Monero Dev Meeting Note Highlights – 2017-05-07
Logs
351 lines
fluffypony1. Greetings
fluffypony2. Brief review of what's been completed since the previous meeting
fluffypony3. Code + ticket discussion / Q & A
fluffypony4. MyMonero-in-tree discussion
fluffypony5. Any additional meeting items
fluffypony6. Confirm next meeting date/time
fluffyponyso let's start with 1. Greetings (aka roll call)
fluffyponyhi
johnalanhi
vtnerdpresent
sgphello!
fluffyponytewinget apologises, he'll be late
ajsSup
endogenico/
rehrarYo
fluffyponyhyc / luigi1111 / ArticMine / othe / smooth / anonimal / binaryFate / dEBRUYNE / dnaleor / gingeropolous / iDunk / IPGlider / Jaquee / jwinterm / kenshi84 / knaccc / luigi1112 / luigi1115 / NoodleDoodle / papalazzarou / pigeons / RedLion[m] / redlion
Jaqueehhelo
pigeons:)
vtnerdalso me
Jaqueemedusa
fluffyponyanyone I forgot
iDunko/
vtnerdoh those are not present whoops
fluffyponylol vtnerd
fluffyponyok so
fluffypony2. Brief review of what's been completed since the previous meeting
fluffyponymerged a bunch PRs
fluffyponykenshi84's GPG key changed
fluffyponyI've confirmed it via sidechannel
fluffyponywe have a new sweepbelow function in the CLI, which you may find useful
fluffyponywe also have a new heavier bias in output selection towards newer outputs
fluffyponymoneromooo can fill us in on that
ArticMineHi
otheoi
fluffyponysmart mining is enabled in the GUI
fluffyponyas in the selection box
moneromoooHmm, I just twiddled the settings for the recent output selection, really. To match some data in the Miller et al paper.
fluffyponywhich is pretty cool
sgpindeed
fluffyponyalso Jaquee has done some work on getting iOS back on track after it borked (visually)
fluffyponywell iOS / mobile
fluffyponywhich brings us to
fluffypony3. Code + ticket discussion / Q & A
Jaqueeyes. and there's some new translations added to gui
fluffyponywe have a number of open PRs
fluffyponywhen tewinget is off his bus he can update us on 0MQ
fluffyponywhich I'd REALLY like to move forward with ASAP
fluffyponyit's been sitting in a holding pattern for ages
fluffyponySnipa: also if you're around maybe you can update us on the testing on that ?
moneromoooI'd like it to be optional, so it can be merged (and thus tested), without causing massive breakage if it does break.
fluffyponyafaik that was the case
Jaqueesounds like a good idea
fluffyponyalso disposable addresses is still hanging around - I think that's pending a review from one of the luigis?
moneromoooAFAIK yes. Also RandomRun had an idea to make it better.
fluffyponyI don't think there's a problem with that hanging around and being improved
fluffyponyas long as the parallel MRL write-up is there
fluffyponyI'd like to discuss 1998
fluffyponythe PR, not the year
fluffyponyat this point in time I'm still swaying towards prevent-user-stupidity-by-default
fluffyponyat the slight inconvenience for a power user / sysadmin who might go "omg really" and then add the flag
fluffyponyI know vtnerd feels the same way, which is why he added it in the first place
fluffyponyI'd be interested in strong arguments for removing the flag
Jaqueewouldnt a text disclaimer be enough?
Jaqueei don't have a strong opinion
fluffyponyJaquee: if you try bind externally and start it without the –confirm-external-bind flag then it refuses to start
fluffyponyand it tells you why
Jaqueeok. apparently hyc started the discussion. Are you around?
fluffyponyI know hyc doesn't like it
fluffyponyvtnerd: has anyone else expressed disdain for it?
vtnerdAFAIK, just the people on that PR and the one referenced
vtnerdand possibly one person in IRC, but they seemed to be questioning why it was necessary (I think)
vtnerdits somewhat low effort to get around it, so most people just add the flag I thnk
vtnerdno one has privately contacted me about it for any reason if that was the question
fluffyponyok
fluffyponyunless hyc comes in I move to close the PR, we can always re-open it later
Jaqueeok with me
fluffyponyok next PR for discussion is 2011
fluffyponymoneromooo had concerns that it was touching consensus critical issues
fluffyponyso/issues/part of the code
moneromoooYes, but it turns out it's actually bypassed when a tx comes from a block. The patch is fine.
