This is the recent traffic on the #SPF-council IRC channel on
irc.pobox.com. Anyone may join the channel, but only council
members can talk.
| --- Sat Nov 25 18:59:17 UTC 2006 --- |
| 18:59 | <Julian> | hi MarkK |
| 18:59 | <MarkK> | good evening |
| 19:00 | <Julian> | <shew> Julian, do you want to add something into the agenda discussing our reaction to the BITs suggestions/requests? |
| 19:00 | <Julian> | shew: Can you clarify, please? (Excuse if I'm being a bit slow...) |
| 19:04 | <SDGathman> | Hello everyone. |
| 19:05 | <Julian> | hi SDGathman |
| 19:05 | <MarkK> | hello stuart |
| 19:05 | <Julian> | SDGathman: Did you read my mail regarding X-Chat for MacOS X? |
| 19:06 | <Julian> | (I sent it just ~15min ago.) |
| 19:06 | <SDGathman> | No, I'll look now. |
| 19:06 | <SDGathman> | I'm using gaim via tightvnc via a Java application. |
| 19:07 | <Julian> | *cough* |
| 19:08 | <SDGathman> | I can't install anything |
| 19:11 | <SDGathman> | I'm fine with this solution. |
| 19:11 | <Julian> | OK. |
| 19:12 | <Julian> | I just sent our agenda for today. |
| 19:12 | <Julian> | It still has to show up in the spf-council archives... *sigh* |
| 19:12 | <Julian> | Oh, there it is: http://archives.listbox.com/1943/200611/0044.html |
| 19:13 | <Julian> | Any wishes for additions or removals? |
| 19:13 | <shew> | I'm fine. |
| 19:14 | <MarkK> | I'm fine, too |
| 19:14 | <Julian> | SDGathman? |
| 19:14 | <Julian> | Oh, I forgot one thing: 3.5. news items for the upcoming news announcement. |
| 19:16 | <Julian> | SDGathman: Are you there? |
| 19:16 | <SDGathman> | yes |
| 19:16 | <SDGathman> | agenda is fine |
| 19:16 | <Julian> | OK, so let's start. |
| 19:16 | <Julian> | 1. Website transition |
| 19:17 | <Julian> | Here's the deal: I'm going to be very busy next week, so I doubt I'll manage to complete the new wizard and "why?" pages by coming Friday. |
| 19:18 | <Julian> | However, Mail::SPF is effectively finished now and I really just need to do those two pages. Therefore I think it would be best if we postponed the website switch for one week. |
| 19:18 | <Julian> | Any comments? |
| 19:19 | <SDGathman> | Can't we switch with old wizard and why? |
| 19:19 | <MarkK> | I see no reason to postpone 1 week extra |
| 19:19 | <Julian> | Well, yes, we could. We would have to link to the relevant parts of the old site, though, and those parts still have links to other parts of the old site. |
| 19:20 | <MarkK> | not to posrtpone, I meant |
| 19:20 | <shew> | I'm fine either way. However, we can always switch before the announcement. |
| 19:20 | <shew> | I'll finally have some time to work on the site myself. (I was unexpectedly busy these last two weeks.) |
| 19:20 | <Julian> | This is not about the announcement (I could live with not mentioning Mail::SPF in it), it's more about the implications of having to link to parts of the old site. |
| 19:21 | <shew> | Then I'm okay with a week's delay. |
| 19:21 | <Julian> | That would be December 8th. |
| 19:22 | <Julian> | (or we could even say Dec 6th or 7th) |
| 19:23 | <shew> | If you're gonig to make a motion on it, i'd suggest the 10th. |
| 19:23 | <shew> | That way we'll have a bit of weekend-time, and cushion. |
| 19:23 | <Julian> | SDGathman, MarkK? |
| 19:24 | <SDGathman> | Disappointed, but if having new wizard is essential for switch, then I guess there has to be a delay. |
| 19:25 | <SDGathman> | Did you say mod_python was available on web server? Should we have a race for the wizard? I'm very busy too, but if the wizard is the holdup ... |
| 19:25 | <Julian> | I don't think that having the _new_ wizard is absolutely essential, but I think it would avoid trouble with the old site, and it would make a better impression on those who actually visit the new site in reaction to the announcement. |
| 19:26 | <Julian> | Well... |
| 19:26 | <MarkK> | yes, better wait until we have the new wizard in place, too |
| 19:26 | <Julian> | Yeah, we could do a race, but any non-Perl solution would probably have a slight disadvantage, because it would somehow have to fit into the site's framework (or at least into the design). |
