I2P_Developers/i2p.www
3
0
Fork 5
Code Issues 27 Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Added meetings 216, 217 and 218

Browse Source
This commit is contained in:
str4d
2013-02-07 00:33:27 +00:00
parent 3a8a6daa32
commit 799f83b8af
6 changed files with 923 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

378
i2p2www/meetings/logs/216.log Normal file
View File

@@ -0,0 +1,378 @@
20:07:05 <hottuna_> Alright, meeting-time?
20:07:27 <str4d> o/
20:08:28 <RN-> tjink so
20:08:41 <hottuna_> eche|on, zzz, dg: ping
20:09:46 <hottuna_> let's wait until 20:15 and see if dg shows up.
20:10:21 <RN-> did everyone read zzz homework assignment?
20:10:36 <hottuna_> yepyep
20:11:28 <RN-> was over my head
20:11:31 <str4d> Okay, looks like the three meeting topics are ugha.i2p, the website revamp and the crypto. Anything else we want to cover?
20:11:50 <hottuna_> I think that is more than enough
20:11:58 <str4d> Alright:
20:12:01 <RN-> read it tho
20:12:04 <str4d> (0) Say Hi.
20:12:11 <str4d> (1) Ugha.i2p
20:12:18 <str4d> (2) Website revamp
20:12:29 <str4d> (3) Crypto discussion
20:12:32 <str4d> (0) Say Hi.
20:12:35 <str4d> Hi!
20:13:00 <RN-> hi
20:13:07 <hottuna_> hello everybody!
20:14:44 <RN-> we waiting 4 zzz and ech?
20:15:21 <hottuna_> I think we can manage until the crypto part
20:15:27 <str4d> eche|on was around an hour ago; zzz tends to speak when he needs to.
20:15:27 <RN-> guess they r at end...
20:15:49 <hottuna_> weltende, welterde, eche|on: ping, re new website
20:15:52 <hottuna_> altight
20:15:58 <RN-> anyone got the baffer?
20:15:58 <str4d> And everyone else can turn up when they do ^_^
20:16:05 <str4d> (1) Ugha.i2p
20:16:05 <hottuna_> So.. ugha?
20:16:08 <str4d> o/
20:16:39 <zzz> here, standing by until 3), if it's reasonbly fast
20:16:52 <hottuna_> Alright, I posted a content-request page last week
20:16:52 <hottuna_> syndie/imule content was requested
20:16:59 <hottuna_> and has as far as I can see been submitted
20:17:18 * str4d can
20:17:29 <str4d> 't actually load ugha right now =P
20:17:41 <str4d> Do we know who runs ugha?
20:18:04 <hottuna_> I don't
20:18:23 <hottuna_> do we have any further ideas about what to change/add to ugha?
20:18:31 <str4d> Because it would be useful to get some proper spam protection if possible.
20:18:38 <eche|on> we do partly know/guess who runs it. but it will not be disclosured here
20:18:47 <eche|on> and owner did not respond yet
20:18:54 <dg> Okay, hey
20:18:57 <str4d> eche|on: fair enough.
20:18:57 <eche|on> ugha.i2p was cleaned from spam
20:19:09 <str4d> eche|on: how much work was that?
20:19:12 <eche|on> and I added a site about iMule and syndie, KillYourTV added a bit more
20:19:30 <eche|on> spam? a lot, it was >200 or even >400 spam messages to be removed
20:19:38 <eche|on> they appeared in 2 years time
20:20:09 <str4d> And just manually removed?
20:20:24 <hottuna_> did it appear over the inproxy?
20:20:35 <dg> I was wondering this
20:20:50 <dg> sorry for being late although I managed to get here :)
20:20:53 <eche|on> yea, str4d, click each spam site, click delete site, click yes, I want to remove, click next spam site
20:21:08 <eche|on> and IMHO it is on INproxy.
20:21:27 <eche|on> yeah, it is
20:21:58 <eche|on> http://ugha.i2p.to/RecentChanges
20:22:01 <hottuna_> alright, maybe it shouldnt be accessible over the inproxy?
20:22:15 <RN-> so... set read omly for inproxy?
20:22:15 <eche|on> maybe someone want to count the "delete" pictures ;-)
20:23:34 <hottuna_> is it possible to notify the admin via the the wiki?
20:23:45 <eche|on> guess not
20:23:48 <hottuna_> a read-only via inproxy rule would probably be good
20:23:51 <hottuna_> ok
20:24:06 <hottuna_> eche|on, but you know who? you could do it?
20:24:28 <eche|on> I cannot do anything on it, I am just a user like anyone else
20:24:43 <dg> The person obviously is not active.
20:24:46 <dg> So.. maybe still no.
20:24:51 <eche|on> all I can do is asking tino (i2p.to owner) to block it.
20:25:18 <hottuna_> is blocking it entirely an acceptable solution?
20:26:01 <eche|on> yes
20:26:05 <dg> not long term
20:26:30 <RN-> I agree with dg
20:26:44 <eche|on> it is a wiki. It needs active administration to remove unwatned content
20:26:44 <hottuna_> i think blocking it is acceptable.. since it only is of use to people who are already using i2p
20:26:57 <eche|on> but as we also have active spammers inside of I2P....
20:26:57 <zzz> tino's not going to take action unless the owner requests it
20:27:04 <zzz> at least, he shouldnt.
20:27:41 <hottuna_> eche|on, could you contact the owner?
20:27:52 <eche|on> currently I visit ugha.i2p daily and remove the spam
20:28:15 <eche|on> hottuna_: I did contact via IRC and email already. now it is time for person to react.
20:28:38 <zzz> if it continues to be an embarassment we can take it out of the router console, whether we have a replacement or not
20:28:41 <eche|on> you know, we´ve seen same problem with forum.i2p already. thats the problem inside of I2P
20:28:48 <hottuna_> regarding blocking from i2p.to?
20:29:02 <eche|on> regarding active admin jobs on it
20:29:25 <hottuna_> ok
20:29:58 <hottuna_> anyway, if you manage to get some response, ask about blocking
20:31:01 <RN-> tino is not only inproxy anymore
20:31:43 <dg> Yeah.
20:32:01 <str4d> Aside from the spam issue, is there any content that ugha should have/needs updated>
20:32:29 <dg> Yes.
20:32:29 <eche|on> I had a look at the russian wiki. Thats a nice nice nice one
20:32:44 <str4d> From /Requests - "More advanced i2p config options and explanations." - hottuna_ you already added some of these, right?
20:32:44 <eche|on> it is really filled with good content and structured. but in russian.
20:32:44 <str4d> eche|on: link?
20:32:53 <hottuna_> what's the url for the russian wiki?
20:33:12 <hottuna_> str4d, yes. And I found a similar list on echelon.i2p
20:33:24 <eche|on> if I find it again...
20:34:10 <eche|on> imho rus.i2p
20:34:56 <eche|on> but more explanation about advanced config is nice
20:34:59 <str4d> Ooh, that *is* a nice wiki.
20:36:25 <eche|on> to sad I am a bit out of time, but if I get the chance, I do a few bits
20:36:32 <RN-> looks like it's using the same nice clean interface as cake why TV on his Cindy page
20:36:42 <dg> is it in english?
20:36:45 <RN-> I'll have to leave in about 10 minutes or less catch up with the rest of the meeting on my scroll back...
20:38:21 <str4d> Are there any other major points about ugha.i2p that need raising?
20:38:36 <hottuna_> no.
20:38:47 <hottuna_> I updated the request site
20:39:50 <str4d> The /I2pRfc page could do with updates, if it is/was ever planned to be authoritative (though the website is probably the better place for specs).
20:40:26 <dg> ugha.i2p has a lot of content which could be added or update
20:40:33 <dg> it seems to have more information about i2p's past and old tech documents than anywhere else
20:41:19 <str4d> Summary so far: spam is (currently) under control but needs active policing; there are numerous old pages that would be good to get updated (a good task for people who like writing).
20:41:34 <hottuna_> agreed.
20:41:41 <str4d> And if possible, the wiki should block edits from the inproxy.
20:41:56 <str4d> Anything else to add before we move on?
20:41:59 <dg> Is that all for the wiki then?
20:42:02 <dg> I don't think so
20:42:52 <str4d> dg: you want to do the honors? ^_^
20:43:11 <dg> Alright :3
20:43:15 <dg> thx
20:43:38 * str4d gets to talk lots in the next topic anyway =D
20:43:53 <dg> Okay, so the website revamp - I feel that the new design headed by str4d (he's doing the backend mostly but some CSS changes) brings a fresh look to i2p and can help refresh people's perspective and first impressions of it
20:44:00 <dg> The current one is rather stale, etc, etc..
