summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorMatt Turner <mattst88@gmail.com>2020-10-18 14:43:43 -0700
committerMatt Turner <mattst88@gmail.com>2020-10-18 14:45:28 -0700
commitc09ed935b714dc54b6e7f700414b1f6cd8085136 (patch)
tree9443d5460b8a2a42a77729db5d26ec216e4a7eed
parentSummary for 20200614 meeting. (diff)
downloadcouncil-c09ed935b714dc54b6e7f700414b1f6cd8085136.tar.gz
council-c09ed935b714dc54b6e7f700414b1f6cd8085136.tar.bz2
council-c09ed935b714dc54b6e7f700414b1f6cd8085136.zip
Log for 20200719 meeting.
License: CC-PDM-1.0 (raw IRC log, not copyrightable) Signed-off-by: Matt Turner <mattst88@gmail.com>
-rw-r--r--meeting-logs/20200719.txt291
-rw-r--r--meeting-logs/20200719.txt.asc10
2 files changed, 301 insertions, 0 deletions
diff --git a/meeting-logs/20200719.txt b/meeting-logs/20200719.txt
new file mode 100644
index 0000000..13a9eeb
--- /dev/null
+++ b/meeting-logs/20200719.txt
@@ -0,0 +1,291 @@
+12:00 <@ mattst88> | hello!
+12:00 <@ gyakovlev> | meeting time! who's chairing?
+12:00 <@ gyakovlev> | !proj council
+12:00 <+ willikins> | (council@gentoo.org) dilfridge, gyakovlev, mattst88, slyfox, ulm, whissi, williamh
+12:01 <@ dilfridge> | right-e-o
+12:01 <@ mattst88> | I can, I suppose. looks like only one thing on the agenda
+12:01 <@ mattst88> | roll call?
+12:01 * | Whissi is here
+12:01 * | slyfox here
+12:01 * | WilliamH here
+12:01 * | dilfridge here
+12:01 * | gyakovlev here
+12:01 <@ ulm> | actually I've a proposal for the agenda
+12:01 * | ulm here
+12:01 * | mattst88 here
+12:02 <@ mattst88> | excellent, looks like everyone's here
+12:02 <@ mattst88> | ulm: okay, what's up?
+12:02 <@ ulm> | can we have an item "Constitute the new Council"?
+12:02 <@ mattst88> | sounds good to m e
+12:02 <@ ulm> | with 3 sub-items
+12:02 <@ ulm> | a) Decide on time of meetings. The previous council had its meetings on the 2nd Sunday of every month at 19:00 UTC
+12:02 <@ ulm> | b) Vote for continuing last Council's workflow considering sending call for agenda items (2 weeks in advance), sending the agenda (1 week in advance) and have the meeting focussed, i.e., have major discussions on -project ML prior to the meeting
+12:02 <@ ulm> | c) Appoint chairmen for this term's meetings
+12:03 <@ slyfox> | sounds good
+12:03 <@ mattst88> | yeah, sounds good to me. any objections?
+12:03 <@ Whissi> | not from me
+12:03 <@ WilliamH> | not from me.
+12:03 <@ dilfridge> | nope
+12:03 <@ mattst88> | great. next topic?
+12:04 <@ ulm> | vote on a) and b)?
+12:04 <@ mattst88> | sure
+12:04 * | mattst88 votes yes for both
+12:04 <@ gyakovlev> | b) goes well with mgorny's question (which should be on agenda irc.
+12:05 * | slyfox yes
+12:05 * | WilliamH votes yes for a
+12:05 <@ dilfridge> | yes for both
+12:06 * | Whissi votes yes for a and b
+12:06 <@ mattst88> | gyakovlev: ah, I wasn't expecting a vote on mgorny's proposal, given that it was an RFC
+12:06 * | gyakovlev yes for a
+12:06 * | ulm yes for a) and b)
+12:06 <@ gyakovlev> | not a voting, a discussion
+12:06 <@ dilfridge> | if we vote all yes for b then it's moot anyway
+12:06 <@ mattst88> | I'd suggest that we accept b) for now and then discuss mgorny's proposal
+12:06 <@ mattst88> | b) is just our baseline
+12:08 <@ gyakovlev> | I abstain voting on b.