moneromoooI OK'd it since.
fluffyponyah ok'
moneromoooWell, wait.
fluffyponystops…hammer time
moneromoooIt's really uneeded (only the wallet bit was wanted). But it's not forkworthy. That said…
moneromoooOlder wallets might create txes which aren't relayed by newer daemons.
moneromoooThat's fairly unlikely, since my code targets 2/3 of max size, but the size approximation is not very precise.
moneromoooThat said, I think it's fine to merge.
hychey. just popped in. reading history
fluffyponyhi hyc !
dEBRUYNERe: 2011, perhaps it also should be dependent on the fee priority level used
fluffyponyplays elevator hold music
hycok, if n0b0dy else cares about that external bind thing then whatever. to me it's redundant
fluffyponyok
hycsince you had to explicitly request a non-localhost address already
fluffyponysure, but you'd be surprised how few people know that 0.0.0.0 exposes everything :-P
endogenic^
hycit d0esn't protect against typos/accidents. it only pisses off people who expect the computer to do as it's told
fluffyponyhyc: view it like a weak password warning
fluffyponyyou can't just expect the computer to accept 1234 as a password
hycyeah, ok…
moneromoooWell, I would…
fluffyponylol
fluffyponymoneromooo is the exception to every rule :-P
fluffyponynow on the GUI side, the only thing I wanted to bounce around is 688
fluffyponytooltips are fine, but if we're going to do some sort of unified help then I would veer towards an overlay that shows once the first time you enter a screen, and can be re-called by clicking the [?] button on the taskbar
fluffyponyhttps://s-media-cache-ak0.pinimg.com/originals/c1/e1/bf/c1e1bfd7fb2770f6745d95af8bf89865.jpg
johnalanlike that style
fluffyponyhttps://s-media-cache-ak0.pinimg.com/originals/43/6e/74/436e746b35142f41d5f9bb8e765963e4.jpg
fluffyponylike that
hycsounds good
johnalan:+1:
Jaqueeproblem is [?] is not around if you use native title bar
fluffyponyJaquee: where else could we add a help button? bottom left?
endogenicone suggestion i'd make for that is to make it c lear to the user they can recall it easily by doing "X" so that they don't fret about having to memorize everything before it's closed
endogenicrecall it -> the help screen
Jaqueei think ^ is good as a start
moneromoooWhere is it on the title bar then, since it's not a WM thing ?
fluffyponyendogenic: agreed
moneromooos/Where/Why/
Jaqueebut some buttons could need longer desriptions
Jaqueelike sweepunmixable and paymentid for example
fluffyponyJaquee: there's enough space in the help overlay, we can use a smaller font to explain them
redlionhow breadwallet on ios handles it when setting up is quite good
fluffyponyor move the help to somewhere where there's space
fluffyponyand use an arrow
Jaqueeyeah. we could find a place for that help button
fluffyponyok - any other PRs that need discussion or can we move on? there's general Q&A shortly
sgpI'd like to merge 261 on monero-site
fluffyponysgp: there's a website meeting after the Kovri one
fluffyponyso we can discuss it then
sgpok
fluffyponyok so
fluffypony4. MyMonero-in-tree discussion
fluffyponyso basically this is about nose-covering and making sure I'm not abusing my position as a maintainer and member of the Monero Core Team
fluffyponycurrently MyMonero has a working API (largely unspecced to be sure), two client implementations (website and app), two server implementations (the live backend and OpenMonero), with a third one coming
fluffyponyI'd like to make sure there is general acceptance and buy-in that the API can be implemented as the general API for lightweight wallets (ie. wallet that use remote viewkey scanning)
hycis it carved in stone now
hycif we need to tweak it we can still do that?
redlionis the license unrestricted?
