開發者會議紀錄 DevMeeting 20170917

  • 本周的v0.11新版發佈因測試跟發佈時間間隔太短了,問題比以往來的多。接下來團隊將會嘗試為測試和新版發佈訂下更嚴格的時程。
    • 硬分叉時間為三月和九月
    • 程式碼則將在六月中與十二月中凍結
    • 這讓新版本在釋出前有更多機會進行測試跟問題的修復
    • 同時也提供更多釋出小數點版本(point release)的機會
  • 因應上述更結構化的步調,較大的更新應該會陸續加入主程式碼中。這些更新都不需要進行分叉。
    • Sub-Addresses
    • Multisig (MRL團隊表示目前看起來還不錯,但還是需要更多驗證,約在2017年十一月底可以完成。)
    • ZeroMQ
  • 需要更多的單元測試(Unit Testing) - 這對新的開發者來說是一個很好的入門方式。
  • Luigi1111 在GUI跟官方網站的工作上都從”候補維護者”轉為”主要維護者”,FluffyPony則轉為候補維護者。
  • 關於MyMonero GUI的部分
    • MyMonero 的Apple 鄧白氏編碼依然還沒收到
    • 為AppStore申請的MyMonero 的Google帳號已經準備好了(但也需要鄧白氏編碼)
    • 兩者都還是有在增加新東西(將iOS的程式碼改寫成JavaScript 的App)
  • MRL 的第二次未來規劃(roadmap)已經準備好讓各方校對了: https://drive.google.com/file/d/0ByPYdEm6O54rZGpTbkcwdWMwcW8/view

以上翻譯來自於紀錄原文: https://monerobase.com/wiki/DevMeeting_2017-09-17

完整會議記錄:
fluffypony> 1. Greetings
fluffypony> 2. Brief review of what’s been completed since the previous meeting
fluffypony> 3. Code + ticket discussion / Q & A
fluffypony> 4. Release timeline: moving releases to 3-months-before-forks or similar
fluffypony> 5. Maintainer: migrating luigi1111 from backup maintainer to lead maintainer on GUI and site
fluffypony> 5. Any additional meeting items
fluffypony> 6. Confirm next meeting date/time
fluffypony> let’s start with 1
_Slack>
serhack> Hi
_Slack>
hrumag> Hi
shillosopher> o/
rbrunner7> Grüezi
_Slack>
rehrar> Yo
i2p-relay> {-lightfighter} hola
imheretolearn> hola
msvb-lab> hello
sarang> yo
surae> guten tag
luigi1111> Oh hai
imheretolearn> ni hao foobar__ has quit (Ping timeout: 260 seconds)
hyc> hi
fluffypony> ArticMine / othe / smooth / moneromooo / tewinget / anonimal / endogenic / gingeropolous / iDunk / JacobJeweler / kenshi84 / knaccc / MoroccanMalinoi- / NoodleDoodle / papa_lazzarou / wallet42 / vtnerd_ / anyone I forgot
ArticMine> Greetings
iDunk> o/
moneromooo> You forgot fluffypony in that list.
fluffypony> lol
surae> 22nd century historians are going to have a hell of a time writing books on the birth of cryptocurrency.
_Slack>
moroccanmalinois> Hi
fluffypony> pigeons sends his apologies, he let me know yesterday that he wouldn’t make today’s meeting
foobar_ ([email protected]/web/freenode/ip.213.225.39.202) has joined
fluffypony> so anything that is wrong we can comfortably blame on him
fluffypony> since he isn’t here to defend himself :-P
fluffypony> ok
fluffypony> let’s move on to 2
fluffypony> so not much has happened since the last meeting
fluffypony> a release or something
fluffypony> nothing major
surae> heh
hyc> I heard we spawned XMRclassic
fluffypony> Monero ABC
surae> i used XMRcash at a coffee shop, but my card was declined :(
fluffypony> inb4 XCash
fluffypony> on the topic of the release, and now is as good a time to talk about it as any
hyc> oooo, I like that one
fluffypony> as you may (or may not) have noticed this release was a little rockier than we’d hoped
rbrunner7> Yeah TinusMars has quit (Ping timeout: 260 seconds)
fluffypony> we’ve known for a while that our testing horizons are too close
TinusMars ([email protected]/web/freenode/ip.84.192.212.30) has joined
fluffypony> a little while ago smooth suggested having MUCH further horizons and sunsets, which I tend to agree with
surae> what are the thoughts on modifying that, then?
fluffypony> so what we’d like to do moving forward is work our horizons of a 6 month schedule, offset from the hard forks by 3 months
fluffypony> in other words
fluffypony> hard forks happen March and September
fluffypony> and code freezes happen mid-June and mid-December
fluffypony> so practically let’s look at the next year as an example
ArticMine> … but will we follow the code freezes?
fluffypony> we have 3 months from now to put stuff in the code, and then we code-freeze in mid-December
imheretolearn> will a 3 month code freeze have a side effect of slowing development?
fluffypony> that means we branch in mid-December
_Slack>
serhack> Okay and then a new release?