+12:08 <@ mattst88> | should we move on to discussing mgorny's proposal in that case?
+12:09 < ulm> | mattst88: can we vote again please, on a and b separately? otherwise this is going to be a mess
+12:09 <@ mattst88> | ulm: sure. votes for (a)
+12:09 * | mattst88 yes
+12:09 <@ dilfridge> | yes
+12:09 * | gyakovlev yes
+12:09 * | ulm yes
+12:09 * | Whissi yes
+12:09 * | WilliamH yes
+12:09 * | slyfox yes
+12:10 <@ mattst88> | votes for (b)
+12:10 * | mattst88 yes
+12:10 * | slyfox yes
+12:10 * | ulm yes
+12:10 * | Whissi yes
+12:10 * | gyakovlev abstain
+12:10 * | WilliamH abstain
+12:11 <@ Whissi> | Could you say a word for my understanding why you two abstain?
+12:11 <@ WilliamH> | mgorny's proposal would completely change b) I like parts of his proposal, but we still have to meet.
+12:11 * | dilfridge yes
+12:12 <@ gyakovlev> | I don't like deferring decision making to a monthly meeting, sometimes it does not make any sense and "back in the day developers waited for half a year for glep editor" is not a good reason to make someone and the distro wait for a decision which will be voted all yes anyway.
+12:13 <@ Whissi> | Thanks for elaborating.
+12:13 <@ gyakovlev> | we need some kind of fast-paced decision approval mechanism
+12:13 <@ WilliamH> | Meetings shouldn't have a lot of discussion I guess, but the meetings are required by glep 39 which has to be changed by a full dev vote.
+12:13 <@ mattst88> | I don't think voting for (b) prevents us from changing later
+12:13 < Whissi> | mattst88: ACK, that was my understanding.
+12:13 <@ ulm> | otoh there's good reason to keep things structured and have the decisions in the meeting's log and summary where they can easily be looked up
+12:14 <@ gyakovlev> | same could be said about bugs
+12:14 <@ WilliamH> | gyakovlev++
+12:14 <@ gyakovlev> | I'm not against structure
+12:14 <@ gyakovlev> | or rules
+12:14 <@ gyakovlev> | I think some rules do not make sense
+12:14 <@ gyakovlev> | like we have a proposed change
+12:15 <@ gyakovlev> | it WILL be accepted
+12:15 <@ dilfridge> | we already have voting via bug if something is urgent
+12:15 <@ gyakovlev> | with 7-0 vote
+12:15 <@ gyakovlev> | why hold it back for 3 more weeks&
+12:15 <@ gyakovlev> | ?
+12:15 <@ WilliamH> | Every other body in Gentoo requires bugs.
+12:15 <@ WilliamH> | We could still mention the bugs in meetings.
+12:15 <@ WilliamH> | so they end up in meeting logs.
+12:16 <@ WilliamH> | We already do that actually.
+12:16 <@ dilfridge> | that is somewhat a contribution of the department of redundancy department
+12:16 <@ gyakovlev> | we can have a formal process of making it a quick vote and long vote(meting) at least.
+12:16 <@ slyfox> | I think in the past we just did not close such off-meeting voting bugs
+12:16 <@ mattst88> | yeah, we could make a point of noting any council decisions made in bugs during the meeting so that it's in the meeting logs
+12:16 <@ dilfridge> | right
+12:17 <@ mattst88> | slyfox: that is to say, that the council wasn't very responsive to such bugs?
+12:17 <@ gyakovlev> | yeah mattst88, that's good suggestion, that way who looks at meeting logs only still can see suggestion. we kinda already do it in open bug section, we can just have "bug decision" item on agenda.
+12:18 <@ WilliamH> | Well, we would just have "open bugs" on the agenda if we move every item to a bug.
+12:18 <@ slyfox> | my impression was that bugs do get processed reasonably fast. but it's just my implression.
+12:18 <@ mattst88> | gyakovlev: right, that seems like a reasonable solution to ensuring that decisions are still recorded in the logs
+12:18 <@ WilliamH> | The agenda would just have "open bugs" and "open floor".
+12:19 <@ dilfridge> | so when and where do we discuss things then?
+12:19 <@ WilliamH> | Or another way to put it is, do we need things on the agenda that aren't in bugs?
