sprint? 452 00:28:41,260 --> 00:28:42,860 Can you explain the steps? 453 00:28:42,860 --> 00:28:45,810 You began to talk about there's a wikipage, but 454 00:28:45,810 --> 00:28:48,570 then what do you need to know, what planning do you need to do 455 00:28:48,570 --> 00:28:51,080 how do you invite people... 456 00:28:51,080 --> 00:28:57,770 [Andreas] Yes, ok. I'm very happy that somebody in our team is doing all the work for this 457 00:28:57,770 --> 00:29:01,590 and I just have to come to these sprints, that's nice. 458 00:29:01,590 --> 00:29:04,360 He is caring for 459 00:29:04,360 --> 00:29:06,360 some hotel, 460 00:29:06,360 --> 00:29:10,680 we have the rule to meet at some place at the beach, 461 00:29:10,680 --> 00:29:14,690 it was in Luebeck, it was in Southport, 462 00:29:14,690 --> 00:29:18,120 in England, and it was close to Kiel. 463 00:29:18,120 --> 00:29:26,720 Find a small hotel, preferable with some better bandwith that we have chosen, 464 00:29:26,720 --> 00:29:28,720 but yes, 465 00:29:28,720 --> 00:29:32,910 and then try to find people, and this person in our team 466 00:29:32,910 --> 00:29:35,470 has a very good ability 467 00:29:35,470 --> 00:29:39,460 to attract people I would never have think about, he's 468 00:29:39,460 --> 00:29:42,360 writing emails to people 469 00:29:42,360 --> 00:29:46,480 "Hey you have this package, and you might be interested in this and that 470 00:29:46,480 --> 00:29:48,480 and please come to our sprint" 471 00:29:48,480 --> 00:29:50,480 and he wrote to fifty people 472 00:29:50,480 --> 00:29:52,480 and twenty or so have come. 473 00:29:52,480 --> 00:29:56,380 This is a good mix of Debian Developers and 474 00:29:56,930 --> 00:30:00,910 other people and this makes some point, but even... 475 00:30:00,910 --> 00:30:04,700 There were even Ubuntu people at these sprints 476 00:30:04,700 --> 00:30:07,870 and even from a derivative of Ubuntu 477 00:30:07,870 --> 00:30:09,870 which is called BioLinux 478 00:30:09,870 --> 00:30:12,360 and the effect of our first sprint was 479 00:30:12,360 --> 00:30:16,490 that these BioLinux people are now using the same Wheezy as we, 480 00:30:16,490 --> 00:30:19,070 so we are working on the same code. 481 00:30:19,430 --> 00:30:25,500 And as a package migrate via Ubuntu to their distribution 482 00:30:25,500 --> 00:30:28,850 and if they have changes we incorporate it and 483 00:30:28,850 --> 00:30:30,850 if we do changes they have it as well. 484 00:30:30,850 --> 00:30:32,850 So this is quite efficient. 485 00:30:32,850 --> 00:30:36,090 Try to find friends for your project. 486 00:30:37,660 --> 00:30:41,920 And then put them in one place, and lock the doors! 487 00:30:41,920 --> 00:30:44,540 Serve them pizza and cola or whatever. 488 00:30:46,720 --> 00:30:49,080 [Asheesh] How long do your sprints last, 489 00:30:49,080 --> 00:30:51,580 and how many have you run, since 2011? 490 00:30:51,580 --> 00:30:54,550 [Andreas] We have run three, once per year, 491 00:30:54,550 --> 00:30:59,390 and they run from Friday afternoon to Sunday evening. 492 00:31:12,610 --> 00:31:17,390 We can also keep on talking about this other mentoring stuff. 493 00:31:18,730 --> 00:31:21,590 Fifteen minutes left, barely. 494 00:31:21,590 --> 00:31:26,480 If you have other ideas how we can get new developers 495 00:31:26,480 --> 00:31:32,140 or how we can create a friendly environment for men and women. 496 00:31:38,670 --> 00:31:44,550 [Franklin] Do you think it would help 497 00:31:44,550 --> 00:31:53,100 to have more Blends for not so specific 498 00:31:53,890 --> 00:31:57,710 users, maybe... 499 00:31:57,710 --> 00:32:02,100 let's put names... What about having your Blend for XFCE desktop 500 00:32:02,100 --> 00:32:06,310 or multimedia desktop? 