_Slack>
hrumag> You could follow the timeline used bu ubuntu
_Slack>
hrumag> by
fluffypony> from mid-December we have some time to test, fix bugs, get the branch stable
fluffypony> (and port fixes to and from master)
fluffypony> with a view to tagging a release as soon as we think there’s some stability
moneromooo> 3 months seems maybe a bit large, but let’s try it. I’m curious to see how much of a pita it’ll be to have to branches.
moneromooo> to have 2
fluffypony> the nice thing is that the release CAN have bugs, because not everyone will upgrade until closer to the hard fork
fluffypony> and so subsequent releases can patch bugs
_Slack>
serhack> Seems good idea fluffypony
floam412 ([email protected]/vpn/privateinternetaccess/floam412) has joined
sarang> agreed
moneromooo> The only thing is that it only leaves 3 months to get range proof changes in… :/
fluffypony> it also means we can have loads of small point releases up to the hard fork, because there isn’t much that changes except bug fixes
moneromooo> or 9. or 12.
_Slack>
hrumag> I totally agree with fluffy
fluffypony> moneromooo: if we don’t hit the mid-December target then leave it for the next one, we have 6 months thereafter
luigi1111> 3 months seems fine to me. The release can be out a while before the fork with no negatives really
fluffypony> yeah exactly
luigi1111> Besides maybe some delay in new features
fluffypony> and it means that for 3-ish months the network is running both stable and newer versions
ArticMine> It is a way safer approach
_Slack>
hrumag> https://wiki.ubuntu.com/ArtfulAardvark/ReleaseSchedule if it can help
fluffypony> most exchanges / pools will stick to the previous stable release
luigi1111> Mooo we’ll just have to get it done by December then:)
fluffypony> some brave ones will upgrade early
moneromooo> ^^
luigi1111> Since like, ms is working and such
moneromooo> -
-
_Slack> Command sent from Slack by serhack:
_Slack> .-.
fluffypony> the real issue is that we have to decide on a new release name within the next 3 months
fluffypony> :-P
knaccc> Iridium Iris
hyc> Lithim lizard braveh4rt has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client)
hyc> lithium
fluffypony> also as pointed out this will mean that some features take a lot longer to go live in the mainnet codebase than we might hope (emergency patches / forks aside)
surae> hypothetical helicopter?
fluffypony> so it means we’re going to have to be patient when things don’t progress as rapidly as we’d like
fluffypony> we \
can* happily deploy stuff on testnet, we’ve had to do massive reorgs on testnet before and it handled it just fine dEBRUYNE grabs torches and pitchforks
fluffypony> and iirc Snipa had some experience setting up a mini closed testnet
luigi1111> I don’t really think 2 ish months is “a lotc
msvb-lab> Dumb question, what’s the name of the current 0.11 release?
luigi1111> “lot”
fluffypony> msvb-lab: Helium Hydra
surae> oh so hypothetical helicotper is right out. :( too bad we missed the possibility for Gluttonous Ganesh?
fluffypony> aka Hail Hydra
msvb-lab> hyc: Then lithium wouldn’t work would it? It doesn’t start with ‘I’.
sarang> Irritable Ibex
imheretolearn> i agrree knaccc iridium iris
hyc> private testnets are easy, just takes 2 nodes pointed at each other –add-exclusive-node
fluffypony> yarp
ErCiccione has quit (Quit: Leaving)
hyc> release names aren’t alphabetical, they’re elements. periodic table.
fluffypony> ^^ amiuhle| has quit (Ping timeout: 264 seconds)
fluffypony> there’s a release naming doc in the meta repo
imheretolearn> iridium iris will have range proofs improvements, larger ring size multi sig and encrypted payment ids. major release
luigi1111> Ugh stop talking about release names
moneromooo> ^
moneromooo> Get on with it :D
fluffypony> let’s move on
fluffypony> 3. Code + ticket discussion / Q & A
fluffypony> I’d like to start merging in the big PRs that have been hanging around
hyc> there’s still a good number of PRs that could be merged soon. should we be planning a point release in the next month or two?
fluffypony> sub-addresses, multisig, 0MQ, etc.