+12:19 <@ dilfridge> | I mean, bugzilla is many things but it's not a discussion medium
+12:20 <@ gyakovlev> | dilfridge: the bug itself or ML seems to be ok.
+12:20 <@ gyakovlev> | bug can be a little disorganized yeah.
+12:20 <@ dilfridge> | we as in the 7 council members
+12:20 <@ dilfridge> | list is usually way to messy
+12:20 <@ mattst88> | WilliamH: right. I would say that we don't need specific agenda topics if everything is listed in bugs, but I would suggest that we simply list the bugs and their titles in the meeting agenda for clarity
+12:20 <@ ulm> | are we trying to change our workflow, just for the sake of change?
+12:21 <@ dilfridge> | ulm: it seems so
+12:21 <@ gyakovlev> | dilfridge: so is IRC, i often loose track of discussion here. I would not on ML.
+12:21 <@ ulm> | seems to me that it mostly works
+12:21 <@ dilfridge> | if we had something really urgent in the past, we usually managed to handle it quickly
+12:21 <+ mgorny> | you can always say that X works if people are forced to use it
+12:21 <@ dilfridge> | and for everything else, having a few days to think about it is a good thing
+12:21 <@ mattst88> | I think the core of the proposal is to move discussion out of the IRC council meetings
+12:22 <@ dilfridge> | also not everyone is around here 24/7
+12:22 <@ gyakovlev> | ulm: I want to change workflow so we don't introduce artificial delays and don't block people who is ready to do work.
+12:22 < WilliamH> | mattst88: I think so too.
+12:22 <@ dilfridge> | ok so here's a compromise proposal
+12:23 <@ dilfridge> | up to now, we had the rule that we did bug decisions only when it was really urgent
+12:23 <@ ulm> | gyakovlev: can you provide an example of the last council term where we had such an artificial delay blocking people?
+12:23 <@ gyakovlev> | dilfridge: yeah, but only urgent. voting should be async by default, and montly only if we defer.
+12:23 <@ gyakovlev> | ulm: yeah, stable arch profile change.
+12:23 <+ sam_c> | ad-hoc meetings can be done when needed though?
+12:23 <+ sam_c> | or is the idea that this should happen, but does not
+12:23 <@ dilfridge> | we change that so we can *also* do bug decisions when everyone agrees (i.e. no discussion needed, unanimous)
+12:24 <@ ulm> | dilfridge: +1
+12:24 <@ dilfridge> | if something is really good and uncontroversial then it can be done fast
+12:24 <@ dilfridge> | if it needs discussion, it needs discussion
+12:24 <@ gyakovlev> | yes!
+12:24 <@ gyakovlev> | and if it's good and there's no reason to block it
+12:24 <@ Whissi> | Wondering if that is really new/a change but... +1 ;)
+12:24 <@ WilliamH> | Well, the way we would do that is put our items in bugs.
+12:24 <@ gyakovlev> | it should be done now
+12:25 <@ gyakovlev> | and not on next meeting
+12:25 <@ WilliamH> | anyone can open a bug and assign it to council.
+12:25 <@ WilliamH> | Once that happens, we decide if it needs discussion.
+12:25 <@ Whissi> | But how do we do it in practice? Example:
+12:26 <@ Whissi> | A creates bug for council. Some council members already start voting... one council members raises concerns and want discussion. What's now?
+12:26 <@ gyakovlev> | so rough idea yeah: a bug, assessment to delay or vote now. if assessment shows it needs more discussion or work - we defer it. but if it's ready - we just vote it.
+12:26 <@ dilfridge> | Whissi: delayed to meeting
+12:26 <@ dilfridge> | as long as the decision is not unanimous it needs to be done in a meeting
+12:26 < veremitz> | add a flag?
+12:27 <@ Whissi> | wfm
+12:27 <@ gyakovlev> | there should be 3 stages imo. 1 - accepted. 2 - ENEEDINFO, discussion. 3- deferred to meeting.
+12:27 <@ dilfridge> | let's not overcomplicate things
+12:27 <@ gyakovlev> | sometimes it should be possible to go from 2 to 1 without going all the way to 3.
+12:27 <@ gyakovlev> | it's not complex, it just common sense =)
+12:28 <+ mgorny> | Whissi: why would you start voting before discussing it?