| 19:26 | <SDGathman> | There are a lot of complaints about braindamage in the old wizard. |
| 19:27 | <Julian> | Yes, I think Alex even stopped handling tickets about the old wizard. |
| 19:27 | <Julian> | :-/ |
| 19:28 | <shew> | There is an option of not having a wizard on the site for a week. |
| 19:28 | <shew> | Well, that would require edits on multiple pages, and we'd probably miss some. |
| 19:28 | <MarkK> | that might give us a wave of critique |
| 19:30 | <Julian> | There's one other thing that might justify postponing the news announcement for some days in particular. |
| 19:30 | <MarkK> | which is? |
| 19:31 | <Julian> | I think we need to debate what to do about the next council elections (whether to do any at all, etc.) and include that in the news announcement. |
| 19:32 | <Julian> | Wayne said he wasn't sure whether electing another council next year would make much sense (we could end up with ~10 people electing 5 representatives, or something like that). |
| 19:32 | <Julian> | I don't think he's right, but I think that needs to be discussed properly. |
| 19:33 | <Julian> | Of course postponing the announcement doesn't imply that we'd have to postpone the website switch (but, well, there's a reason for that, too -- see above). |
| 19:33 | <Julian> | What do you think? |
| 19:35 | <shew> | As far as elections, we have no choice in the matter. Any change of election rules should (imho) be voted on by the same people who would vote for the council. |
| 19:35 | <MarkK> | I think wayne may be right; but we could check the 'animo' for it in discuss |
| 19:36 | <Julian> | I think you're both right, and we should at least ask on spf-discuss what people think. |
| 19:36 | <shew> | Julian: Are you suggesting disbanding the council? |
| 19:36 | <shew> | That is, that we shouldn't really have one? |
| 19:36 | <Julian> | shew: No. I think we should make "No other candidates -- Disband the council" as an option like last year. |
| 19:37 | <Julian> | So if less than 5 get elected, then the council would be disbanded. |
| 19:37 | <shew> | Ahh. Okay |
| 19:38 | <Julian> | Actually, it read "None of those remaining, prefer to hold new elections". |
| 19:39 | <Julian> | Not sure why it didn't read "None of those remaining, prefer to disband the council". Doing another election round usually doesn't make much sense -- people rarely change their mind within a very short amount of time. |
| 19:41 | <shew> | Well, regardless of the website delay or not...election work has to go on. |
| 19:43 | <Julian> | OK, then I'll make this a motion: Postpone the website switch and news announcement. The old and new websites shall be switched around on or before 2006-12-10. The switch shall be publicly announced on the same day the website switch happens. |
| 19:44 | <shew> | 19:43u seconded |
| 19:44 | <Julian> | Votes for 19:43u? |
| 19:45 | <SDGathman> | 1943u ok |
| 19:45 | <MarkK> | 1943u: yes |
| 19:45 | <Julian> | 1943u: yes |
| 19:46 | <shew> | 1943u: yes |
| 19:46 | <Julian> | OK. I'll do my best to expedite the switch. |
| 19:46 | <Julian> | 2. The test suite |
| 19:47 | <Julian> | I think we should release the test suite soon. However we shouldn't release it with a version number of "1.0" or something but use a date-based numbering scheme instead. E.g., "2006.12" or something. |
| 19:47 | <Julian> | That way we can make another release without looking bad. |
| 19:47 | <SDGathman> | Are we going to vote on the ip6/ip4 controversy? |
| 19:47 | <Julian> | Comments on the general idea of releasing the test suite soon and on version numbering? |
| 19:48 | <Julian> | SDGathman: I'm not going to make a motion. But we can vote if someone else does. |
| 19:48 | <Julian> | In any case, I am NOT prepared to debate the "ip6"/IPv4 issue tonight. :-) |
| 19:48 | <MarkK> | yes, I was just gonna ask: what about the ipv4/ipv6 issue (ambiguity) we duscussed last week? |
| 19:49 | <Julian> | I'm just too tired to have any real chances of convincing anyone tonight. :) |
| 19:49 | <Julian> | But if someone wants to make a motion, feel free. |
| 19:49 | <shew> | I like the "2006.11" type scheme. |