20:44:11 <dg> I think that we should look into what needs completing in order to push it live
20:44:34 <str4d> What *must* be completed before pushing live:
20:44:37 <dg> Minor issues can be worked on when it's out there so the blockers we need to consider here?
20:44:48 <str4d> - translation tagging
20:45:01 <str4d> (well, not *must* but most at the very least)
20:45:17 <str4d> - checking that all site-internal links are updated and valid
20:45:36 <str4d> That's basically it.
20:45:56 <hottuna_> how is translation tagging done?
20:46:07 <str4d> I've already started on that, and have covered most of the site pages (if you leave out the docs, which are large on their own)
20:46:22 <dg> Latter isn't too hard. There's tools for it IIRC but I can go around clicking (take one for the team ;) if push comes to shove.
20:46:33 <dg> Explain translation tagging?
20:46:40 <str4d> hottuna_: Jinja2 template tags
20:46:40 <str4d> And gettext PO files
20:47:05 <str4d> <h2>{% trans %}A Gentle Introduction to How I2P Works{% endtrans %}</h2>
20:47:08 <str4d> <p>{% trans -%}
20:47:08 <str4d> I2P is a project to build, deploy, and maintain a network supporting secure and anonymous
20:47:08 <str4d> communication. People using I2P are in control of the tradeoffs between anonymity, reliability,
20:47:11 <str4d> bandwidth usage, and latency. There is no central point in the network on which pressure can be
20:47:11 <str4d> exerted to compromise the integrity, security, or anonymity of the system. The network supports
20:47:11 <str4d> dynamic reconfiguration in response to various attacks, and has been designed to make use of
20:47:11 <str4d> additional resources as they become available. Of course, all aspects of the network are open and
20:47:11 <str4d> freely available.
20:47:15 <str4d> {%- endtrans %}</p>
20:48:17 <str4d> The tagged blocks get extracted into a messages.pot which can then be translated like the routerconsole is.
20:48:36 <str4d> That's another task that I think *must* be done before launch:
20:48:57 <str4d> - Migrate old translated pages (e.g. /how_intro_fr) to PO files
20:49:53 <hottuna_> ok
20:49:56 <hottuna_> whats the mtn repo name?
20:50:04 <hottuna_> alright
20:50:08 <str4d> That one I can't do much about =P I've migrated one page as a test, but I can't verify the accuracy of the old translations (especially as there was nothing to keep things in sync between the static pages)
20:50:12 <str4d> i2p.www.revamp
20:51:02 * str4d starts up the test site again
20:52:33 <str4d> Okay, http://vekw35szhzysfq7cwsly37coegsnb4rrsggy5k4wtasa6c34gy5a.b32.i2p/en/ is back up.
20:52:44 <iRelay> Title: I2P Anonymous Network (at vekw35szhzysfq7cwsly37coegsnb4rrsggy5k4wtasa6c34gy5a.b32.i2p)
20:52:59 <str4d> Something else I've done is added mobile support to the website - you can see it by narrowing your browser window below 768px
20:53:34 <dg> What are we doing about blog/
20:53:34 <dg> ?
20:53:45 <str4d> dg: what do you mean?
20:53:52 <str4d> (In what regard?)
20:54:04 <dg> Who will be blogging and how will we set it up? When will we blog also? :)
20:54:43 <str4d> At present the blog just contains the (old) release posts and the (much older) status posts.
20:54:54 <str4d> At the very least there will be the release posts as normal.
20:55:50 <str4d> That's a later issue though - we need to actually get the site finished first!
20:56:09 <hottuna_> agreed
20:56:20 <str4d> Ticket #807 does have a few things in it which would be good to get done, but are not blockers
20:56:32 <iRelay> http://trac.i2p2.i2p/ticket/807 - (accepted enhancement) - Revamp of website
20:56:44 <str4d> They are somewhat spread out through the ticket, but some are:
20:57:02 <str4d> - fill out /about/glossary
20:57:21 <str4d> - improve blog/meetings layout and styling
20:58:17 <str4d> - fix or replace the theme
20:58:36 <hottuna_> re translation tagging: is """{{ _('Friends of I2P') }}""" tagable in a straight forward manner
20:59:03 <str4d> hottuna_: That already is tagged.
20:59:26 <hottuna_> just curious about syntax
20:59:29 <str4d> (That's the more compact notation)
20:59:39 <hottuna_> aah
20:59:42 <str4d> {{ }} inserts the result of the contained Python method
20:59:53 <str4d> _() is the gettext call in Python
21:00:00 <str4d> (well, the one that is imported into Jinja2
21:00:03 <str4d> )
21:00:19 <hottuna_> thanks
21:00:34 <str4d> {% trans %}{% endtrans %} is a more verbose tag, but it's the Jinja2 tag and supports any content between the tags.
21:00:49 <str4d> (whereas the _() one can't contain e.g. '
21:00:52 <hottuna_> what is left to tag?
21:01:13 <str4d> hottuna_: check the mtn log for details of what has been tagged, but IIRC:
21:01:44 <str4d> - get-involved/guides (I've tagged ides and dev-guidelines there)
21:01:55 <str4d> - misc/*
21:01:58 <str4d> - docs/*
21:02:09 <str4d> And then any blog posts that we want translated.
21:03:06 <str4d> (I've already migrated and tagged the 0.9.4 and 0.9.3 posts, and future posts can be tagged as well; earlier ones can be tagged as/when people can be bothered)
21:04:17 <str4d> Okay, we do need to get a move on in the meeting.
21:05:18 <str4d> Summary: site revamp is almost ready, help is appreciated getting the rest of the site tagged for translation and url-checked (can be done simultaneously) (thanks hottuna_ for offering to help (I assume that's what you are doing?))
21:05:45 <str4d> And other text/layout changes are appreciated but not blocking.
21:06:31 <str4d> Oh: and if anyone wants to get started on translating the pages (using the old translated pages as reference or for copy-paste), *please do so*.
21:06:34 <str4d> Anything else?
21:06:49 <hottuna_> ill have a look at tagging
21:07:48 <str4d> hottuna_: thanks. Leave get-involved/guides to me, as I've already started in there.
21:08:43 <str4d> dg: are you keeping an eye on the meeting (timeliness)?
21:09:02 <dg> oh, sorry
21:09:14 <dg> So we're done with website/
21:09:41 <dg> Crypto time :-D
21:10:16 <dg> Let me dig up the relevant topics
21:10:16 <dg> One moment
21:11:28 <dg> http://zzz.i2p/topics/1328 + http://zzz.i2p/topics/715
21:11:38 <iRelay> Title: zzz.i2p: Meeting [22nd January] (at zzz.i2p)
21:12:10 <dg> TL;DR: We need to be discussing which components of the i2p router need to be changed in order of priority (or as zzz put it, "to talk generally about which uses are more vulnerable than others"
21:12:10 <dg> )
21:12:17 <dg> (for the DSA change)
21:12:45 <dg> It's an apt time to discuss any other crypto changes that could be thrown in but right now, we should stick to what zzz suggested as it's a masssive rabbithole
21:12:52 <hottuna_> like noted in the tor cipher migration document we should strive to do changes where they are the most important and not necessarily the easiest
21:13:26 <dg> (https://gitweb.torproject.org/torspec.git/blob_plain/34ecac0fbac7f476bfcbf813767721fada62c17e:/proposals/ideas/xxx-crypto-migration.txt)
21:15:55 <hottuna_> in my mind the most important areas are those using potentially weak ciphers for longterm keys
21:16:39 <dg> hottuna_: I'm no crypto expert (and as such I'll stay out unless I know something) but aren't the longterm keys also the keys which could cause a flag day?
21:17:12 <hottuna_> changing most ciphers would cause a flag day
21:17:31 <dg> I was thinking all destinations being fucked
21:17:38 <dg> so yeah
21:17:41 <hottuna_> well basically
21:18:03 <hottuna_> i dont see a way around destinations being wrecked
21:19:03 <hottuna_> Im don't have a list of places where long-term keys are used
21:19:22 <hottuna_> but such a list and the corresponding cipher used should be created
21:21:04 <str4d> Agreed. We should also rank their perceived vulnerability.
21:21:11 <str4d> (This would make a good wiki page on Trac)
21:21:19 <hottuna_> yes.
21:22:02 <hottuna_> we should also create a list of ciphers that have been proven as safe (by the test of time) and are otherwise viable for us
21:22:17 <str4d> Section 2 of the Tor page basically applies to us as well.