501 00:32:06,740 --> 00:32:08,860 [Andreas] Yes, I think that is a good idea, 502 00:32:08,860 --> 00:32:12,810 because you say "not so specific" 503 00:32:13,230 --> 00:32:19,210 you need to decide how narrow you make the focus of your Blend 504 00:32:19,210 --> 00:32:24,090 For instance, in Debian you have a GIS team and an OpenStreetMap team 505 00:32:24,090 --> 00:32:27,770 and both have common packages and both have separate 506 00:32:27,770 --> 00:32:30,280 I don't think this is a good idea, you need to 507 00:32:30,280 --> 00:32:32,280 get some critical mass, 508 00:32:32,280 --> 00:32:36,900 to have ten developers, and not three there and five there. 509 00:32:36,900 --> 00:32:38,900 And then 510 00:32:38,900 --> 00:32:42,650 it could fly. For instance, if I'd started with Medicine, nobody 511 00:32:42,650 --> 00:32:45,940 immediately connects biology with medicine, but 512 00:32:45,940 --> 00:32:47,940 biology was a lot of stuff 513 00:32:48,060 --> 00:32:52,710 it was there, we used that. And so this build some creative base and then 514 00:32:52,710 --> 00:32:54,710 other things came around. 515 00:32:54,710 --> 00:33:02,320 Try to make it large enough to last some time, to get some critical mass. 516 00:33:05,040 --> 00:33:09,190 [Asheesh] I think that the idea of Blends is really interesting, it sounds almost like 517 00:33:09,190 --> 00:33:13,050 a collection of people who use Debian 518 00:33:13,050 --> 00:33:17,390 who care about some topic or packages, and the fact that there is a Blend 519 00:33:17,390 --> 00:33:19,390 is almost separate 520 00:33:19,390 --> 00:33:23,120 from the fact that it's just people interested in a topic or packages, 521 00:33:23,120 --> 00:33:28,630 there's no reason you couldn't have like Debian e-mail sprint, or something. 522 00:33:28,630 --> 00:33:34,080 [Andreas] You can do these sprints on any topic, yes. 523 00:33:35,760 --> 00:33:39,340 [Asheesh] The thing I'm trying to think through is that 524 00:33:39,340 --> 00:33:43,480 what you're talking about Blends it just makes me think some 525 00:33:43,480 --> 00:33:45,480 notion of teams 526 00:33:46,190 --> 00:33:50,450 and I guess I'm struggling to see the difference between Blends and teams 527 00:33:50,450 --> 00:33:52,450 and maybe we should try harder 528 00:33:52,450 --> 00:33:55,990 to make more topical collection of packages being in a team. 529 00:33:56,070 --> 00:34:01,990 [Andreas] The Blend is actually a team working together on a certain user-oriented topic 530 00:34:01,990 --> 00:34:06,400 It is user-oriented for certain 531 00:34:08,030 --> 00:34:13,140 for certain work fields, and you have some tasks in it and 532 00:34:13,140 --> 00:34:16,580 my idea was just to attract 533 00:34:16,580 --> 00:34:20,180 people who are not necessarily geeks 534 00:34:20,180 --> 00:34:23,650 but just want to do their normal work with Debian. 535 00:34:26,930 --> 00:34:33,280 I assumed that we would have way more Blends and I pushed 536 00:34:33,280 --> 00:34:35,790 the principle on people 537 00:34:35,790 --> 00:34:42,040 I wrote this stuff for the DebiChem team, all the frameworks I put on and they said "that's nice" 538 00:34:42,040 --> 00:34:43,350 And they use it now. 539 00:34:43,350 --> 00:34:48,170 And there's also the Debian Science Blend which is actually... 540 00:34:48,170 --> 00:34:52,710 actually we should have Debian Mathematics, Debian Physics, Debian 541 00:34:52,710 --> 00:34:56,030 Electronics or so, but there is nobody 542 00:34:56,030 --> 00:34:58,770 really engaged in 543 00:34:58,770 --> 00:35:03,220 separating things out and make a real team around these topics 544 00:35:03,220 --> 00:35:05,860 We would need such people who are doing this. 545 00:35:06,220 --> 00:35:10,080 [Asheesh] If I understand correctly, the core concept of Blends 546 00:35:10,080 --> 00:35:14,840 is that Debian when installed contains all of these packages, is it right? 547 00:35:14,840 --> 00:35:17,440 [Andreas] Yes, the Blend is completely integrated into Debian 548 00:35:17,440 --> 00:35:21,210 [Asheesh] But it's not just that it's integrated, there is an installable 549 00:35:21,210 --> 00:35:22,980 a sub-installable Debian... 