+12:28 <+ mgorny> | the whole point is to have discussion on the ml
+12:28 <@ mattst88> | It seems that this doesn't address what I thought was the core of the proposal -- to have the discussion outside of the IRC meeting
+12:28 <@ Whissi> | mgorny: As outlined, anyone can file bugs for council all the time.
+12:29 <@ WilliamH> | So: 1) a bug gets opened and assigned to council. 2) it gets announced on the ml by a council member. 3) discussion may or may not happen on the ml.
+12:29 <+ mgorny> | Whissi: this statement of fact doesn't address any point of the proposal
+12:29 <+ mgorny> | Whissi: no, 1) discussion is started on the ml, 2) bug gets opened -> Council member need to join the discussion
+12:29 <@ WilliamH> | 4) we go from there to decide whether to vote on the bug or in a meeting.
+12:30 <@ WilliamH> | mgorny: Hmm, that might work too.
+12:31 <@ mattst88> | Evidently we need to discuss this/these proposals more. should we try discussing them in the manner mgorny suggests, as a trial run? :)
+12:31 <+ mgorny> | 3) the upcoming meeting's chair could be made responsible for establishing when the discussion is finished and voting should start
+12:31 * | dilfridge get some trial rum
+12:31 <@ Whissi> | mgorny: But what will happen if someone create a bug first? Or if there is a discussion, someone loses patient and creates a bug to 'force' a decision (=progress)? Can this end discussion? Many questions... but dilfridge answer was clear IMHO. Any concerns will block fast vote and delay for next real meeting which should be OK.
+12:32 <+ mgorny> | Whissi: then you tell that someone to start a discussion
+12:32 <@ WilliamH> | It wouldn't bother me personally if someone made a bug first.
+12:33 <@ dilfridge> | I think the main point we are (willing to address) addressing here is to handle uncontroversial things fast.
+12:33 <@ dilfridge> | Like, a finished GLEP, where the implementation is ready, and we just need to stamp it final.
+12:33 <+ mgorny> | and avoid creating more work for people by sticking to inconvenient rules
+12:33 <@ gyakovlev> | dilfridge: that was my suggestion/proposal mostly. mgorny's is a bit different.
+12:33 <+ mgorny> | like waiting 3 weeks for council item collection mail
+12:33 <+ mgorny> | which sometimes gets forgotten
+12:34 <@ mattst88> | many legislative bodies pass legislation by voice vote (without registering individual votes) and it just takes one person to say 'no'
+12:34 <@ Whissi> | I think the point raised was, "Whenever we don't agree, something can't take the fast road"... and that's something we all can live with, given that there wasn't a critical problem we didn't handled well in last 3 years...
+12:34 <@ mattst88> | which dilfridge's proposal addresses
+12:34 <@ dilfridge> | Yes, and I would like to point out that there is a big difference between "discussion among council" and "open discussion"
+12:34 <@ Whissi> | ACK
+12:35 <@ dilfridge> | having a good chance to do the former doesnt exclude the latter, but still has separate value
+12:35 <@ mattst88> | fair point
+12:35 <@ gyakovlev> | dilfridge++
+12:36 <@ mattst88> | okay, I suggest that we vote on dilfridge's proposal. dilfridge: would you restate it for the vote?
+12:36 <@ dilfridge> | on it
+12:36 <@ WilliamH> | I do think we can substitute the call for agenda items with: If you want the council do discuss or vote on something, open a bug assigned to council."
+12:37 <@ WilliamH> | Then we handle the items from there.
+12:37 <@ mattst88> | WilliamH: I would worry a little that many on gentoo-project@ would not notice a new bug assigned to council and the open discussion would be limited as a result
+12:38 <@ mattst88> | Perhaps a mail to gentoo-project@ should accompany the new bug
+12:38 < veremitz> | much easier to trace
+12:38 <@ dilfridge> | Proposal: "In addition to the already existing rule that urgent decisions can be handled between council meetings via a bug, we will also handle uncontroversial decisions between council meetings via a bug. Anyone can file the bug following discussion on the mailing lists. If the council members do not agree unanimous, discussion and vote during a council meeting needs to take place."
