1 17:57 < pere> as always, I refuse to start the on the agenda before a summary writer is found, so it would be great with a volunteer soon.
2 17:58 -!- vagrantc
3 17:59 -!- h01ger
4 18:00 < pere> So, lets try to get started.
5 18:00 < pere> can everyone present indicate so using /me = Full Name for the summary writer?
6 18:00 * pere = Petter Reinholdtsen
7 18:00 * sepski = Ronny Aasen
8 18:00 -!- pere
9 18:01 * vagrantc = Vagrant Cascadian
10 18:01 * jever = Jürgen Leibner
11 18:01 * danielsan = Daniel Hess
12 18:01 * klausade is on/off
13 18:01 < pere> anyone capable of writing the summary?
14 18:01 < pere> can everyone present indicate so using /me = Full Name for the summary writer?
15 18:02 * klausade = klaus ade johnstad
16 18:02 * h01ger = Holger Levsen
17 18:03 < pere> in my view, if no summary is written, and thus not posted to the list, the results from this meeting will be forgotten and we can just as well not hold it. :)
18 18:03 < pere> so writing the summary is important for our collective memory to work.
19 18:04 < pere> anyone?
20 18:04 * h01ger changed topics a bit (basically put rc1+2 on the agenda as such)
21 18:04 < pere> h01ger: ack.
22 18:04 < sepski> It's allways difficult to find a summary writer. i'll do it today then.
23 18:04 -!- h01ger
24 18:04 < pere> sep: great.
25 18:05 -!- h01ger
26 18:06 < vagrantc> so, it was announced....
27 18:06 < pere> Yes, RC1 is out, and we are waiting for feedback. :)
28 18:06 * h01ger cheers sepski for announcing a new release! :)
29 18:06 < pere> yes, that was great. I was busy elsewhere, and just as I sat down to check status, I discovered the announcement. :)
30 18:07 < sepski> I only had time to test that it installed. and that i could login, i did not have time to test services
31 18:07 < pere> sepski: the services are supposed to be tested by the test suite. :)
32 18:07 < sepski> the test suite reported no errors :)
33 18:07 < pere> but as far as I can tell, the DVD is just fine. Have not tested the netinst CD.
34 18:08 < h01ger> we have 47 bugs as "feedback", there are only 3 left which need confirmation. and only 4 of prio 1 +2 :)
35 18:09 * h01ger added "souce media" to the agenda
36 18:09 < vagrantc> does LTSP install from the CD ?
37 18:09 < sepski> klausade, have you tested joining the samba domain with a windows machine since oslo ?
38 18:09 < sepski> vagrantc, yes
39 18:09 < vagrantc> great :)
40 18:09 < sepski> if you choose the ltsp server profile
41 18:10 < pere> vagrantc: yes, but the desktop installed on the server is a bit reduced without network available.
42 18:10 < klausade> sepski: sorry no, no windows machine within 100 miles available to me :-)
43 18:10 < pere> I plan to modify the munin setup to use the information from sitesummary to genererate the list of hosts to monitor.
44 18:10 < vagrantc> pere: so, that's an issue with the CD generally.
45 18:10 < pere> vagrantc: yes.
46 18:11 < sepski> h01ger, suggested making the cd a pure netinstaller,. and reducing it's size. if we are only going to release CD1 then i think that's a good idea
47 18:11 < sepski> klausade, i still fear we have a issue with joining windows machines into samba
48 18:11 < h01ger> pere, i saw that diff. great!
49 18:12 < pere> I'm also working on a tool to automatically increase lvm partitions when they fill up, but am not sure that will be ready before release.
50 18:12 < h01ger> :)
51 18:13 < h01ger> so, are we making the cds pure netinstall cds? or keep one with 650mb? or do we create 5 (or as many we need for the stuff on the dvd?)