fluffypony> hyc: yes - I had to first merge the PRs that were already merged in the branch, which I finished a couple of days ago
moneromooo> For the record: multisig is not ready yet. It’s got a DO NOT MERGE sign.
fluffypony> ok
fluffypony> moneromooo: is it waiting on code or MRL stuff?
luigi1111> yes
knaccc> I just wanted to point out, in case it had slipped through the cracks, that kenshi84 has partially implemented subaddresses in the GUI but there are wireframe mockups he wasn’t able to implement because he’s not used Qt much before. The mockups suggested to be implemented are here https://github.com/monero-project/monero-core/pull/793#issuecomment-314477091
moneromooo> It’s waiting on me making core tests for it, and cleaning up.
moneromooo> And the core tests are a massive pita.
fluffypony> moneromooo: ok cool
fluffypony> moneromooo: is it worth looking at a different test framework?
luigi1111> all of these things are not forking, so could be added whenever
moneromooo> Not really, core tests are great, powerful, and I love them. I also hate them.
moneromooo> Once they work, they’re really great.
fluffypony> knaccc: ok cool, hopefully someone can pick that work up
luigi1111> n-1 only recently started working, so I assume it needs some testing etc
surae> sarang and i are nearing completion on both the multisig paper, which requires new security definitions, and the sub-address paper (and that scheme looks pretty good!). the multisig stuff, if you held a gun to my head, i would say is secure, but until the proofs are written up, i don’t want to say anything about it definitively
fluffypony> knaccc: else if we need to find a QtQuick developer and pay them we can afford to do so
hyc> so should we continue to PR to the release branch in addition to master?
ErCiccione ([email protected]) has joined
surae> we anticipate both of those papers to be done as MRL bulletins by end of November.
moneromooo> I hope not.
fluffypony> hyc: no
fluffypony> release branch is for bug fixes only
fluffypony> surae: ok great
hyc> ok _operative ([email protected]) has joined
i2p-relay> {-lightfighter} is there a README for writing new tests? if not could one be added?
moneromooo> No. Sure.
fluffypony> @lightfighter no there isn’t, but it’s something you could tackle if you’re up for it
fluffypony> on the topic of tests
moneromooo> But really, I start from an existing one.
fluffypony> if anyone is unfamiliar with the codebase and wants to learn, adding tests is literally the best way
fluffypony> there are tons of things that lack unit tests
moneromooo> Unit tests are very easy to add.
fluffypony> need moar test coverage plx
surae> unit tests are. so. freaking. helpful.
moneromooo> Though they might need API changes to expose bits and bobs.
hyc> in that regard, might be nice to appoint a release/test manager, responsible to either run tests or collect test results from the community
fluffypony> hyc: good idea
parasew> will definitely start implementing tests, its in a way comforting :)
fluffypony> I wonder if we couldn’t build a little web API that people could submit test results to
Guest66881 is now known as expez
i2p-relay> {-lightfighter} fluffypony: i’d be happy to start working on test docs. is the GTest primer a good place to start?
moneromooo> https://github.com/monero-project/monero/issues ?
vtnerd_> boost has some such framework, but I dont know much about it
parasew> on another note: in the current gui-release, the mining has issues – other people also reported at monero meetup.
vtnerd_> various members run machines on various processors / platforms amiuhle| ([email protected]) has joined
i2p-relay> {-lightfighter} also agree that writing tests is best way to learn codebase (my main motivation for doing so)
fluffypony> parasew: then they must open an issue on GitHub
dEBRUYNE> parasew: Already reported on monero-core github repository
parasew> fluffypony ofc, will pass that on.
parasew> dEBRUYNE which issue?
dEBRUYNE> https://github.com/monero-project/monero-core/issues/877
parasew> dEBRUYNE ty, will also have a look!
fluffypony> let’s give some thought to a reportable web API that can collate test results, we’ll figure out a nice way of doing it
hyc> a web tool sounds like a great idea. I don’t know of one at the moment.
dEBRUYNE> hyc: for the GUI release I created a reddit thread with buildbot bins and compiling instructions
dEBRUYNE> Subsequently I put up all reports on Github
dEBRUYNE> But it’d be preferable if we have a more streamlined process
hyc> indeed
dEBRUYNE> (to report issues that is)
fluffypony> agreed
_Slack>
rehrar> Will look into it
fluffypony> we’d probably have to roll our own tool for this, let’s see
moneromooo> Anyone who knows about DNS/DNSSEC btw ? We need to fix how we do that apparently, but I know nothing aobut DNS :/
moneromooo> Thoise are the failing unit tests.