550 00:35:22,980 --> 00:35:26,400 [Andreas] Well, it's installable as metapackages, 551 00:35:26,400 --> 00:35:30,820 we would like to have something to say about 552 00:35:30,820 --> 00:35:32,820 create a DVD just with the Blends, 553 00:35:32,820 --> 00:35:37,310 we have the idea, asked once on IRC, 554 00:35:37,310 --> 00:35:41,420 you could put it on tasksel, then just select your Blend 555 00:35:41,420 --> 00:35:46,960 there is an open bug reported, from 2003 or so, which is tagged "wontfix" 556 00:35:48,190 --> 00:35:51,210 [Asheesh] It sounds to be like 557 00:35:51,210 --> 00:35:55,610 - and I haven't thought through this carefully so it just may be crazy - 558 00:35:55,610 --> 00:35:58,740 that we should just rename Blends to Topics. 559 00:35:59,450 --> 00:36:04,080 [Andreas] Yeah, the naming topic... You have not seen my talk, 560 00:36:04,080 --> 00:36:08,670 the name of Blends changed over the time, I can repeat it if we have time. 561 00:36:08,670 --> 00:36:12,410 It was first Debian Internal Projects, and we wanted to split from 562 00:36:12,410 --> 00:36:16,120 this technical projects to user oriented, this was 563 00:36:16,120 --> 00:36:18,120 somehow logical, and then 564 00:36:18,120 --> 00:36:21,410 somebody suggested Custom Debian Distributions, 565 00:36:21,410 --> 00:36:23,890 which also was a total failure because, if you heard the... 566 00:36:23,890 --> 00:36:26,480 You are smiling, you know! 567 00:36:26,480 --> 00:36:29,920 Well, I suggested 568 00:36:29,920 --> 00:36:33,350 Debian Internal Solution, or something like this 569 00:36:33,350 --> 00:36:35,350 which sounded not 570 00:36:35,350 --> 00:36:37,350 that bad in my opinion but 571 00:36:37,350 --> 00:36:42,120 once I did this, people write that... I've never seen these people before, I've never 572 00:36:42,120 --> 00:36:46,970 seen any code from them, they're not supposingly on the mailing list but 573 00:36:46,970 --> 00:36:50,660 once there was a name we wanted an opinion on 574 00:36:50,660 --> 00:36:53,100 they raised their hands, they said I have an opinion 575 00:36:53,100 --> 00:36:56,790 we should not use this name, we should use something else. 576 00:36:56,790 --> 00:36:59,690 And then there was some voting and 577 00:36:59,690 --> 00:37:01,690 now it's called Blends. 578 00:37:01,690 --> 00:37:09,500 I don't really want to go this street twice. 579 00:37:09,500 --> 00:37:11,500 [laughter] 580 00:37:11,500 --> 00:37:15,900 [Franklin] I personally see quite some 581 00:37:15,900 --> 00:37:18,880 use for Blends for 582 00:37:18,880 --> 00:37:21,360 having people be able to provide 583 00:37:21,360 --> 00:37:27,490 multiple variance of Debian for security 584 00:37:27,490 --> 00:37:32,040 maybe a live CD for security purposes 585 00:37:32,040 --> 00:37:34,920 and some of them might be overlapping but... 586 00:37:36,420 --> 00:37:41,990 [Andreas] There's also now overlap: we have overlap between DebiChem and Debian Med, so there is overlap 587 00:37:41,990 --> 00:37:45,160 [Franklin] and I think it's fine to have them overlapping 588 00:37:45,160 --> 00:37:49,950 it just is a way of having people not forking Debian 589 00:37:49,950 --> 00:37:55,440 just for specific needs, it comes to have people working inside Debian 590 00:37:55,440 --> 00:38:00,520 and contributing inside Debian rather than forking and having their own 591 00:38:00,520 --> 00:38:07,060 [Andreas] For me is... A Blend is kind of making a Debian derivative the right way 592 00:38:07,060 --> 00:38:09,060 by staying inside Debian 593 00:38:09,060 --> 00:38:11,060 and maybe you finally can do 594 00:38:11,060 --> 00:38:17,180 some nice background, and release something that's your product but it's inside Debian. 595 00:38:17,680 --> 00:38:22,210 [Asheesh] I think the key concept you have about keeping the work inside Debian is well taken. 596 00:38:22,210 --> 00:38:25,880 For me, I kind of didn't realized that Blends... 597 00:38:26,260 --> 00:38:31,850 I thought that the point of Blends was to create non-derivative derivatives. 