52 18:13 < pere> regarding the non-free stuff, the easy way out is to just drop them. I believe h01ger's analysis is based on flawed assumtions at least regarding java, because if that was true, debian could not distribute gcc and java together. but I am not prepared to spend time investigating nor convincing h01ger, so the easy way out is just to drop the packages from the DVD.
53 18:14 < sepski> making them a smaller netinstall should be pretty easy and not break much
54 18:14 < pere> h01ger: the nice thing about having most dependencies on the CD is that it is actually useful without network available.
55 18:14 < h01ger> pere, ack. can you send a short mail in reply? (re: non-free)
56 18:15 < pere> h01ger: will do.
57 18:15 < h01ger> pere, and where would you cut? at 650mb?
58 18:15 < danielsan> what if the edu-cd has all packages which are different from debian etch one could use it together with debian etch cds to create a mirror (without internet)?
59 18:15 < h01ger> (pere, because i'm also happy to keep them if easily convinced :))
60 18:15 < pere> h01ger: we earlier decided on 700 MiB, I believe. It was implemented last year. but 650 is a safer bet for older writers. not sure if so old writers are a problem yet. :)
61 18:15 < sepski> danielsan, yes
62 18:15 < klausade> new bug #1165
63 18:16 < h01ger> danielsan, yup
64 18:16 < mannix> hi, i burned the etch cd from here http://ftp.skolelinux.org/cd-etch/, but it's not booting
65 18:16 < h01ger> danielsan, the edos project has a tool for that (too)
66 18:16 < danielsan> cool
67 18:16 < h01ger> mannix, did you check the md5sum?
68 18:16 < mannix> after the burn?
69 18:17 < mannix> or after dwnld?
70 18:17 < sepski> mannix, both
71 18:17 < h01ger> danielsan, http://debian.edos-project.org/xwiki/bin/Main/
72 18:17 < mannix> no, i did not
73 18:17 < sepski> mannix, works for me.
74 18:18 < h01ger> pere, 650 was rather $cdsize to me :) sorry..
75 18:18 < pere> did anyone test the amd64 or powerpc CDs? still no entries showing up on popcon.skolelinux.org yet. :/
76 18:18 < pere> h01ger: I'm not sure if we want to push the CD size limit when we also release DVDs. After all, the CDs are for those without DVDs and thus old hardware.
77 18:18 * h01ger was too busy for testing but will test rc1. and then keep/make popcon send :)
78 18:19 < h01ger> pere, right
79 18:19 < mannix> so, you just dwnld cd1 and burn as iso 9660 (joliet) with md5sum?
80 18:19 < h01ger> mannix, yes.
81 18:19 < pere> lets stick to development talk, and do the CD burn debugging after the meeting.
82 18:19 < mannix> ok, i used same app and settings for kubuntu feisty fawn and debian etch netinsall w. no probs
83 18:19 < sepski> mannix, the iso is the cd. you cant burn it as iso/joliet since that's allready decided on the iso file. just wodim file.iso
84 18:19 < h01ger> mannix, check the md5sum of the downloaded file and compare it with the one on the website. you can also check the md5sum of the burned media, if it doesnt work
85 18:20 < vagrantc> keeping the CD size at 650 seems wise, since the DVD is available as well.
86 18:20 < h01ger> the dvd is too full, so we need to create an extra source dvd?
87 18:20 < mannix> ok, i will try tjhat
88 18:20 * danielsan was busy writing the howto and test what to change to build multiarch dvds: http://lou.ping.de/~daniel/cd-etch-test-multi-dvd/
89 18:20 < sepski> so the CD stays at 650 and we dont do nothing with that. Will we remove CD2 ?
90 18:20 < pere> h01ger: it is too full? half of it is just padding.
91 18:20 < sepski> it's not beeing built anymore
92 18:20 < h01ger> right. cd2 should be removed from etch, etch-test, -powerpc and amd64
93 18:21 < h01ger> and lets stay at 650 :)
94 18:21 < h01ger> pere, then lets remove the "padding". i thought it was all useful suggests ;)
95 18:21 < pere> yes, everyone seem to agree on 650.