dEBRUYNE> I suppose though that if we’d create something like that someone still has to create issues on github
surae> sarang? know anything about DNS?
hyc> DNSSEC in particular
sarang> Protocol? Some. Implementations? Negatory
hyc> we need to find out why libunbound is complaining now
moneromooo> The complaint’s likely unrelated, and due to the openssl config.
moneromooo> (assuming you mean the “invalid algorithm” or so)
rbrunner7 has quit (Read error: Connection reset by peer)
fluffypony> could be openssl related IRC-Source_14236 ([email protected]/web/cgi-irc/kiwiirc.com/ip.212.186.179.187) has joined
hyc> yes. I’ve only hit that in the past with a crippled Android libssl
hyc> so I’m surprised to see it on an up-to-date desktop build
fluffypony> I’m also open to dumping unbound in favour of another library
fluffypony> we literally only use it for DNS / DNSSEC
rbrunner7 ([email protected]) has joined
moneromooo> Guess I’ll have a look to see what libraries there are for this.
parasew> is there an irc log online somewhere? (i came a bit late)
fluffypony> ok let’s do that, Unbound has been a PITA for a while
moneromooo> Or… leave it till someone who knows about DNS pops in…
fluffypony> parasew: it’ll be posted up after
surae> parasew the minutes will be posted after the meeting
fluffypony> ok let’s move on
_Slack>
serhack> ;) rbrunner ([email protected]) has joined
parasew> thanka fluffypony, surae i just wanted to catch up now, but its ok :(
fabian has quit (Remote host closed the connection)
fluffypony> 4. Release timeline: moving releases to 3-months-before-forks or similar
fluffypony> we already covered that
fluffypony> so moving on
fluffypony> 5. Maintainer: migrating luigi1111 from backup maintainer to lead maintainer on GUI and site
moneromooo> Does he agree to ? :)
fluffypony> well, moneromooo, luigi1111 has gone full-time crypto since he’s made millions from the Monero ICO
dEBRUYNE> No we forced him :P
hyc> lol
fluffypony> so he can merge from atop the mountain in his Lambo
surae> i hear Lambos work real great on mountains
_Slack>
rehrar> Congrats Luigi. An example to us all.
fluffypony> does anyone have an objection to me moving to backup maintainer on the GUI and the site, and luigi1111 moving to lead maintainer? I’d probably still do GUI releases since they’re done in conjunction with the CLI rbrunner7 has quit (Client Quit)
_Slack>
serhack> A little fun fact
_Slack>
serhack> Oh great Luigi
hyc> Sounds like a good move
fluffypony> or we’ll do it together or something
_Slack>
rehrar> Will Luigi just merge the site repo, or also put it on the server?
_Slack>
serhack> Congratulations
hyc> but that reminds me we need to move forward on deterministic builds
fluffypony> rehrar: pulling on the site is still manual in case the repo gets compromised
_Slack>
rehrar> Right.
surae> is this marking a fundamental shift in the power dynamics of Monero?
_Slack>
rehrar> So will he do that manual task too, or is that still you?
fluffypony> which would be embarrassing, but not as embarrassing as the site being compromised via the repo
fluffypony> rehrar: still me, but it takes me ten seconds
fluffypony> so he’ll ping me
_Slack>
rehrar> K.
fluffypony> surae: yes, we’re appointing luigi1111 as CTO (Chief Trolling Officer), whilst I focus on my role as CEO (Chief Entertainment Officer)
fluffypony> I can’t do both any more, it’s too much for one troll
surae> oh man, can luigi make fancy videos of people wistfully walking through the wheat? i’m not sure if anyone can do that like you. :( (moving on…)
fluffypony> lol
fluffypony> ok
fluffypony> 5. Any additional meeting items
imheretolearn> globee or mymonero updates?
scoobybejesus> Mooo and other FFS to fund, perhaps
hyc> app store progress?
_Slack>
serhack> Ffs?
fluffypony> MyMonero Google account is ready
fluffypony> waiting on our D-U-N-S number for Apple
fluffypony> or as endogenic joked, we’re waiting on our DUNCE number
Olufunmilayo raises hand
_Slack>
serhack> How many monero have we collected for ffs? anonimobile ([email protected]/tor-sasl/anonimobile) has joined
fluffypony> scoobybejesus: huh?
dEBRUYNE> I think scoobybejesus suggests to discuss the FFS proposals (if we want to)
fluffypony> serhack: open them and check
fluffypony> Olufunmilayo: don’t ask to ask, just ask :)
scoobybejesus> Really just bringing them to light. Yes.