598 00:38:31,850 --> 00:38:39,630 Like an installable Debian Med image 599 00:38:39,630 --> 00:38:41,900 or whatever the package is called, 600 00:38:41,900 --> 00:38:43,900 So it seems like we should try 601 00:38:43,900 --> 00:38:47,130 to run around all of the users of Debian and 602 00:38:47,130 --> 00:38:49,130 who have friends who are DDs 603 00:38:49,130 --> 00:38:53,350 and say "Is there a topic in Debian you care about? Create a Blend today!" 604 00:38:53,350 --> 00:38:56,090 [Andreas] This is what I'm doing since 2003. 605 00:38:56,090 --> 00:38:58,200 Ten years of doing this... 606 00:38:58,200 --> 00:39:00,200 [Asheesh] I'm sorry that I'm so late to the party! 607 00:39:00,200 --> 00:39:02,200 [laughter] 608 00:39:02,200 --> 00:39:04,440 [Andreas] This is actually my point 609 00:39:04,440 --> 00:39:06,440 I was with this Blends 610 00:39:06,440 --> 00:39:09,740 topic, I was degraded to the second talkroom and there were 611 00:39:09,740 --> 00:39:12,430 basically people that are medicine or so related 612 00:39:12,430 --> 00:39:14,430 and other people think "well this stuff 613 00:39:14,430 --> 00:39:15,780 it's not interesting for me" 614 00:39:15,780 --> 00:39:17,090 And now I'm here for 615 00:39:17,090 --> 00:39:19,090 getting a larger team and people... 616 00:39:19,090 --> 00:39:21,900 I get more attention perhaps this way. 617 00:39:21,900 --> 00:39:24,920 On my next DebConf, in my talk 618 00:39:24,920 --> 00:39:30,410 on Blends I will find a funny title, and talk about it 619 00:39:30,410 --> 00:39:32,410 and let's see! 620 00:39:32,410 --> 00:39:36,570 Any other question? We have five minutes left. 621 00:39:36,570 --> 00:39:43,280 [Phil] Is it easy to discover the metapackages that make up a Blend, 622 00:39:43,280 --> 00:39:44,840 at the moment? 623 00:39:44,840 --> 00:39:47,270 [Andreas] I hope so, 624 00:39:47,270 --> 00:39:49,410 there's a common prefix also 625 00:39:49,410 --> 00:39:53,840 [Phil] Ok but... 626 00:39:53,840 --> 00:39:57,510 do we have some submenu on the expert install or whatever? 627 00:39:57,510 --> 00:40:01,850 [Andreas] Well, we could have a lot of them but the effect 628 00:40:01,850 --> 00:40:05,880 that is that tasksel stuff was not accepted 629 00:40:05,880 --> 00:40:07,880 [Phil] Well I'm not talking about tasksel, I'm talking about the 630 00:40:07,880 --> 00:40:09,880 initial install 631 00:40:09,880 --> 00:40:11,880 [Andreas] The installer or whatever. 632 00:40:11,880 --> 00:40:17,240 [Phil] You could... you can say which default desktop you want installed, 633 00:40:17,240 --> 00:40:23,310 you can have a parallel menu on the first bit of the Debian Installer 634 00:40:23,310 --> 00:40:26,500 with the menu being constructed automatically 635 00:40:26,500 --> 00:40:28,500 [Andreas] Whatever way 636 00:40:28,500 --> 00:40:30,500 you're suggesting, I'm all for it! 637 00:40:31,920 --> 00:40:35,340 I can't really show this here now but... 638 00:40:35,340 --> 00:40:37,340 [Phil] The point with that is that you could 639 00:40:37,340 --> 00:40:40,300 probably include udebs in your Blend 640 00:40:40,300 --> 00:40:42,300 and change the nature of the installer at the same time 641 00:40:42,300 --> 00:40:44,300 [Andreas] Most probably but 642 00:40:44,300 --> 00:40:48,980 we are lacking the technical manpower on this. 643 00:40:49,350 --> 00:40:52,280 So if anybody have some ideas, I'm all for it. 644 00:40:52,280 --> 00:40:54,280 There's also a frequent question 645 00:40:54,280 --> 00:40:57,480 "Do you have a DVD?" and I say "Well, 646 00:40:57,480 --> 00:40:59,480 this set of Debian 647 00:40:59,480 --> 00:41:04,290 DVDs is actually our work but people want this one." 648 00:41:04,290 --> 00:41:10,330 This is a Debian Med or whatever, so you can do it easily by 649 00:41:10,330 --> 00:41:14,370 using Debian Live and put all the metapackages on it, then you have it. 650 00:41:15,240 --> 00:41:20,280 Or you can do some other clever trick, but this would be nice if somebody would do this. 651 00:41:20,280 --> 00:41:30,650 What I'm always doing these webpages - which has now vanished because network is goin - contain everything 652 00:41:30,650 --> 00:41:34,500 translated with popcon stuff, 653 00:41:34,500 --> 00:41:36,500 ... maybe it's working again ... 