| 19:50 | <MarkK> | me too |
| 19:51 | <Julian> | Of course that shouldn't relieve us from trying to get it right the first time. ;-) |
| 19:51 | <Julian> | Actually, I think the test suite is very good and correct already. |
| 19:51 | <Julian> | I just haven't had the time so far to manually review all the "OK"s that the Mail::SPF test driver spits out. |
| 19:52 | <Julian> | And there might still be additions to the test suite later that test more stuff. |
| 19:52 | <SDGathman> | To summarize ip4/6 controversy: shouild ip4 connections on an ip6 interface be considered ip4 only (not match any ip6 mech) or both ip4 and ip6 (match approprate ip6 as well as ip4 mechs)? |
| 19:52 | <Julian> | Good summary. |
| 19:53 | <SDGathman> | I added a simple test for pasting TXT records properly. Will commit tomorrow night. (This meeting is the only work I'm doing from friends house). |
| 19:53 | <Julian> | The difference between both views is that (1) IPv4/IPv6 address spaces are either distinct or IPv6 space is a super set of IPv4 space, and (2) what practical implications are there of doing it one way or the other? |
| 19:54 | <Julian> | But I suggest that we don't debate this right now. |
| 19:54 | <SDGathman> | Ip4 policies are unaffected (that don't have any ip6 mechs). |
| 19:54 | <shew> | I agree with Julian. This isn't appropriate for a council meeting. |
| 19:55 | <SDGathman> | Ok, then should that test be removed for 2006.11 release? |
| 19:55 | <shew> | (I have seen the email controversy, but I haven't followed all the arguments through, and I won't be able to do so during the space of a meeting.) |
| 19:55 | <MarkK> | The longer I think about it, the more I lean towards treating it as just an ipv6 address; regardless of the fact that it is a superset of te ipv4 space, if the address appears in ipv6, it seems reasonable to treat it as ipv6. |
| 19:55 | <MarkK> | but we can discuss it another time :) |
| 19:55 | <SDGathman> | Or at least downgraded to accepting both results? |
| 19:55 | <Julian> | SDGathman: If we really want to release the test suite in 2006-11 (as opposed to 2006-12), then the test should probably be removed or disabled, yeah. |
| 19:55 | <shew> | If that's the only holdup for a 2006.11 release, I would say to remove it for the moment yes. |
| 19:57 | <SDGathman> | MarkK: "just ipv6" is not an option. Possible interpretations of RFC are "ip4 only" or "both ip4 and ipv6" for the ::FFFF:ip4 connections on an ip6 interface. |
| 19:57 | <Julian> | But I would probably defer to the "ip6 never matches IPv4" crowd anyway because I can see the implications it has on existing implementations. It's just one of those things that, if it was changed in a future revision of SPF, would confuse people a lot. |
| 19:57 | <SDGathman> | I will modify test to accept both interpretations for the time being and commit tomorrow night. |
| 19:58 | <Julian> | OK |
| 19:58 | <Julian> | So then we can release rfc4408-tests.yml 2006.11 within the next few days, right? |
| 19:58 | <SDGathman> | What do existing implementations do? Which implementations handle ip6? |
| 19:59 | <Julian> | To be honest, I have no idea. |
| 19:59 | <Julian> | Mail::SPF supports "ip6". |
| 20:00 | <Julian> | BTW, I haven't been able to complete the YAML->XML converter and the RNG schema validator for the test suite yet. |
| 20:00 | <Julian> | But that may not be a priority. |
| 20:00 | <Julian> | (RNG = Relax NG, a simple schema language) |
| 20:01 | <Julian> | So then we can release rfc4408-tests.yml 2006.11 within the next few days, right? |
| 20:01 | <SDGathman> | Julian: yes |
| 20:02 | <Julian> | OK. Any other thoughts on the issue before we move on? |
| 20:02 | <SDGathman> | Where will it go on web page? |
| 20:02 | <MarkK> | Stuart, yes, you're right: so I guess we'd have to accept both ipv6 and ipv4 then; (hmm, not all that happy with that either); |
| 20:02 | <SDGathman> | MarkK: then you are on Julian's side. |
| 20:03 | <Julian> | SDGathman: I think it should have its own page (I think it already does), which would be linked from the "Implementations" and "Specifications" page (and perhaps other places). Sounds good? |