96 18:21 < h01ger> suggests:
97 18:21 < h01ger> i ment
98 18:21 < pere> h01ger: nope, our task packages fill up to around 2.2 GiB. The rest is padding based on the debian popularity-contest usage numbers.
99 18:22 < h01ger> ic
100 18:22 < pere> just check location of the last entry from the wanted list in the cdspacelist file.
101 18:22 * h01ger would 3> a combined source+i386 dvd :)
102 18:22 < pere> it will tell you how much "our" packages uses on the DVD.
103 18:22 < h01ger> <3 even
104 18:23 < pere> yes, that would solve some logistical problem for us.
105 18:23 < h01ger> pere, can you remove the cd2s please?
106 18:23 < h01ger> but i would also be happy about a binary and a source dvd. and maybe more happy then missing good software out
107 18:24 * h01ger checks size
108 18:24 < h01ger> a bit more than 2355200 KiB
109 18:25 < pere> I suspect it is most easy for us to make a separate source dvd, and suggest we use that if it is already working out of the box with debian-cd.
110 18:25 < vagrantc> on the other hand, forcing people to download 4+GB when they only really need slightly more than half that seems ... wasteful.
111 18:25 < h01ger> i can try to make happen this and break the etch-test-dvd this week :) but i'm fine if someone else does..
112 18:25 < h01ger> pere, afaik debian-cd supports both
113 18:25 < pere> h01ger: reat!
114 18:26 * h01ger nods vagrantc
115 18:26 < pere> vagrantc: I agree. on the other hand, if you are stuck on the savanna in Africa, it would be great if the source was included. :)
116 18:26 < vagrantc> i'd say keep the source CD separate and drop the padding ...
117 18:27 < jever> vagrantc: I agree
118 18:27 < vagrantc> pere: depends on if they're downloading the whole thing or not.
119 18:27 < vagrantc> or if they even want the sources.
120 18:27 < pere> I tried dropping the padding, but the variable I changed didn't have any effect. sepski believe it only have an effect when building custom DVDs, and chaning the build method will require changes in the install scripts.
121 18:27 < pere> s/chaning/changing/
122 18:27 < vagrantc> hrm.
123 18:27 < sepski> you must set disktype=CUSTOM,
124 18:28 < sepski> and if that changes the /cdrom/.disk/disk_type then we must detect dvd media in a different way
125 18:28 < vagrantc> well, we should define what we want, and then figure out the technical barriers along the way and the compromises we need to make
126 18:28 * h01ger nods vagrantc
127 18:29 < pere> yeah. Personally I would like a dual-boot amd64/x86 install DVD, with the live CD and the source included. :)
128 18:29 * h01ger volunteers to research this and suggests to discuss next week again. and work on a seperate source dvd now
129 18:29 < sepski> nothing that's impossible to do. but we are on rc1... and want to release in may :)
130 18:29 < h01ger> pere, yay :)
131 18:29 < pere> yes. so I believe we want the easy source solution for now.
132 18:30 < pere> without many changes to the installation system. to easy to break something in the process, or forget something.
133 18:30 < sepski> but we should definitiftly define what we want to have. so it's easier to aim for in lenny
134 18:30 * h01ger nods
135 18:30 < vagrantc> or if it's not too much trouble, a 3.0 post-release
136 18:30 < pere> are we done with status of rc1? is bugzilla complete with all the blockers for the release?
137 18:31 * h01ger is not sure if all blockers have p1 or p2 - but thats nothing we need to discuss in detail now
138 18:31 < sepski> pere, i hope test some services this week,
139 18:31 < sepski> espoecialy samba.
140 18:31 < pere> good. I hope those using debian edu today know that this is the time to provide feedback. :)
141 18:31 < h01ger> but we should aim to reach 0 p1 and p2 for next weekend? the weekend after (=trondheim)?