21:22:20 <hottuna_> that list should include asymetric
21:22:55 <zzz> sounds good
21:23:11 <hottuna_> asymmetric* encryption, symmetric encryption, signatures and hmac ciphers that we trust
21:23:49 <zzz> how_cryptography page is a good reference
21:24:32 <hottuna_> str4d, did you start a wiki page or should I?
21:24:40 * str4d is doing so now
21:25:00 <str4d> /Crypto/CurrentSpecs sound alright?
21:25:09 <str4d> (For the summary table)
21:25:09 <hottuna_> sure
21:25:16 <zzz> DSA is a nice place to start analysis because it's easy to understand, and it's on the surface the weakest
21:26:15 <hottuna_> yes
21:27:01 <hottuna_> as for what is used where and what time periods which keys are used for I dont know much
21:28:56 <zzz> the OP on http://zzz.i2p/topics/715 has a list
21:29:03 <zzz> ~8 places we use DSA
21:29:05 <iRelay> Title: zzz.i2p: DSA 1024/160 Replacement (at zzz.i2p)
21:29:40 <hottuna_> the one with the longest validity is routerinfo?
21:30:23 <str4d> || '''Aspect/Location''' || '''Cipher used''' || '''Cipher details''' || ''' Perceived vulnerability''' || '''Comments'''
21:30:30 <str4d> Anything else that needs to go into the table?
21:30:30 <zzz> maybe dest. which isn't listed.
21:31:12 <zzz> theres both a dest key and a leaseset key I think the dest signs the leaseset and the leaseset key is unused
21:31:38 <hottuna_> str4d, validity period
21:32:24 <zzz> wouldnt be the end of the world to have a RI flag day but throwing out all 2500 in hosts.txt is another story
21:32:38 <str4d> Hmm... maybe the Perceived vulnerability / validity should be in a separate table then.
21:33:07 <zzz> datagrams is a problem, dests is a problem
21:33:22 <hottuna_> throwing out hosts is a huge issue. but it is also the most vulnerable key in my mind
21:34:37 <zzz> for each case we have to go farther though. not just how easy to break but what's the threat model / consequence.
21:35:08 <hottuna_> yes. maybe link to a separate page for each case?
21:35:26 <str4d> http://trac.i2p2.i2p/wiki/Crypto/CurrentSpecs now exists and has some basic content
21:35:33 <iRelay> Title: Crypto/CurrentSpecs – I2P (at trac.i2p2.i2p)
21:36:09 <zzz> and put that in perspective gven the size of the net, etc. e.g., we currently have a guy that claims he can shutdown an eepsite for 23 1/2 hours a day.
21:37:13 <hottuna_> christoph1, ?
21:37:25 <dg> Yikes.
21:37:28 <str4d> Mmm.
21:37:35 <dg> How does that work?
21:37:58 <hottuna_> eclipse attack on our floodfills
21:38:01 <christoph1> use enough precomputed routerinfos, put 10 bad nodes near the target hash block lookup
21:38:20 <lillith> why is it not 24 hours?
21:38:35 <christoph1> because midnight is a bit tricky
21:38:46 <christoph1> you can use another 10 to put in place for tomorrow
21:39:05 <christoph1> but there's still a period around the keyspace rotation where things are unstable
21:39:22 <lillith> so the router gets half an hour where the floodfills are uncertain?
21:39:33 <christoph1> (client can hit one of the good nodes by chance because it doesn't know all attackers jet
21:39:52 <str4d> The keys for the next day can be known in advance, so positioning malicious nodes could be planned in advance, no?
21:39:59 <christoph1> jep
21:40:22 <christoph1> still it seems around rotation it is somewhat unstable
21:40:49 <str4d> Anyway, this is somewhat off-track for this topic (sorry christoph1)
21:41:05 <christoph1> ack
21:43:08 <str4d> Okay, does anyone want to work on getting http://trac.i2p2.i2p/wiki/Crypto/CurrentSpecs filled out?
21:43:14 <iRelay> Title: Crypto/CurrentSpecs – I2P (at trac.i2p2.i2p)
21:43:26 <zzz> dg, please keep us on track, not drag us off it :)
21:43:42 <hottuna_> str4d, yeah. I just managed to log in :P
21:44:01 <str4d> Maybe we should quickly clarify what exactly we want on that page (my column headings are rather generic)
21:44:36 <dg> zzz: sory ;)
21:44:59 <str4d> First table: a summary of the crypto used in the router. Name, validity period, vulnerability... key length? Prime strength?
21:44:59 <zzz> m yfault too
21:45:48 <str4d> Second table: a list of every point in the router where crypto is used. Location and cipher name (of course). Usage details? What is important to know here?
21:46:27 <str4d> We can probably elaborate on separate pages for the second table if necessary (link the location name to a subpage)
21:47:41 <hottuna_> str4d, added subpage
21:48:06 <str4d> IMHO this should be a page that someone can glance at and understand the current state-of-play (whereas the site docs are the full specs)
21:48:32 <str4d> hottuna_: ah, I get what you mean by validity period now.
21:48:39 <hottuna_> :)
21:50:20 <str4d> hottuna_: there's already an entry for destinations - LeaseSet signing
21:50:29 <hottuna_> oh
21:50:29 <hottuna_> sorry
21:50:36 <str4d> (For the DSA part at least - I think you're thinking there of the encryption)
21:51:56 <str4d> Also, I'd call it "Security timescale" rather than "Validity period"
21:52:38 <hottuna_> yep
21:52:38 <zzz> FYI for everybody else - every RI and Dest has two keys, one for encryption and one for signing
21:53:11 <hottuna_> ok
21:53:11 <hottuna_> why?
21:53:32 <zzz> ElG was deemed far too slow for signing
21:54:44 <str4d> This might be a silly question, but how are the two keys "linked" verifiably?
21:55:23 <zzz> for both RI and Dest, the Hash covers both keys + the (usually null) Certificate
21:55:23 <hottuna_> a public key is derived from the private key
21:55:51 <zzz> change any of the 3 and you change the hash.
21:56:13 <str4d> Ah, k (you mean the Destination hash?)
21:56:23 <str4d> (i.e. the B64)
21:56:26 <zzz> yes
21:56:53 <str4d> Okay... the problem with upgrading the Destination crypto makes much more sense now...
21:56:59 <zzz> and for Dests, change any of the 3 and you need a new hosts.txt entry
21:58:34 <zzz> and (hint) non-null certs may be the path to upgrades w/ (partial) compatibility, i.e. not breaking gravity. That's what's covered further down in topic 715
21:59:39 <str4d> Yeah - that enables both to work alongside each other.
22:00:09 <str4d> But it still means that the end-to-end crypto for the old Destinations is untouched.
22:00:52 <str4d> The point where the Dest crypto key is most important is the leg between the OPEP and IBGW, right?
22:01:26 <zzz> not sure
22:01:53 <zzz> other complication is there used to be two layers of end-to-end crypto, one in the router and one in the client, and some keys are now unused
22:02:32 <zzz> ditto w/ signing keys... one was for LS revocation and is unused
22:02:46 <zzz> so that's another opportunity, maybe
22:03:29 <str4d> http://www.i2p2.i2p/how_intro seems to indicate that the ElGamal/AES+SessionTags is used for end-to-end router encryption.
22:04:37 <zzz> crypto is much harder to discuss than signing. theres the ElG wrapping the AES and the Tags, together with the DH exchange.
22:05:35 <str4d> Yes. But as far as e.g. LeaseSets go, we probably need to discuss both in tandem, no?
22:05:46 <zzz> I'd suggest not even trying to get into the crypto side today.
22:05:53 <str4d> Not today, no.
22:06:00 <zzz> maybe, maybe not
22:06:03 <str4d> So, back on topic *derp*
22:06:30 <zzz> you change one key, you change the hash. But as the Tor doc says, don't try to change everything just because you're changing one thing
22:06:33 <str4d> What is the issue with Datagram signing?
22:07:12 <zzz> it's using our signing algorithm, i.e. DSA. Which we use to sign everything. (including suds)
22:07:54 <zzz> which also isn't on the list on topic 715, and might be the longest-lived key of all
22:09:04 <str4d> Right, but the specific problem I'm guessing with Datagrams is ensuring that routers can still talk to each other
22:09:04 <str4d> ?
22:10:00 <zzz> right. change signing and you break all RI and LS lookup, and all signed end-to-end communication
22:10:51 <zzz> because almost everything is signed
22:11:41 <str4d> So really the only way to move forward with upgrading the signing algorithm is to ensure that every place it is used can handle multiple signing algorithms?
22:12:27 <str4d> The problem then becomes knowing what versions are supported by a router (and the partitioning problems from the Tor doc are relevant here).