| 20:03 | <Julian> | http://new.openspf.org/Test_Suite |
| 20:03 | <MarkK> | What does Mail::SPF do in that regard? |
| 20:03 | <Julian> | Currently it considers ::ffff:n.n.n.n both an IPv4 and an IPv6 address. |
| 20:04 | <Julian> | (which can match both "ip4:n.n.n.n" and "ip6:::ffff:n.n.n.n") |
| 20:04 | <MarkK> | Ok |
| 20:05 | <SDGathman> | I would consider matching what most implementations do. |
| 20:05 | <Julian> | Does something need to be decided here right now, or should we move on? |
| 20:06 | <MarkK> | or make it a switch, perhaps? |
| 20:06 | <MarkK> | ok, lets move on |
| 20:06 | <SDGathman> | Move on |
| 20:06 | <Julian> | 3. Mail::SPF |
| 20:06 | <Julian> | I added this to the agenda only because it's one of our three current sub-projects. The status is as follows: |
| 20:08 | <Julian> | Mail::SPF itself is 99.5% done (it just needs some documentation fixes and very few other adjustments). |
| 20:08 | <Julian> | I adopted `spfquery` from Mail::SPF::Query and rewrote most of it to work with Mail::SPF instead. That's finished, too. |
| 20:08 | <MarkK> | Personally. I look rather forward to Mail::SPF. |
| 20:09 | <Julian> | I still need to convert `spfd`. That's non-essential. |
| 20:09 | <Julian> | I'm going to make one last pre-release (2.000_003) to CPAN within the next few days and ask for documentation review and some testing on spf-devel and perhaps spf-discuss. |
| 20:10 | <Julian> | Then the final 2.001 release won't be far behind. (2.000 was the initial "release" uploaded by Shevek in 2005, which was in a really, really bad shape). |
| 20:10 | <Julian> | OK, that's it for Mail::SPF right now. If you have any urgent questions, type "..." quickly. |
| 20:11 | <MarkK> | I'll test it, and if there's any issue with it, I'll let you know. |
| 20:12 | <Julian> | (2.000 didn't work right. A lot was missing, and some things weren't designed in a very flexible way. And the entire thing wasn't even dreaming of complying with RFC 4408. It really was an alpha release.) |
| 20:12 | <Julian> | OK, no questions, so let's move on- |
| 20:12 | <Julian> | 4. How to react to BITS' desire for some kind of receiver policy? |
| 20:13 | <Julian> | shew asked to discuss this issue briefly, and in a private channel because he wasn't sure what level of privacy was still required by BITS. |
| 20:13 | <Julian> | Shall we go to #spf-private for a few minutes? |
| 20:13 | <MarkK> | ok |
| 20:13 | <shew> | ok |
| 20:36 | <MarkK> | back |
| 20:36 | <SDGathman> | back |
| 20:36 | <shew> | In #spf-private we discussed the need for a BCP-type document for those concerned with receiver policy. I volunteered to edit such a BCP document or white paper, soliciting input from spf-discuss. |
| 20:37 | <Julian> | Awesome! Thanks, shew! |
| 20:37 | <Julian> | I think we can skip the remaining agenda item (5. The RFC errata) and conclude the meeting. |
| 20:37 | <shew> | The BCP/white paper(s) would include recommendations on both publishing records and implementing spf checking. |
| 20:37 | <Julian> | Motion: conclude this meeting. |
| 20:37 | <shew> | 20:37: seconded. |
| 20:37 | <MarkK> | 2037u: yes |
| 20:37 | <shew> | 2037u seconded |
| 20:38 | <SDGathman> | 2037u yes |
| 20:38 | <shew> | 2037u yes |
| 20:38 | <Julian> | Votes on 2037u? |
| 20:38 | <Julian> | 2037u: yes |
| 20:38 | <Julian> | The meeting is concluded. Thanks to everyone! |
| 20:38 | <Julian> | SDGathman: Please extend our apologies to your hosts! |
| 20:38 | <SDGathman> | No problem. |
| 20:39 | <SDGathman> | bye everyone |
| 20:39 | <Julian> | SDGathman: Bye! |
| 20:41 | <MarkK> | Julian, when will you upload MAIL::SPF to CPAN? |
| 20:42 | <Julian> | There's already a 2.000_002 pre-release, but I think it still has some problems, such as a few missing features (but IIRC, it already conforms to the test suite). |
| 20:42 | <Julian> | I'll try to get the 2.000_003 pre-release out tomorrow. |
| 20:42 | <MarkK> | great! Then I can finally get rid of the old one. :) |
| 20:43 | <Julian> | MarkK: --> #spf |
| 20:43 | <MarkK> | yes |