+12:38 <@ dilfridge> | s/unanimous/unanimously/
+12:38 * | gyakovlev yes
+12:39 * | Whissi yes
+12:39 * | dilfridge yes
+12:39 * | slyfox yes
+12:39 * | ulm yes
+12:39 * | mattst88 yes
+12:39 * | WilliamH yes
+12:39 <@ mattst88> | excellent
+12:39 <@ gyakovlev> | beautiful ^
+12:40 <@ mattst88> | mgorny: let's continue the discussion of your proposal on the mailing list. okay with you?
+12:40 <+ mgorny> | sure
+12:40 <@ WilliamH> | I like the idea of dropping the call for agenda items and all agenda items being in bugs.
+12:41 <@ mattst88> | okay, the only bug that I see that needs council involvement is bug 729062
+12:41 < willikins> | mattst88: https://bugs.gentoo.org/729062 "Services and Software which is critical for Gentoo should be developed/run in the Gentoo namespace"; Gentoo Council, unspecified; CONF; jstein:council
+12:41 <@ WilliamH> | That's probably ml discussion though.
+12:41 <@ dilfridge> | I dont even know what he means with namespace
+12:41 --> | prometheanfire [~promethea@gentoo/developer/prometheanfire] has joined #gentoo-council
+12:41 --- | ChanServ sets modes [#gentoo-council +v prometheanfire]
+12:41 <@ Whissi> | I can explain that.
+12:42 <@ ulm> | dilfridge: infra, I guess
+12:42 <@ mattst88> | I'm loathe to tell developers who have done the work to create useful software that because they've created useful software that they now must move it to gentoo infrastructure
+12:42 <@ Whissi> | We all know pkgcore is used in CI.
+12:42 <@ mattst88> | I expect that the topic of the bug is pkgcore/pkgcheck/nattka
+12:42 <@ Whissi> | CI will break mirrors on errors.
+12:42 <@ Whissi> | Image there is no a problem or something should change.
+12:42 <@ Whissi> | In *theory* there are doormans who could block something like that
+12:43 <@ dilfridge> | well luckily git is a *distributed* version control system
+12:43 <@ Whissi> | because they are the only people with access
+12:43 <@ dilfridge> | so every clone can replace the original in an emergency
+12:43 <@ Whissi> | Having this in Gentoo namespace, i.e. https://gitweb.gentoo.org/ would allow anyone with access (like most Gentoo devs) to push changes
+12:43 <@ Whissi> | dilfridge: No. You can't do that in an emergency. Like you would have to change deployment and stuff like that
+12:43 <@ ulm> | dilfridge: if the source is available
+12:44 <@ ulm> | in the past, we also had the problem that the source was not public at all
+12:44 <@ ulm> | e.g. previous stablebot
+12:44 <@ Whissi> | So this should be done from the beginning so there isn't a problem later in case this will become a problem.
+12:44 <@ mattst88> | ulm: true, and we should not allow that to happen in the future
+12:44 <@ Whissi> | At least that is the idea of that motion.
+12:44 <@ gyakovlev> | ulm: stablebot is a perfect example of why non-open-source but free FW should not be used.
+12:45 <@ gyakovlev> | but the proposal speaks about open-source projects, so it does not apply really.
+12:45 <@ gyakovlev> | s/FW/SW/
+12:45 <@ slyfox> | there was a request to move discussion to the ML in the bug. it did not happen. shoud it?
+12:45 <+ mgorny> | this would be worth discussing *if* infra deployed gitlab or equivalent
+12:45 <@ dilfridge> | yes
+12:45 <@ Whissi> | This wasn't about gitlab...
+12:45 <@ dilfridge> | well, I guess we won't do ee anyway
+12:45 <+ mgorny> | getting repos on infra right now is painful
+12:46 <@ Whissi> | Asking infra to create a new git repo is painful?
+12:46 <@ mattst88> | Whissi: he's saying that he wouldn't object to moving things to git.gentoo.org if git.gentoo.org were GitLab, presumably because he likes having built-in CI, etc
+12:47 <@ Whissi> | You are free to mirror to github so you can use whatever github offers you.
+12:47 <@ dilfridge> | err
+12:47 <@ ulm> | we cannot rely on github which is nonfree
+12:47 <@ ulm> | and F-rated :/
+12:48 <@ WilliamH> | Someone was looking at gitea but I don't know what's going on with it.