22:12:30 <zzz> but then every dest would need two sets of tunnels, one for old and one for new, afaik
22:12:49 <zzz> there's two kinds of compatibility to consider.
22:13:19 <str4d> That's a good point >_<
22:13:42 <zzz> 1) "network" compatibility, i.e. can the RIs and LSs be stored and retrieved, can msgs get thru tunnels, even if the ffs or participants are down-rev;
22:14:21 <zzz> 2) end-to-end compatibility, can A talk to B. For that, seems like both A and B need to support the same things
22:15:43 <str4d> 2) is "easy" to handle for direct router-to-router communication, as the router versions are public knowledge. What about end-to-end communication?
22:17:24 <zzz> the other thing is an RI has a whole Properties in it, we can put whatever flags we want in there
22:17:27 <str4d> Where would a router need to look to determine if another router (such as an eepsite server) supports the new signatures?
22:17:30 <zzz> nothing like that for LS
22:18:01 <zzz> certs is the magic
22:18:48 <zzz> in a cert we can spec both crypto and signing algo, and store the extra bytes if it doesnt fit in the first 384
22:18:59 <zzz> again, that's the topic 715 stuff
22:19:53 <zzz> the cert has to start at byte 385 to not break 1)
22:20:54 <zzz> is that about enough for today? got out of this what you wanted?
22:21:09 <hottuna_> i think this is a beginning
22:21:34 <hottuna_> more specific issues and solutions cna be discussed and the wiki page used as an aid
22:23:50 <str4d> zzz: it's a good start - thank you =)
22:24:24 <zzz> lots of work ahead...
22:24:39 <str4d> Yes, but we have to start somewhere ^_^
22:24:54 <hottuna_> str4d, pushed tags for monotone.html
22:25:05 <zzz> I had one more topic for the mtg but only if welt welterde weltende is around
22:25:26 <str4d> hottuna_: the one under get-involved/guides? I'll drop the ones I'd started putting in then ^_^
22:25:37 <hottuna_> yes
22:26:00 <hottuna_> alright, are we done then?
22:26:11 <dg> I'd say so?
22:26:15 <str4d> I'd like to add a random point:
22:26:18 * dg had nothing to chime in with
22:26:21 <dg> not a crypto god
22:27:08 <str4d> I'd like to congratulate sponge on his efforts with Android - stock I2P now successfully runs on Android devices.
22:27:46 <str4d> And initial reports seem to indicate better performance and lower battery usage than I2P-Android
22:27:53 <hottuna_> that's quite the feat
22:28:04 <hottuna_> well done sponge
22:28:16 <hottuna_> i've gotta go now
22:28:23 <hottuna_> dg, will you strat the thread for next week?
22:28:27 <dg> spogne has done extremely well
22:28:56 <dg> Will do. Topics? Seems crypto needs to be a recurring topic for the next few weeks. :)
22:29:03 <dg> I should be here on time next week also
22:29:47 <str4d> If we can get the revamp tagged by then, we could potentially go live with the new site (though I would prefer to get actual translations in first)
22:30:18 <str4d> (Also depends on welterde being around)
22:30:25 <hottuna_> str4d, i think actual translations will take a very long time
22:30:52 <hottuna_> alright, nn ppl
22:30:59 <str4d> hottuna_: complete translations, yes. But there are already-translated pages (see www.i2p2/pages/translations) which would be quick to migrate.
22:31:07 <str4d> (For people who understand the language)
22:31:14 <str4d> o/ hottuna_
22:31:45 * str4d *baf*s the meeting closed.

18
i2p2www/meetings/logs/216.rst Normal file
View File

@@ -0,0 +1,18 @@
I2P dev meeting, January 22, 2013 @ 20:00 UTC
=============================================
Quick recap
-----------
* **Present:**
christoph1,
dg,
eche|on,
hottuna,
lillith,
RN,
str4d,
zzz
* **Next Meeting**
The next meeting is scheduled for Tuesday, January 29 @ 20:00 UTC (8:00PM)

96
i2p2www/meetings/logs/217.log Normal file
View File

@@ -0,0 +1,96 @@
20:08:14 <dg> so hi all
20:08:33 <dg> we didn't really talk much about the agenda but we have some big stuff to discuss anyhow so no big deal
20:08:55 <dg> first topic is the new website design and the blockers for getting it live
20:08:58 <dg> so i'll hand over to str4d
20:09:09 * dg passes metaphorical spotlight to str4d
20:09:20 <str4d> Hi!
20:09:27 <str4d> </ocd>
20:10:07 <str4d> Right, well I haven't had as much time over the last week to work on it, but the site *can* go live if desired.
20:10:26 <str4d> The blockers (that I see) are tagging the documentation for translations.
20:10:37 <str4d> Oh, and fixing URLs - that actually is a blocker.
20:10:48 <str4d> (I've been doing them at the same time).
20:11:45 <str4d> I've done the rest of the site - we just need to go through every page of the documentation and repeat what I've been doing: add translation tags to each paragraph/heading/list item/whatever, and check/correct the site-internal URLs.
20:12:29 <str4d> The other bit that I consider a blocker is getting the old translations migrated to the new format, i.e. going through the old faq_lang.html etc. pages and copying their text into the .po files as appropriate.
20:12:55 <dg> Translation tagging IMHO isn't a true blocker but something we need to finish soon, I don't know how you guys feel about going live without translations fully done
20:13:14 <str4d> (I've done 1.2 pages for de and 1 page for es as an example, but it really needs to be someone who understands the language, so that the copy-paste can be verified)
20:14:00 <str4d> (since there is no guarantee that the two individual pages have the same content)
20:15:52 <str4d> Design-wise, I'm starting to be more inclined to just leave the modified duck's theme there; it's had praise from quite a few people, including a designer. I'm personally still not happy with it (e.g. I find it a bit hard to read the long text passages in the documentation), but I wouldn't call it a blocker.
20:16:09 <dg> I like it more or less
20:16:40 <str4d> Oh - one bad thing currently is that the mobile CSS I added seems to not work on mobiles *derp*
20:16:52 <str4d> (Though it works quite nicely on a narrowed desktop browser)
20:19:26 <str4d> That's really about it for the site, as far as I can think of - it's functioning nicely, and most of the groundwork is in place.
20:19:41 <dg> I'll go check out for any broken links later
20:19:52 <dg> I found some previously
20:20:17 <str4d> dg: my test site is running the latest version, so you can check there.
20:20:32 <str4d> The only broken links should be in /lang/doc/*
20:20:35 <dg> let me dig up the ticket
20:20:38 <str4d> Every other page should be fine.
20:20:45 <str4d> It's ticket #807
20:20:52 <dg> saved me some time :)
20:20:59 <iRelay> http://trac.i2p2.i2p/ticket/807 - (accepted enhancement) - Revamp of website
20:20:59 <dg> want to revisit this next week then?
20:23:38 <dg> Found one.
20:23:46 <dg> http://vekw35szhzysfq7cwsly37coegsnb4rrsggy5k4wtasa6c34gy5a.b32.i2p/en/docs/how/networkcomparisons from http://vekw35szhzysfq7cwsly37coegsnb4rrsggy5k4wtasa6c34gy5a.b32.i2p/en/docs/how/garlic-routing
20:23:49 <str4d> Oh, and /lang/misc/* hasn't been looked at either.
20:23:56 <iRelay> Title: Garlic Routing - I2P (at vekw35szhzysfq7cwsly37coegsnb4rrsggy5k4wtasa6c34gy5a.b32.i2p)
20:24:06 <dg> Should be /en/comparison
20:24:12 <dg> sorry for pasting whole thing, won't do it again
20:24:15 <dg> eyerape
20:24:22 <str4d> dg: yep, that is a side-effect of my hypenating all the double-worded pages.
20:24:30 <str4d> Oh, that too.
20:25:09 <dg> /en/docs/how/garlic-routing has /en/docs/how/elgamalaes, should be /en/docs/how/elgamal-aes
20:25:16 <dg> i'll put them all into the ticket later
20:25:23 <dg> <+dg> want to revisit this next week then?
20:26:01 <str4d> I'm happy to. Anyone else want to comment? ^_^
20:27:18 <Meeh> http://meeh.i2p/viewmtn/viewmtn.py/
20:27:49 <dg> internal server error on http://meeh.i2p/viewmtn/viewmtn.py/branch/changes/i2p.i2p
20:28:45 <Meeh> yepp
20:28:48 <Meeh> working on it
20:28:55 <dg> I'll take it as a no, anyway
20:28:58 <dg> so onto .. crypto?
20:29:18 <str4d> One last question then: given the apparent level of apathy from people in here regarding copying over the old translations, would it be a better idea to just abandon them and upload the blank .po to Transifex?