fluffyponyand that MyMonero-written or MyMonero-derived code is generally acceptable to be merged into the source tree (ie. the open-source backend implementation that vtnerd is working on)
fluffyponyredlion: BSD 3-clause
fluffyponyhyc: as long as mWo12 changes it, and we match the changes in the live backend and the new backend then yse
fluffyponyyes
fluffyponywe can make any changes, and we WILL make changes to make it smarter
moneromoooIf it's beneficial to monero and it works fully by itself without needing proprietary gunk, then I'm OK with it.
fluffyponyeg. tx history comes in raw, instead of paginated
fluffyponyso that needs to change
hyc+1 moneromooo
fluffyponymoneromooo: yeah the new backend will use LMDB instead of mysql
fluffyponyso it will be unencumbered in the source
ArticMineAs long as there are no proprietary dependencies I am fine
hycI like it even more now ;)
johnalanI think it beneficial too
moneromoooMaybe a separate repo (similar to monero-core) might be best, but that's details.
johnalan*its
moneromoooit's
iDunkit's
jollymortcan't wait to run a mymonero node myself!
vtnerdalso the current "primary" wrapper around the DB is actually C, so theres that for you guys
fluffyponymoneromooo: I thought about that, but it's a single daemon that should exist in the repo alongside the wallet RPC etc.
hycdoesn't it supersede wallet-rpc?
fluffyponynow
fluffyponyhyc: no
fluffyponywallet-rpc is good for integration, this isn't
johnalanthere is obviously an element of centralisation, but it’s nearly impossible to avoid
fluffyponyalso on this topic
fluffyponyJaquee has begun working on client integration in the CLI and GUI
moneromooo"client integration" ?
vtnerdyou mean for light-wallets?
fluffyponythat will mean that both CLI and GUI will be able to run in lightweight / remote-scanner / MyMonero mode
fluffyponymoneromooo: as opposed to implementing the server protocol
hycsounds good
moneromoooOh, mymonero client integration ?
fluffyponymoneromooo: let's call it something else
moneromoooThat went pretty damn fast :D
fluffypony"lightweight wallet"
jollymortit's not really centralization if any monerod acts as a server
hycbut I'm still missing why we need old wallet-rpc if this mymonero api exists
jollymortit's literally my monero :)
fluffyponyhyc: wallet-rpc is completely different
johnalanso the core GUI will be able to interact with MyMonero backend too?
vtnerdfor people that want to run VPS node but keep their viewkey ?
moneromoooYes, would be nice to see what bits are needed where, and the actual API (even if roughly).
fluffyponyit provides an API for integrators
fluffypony@johnalan yes
fluffyponyso basically
johnalanis this needed with the MyMonero Desktop wallet?
ArticMineWith what as the backed / server
moneromoooThat can be posted later though, :49 now.
ArticMinemonerod?
fluffyponylightweight wallets will have 3 server options:
fluffypony1. OpenMonero
fluffypony2. the new in-source backend that vtnerd is working on
fluffypony3. the live MyMonero backend
fluffyponyit will also have multiple client options:
hycafaik the main difference btw an ordinary wallet and mymomero is you tell mymonero your viewkey
fluffypony1. OpenMonero's web wallet (clone of the current MyMonero web wallet)
hycand the ordinary wallet has all your keys
fluffypony2. the MyMonero applications
fluffypony3. monero-wallet-cli
fluffypony4. monero-wallet-rpc
fluffypony5. the Monero GUI
fluffyponyhyc: monero-wallet-rpc can still use this on the backend
fluffyponyso it's unrelated
hycok
ArticMineok
jollymortabout #2011 - you could modify it to (median)+0.6% for it to be mine-worthy, or even have the wallet check for fee setting and then it would be matched like 1: +0.6%, 2: +2.4%, 3: +12%, 4:+100%
fluffyponyalso this will mean that the GUI / CLI may end up supporting the MyMonero 13-word seed derivation by virtue of the integration effort
fluffyponydoes anyone have a fundamental issue with that ?