654 00:41:40,220 --> 00:41:43,350 No, no network. Whatever, I don't know. 655 00:41:47,980 --> 00:41:52,960 Maybe it's back now... Yes, it's back. 656 00:41:57,590 --> 00:41:59,740 What I'm advertising is 657 00:41:59,740 --> 00:42:03,580 we have all these translated descriptions what belongs to the Blend 658 00:42:03,580 --> 00:42:06,210 so if this colleague of mine 659 00:42:06,210 --> 00:42:09,000 (I made it also for my colleagues who are working on 660 00:42:09,120 --> 00:42:10,210 a different operating system 661 00:42:10,210 --> 00:42:13,090 and using sometimes Free Software, while not being aware of this) 662 00:42:13,090 --> 00:42:20,760 We have all these applications, you have screenshots 663 00:42:23,810 --> 00:42:28,850 ... you should have screenshots on the right...yeah! we have screenshots 664 00:42:31,500 --> 00:42:36,180 Ok, you can upload screenshot if it is missing, you can translate, 665 00:42:36,180 --> 00:42:41,710 here's just a German locale to see that it's really translated, and everything of this come from UDD 666 00:42:41,710 --> 00:42:44,950 so the only thing you do to structure your Blend is say 667 00:42:44,950 --> 00:42:47,700 Depends: packagename, Depends: packagename 668 00:42:48,620 --> 00:42:52,230 and everything is done with this framework 669 00:42:52,230 --> 00:42:57,570 we have dedicated bugpages, so that if a biologist wants to help Debian to fix a bug 670 00:42:57,570 --> 00:43:02,970 he can go to the Biology topic and find the bugs in this. And we have other tools 671 00:43:02,970 --> 00:43:06,340 which are ready for any topic. 672 00:43:06,340 --> 00:43:08,420 It just needs to be used. 673 00:43:08,420 --> 00:43:16,370 I will run other BoFs, Debian Multimedia, Debian GIS and Debian Games and try to do it. 674 00:43:20,220 --> 00:43:23,500 At DebConf11 I was sitting at the Debian Games BoF 675 00:43:23,500 --> 00:43:28,960 and it was 45 minutes and I started from scratch and I made for them a template of framework 676 00:43:28,960 --> 00:43:30,960 up to this webpages 677 00:43:30,960 --> 00:43:35,420 There was no change until now to put another 678 00:43:35,420 --> 00:43:37,760 games or make some other task... 679 00:43:38,280 --> 00:43:41,490 I'm lacking the knowledge about games 680 00:43:41,490 --> 00:43:44,080 this knowledge needs to be put into the framework 681 00:43:44,080 --> 00:43:46,340 and perhaps we'll make sometimes out of this or 682 00:43:46,930 --> 00:43:50,060 try to re-write Debian Junior or so 683 00:43:54,950 --> 00:43:58,780 [Miriam] The problem with games is that they are so different 684 00:43:59,030 --> 00:44:03,230 that you would probably need to compartmentize that because 685 00:44:03,230 --> 00:44:07,630 for example is not the same 686 00:44:07,630 --> 00:44:09,630 kind of people who play 687 00:44:09,630 --> 00:44:11,630 Quake and play 688 00:44:11,630 --> 00:44:13,630 I don't know, chess. 689 00:44:15,100 --> 00:44:18,110 Probably in Debian Junior or 690 00:44:18,110 --> 00:44:22,800 or how we decide to call it, it might be 691 00:44:23,600 --> 00:44:26,400 quite easier because the age range is different 692 00:44:26,400 --> 00:44:28,400 but we would also might need to 693 00:44:28,400 --> 00:44:30,400 separate between different... 694 00:44:32,180 --> 00:44:35,140 [Andreas] I'm sorry if I interrupt you, but that's not really the point 695 00:44:35,140 --> 00:44:38,300 we have hospital information systems, biology, medical imaging 696 00:44:38,300 --> 00:44:41,930 this is as far and different things like 697 00:44:41,930 --> 00:44:45,970 Quake and Frozen Bubble. 698 00:44:45,970 --> 00:44:47,970 We just need to 699 00:44:47,970 --> 00:44:49,970 find the right category. 700 00:44:49,970 --> 00:44:58,390 [Miriam] Is there a way to make like tags, or something like that, to filter 701 00:44:58,390 --> 00:45:03,510 similar packages, like for example, I'm interested in this kind of game develooment 702 00:45:03,510 --> 00:45:06,520 [Andreas] We can do it. 703 00:45:07,620 --> 00:45:16,030 Time's over. Thank you everybody. 704 00:45:16,550 --> 00:45:20,990 [applause]