20:29:44 <dg> Isn't Transifex reasonably active?
20:29:51 <str4d> The .po file will end up on Transifex anyway; I had just planned on it containing the old translations as a bit of a head start.
20:30:05 <dg> I'd say yes
20:30:48 <str4d> Any other votes?
20:32:38 <str4d> If we go with that, then I'll try and tag as many more pages as I have time for, and then I'll create .po files for the current website languages. Then someone with Transifex access can create an i2p.www subprojcet and upload them (KillYourTV?)
20:32:50 * psi reads scrollback
20:33:34 <psi> votes on a hackfest?
20:35:05 * psi reads more scrollback
20:39:01 <str4d> ...
20:39:04 * str4d doesn't like assuming, but assumes that no one has objections to that line of action.
20:39:04 <dg> too quiet >.>
20:39:04 <str4d> In that case, the old translations will be discarded from i2p.www.revamp (or maybe just left in the branch for now - if a translator really wants to they can find them for reference).
20:39:04 <psi> no formed opinion
20:39:14 <SkinSystem> hi, i heard k0e touches himself whilst watching granny pr0n
20:39:21 <dg> uh?
20:39:32 <psi> >.> #i2p-chat
20:40:45 <str4d> Right, that's it from me on the website.
20:49:59 <dg> well..
20:50:11 <dg> IDK if wait(str4d)
20:51:33 <dg> AFK, back soon
21:03:55 <hottuna> zzz, did you see the update of http://trac.i2p2.de/wiki/Crypto/CurrentSpecs ?
21:03:58 <iRelay> Title: Crypto/CurrentSpecs – I2P (at trac.i2p2.de)
21:07:36 <zzz> yeah you guys are doing a great job
21:08:13 <hottuna> do you think we are using asymmetric ciphers where they aren't needed?
21:08:23 <hottuna> and could be replaced by symmetric alternatives?
21:08:38 <zzz> I doubt we're using anything where it isn't needed
21:09:04 <hottuna> good
21:09:15 <zzz> the web page crypto section needs the symm/asymm crypto added ofc, right now it's only sigs
21:09:46 <hottuna> yeah. But one issue at a time
21:09:53 <hottuna> I think the signs are the most vulnerable anyway
21:10:07 <zzz> right. just as a one-liner placeholder is all I meant
21:20:19 <hottuna> zzz, sud signing should be fairly do-able as far as implementation and not having a flag day goes?
21:20:41 <hottuna> do-able as in possibly to change cipher for.
21:23:58 <dg> back
21:26:53 <zzz> i guess. we did su2 w/o incident, so why not su3, su4, ...
21:28:38 <hottuna> but before we should decide what cipher is the most viable
21:36:25 <dg> okay
21:36:33 <dg> since str4d_afk is gone.. uh
21:36:51 <dg> zzz, hottuna, do you want to discuss anything further or should we close the meeting? It seems like nobody else is around
21:40:29 <dg> a'ight
21:40:35 * dg bafs the meeting closed
21:41:11 <hottuna> thanks dg :)
21:41:55 <dg> str4d went MIA
21:45:03 <dg> anyway, np
23:00:32 <str4d> Sorry, connection went down and then I was AFK

14
i2p2www/meetings/logs/217.rst Normal file
View File

@@ -0,0 +1,14 @@
I2P dev meeting, January 29, 2013 @ 20:00 UTC
=============================================
Quick recap
-----------
* **Present:**
dg,
hottuna,
str4d,
zzz
* **Next Meeting**
The next meeting is scheduled for Tuesday, February 5 @ 20:00 UTC (8:00PM)

398
i2p2www/meetings/logs/218.log Normal file
View File

@@ -0,0 +1,398 @@
19:59:45 <dg> str4d: thoughts wrt IPv6 in mtg?
20:01:00 <str4d> Bring it up if you want. I know there are real issues that are holding that back, but it could be a good idea to verify what they currently are.
20:01:03 <str4d> I'll also be bringing up the website revamp.
20:01:14 <dg> Sweet.
20:01:22 <str4d> (So poke that in somewhere ^_^)
20:01:33 <dg> Will do :)
20:01:36 <dg> We'll put crypto last, if at all
20:01:44 <dg> I don't want to kill the movement but I don't know where we are going wit hit
20:01:55 <dg> I don't know enough about it to try .. push for it
20:05:01 <str4d> Yeah, last is a good idea, as long as the earlier sections of the meeting stay on track.
20:05:28 <str4d> And we already are in a much better place - the crypto review page has some interesting data (provided by "guest" so I have no idea who ^_^)
20:05:41 <dg> I was thinkign about that
20:14:16 <str4d> dg: ain't it meeting time now?
20:14:48 <dg> oh yeah
20:14:55 <dg> okay then
20:14:58 <darrob> i was going to ask that. i'm confused to see you *plan* a meeting at 08:05.
20:15:36 <dg> the agenda today is:
20:15:36 <dg> (0) - Hi!
20:15:36 <dg> (1) - Primary domain
20:15:36 <dg> (2) - IPv6 progress, blockers
20:15:39 <dg> (3) - Site revamp
20:15:42 <dg> (4) - Crypto
20:15:57 <dg> didn't realize the time
20:16:20 * str4d sends a CTCP TIME message to dg
20:16:34 * str4d watches it bounce off the I2P IRC tunnel.
20:16:41 <hottuna> Hello everybody!
20:16:48 <str4d> Hi!
20:16:55 <dg> We'll start off then?
20:17:25 <dg> So, (1) - primary domain
20:17:28 <dg> (also, hi)
20:17:46 <dg> The current primary domain (used in most places) is i2p2.de
20:17:49 <dg> There's a few problems with it
20:17:53 <str4d> www.i2p2.de
20:18:00 <str4d> (not i2p2.de)
20:18:00 <dg> (1) - It's not attractive or memorable.
20:18:11 <dg> (2) - SEO issues due to the country based TLD
20:18:29 <dg> (3) - Conflicts with other domains (we use geti2p.net on stickers, for example)
20:19:03 <eche|on> I do use www.i2p2.de on I2P stickers
20:19:06 <hottuna> (3) was my fault, it's just a more memorable domain
20:19:13 <dg> I'm proposing we use geti2p.net for all matters from now on but keep the old sites up with a redirect or some SEO magic to tell sites that we've moved without a full redirect
20:19:16 <str4d> (2) isn't a direct problem for Google at least.
20:19:54 <dg> i2p2.net is the better out of the pick though, there's ones like i2p-projekt.de that IMHO are too obscure for project usage
20:20:01 <dg> geti2p.net is kind of motivational too
20:20:16 <str4d> Their support page on geolocation says that geotargeting doesn't specifically affect page rank unless the search is limited by country.
20:20:40 <str4d> But then they also say that they use the geotargeting along with a heap of other things they look at to decide on regionalization.
20:20:43 <hottuna> I prefer geti2p.net, but I think we would need to have an official vote to change it
20:20:49 <KillYourTV> iRelay: web title i2p2.net
20:20:52 <dg> I concur, hottuna
20:20:55 <iRelay> Home - Contractor Safety Program Manuals - Downloadable and Turn-Key Custom-Written
20:21:23 <dg> i2p.net would be ideal, simple & the original domain but we can't have that for years. No promise we'll get it when it expires, either.
20:21:34 * KillYourTV was curious what was there, typo or not
20:22:01 <str4d> What about i2p.com and i2p.org?
20:22:16 <str4d> (Or one of the other generic TLDs?
20:22:16 <str4d> )
20:22:35 <str4d> Were they long-time registered?
20:23:10 <dg> Believe so. AFAIK, we were lucky to get .net when the project started but I'm not sure.
20:23:10 <eche|on> 1. i2p.net is registered til 2016 and will be grabbed by some machines after wards
20:23:21 <eche|on> 2. do NOT change the domain again. Thats a pain in the ass. really.
20:23:39 <eche|on> user credability is not to be raised with a domain change
20:23:42 <dg> It's not going to be an emergency change this time.
20:23:50 <dg> We already have 5 or so domains floating about
20:23:57 <eche|on> it was a looong 3 year phase to get www.i2p2.de into the game after i2p.net was dead
20:24:08 <str4d> eche|on: that's because you didn't have access to i2p.net
20:24:15 <dg> totally different circumstances
20:24:27 <dg> (and a totally different process)
20:24:30 <str4d> With access to both the old and new domains, either leave the content at the old one, or (better) do a 301 redirect.