_Slack>
serhack> :+1::skin-tone-2:
surae> MRL is about to put out our second roadmap (which includes an update to the first one). hyc already provided some copy-editing comments. we are currently looking for 1) typos and copy-editing stuff, but more importantly 2) any projects the developers feel should be included that are not being included, and 3) any inaccuracies in the statements we’ve made already.
surae> https://drive.google.com/open?id=0ByPYdEm6O54rZGpTbkcwdWMwcW8
sarang> We’ll wait while everyone reads all 12 pages
surae> PM us with edits and comments. sarang and i both have edit access
fluffypony> surae: only thing I’d like to add there in section 7 is a note about zk-STARKs being preferable since no trustless setup, but memory / CPU requirements are too high atm, so maybe we just stick to zk-SNARKs in the interim
moneromooo> We’ll have to decide on an ordering for merging stuff, because there’s a glut again, and that means conflicts.
moneromooo> How about 0MQ first ?
Olufunmilayo> fluffypony, nvm, now that I typed it up and re read scroll back it has already been addressed
surae> fluffypony: zk-starks aren’t published yet and afaik only the creators have even pseudocode. currently, unless i’m wrong, zk-starks are a myth, and until i see some proofs otherwise, i’m going to pretend like they aren’t available yet. but i can mention them as a “next step” under section 7
hyc> yeah, let’s go with older PRs first
hyc> although for gdb sanity I would like threadpool to go in soon.
Jaquee> shouldnt threadpool be added to release branch?
sarang> On a meta-MRL note, I’d like to offer my thanks to everyone who supported the FFS to bring me aboard
foobar_ has quit (Quit: Page closed)
msvb-lab> moneromooo: I think it’s pretty great we’re integrating 0MQ, but I’m not that qualified to vote on PR arangement.
fluffypony> surae: sounds good
fluffypony> moneromooo: I agree - 0MQ and DRY PRs, and then threadpool?
moneromooo> OK
hyc> Jaquee: not sure threadpool counts as a bugfix
dEBRUYNE> moneromooo: Did you manage to provide feedback to tewinget?
moneromooo> Many times.
dEBRUYNE> I think he asked for that, but you were busy with the release
surae> also, the section on untraceability in monero… i need some opinions on that, too, but i want people’s fresh impressions of that section. i had a certain intention when i was writing it and i want to see how well it came across
tewinget> sorry, was slightly afk
moneromooo> Is there a list of things he wanted feedback on ?
dEBRUYNE> tewinget ^
tewinget> saw my screen blink
fluffypony> Jaquee: I agree with hyc - let’s avoid merging anything that could be wonky if we can, release seems to be reasonably stable
Jaquee> all right
Jaquee> we can live with a little unresponsiveness i guess
Jaquee> in gui ^^
moneromooo> er… did you check it fixed your problenm in the first place ? PuebloZag has quit (Ping timeout: 260 seconds)
moneromooo> I’m pretty sure it won’t.
Jaquee> no i havent
hyc> it’s worth checking. I also see responsiveness issues in current 0.11, monerod spawns too many threads during verification
aerbax> With regards to the MyMonero AppStore deployments. If Google/Android is ready, are there any ongoing issues that would prevent a release?
dEBRUYNE> I think they want to release all platforms in conjunction
Jaquee> are we using the same DUNS/account for monero apps?
fluffypony> Jaquee: no, completely separate one
fluffypony> have to have that distinction
cardboardoranges has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client)
parasew> are there plans of putting the GUI into apple appstore?
fluffypony> both are waiting on D-U-N-S numbers
Jaquee> ok
fluffypony> parasew: yes, check previous meeting logs
parasew> fluffypony ok, will do :)
msvb-lab> What’s the main reason for one mobile platform release to wait until others are ready?
fluffypony> Jaquee: we registered “Monero Distribution Company” as the name * xpto ([email protected]/web/freenode/ip.206.225.72.22) has joined
_Slack>
serhack> Fluffypony: I have an android phone, can I test mymonero ? Can you send me an apk?
fluffypony> msvb-lab: on the Monero GUI side, nothing, we can push it into the Google Play store whenever
moneromooo> Are we done with the dev stuff ?
fluffypony> on the MyMonero side we’re still busy adding some stuff
msvb-lab> Additional meeting items: Might be worth mentioning that both mattermost.getmonero.org and taiga.getmonero.org are quite bad ans.
fluffypony> mostly porting stuff from the iOS app to the JS app
hyc> coming up to top of the hour
fluffypony> thanks hyc
fluffypony> msvb-lab: we can discuss that after
fluffypony> ok so next meeting in 2 weeks
fluffypony> 1st October