ArticMineno
fluffyponyI mean, I do, because I don't want to be abusing my position, but it is what it is :-P
jollymortdidn't you deprecate 13-word?
moneromoooDid you not say the 13 word seed was going to be obsoleted ?
endogenicjollymort: working on it
johnalanno
endogenicbut client still needs to be able to read 'em
redlionelectrum/mycelium support a few different seed lengths iirc
redlionworks well
jollymortalso luigi was playing around with an idea for 17-word, integrating creation height in it etc
fluffyponymoneromooo: it's import only
fluffyponynot create
knacccdoesn't it put a huge load on mymonero when someone asks it to scan the blockchain from zero with their view key? How long does mymonero take to scan the entire blockchain?
moneromoooAnyway, I'm fine with that as presented.
hycthat all sounds like a win to me. people have been whining about not being able to import their 13-word seed into regular CLI wallet
shuannelsonso monero-wallet-cli/monero GUI will not be able to create light-wallets?
fluffyponyknaccc: yes it does - about 10 minutes
jollymortyeah import only sounds lovely
ArticMineIf we are setting the stage for a competitive market based upon FLOSS then I am fine with it
vtnerdI do have the ASM code working, so hopefully that will tighten up some too (altough there is something else blocking that)
fluffyponyshuannelson: yes they will
fluffyponybut with 25 word seed, not 13
fluffyponywe have 7 minutes left - so I'd like to move on to the last item
shuannelsonawesome!
fluffyponywe can discuss MyMonero more after the meeting
redlion@shaunnelson, I think it's just that the CLI/GUI won't create 13-word seeds, but will accept already created ones
hycyeah sounds fine
fluffypony5. Any additional meeting items
knaccc10 mins is quite a speedup vs downloading the entire blockchain, so sounds awesome.
jollymortany thoughts on future of penalty/blocksize? i kind of left the research open-ended
hyc^^ get a faster CPU and it'll be quicker ')
redlionDoes anyone have a working monero-core or mymonero build on ios currently? I've been fiddling around and I can't seem to get either properly functional on the sim/device, though I may be missing something
fluffyponylol hyc
endogenicredlion: pls come join #mymonero but yes i do :)
Jaqueeredlion: i have. it has some nasty bugs but it's running
redlionok thanks, I'll talk to you after this
hycbtw iOS still limits process VM size to 4GB so we won't be running monerod native on iOS any time soon
fluffypony@jollymort let's discuss it after the meeting, or maybe next week - there are 2 more meetings to go tonight :)
fluffyponyand that's a large topic
jollymortsure, another time
redlionthanks jaquee, are there any build instructions or a (sort of) working build posted somewhere?
fluffypony6. Confirm next meeting date/time
fluffyponyMay 21
fluffyponyday before Consensus
hyccool
hycoh. this week I expect to have wolf miner fully ported to Android, with GPU support too
fluffyponyendogenic can come to my hotel and we can do the meeting together :-P
endogenicoooh
tewingetanyway, I have the daemon's side of the code rebased and nearly ready to PR and merge. I mean, it could be merged now, but I should clean it up a little/address a few more of the comments on the existing PR first.
tewingetthe wallet side of things will be based on that, and won't take too long. I just thought it made sense to separate it into two PRs (and rebase while I'm at it because why not?)
fluffyponysuweet
fluffyponyjust check the meeting logs for the bit from moneromooo about it
tewinget(the wallet stuff is still "done already", but as with the daemon side there are comments/suggestions to address as I rebase it as well.)
tewingetat any rate, I plan today to finish with the cleanup of the daemon side of things, close the existing PR, and open a new one for the daemon that should be mergeable.
fluffyponygreat stuff
fluffyponypigeons: did you see the 96boards thing?
tewingetfluffypony: sorry I didn't respond right away to your pinging on the github PR, but when I said it was already rebased I meant on a different branch, as I'm leaving that branch up (and separate) until I finish rebasing.
fluffyponyok cool
moneromoootewinget: is the 0MQ stuff deselectable if needed (so if it somehow breaks, you can run the wallet with the existing JSON comms) ?
moneromooowallet/daemon
tewingetmoneromooo: I'll make it so when I rebase the wallet side of things.
moneromoooExcellent, thank you :)
Sources and notes
- Meeting log: Overview and Logs for the Dev Meeting Held on 2017-05-07, getmonero.org. The official post carried the raw log and pointed readers to Hello Monero for the overview.
- Original Hello Monero URL slug recovered from the Wayback Machine index; the original page content did not survive, so this reconstruction carries the log as archived by the Monero Project.