20:24:45 <dg> it's not like we're recovering from such a major blow this time, just changing the default domain we refer to
20:25:50 <zzz> Back in the day, com/net/org were the best. But not true now, plus the whole issue of the U.S. can and does shut those down as they please. You really want to go back to a .net?
20:26:24 <dg> It's not like the U.S. can't get .de. If keeping things safe was as simple as being foreign, we wouldn't need I2P.
20:26:49 <hottuna> i2p.{aero|af|ag|ai|am|asia|bz|cat|cm|coop|cx|gg|gr|gs|gy.|hk|hn|ht|im|io|je|ki|la|lc|li|lt|lv|me|mg|mn|ms|mx|name|nf|nu|pt|re|tel|tv} are available
20:27:07 <dg> the "2" makes it unclean which bugs me
20:27:26 <dg> hottuna: i2p.io sounds groovy
20:27:41 <eche|on> I want i2p.xxx
20:27:44 <hottuna> the 2 is nonsensical for any newbie
20:28:11 <str4d> i2p.coop
20:28:14 <dg> I believe in the CCC talk, the guy speaking actually called i2p "i2p2"
20:28:17 <str4d> i2p.coup? ^_^
20:28:18 <christoph3> the list gets probably a lot shorter by removing inacceptably expensive domains ;-)
20:28:24 <zzz> dg you're greatly underestimating the difference in US-shutdown-ability of .de vs .net
20:28:24 <dg> "to i2p or not to i2p" (or something)
20:28:55 <str4d> zzz: i2p.net is out of the question anyway =)
20:29:09 <dg> zzz: really? I'm not so sure. If the US wanted it down, they could surely exercise their US-shutdown-abilities.
20:29:09 <hottuna> zzz, agreed. Also a valid point.
20:29:24 <dg> str4d: but geti2p.net is the alternative I proposed at least, so I guess he is referring to that
20:29:27 <hottuna> until now only us domestic domains have been taken down
20:29:57 <eche|on> even .at domains were not taken down though being spam and virus spreading domains
20:30:17 <christoph3> zzz is probably not talking about worst-case where the US really really really wants to shut it down
20:30:28 <christoph3> but just would like to
20:30:31 <str4d> Since this was started off by geotargeting issues: what are the actual generic TLDs?
20:30:31 <dg> Pretty sure if the US started shouting it was terrorist related, it could be taken down too
20:31:12 <hottuna> how are .org domains controlled?
20:31:12 <str4d> Ooh, just found a list.
20:31:12 <eche|on> str4d: a lot of them,also including some local russian, asian, chinese... one
20:31:19 <str4d> https://support.google.com/webmasters/bin/answer.py?hl=en&answer=1347922
20:31:34 <str4d> Google treats all of those ^ as gTLDs.
20:31:38 <eche|on> http://de.wikipedia.org/wiki/Top-Level-Domain
20:31:41 <iRelay> Title: Top-Level-Domain – Wikipedia (at de.wikipedia.org)
20:31:49 <christoph3> .org was verizon just like .com and .net wasn't it?
20:31:52 <eche|on> i2p.int would be nice.
20:31:59 <eche|on> and it would be possible.
20:32:18 <eche|on> or a .nato?
20:33:05 <hottuna> int sounds nice, but is controlled by IANA which is us government
20:33:27 <str4d> hottuna: aren't all TLDs eventually controlled by IANA?
20:33:46 <str4d> (excluding the new raft of TLDs that companies are allowed to buy now)
20:34:04 <dg> new proposal: we buy .i2p
20:34:11 <hottuna> more or less directly? .int being directly?
20:34:11 <hottuna> :P
20:34:14 <hottuna> i2p.i2p?
20:34:21 <dg> project.i2p
20:34:36 <dg> download.i2p
20:34:39 <darrob> hottuna: obviously the main page would have to be www.i2p2.i2p. :)
20:34:54 <hottuna> :P
20:35:12 <str4d> But I just went to all this effort of double-tagging the entire website revamp >_<
20:35:59 <hottuna> str4d, I did some translations. About 70k words left though.
20:37:10 <str4d> There were about 70k words to start with =P
20:37:13 <darrob> how immediate is this US shutdown threat though? instead of worrying about it we could just wait and see. worst case: we'd have to go piratebay-crazy with proxies and media attention and all.
20:37:36 <eche|on> so far it is a no brainer, as tor is still alive.
20:37:47 <dg> oh. yeah. Tor, of course.
20:38:19 <darrob> dg: go on?
20:38:38 <hottuna> realistically we will still have the mirrors, so a specific tld being lost isnt much of an issue
20:38:53 <dg> darrob: about Tor??
20:38:56 <dg> darrob: as for the discussion, I think we're still talking
20:39:03 <dg> So, do we want to put this down to a vote?
20:39:18 <dg> If so, what would be the options?
20:39:18 <dg> (and when?)
20:39:22 <hottuna> i would like geti2p.net to be an option
20:39:28 <dg> We could do it now if you guys wanted, I don't know if we have enough people
20:39:35 <KillYourTV> +1 for new TLD. Often they're a steal at $185,000 for registration then $6,250 each quarter. Surely we have enough funds available for that.
20:39:35 <hottuna> next meeting?
20:39:50 <dg> Hell yeah!
20:39:53 <dg> hottuna: sure
20:39:53 <hottuna> and announce it on the forums
20:39:53 <darrob> dg: sorry, never mind. your tor comment was not clear to me but it's irrelevant now.
20:40:04 <eche|on> KillYourTV: ^^ oerfect solution, but hosting is another topic.
20:40:35 <hottuna> would anyone like to propose a change to a domain other than geti2p.net?
20:40:38 <darrob> KillYourTV: we'd just need to sell our bitcoins at the right time.
20:41:00 <dg> hottuna: I have none but then again, I don't know the other ones we have.
20:41:55 <str4d> i2p.io sounded quite good, but .io is still country-specific
20:42:42 <dg> ideal world: we would have i2p.org
20:43:34 <hottuna> ok, then we have 2 alternatives. i2p.io and geti2p.net
20:43:45 <str4d> i2p.int?
20:43:48 <str4d> i2p.info?
20:43:56 <str4d> i2p.jobs =P
20:43:59 <dg> is i2p.info available?
20:44:18 <dg> i2p.transformers?
20:44:30 <eche|on> i2p.onion
20:44:45 <hottuna> i2p.info is taken
20:45:00 <str4d> i2p.co - that's a country one that is treated (by Google at least) as generic.
20:45:23 <hottuna> i2p.co is taken
20:45:30 <str4d> i2p.me?
20:45:37 <dg> i2p.me is cool.
20:45:58 <str4d> (another gccTLD)
20:46:09 <darrob> and cool is gimmicky if i may say so. i'd rather see i2p2.org or something.
20:46:09 <hottuna> suggestions: i2p.io i2p.int i2p.me geti2p.net
20:46:28 <dg> I want to drop the "2".
20:46:31 <str4d> darrob: I'd rather see i2p.something
20:46:53 <str4d> Why wasn't i2p.de chosen originally? Taken?
20:47:01 <str4d> s/originally/at the time/
20:47:04 <iRelay> str4d meant: Why wasn't i2p.de chosen at the time? Taken?
20:47:12 <hottuna> i2p2.org taken
20:47:16 <hottuna> i2p.de taken
20:47:31 <str4d> Mmm.
20:47:50 <KillYourTV> i2p.ispent185000dollarsandalligotwasthisstupidtld
20:48:00 <str4d> I agree that we don't want something gimmicky.
20:48:07 <str4d> It needs to make sense.
20:48:22 <dg> cool == workable in this context
20:48:24 <dg> (too)
20:49:13 <hottuna> i2p.int is available btw
20:50:52 <hottuna> dg, will you add i2p.io i2p.int i2p.me geti2p.net to the alternatives of the vote?
20:51:18 <dg> hottuna: Yup. Making a note of it now.
20:51:27 <hottuna> danke
20:54:10 <str4d> Anything else here, or next topic dg?
20:54:40 <dg> was writing up the topic
20:54:40 <dg> yeah, next
20:54:43 <dg> Alright, IPv6!