142 18:31 -!- H
143 18:31 -!- pere
144 18:31 < h01ger> minutes for the summary: ^^
145 18:32 < h01ger> 1138 nor P1 All email@example.com ASSI We need to check and act on the non-free packages on the DVD
146 18:32 < h01ger> 1146 maj P2 x86 firstname.lastname@example.org ASSI Need to tune slapd.conf to allow admins to create users, ...
147 18:32 < h01ger> 1149 nor P2 Oth email@example.com ASSI we want source CDs and DVDs
148 18:32 < h01ger> 1156 maj P2 Oth firstname.lastname@example.org ASSI Start Page is different beetween English and other languages
149 18:32 < h01ger> are the four blocking bugs
150 18:32 < pere> yes. this week will be very busy for me, so I am not sure I will be able to do much. :(
151 18:32 < h01ger> plus the status of debian-edu-doc is not really releaseworthy, but could also be fixed with 3.0r1
152 18:33 < pere> the start page need someone with real school experience to provide feedback, and then work from the translators.
153 18:33 < pere> I assume we all agree that the english page is the master page?
154 18:33 < pere> who is working on the slapd.conf stuff?
155 18:34 < h01ger> yup. but finetuning/huge improvements also can happen in a 3.0r1
156 18:34 < pere> yeha
157 18:34 < h01ger> slapd/lwat not
158 18:34 < pere> as far as I can see, no-one is on the non-free bug, but I am willing to throw the packages of the DVD to solve it.
159 18:34 < h01ger> (or also, but :)
160 18:35 < pere> h01ger: is working on the source build
161 18:35 < pere> and no-one is working on the slapd and web page stuff.
162 18:35 * h01ger is also "happy" to work on the non-free stuff if pere replies shortly :) ie asking the other ftpmasters..
163 18:35 < pere> h01ger: pere already replied. :)
164 18:35 < h01ger> woow \o/
165 18:36 < h01ger> and the start page is something almost everyone can improve. then we need translators..
166 18:36 < h01ger> we need to inform the people mentioned in knuds press release :)
167 18:36 < sepski> i have edited some of the easy stuff. like webmin->lwat
168 18:37 < pere> I guess we should just tell the translators to update now, and leave the fine tuning for later. is the wlus link removed from the web page? it should be.
169 18:37 < pere> good.
170 18:38 < pere> is working local device support in ltsp a requirement? klausade said earlier today that it only worked for root.
171 18:39 < sepski> i belived i had it working for regular users when i wrote the howto
172 18:39 < pere> I suspect it is a requirement, especially since it is know to be working.
173 18:39 < h01ger> "Terra supports USB pens, CD/DVD_ROMS and audio out of the box." - from the press release draft :)
174 18:39 < pere> yeah.
175 18:39 < pere> well, audio isn't enabled on thin clients, by design.
176 18:40 < pere> we can of course turn it on by default, but that would be against recommendations we got from schools a long time ago.
177 18:40 -!- schasi
178 18:40 < pere> on the other hand, it is just a question of who should have to edit the config.
179 18:40 * h01ger will write mail to all people there to ask for translations of the internal startpage (and to make them aware that we will release RSN)
180 18:40 < pere> do we want sound working out of the box?
181 18:40 < h01ger> pere, document it
182 18:41 < pere> I believe it is already documented.
183 18:41 < h01ger> where? :)
184 18:42 < h01ger> klausade, if you have a usb local device bug, please make it priority 2 :)
185 18:42 < pere> google isn't my friend. not able to find it.
186 18:43 * h01ger thinks sound should be turned off but it will be a faq how to turn it on. or how to use italc to control it
187 18:43 < pere> <URL: http://wiki.skolelinux.no/Dokumentasjon/ITIL/Konfigurasjonstyring > got it in english, at least.
188 18:43 < pere> s/english/norwegian/
189 18:43 * vagrantc was about to say something
190 18:43 < pere> vagrantc: do you know about english documentation?
191 18:44 < h01ger> vagrantc, ?