+12:48 <@ Whissi> | Like I remember old Ruby web stuff Gentoo was/is still using... when graaf helped here... he added CI (travis) and therefore mirrored everything to GitHub. Worked without any problems.
+12:49 <@ Whissi> | But origin is still on git.gentoo.org...
+12:49 <@ WilliamH> | Whissi: as ulm said, github is non-free so it can never be primary due to our social contract.
+12:49 <@ Whissi> | WilliamH: Read what I just wrote. It isn't the primary...
+12:49 <@ Whissi> | I agree that primary should be git.gentoo.org...
+12:49 <@ mattst88> | I think we're getting massively off track with the mentions of github/non-free/social contract
+12:49 <@ dilfridge> | the soci
+12:49 <@ dilfridge> | oops
+12:50 <@ slyfox> | so what do we do with the bug? :)
+12:50 <@ gyakovlev> | ask to paint
+12:50 <@ Whissi> | Maybe we can delay it until next month and do some more discussion on ml and come up with a formulated motion for next meeting.
+12:50 <@ WilliamH> | Whissi++
+12:51 <@ mattst88> | works for me
+12:51 <@ dilfridge> | wfm
+12:51 <@ mattst88> | do we have any other topics?
+12:51 <@ slyfox> | who will create the mail thread?
+12:51 <@ Whissi> | slyfox: I'll do tomorrow
+12:51 <@ slyfox> | thank you!
+12:51 <@ WilliamH> | Should we figure out chairs for the term?
+12:51 <@ mattst88> | oh, do we want to pick meeting chairs?
+12:51 <@ Whissi> | (or after I got reply from jstein... )
+12:52 * | ulm could do september and october
+12:52 <@ Whissi> | But I'll take care of that
+12:52 <@ slyfox> | is Aug already on mattst88?
+12:52 <@ mattst88> | I suppose if I do next month then I'll have done my two? :)
+12:52 <@ mattst88> | sure, I'll do August
+12:52 * | gyakovlev can take nov & dec
+12:53 <@ slyfox> | i can do jan / feb
+12:53 * | dilfridge march & april
+12:53 * | WilliamH can do May and June
+12:53 * | Whissi can do Jul
+12:53 <@ ulm> | heh :)
+12:53 <@ mattst88> | cool, done
+12:53 <@ slyfox> | \o/
+12:53 <@ mattst88> | I'll update the wiki
+12:54 <@ mattst88> | other topics?
+12:54 <@ mattst88> | open floor?
+12:54 <@ slyfox> | sounds good
+12:55 <@ mattst88> | motion to adjourn? :)
+12:55 <@ slyfox> | \o/
+12:56 <@ dilfridge> | \o/
+12:56 <@ Whissi> | \o/
+12:56 <@ WilliamH> | :-)
+12:56 <@ ulm> | :)
+12:56 <@ gyakovlev> | \o/
+12:56 <@ mattst88> | \o/
+12:57 <@ mattst88> | we're done :)
+12:57 <@ mattst88> | thanks all!
+12:57 <@ Whissi> | Thank you for chairing.
+12:57 < gyakovlev> | mattst88: don't forget to save the log, thanks for chairing.
+12:57 <@ ulm> | thanks for chairing
+12:57 <@ mattst88> | twas nothing!
+12:57 <@ Whissi> | Wait until you start working on the summary ;-)
diff --git a/meeting-logs/20200719.txt.asc b/meeting-logs/20200719.txt.asc
new file mode 100644
index 0000000..0922348
--- /dev/null
+++ b/meeting-logs/20200719.txt.asc
@@ -0,0 +1,10 @@
+-----BEGIN PGP SIGNATURE-----
+Version: GnuPG v2
+
+iNUEABYKAH0WIQSzlptPDvl9ch5jjr2cglpmBdQLvgUCX4y2mF8UgAAAAAAuAChp
+c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0QjM5
+NjlCNEYwRUY5N0Q3MjFFNjM4RUJEOUM4MjVBNjYwNUQ0MEJCRQAKCRCcglpmBdQL
+vj1hAP92J5dMsc89giyfWPMkp333KuuTpAgJoeYr7AitsIKilgEA+9fPNrDz0NWn
+t391MbJQVoWy4WQsDROdYiOUecKx/QY=
+=tbZ5
+-----END PGP SIGNATURE-----