20:54:53 <dg> From what I understand, welt* was working on this and then.. ??? happened
20:55:04 <dg> I don't know the blanks, maybe someone can fill them in
20:55:21 <dg> There's a bounty right now for IPv6 support in I2P
20:55:42 <dg> It's <20>100 EUR
20:55:45 <hottuna> http://zzz.i2p/topics/109
20:55:48 <dg> http://www.i2p2.i2p/bounty_ipv6
20:55:55 <iRelay> Title: zzz.i2p: IPV6 TODO (at zzz.i2p)
20:55:56 <iRelay> Title: Bounty I2P IPv6 native - I2P (at www.i2p2.i2p)
20:56:04 <dg> I think after/alongside the crypto doup, this is important
20:56:43 <dg> I could do with zzz or welt*
20:56:57 <dg> It'd be fantastic if welt* took it up again
20:59:01 <dg> Meeh is seemingly OK with helping
20:59:12 <dg> If zzz is going to, he needs SSH
20:59:42 <Meeh> I can provide zzz with SSH
20:59:45 <Meeh> to a dev server
21:00:09 <Meeh> we should also start using mailing lists now that it's up :)
21:00:20 * dg nods
21:00:44 <dg> I'm not aware of mailing list etiquette but I can start posting there
21:00:51 * dg crickets
21:01:54 <str4d> Meeh: is it I2P-internal as well, or should we use @i2pmail.org?
21:02:17 <dg> i2p-internal I assumed but..
21:02:20 <Meeh> @i2pmail.org for now... haven't figured out howto bring it inside i2p yet
21:03:16 <str4d> Meeh: you'd need some sort of rewriting. Maybe talk to postman?
21:03:36 <Meeh> yepp, I will
21:04:34 <dg> Should we move on or wait?
21:04:49 <str4d> Back to IPv6, zzz provided a nice summary in the dev forum post
21:05:20 <str4d> It sounds like the next step is to compare and contrast the three possible options.
21:05:43 <str4d> Do we want to do that here/now? In the dev thread? On a Trac wiki page like the crypto one?
21:06:17 <dg> Trac, I say
21:06:52 <Meeh> dg, str4d :
21:06:52 <Meeh> 22:05:27 <postman> if the mailinglist server resides in the internet
21:06:52 <Meeh> 22:05:40 <postman> all people use their @i2pmail.org adress for signup
21:06:52 <Meeh> 22:06:14 <postman> teh mails will be automatically rewritten when transferred to / from internet
21:06:52 <Meeh> and i2p
21:07:34 <Meeh> 22:07:02 <postman> ( i think it would be best to have a ml server public)
21:07:56 <Meeh> 22:07:16 <postman> so people without mail.i2p adresses can subscribe
21:08:12 <Meeh> so I suggest current setup, since many that's not on i2p can join the list
21:08:37 <str4d> Meeh: alright.
21:08:54 <dg> <+dg> Should we move on or wait?
21:08:54 <str4d> Oh, and also, if the domain name changes, what happens to the list email?
21:09:04 <dg> CNAME the MX?
21:09:14 <str4d> Ah, true.
21:09:25 <str4d> dg: move on, I think. I don't think we are going to have in-depth IPv6 discussions today,.
21:09:40 <dg> Yeah.. seems that way. I'm dubious about crypto too? :|
21:09:51 <dg> Okay, website revamp!
21:10:07 <dg> str4d has been chugging away at tagging and eliminatin the blockers
21:10:14 <dg> I think we're getting closer to being able to put it live
21:10:17 * dg hands mic to str4d
21:10:50 <str4d> Okay, the site is edging closer and closer to being ready.
21:11:02 <str4d> I've nearly tagged the entirety of docs/
21:11:36 <str4d> I'm just working on docs/spec/* now, and then docs/discussion/* (though I'm not sure if that one is needed?)
21:12:03 <str4d> docs/spec/ is somewhat hard to fully tag though, as there is a lot of <pre>-formatted stuff which translating will just break.
21:12:36 <str4d> URL-wise, aside from docs/spec/* and docs/discussion/* the entire site should have working URLs.
21:12:51 <str4d> If someone wants to test that (manually or automated) that would be good.
21:13:37 <str4d> Structurally, I don't think there is much more needed for the site - can anyone think of anything?
21:14:55 <hottuna> No
21:15:06 <hottuna> and if flaws are found they can be fixed
21:15:24 <dg> I don't think there's any.. let alone critical
21:15:31 <str4d> Design-wise, I'm still not entirely happy, but that's not critical.
21:15:53 <str4d> The mobile CSS I added *does* have flaws, but it is far easier to use/read than the desktop CSS.
21:16:27 <str4d> So I think the only thing left is to wait for the translations to start coming in.
21:17:10 <str4d> (I'd like to at least have some of the main pages translated before launch, so that Google doesn't think I'm lying to it in the sitemap.xml)
21:19:52 <str4d> And here's an idea: If/when we decide on a new domain, we launch the new site there, so we can test it live, and then inform Google/users via 301 redirect that the new site is the new domain?
21:21:03 <hottuna> the translation is a very big job
21:21:03 <hottuna> one that will take a lot of time
21:21:03 <hottuna> why not update sitemap.xml when translations are available?
21:22:56 <str4d> hottuna: the sitemap is somewhat recursively generated.
21:22:59 <hottuna> Would that be better?
21:23:02 <hottuna> ok
21:23:28 <str4d> It's to do with the whole reflang thing that Google requires.
21:23:51 <str4d> Every page in the sitemap has to list as a lang link every other language page, including itself.
21:24:06 <str4d> And that has to be repeated for each lang variant of a page.
21:24:25 <str4d> I've already split it up into a /sitemapindex.xml and /lang/sitemap.xml pages.
21:24:41 <dg> Would Google really care?
21:24:41 <hottuna> but we're always going to lack translations for some languages?
21:24:45 <str4d> But we also have a Language dropdown, which Google is going to find.
21:25:23 <str4d> hottuna: the list of Languages on the revamp is taken from the current site - there are partial translations for all of them.
21:25:26 <str4d> dg: no idea really.
21:25:59 <dg> It could take a long time to get us translations
21:26:06 <str4d> The sitemap language thing is there so that Google doesn't use the pages themselves to determine language (as in that case, every language variant would be called "English" currently)
21:26:33 <str4d> Hmm... I'll go check whether Google defers to the sitemap entirely, or if it still uses its own observations.
21:26:39 <str4d> I could just be needlessly worrying.
21:26:47 <hottuna> dg: I think it will, historically that has been the case for most languages.
21:27:02 <hottuna> maybe str4d, but ultimately this is your call
21:27:06 <str4d> And if having "lang-specific" pages which aren't translated isn't going to affect our pagerank, then it's not a blocker.
21:29:44 <dg> Hm, ok
21:31:22 * dg waits
21:32:56 <str4d> Okay, just did some reading, and it looks like Google defers to the sitemap
21:33:25 <str4d> Since one usage is translating only the template of the page, and leaving the content in a single language (like forum posts).
21:33:44 <str4d> So, translations are not a blocker, and can be updated as we go.
21:33:51 <str4d> In which case, the site could go live, now.
21:33:51 <dg> Any other bugs?
21:33:58 <dg> Oh. Sweet.
21:34:01 <str4d> (Well, now + time it takes to check all remaining URLs)
21:34:04 <dg> Anyone got any views on putting it live?
21:37:48 <hottuna> alright. if now is an option, i would vote for now
21:37:59 <dg> me too
21:38:34 <str4d> Putting it live requires some help from whoever is running the server (welterde IIRC)
21:38:45 <str4d> And the mirrors.
21:39:20 <hottuna> alright, at next sighting of the welterde/weltende, let's ask?
21:39:34 <str4d> Propagating back to i2p.www is simple. But the server ops need to run the setup script and then configure their WSGI setup to use the virtualenv.
21:40:15 * str4d will prop i2p.www onto the revamp now so that merges etc. can be handled now.
21:42:53 <dg> Huzzah
21:43:12 <hottuna> Yep
21:46:43 <str4d> Okay, so that's the revamp done. dg?
21:46:50 <dg> str4d: Yup.
21:47:21 <dg> I was avoiding moving on
21:47:22 <dg> due to the IPv6 discussion (or lack of..)
21:47:25 <dg> Wanna call it a night?
21:47:25 <dg> I think it's the best choice
21:47:33 <hottuna> Sure
21:47:59 <str4d> Yeah, 1.75 hours can be enough for today
21:48:10 <dg> Heh.
21:48:10 <dg> Alright,
21:48:13 * dg bamfs the meeting closed
21:49:21 <dg> http://zzz.i2p/topics/1343
21:49:28 <iRelay> Title: zzz.i2p: Meeting [6th February] (at zzz.i2p)
21:50:31 <hottuna> the next meeting is the 12/2 not 6/2
21:50:42 <hottuna> wut
21:50:42 <hottuna> ok
21:50:49 <hottuna> or am I confused
21:51:47 <dg> my mistake
21:52:10 <dg> ;_;
21:55:01 <iRelay> <str4d@freenode> ping weltende
23:39:54 <zzz> um, was it just me, or did the meeting totally skip the final decision that changing our URL was definitely a good idea?