192 18:44 * vagrantc doesn't understand the question
193 18:44 < sepski> pere, is that _all_ that's needed ? no chaning sound deamons in kde and applications ?
194 18:44 < pere> sepski: not sure, but I do not believe so.
195 18:44 < pere> some apps might have to be told to use esd.
196 18:45 < vagrantc> we should switch to pulseaudio at the last minute :)
197 18:45 < sepski> vagrantc, :)
198 18:45 < h01ger> :)
199 18:45 < sepski> pere, i can add that SOUND=Y commented in the lts.conf file.
200 18:46 < pere> sepski: I believe we are not installing a lts.conf file. and I believe we should not do it either, to avoid upgrade problems.
201 18:46 < pere> vagrantc: perhaps not. :)
202 18:46 < sepski> pere, we are (i am) in order to get network swap to work out of the box
203 18:46 < pere> sepski: good. I thought it was enabled already?
204 18:47 < sepski> it's copied from lts.conf.dist
205 18:47 < pere> sepski: ah, right.
206 18:47 < pere> 13 minutes left, and two points on the agenda. should we move on?
207 18:47 < h01ger> yes
208 18:47 < pere> I conclude that we will need a rc2 before release.
209 18:47 -!- pere
210 18:48 < pere> h01ger: you seem to know most about the doc status?
211 18:48 < h01ger> the package debian-edu-doc is for the first time (?) included in debian-edu, but no documentation builds from source.
212 18:49 < h01ger> that is, the architecture document, the newdriftbook and...
213 18:49 < pere> h01ger: it isn't for the first time. it was included for a while in the woody version.
214 18:49 < sepski> h01ger, is this documentation linked from http://tjener.intern ?
215 18:49 < h01ger> the ikt book
216 18:49 < h01ger> sepski, no
217 18:49 < h01ger> pere, ic
218 18:49 < sepski> h01ger, can it be ?
219 18:49 < h01ger> cool :)
220 18:49 < h01ger> sepski, sure. it should
221 18:49 < sepski> or it is all over the place ?
222 18:49 < h01ger> /usr/share/doc/debian-edu-doc/
223 18:50 < h01ger> which should be exported by apache too
224 18:50 < h01ger> but currently all (?) documents are outdated
225 18:50 < pere> h01ger: I have vague memories of debian having some documentation framework for browsing the available documentation. do you know about this? do our package hook into it?
226 18:50 < h01ger> pere, no.
227 18:51 < pere> I suspect all documentation is slightly outdated, as no-one has updated it for the etch release, naturally. :)
228 18:51 * h01ger also has vague memories about that, but thinks thats not wanted: we dont want to expose the whole system config (=/usr/share/doc) to world.
229 18:51 < h01ger> so just serving /u/s/d/d-e-doc is IMHO better
230 18:52 < pere> ok.
231 18:52 < jever> ack
232 18:52 < sepski> h01ger, i added a link on the english page. ill export it in apache later
233 18:52 < h01ger> and making them build from svn is the next step on my plan. i need your help there :)
234 18:52 < pere> h01ger: did you post to the list with info on what documentation is included?
235 18:52 < sepski> h01ger, atlest this way the translators pick it up
236 18:52 < h01ger> debian/rules shows what is checked out from svn. then run make and fix the failure :)
237 18:53 < h01ger> currently debian/rules just copies whats in svn. and thats partly outdated for translations / doesnt allow new translations
238 18:53 < h01ger> pere, i didnt post that yet. i planned to do so todo, but fever stopped me
239 18:53 < pere> ok
240 18:53 < pere> I hope some of the readers can verify that we remembered all the doc, or point us to some stuff we forgot.
241 18:54 < pere> h01ger: did you clear the license issue?