23:40:30 <zzz> I saw some pros and cons on changing and then 'lets post a vote for which new one to pick'
23:40:50 <iRelay> <jenkins@kytv> Project Syndie HSQLDB2 build #9:SUCCESS in 15 sec: http://eotfca7qexthbireor6ae7g4hbj5hwuhe4gkzxdx3l3g2t5gzn7q.b32.i2p/job/Syndie%20HSQLDB2/9/
23:41:54 <zzz> nor did I see anybody volunteer to do all the work req'd to implement such a change
23:45:25 <zzz> or is the (unstated) plan to vote on the best name first, and only then decide whether to do it and how?
23:45:29 <zzz> ^^ dg
23:45:55 <dg> Correct, sorry if it wasn't clear
23:46:03 <dg> I'll outline more next time
23:46:49 <zzz> might be me, I was just skimming.
23:46:55 <zzz> so was there a decision made or not?
23:47:37 <dg> We're having a vote next week
23:48:03 <zzz> a vote on what? whether to change, or what to change it to?
23:48:33 <dg> What to change it to
23:48:56 <zzz> so was there a decision made or not on whether to change it at all?
23:48:59 <dg> We could certainly have people abstain though
23:49:06 <dg> which would be a "no change"
23:49:42 <dg> We were in agreement that a change would be nice but unable to agree on what *to*
23:49:52 <zzz> is the decision to change at all a) already made or b) deferred until after the vote?
23:50:14 <dg> b)
23:50:54 <zzz> that wasnt clear to me either from the mtg or from http://zzz.i2p/topics/1343
23:51:01 <iRelay> Title: zzz.i2p: Meeting [12th February] (at zzz.i2p)
23:51:18 <str4d> zzz: What I gathered is that, of the people present at the meeting, the majority were for a change.
23:51:25 <zzz> and to me, imho "would be nice" is far far short of "it's definitely a good idea AND we have volunteers to do the work"
23:51:28 <str4d> But, not all relevant people were at the meeting.
23:51:38 <dg> (e.g. you)
23:52:04 <dg> welt wasn't there either, I may be forgetting someone but was anyone else missing?
23:52:37 <zzz> sure, I was in and out, my bad. But my review of the logs is the pros were saying 'would be nice' and the cons had significant objections.
23:53:04 <str4d> zzz: I don't recall significant objections. Care to elaborate?
23:53:07 <zzz> but dg says it hasn't been decided yet,
23:53:25 <str4d> The only major one was eche's about linkage loss, but that is easily solved with a 301 redirect
23:53:28 <dg> The big one was from ech, saying it would cause big issues
23:53:31 <dg> and that was't true
23:53:45 <dg> s/was't true/debunked
23:53:52 <str4d> (So can't really be fairly compared to the i2p.net -> i2p2.de migration, which *was* an issue)
23:53:58 <zzz> <eche|on> 2. do NOT change the domain again. Thats a pain in the ass. really.
23:53:58 <zzz> <eche|on> user credability is not to be raised with a domain change
23:54:09 <zzz> there's the US-controlled issue I raised
23:54:19 <zzz> there's the amount of work required, which nobody raised
23:54:30 <str4d> "pain-in-the-ass" was referring to above.
23:54:46 <dg> There's links on the main site (how much of that can be done with regex?), some in the source code (same again?)
23:54:46 <dg> What else?
23:54:53 <zzz> there's the SEO issue, which you apparently think is managable, but I'm not so sure
23:55:16 <str4d> User credability - that is rather wishy-washy and IMHO changing from i2p2.* to i2p.* is going to make it easier for users.
23:55:16 <zzz> there's links in our docs and code
23:55:35 <str4d> Links on the site are all relative, so no issues there.
23:55:35 <dg> s/http://www.i2p2.de/http://geti2p.net/
23:55:38 <iRelay> Title: Not found - I2P (at www.i2p2.de)
23:55:38 <dg> s/http://www.i2p2.de/http://geti2p.net/g
23:55:41 <iRelay> Title: Not found - I2P (at www.i2p2.de)
23:55:50 <str4d> Docs and code and SEO - again, a 301 redirect solves all that.
23:56:05 <dg> 301 is perm redirect, yeah?
23:56:08 <str4d> (The docs and code of course should be updated, but the links would not break)
23:56:11 <str4d> Yes.
23:56:21 <dg> If so, I'd assume that's what it was created for - that kind of thing anyway.
23:56:21 <str4d> It's what the revamp has for all the old page locaions
23:56:24 <zzz> if it's not decided yet, (and dg says it isn't) then we can talk about it next week. my reading from the logs and zzz post was that it was decided, but apparently not
23:56:42 <str4d> e.g. /debian.html --[301]--> /en/download/debian
23:56:57 <str4d> zzz: yep, not decided yet, this was just getting the ball rolling.
23:57:04 <zzz> ok
23:57:19 <zzz> also a lot of this requires welt's tme which is in short supply
23:57:24 <str4d> Nothing is going to happen without informed consensus.
23:57:44 <zzz> ok great. guess I misread the logs
23:57:55 <str4d> Yeah, which is why I'd like to sort out the revamp and push it live at the same time.
23:58:29 <str4d> Or rather, I'll keep on working on site stuff and whenever welt turns up we can push it live.
23:59:15 <dg> I'll edit the post to make it clear
23:59:18 <zzz> maybe. trying to change 6 things at once increases the risk too. I'd recommend incrementalism
00:00:16 <zzz> also beware of becoming overly google-focused as there are other search engines too, who may behave differently
00:00:31 <zzz> or at least more slowly
00:00:34 <str4d> zzz: mmm, which is why I suggested putting the revamp on the "new" url to test it.
00:00:42 <str4d> And then apply the 301.
00:00:53 <str4d> (Assuming the url is changed)
00:01:30 <str4d> Yeah, I'll do some checking up of other search engines, but Google seems to use reasonably standard technology (e.g. the sitemap stuff)
00:02:05 * str4d is currently checking and fixing old/broken URLs.
00:03:24 <dg> Edited my post. http://zzz.i2p/topics/1343
00:03:31 <iRelay> Title: zzz.i2p: Meeting [12th February] (at zzz.i2p)
00:05:36 * dg waits for ech to come back
00:05:43 <dg> That'll be tomorrow, right?
00:08:17 <zzz> re: incrementalism, e.g. you may wish to consider using 302s for a few days on the live site until you're sure it's all good before switching to 301s, as you really don't want to mess up 301s
00:08:28 <str4d> Mmm, yep.
00:08:40 <str4d> Do you mean for the legacy pages as well?
00:09:16 <zzz> maybe. something to think about
00:09:43 <str4d> I'm pretty sure all the old pages are good (I manually checked most of them)
00:09:53 <dg> As for domains, https://twitter.com/i2p/status/298485275053666304 - I can't be the only one who thinks that's messy
00:10:00 <str4d> But probably worth checking again at some stage.
00:10:28 <dg> str4d: No reply from zab.
00:12:36 <iRelay> <jenkins@kytv> Starting build #43 for job Cobertura Coverage Reports
00:13:41 <zzz> not saying it's a bad idea to switch (and if we do, geti2p.net is the only choice). Just that we do it purposefully if we do. You made a proposal and held discussion but never called for a decision, or even identified how or when a decision would be made.
00:14:59 <dg> <+hottuna> I prefer geti2p.net, but I think we would need to have an official vote to change it
00:15:02 <dg> <+dg> I concur, hottuna
00:15:13 <dg> only geti2p.net?
00:15:20 <dg> So fuck the vote and just have a y/n?
00:15:27 <dg> that's fine but tell me now :p
00:16:10 <zzz> so the vote to change it and the vote on what to change it to got conflated
00:17:01 <zzz> "only choice" just imho
00:17:12 <dg> Everyone attending was up for changing it (ech was the only one who disagreed and he.. went quiet.. I don't know what that means, no?) so I skipped that step
00:20:04 <zzz> well, run the meeting how you want, but not sure you want to discount people who only object once, especially if you don't have an obvious call for the end of discussion and a decision
00:20:13 <dg> no, i see your opint
00:20:16 <dg> point*
00:20:26 <dg> i do need to outline things more

19
i2p2www/meetings/logs/218.rst Normal file
View File

@@ -0,0 +1,19 @@
I2P dev meeting, February 5, 2013 @ 20:00 UTC
=============================================
Quick recap
-----------
* **Present:**
christoph3,
darrob,
dg,
eche|on,
hottuna,
KillYourTV,
Meeh,
str4d,
zzz
* **Next Meeting**
The next meeting is scheduled for Tuesday, February 12 @ 20:00 UTC (8:00PM)