242 18:54 < h01ger> the ITIL book has a unclear licence situation (see debian-edu-doc/debian/README and debian-edu-doc/todo in svn)
243 18:56 < h01ger> the kursbook is just gfdl without specifying "no invariant sections" in the licence without specifying that it has no invariant sections - so i will probably include it in the next upload (awaiting feedback from other ftpmasters)
244 18:57 < pere> right.
245 18:57 < pere> ok. 3 minutes left. next point.
246 18:57 < h01ger> next point?
247 18:57 -!- pere
248 18:57 < sepski> pere, i added a agenda point
249 18:57 < h01ger> 5. courier imap ?
250 18:57 < pere> in one week?
251 18:57 < pere> sepski: right. did not notice that.
252 18:58 < sepski> i added it since i noticed that courier imap only run the ssl version
253 18:58 -!- pere
254 18:58 < sepski> but i assumed most clients used TLS today and then they do conenct to the default port first
255 18:58 < pere> sepski: as long as it is impossible to get passwords sent in clear text of the net accepted, I believe tls can be fine.
256 18:59 < h01ger> do(es) our client(s) per default use tls? :)
257 18:59 < pere> I doubt it
258 18:59 < sepski> what clients ?
259 18:59 < h01ger> kmail
260 18:59 < pere> not sure if our clients (kmail?) get any config at all at the moment.
261 18:59 < h01ger> icedove
262 18:59 < pere> icedove isn't installed by default, I believe.
263 18:59 < sepski> icedove has tls if available as its default i belive.
264 19:00 < h01ger> http://bugs.skolelinux.no/show_bug.cgi?id=1013 is related...
265 19:00 < h01ger> if we say "firefox^wiceweasel" we should also have icedove on the dvd
266 19:00 < pere> I do not understand why the client ask repetedly for passwords. I would expect it to ask only once per login.
267 19:01 < h01ger> so what we want is clear, next topic? (=next meeting when?)
268 19:01 < pere> sound to me like a broken client.
269 19:01 < pere> ok. next meeting
270 19:01 -!- pere
271 19:01 < vagrantc> -1 minutes remaining
272 19:02 < h01ger> next monday, 18 utc?
273 19:02 < pere> both sunday and monday work for me. monday is a national holiday here i norway.
274 19:02 < h01ger> or tuesday, 19 utc?
275 19:02 < pere> busy tuesday.
276 19:03 * vagrantc is universally uncertain of anything
277 19:03 < pere> (most probably, at least. depend on the topic for osllo linux user group.
278 19:03 < h01ger> sunday, 17 utc
279 19:03 < h01ger> ?
280 19:03 < pere> is there any way we can get white to show up?
281 19:04 < h01ger> white, that would be nice+useful :)
282 19:04 < h01ger> its 7 am there, now. i believe. (summertime...)
283 19:04 < pere> I suggest we go with sunday or monday, and try to pick a time that work for white too. anyone got time to try to make that happen?
284 19:04 < pere> if not, we just pick a time now.
285 19:04 < h01ger> mention him in the log?
286 19:05 < h01ger> white must make it happen ;)
287 19:05 < pere> good idea. :)
288 19:05 < pere> backup plan monday 18:00 utc?
289 19:05 * h01ger sends him an email
290 19:05 < pere> and we plan a new release next saturday?
291 19:06 < sepski> yes
292 19:06 < h01ger> sunday?
293 19:06 < sepski> h01ger, saturday if we manage, sunday if we dont
294 19:06 * h01ger suggests to release on sunday to be able to work on it on saturday
295 19:06 < sepski> nice to have a backup day :)
296 19:07 < h01ger> also right
297 19:07 < pere> yeah, lets ame for saturday, and fall back to early sunday if we don't.
298 19:07 < pere> h01ger: don't worry, we have to work hard on saturday with testing and fixing to release on sunday. :)
299 19:08 < h01ger> no worries :)
300 19:08 < pere> ok. thank you all for joining this meeting. I hereby close it. :)
301 19:08 < h01ger> \o/
302 19:09 * sepski notices the bug count increasing during the meeting
303 19:09 -!- pere