Monero Dev Meeting Note Highlights – 2017-05-21
Logs
351 lines
endogenicwe have to do the dev meeting really quietly so the pony can get some sleep
moneromooo1. Greetings
moneromooo2. Brief review of what's been completed since the previous meeting
moneromooo3. Code + ticket discussion / Q & A
moneromooo4. Any additional meeting items
moneromooo5. Confirm next meeting date/time
moneromooo(no relay bot btw)
moneromoooSo, since last time, I'm not sure what's been merged. Not much I think ? There was the key image flaw made public too.
moneromoooI'm working on abstracting network addresses. Mostly works, but serialization is being a massive pain.
moneromoooAnyone else want to say anything about what they're working on ?
vtnerdeventually I hope to have the amd64-51-30k speedups in wallet2.cpp (assuming fluffy approves)
vtnerdand improve the hex -> bin performance would be nice too (not nearly ready for a merge though)
moneromoooWhat is that short for ?
vtnerdoh sorry, the ASM for wallet scanning
moneromoooWhat are the 51 and 30k short for ?
vtnerdits already written so that it can be removed by cmake, and does not touch the consensus code at all
vtnerd51-bit limbs and 30k tables
vtnerd30k table singular, sorry
moneromoooAnyone else doing something nice ? hyc, LMDB wallet code maybe ? :)
moneromoooNobody else ? tewinget ?
vtnerdeveryone is out spending bytecoins
moneromoooAlright. Ticket discussion. The only thing I know is people reporting sync stopping. Some of them are user error, some are corrupt db, but I think there may be something else. I'd need to see good logs for both sides though.
moneromoooAnyone wants to say anything about pending bugs ?
iDunkTesting related… your 1982 works nicely, with no apparent issues after several days. That on a daemon with none to very light load, fwiw.
moneromoooCool, thanks.
moneromoooAbout pending bugs: we're still waiting for finalization of the disposable address theory, IIRC RandomRun and knacc were discussing a change recently.
moneromoooknaccc: any comment about this ?
iDunkI'm syncing on Windows right now, albeit with non-default settings. Will report if it gets wedged.
vtnerdchange to which portion, disposable addresses or sub-addresses?
vtnerdor I guess you specifically said disposable address
moneromoooI don't remember. It's 1753.
moneromoooI did, but not on purpose :)
moneromoooI have a feeling it'll be a fast meeting.
moneromoooluigi1112: anything new about smaller range proofs ?
endogeniche's out spending bytecoin too
jollymorthappened to me today (sync); it just hanged; killed the process and restarted monerod; continued seamlessly
jollymortcan re-sync with different loglvl to see if it happens again
moneromoooNext time, don't kill it before you've looked at the thread stack traces.
moneromoooAnything else someone wants to say in the meeting ?
rehrarKovri test site? :D Not super dev-y, but nobody else says anything.
moneromoooKovri meeting is in 48 minutes.
rehrarthere is no kovri meeting today
moneromoooaw… then feel free.
moneromoooanonimal: around ?
luigi1112moneromooo sorry no. will update soooooon
moneromoooty
moneromoooHmm.
luigi1112next meeting will be exciting :-)
moneromoooWell, no test site either then.
unknownidsoh baby i know what that means
unknownidsi think
moneromoooNext meeting: in two weeks, same time.
moneromoooDone. The only good meetings are short meetings.
luigi1112am I doing this right
luigi1112ann of ann
jollymortmooo, how to use stacktrace?
moneromoooOh he's back now :)
moneromooojollymort: gdb /path/to/monerod pidof monerod
moneromooothread apply all bt
jollymortso i can do it while it's running
moneromoooYes.
jollymortno need to start it in some special way?
jollymortok
jollymorti'll set loglvl to max also
jollymortand start sync from 0
moneromoooI wouldn't.
jollymortok, so just restart and wait for it to hang?
moneromooo3 and 4 are toop spammy for being much use in general usage.
moneromoooYes.
moneromooorehrar: free free if you have something to say/show.
rehrarAnyways, Monero test site here: http://45.32.161.183/ mobile responsive now and a brand new mobile navigation.
rehrarKovri test site: http://45.32.171.42/ completely integrated into the Jekyll. The custom CSS framework was ported without issue. This bodes well for my further work on getmonero.org.
moneromoooLooks clean.
Jaqueehi. a bit late to meeting. i'm working on lightwallet integration in cli and gui
Jaqueeor wallet2 mostly
moneromoooYou mean mymonero ?
rehrarWe scrubbed it down.
Jaqueeyeah. and openmonero
Jaqueeshould work with both backends
Jaqueeand should be switchable from full wallet to light wallet. vtnerd. would you consider implementing input/output import in mymonero api?
Jaqueewould make it possible to switch from full wallet to light wallet without a rescan
endogenicnot clear how the mobile experience would go
endogenicjust an initial impression
Jaquee?
Jaqueeexperience of what?
endogenicas in how user gets wallet file on their mobile and how they pick it
vtnerdI don't know what you mean Jacquee
vtnerdyou want to import an existing wallet quickly from a local DB to a mymonero backend ?
vtnerdand also the other direction ?
luigi1112that would be cool. maybe :-)
Jaqueeyeah. if i have a full wallet on my desktop and want to use the same wallet on my mobile. I could convert my full wallet to lightwallet by sending my inputs/outputs to mymonero/openmonero backend. And then login as usual on my phone
Jaqueewith seed
Jaqueewithout paying the 1 xmr rescan fee
Jaqueethat was my idea
Jaqueeso the magic happens in desktop wallet.
Jaqueeon mobile you only login as usual
vtnerdyeah, but its doable but I would have to think about it a bit
vtnerdthe backend would still want to verify correctness, so its not stomping on someone elses coins, etc
vtnerdbut it would be doing much less work
Jaqueeok. yeah backend could verify with view key
Jaqueethat's probably best
vtnerdyes. but it would only be looking at the set of outputs specified in the RPC call which will be much faster
Jaqueeexactly
endogenicJaquee: ic
vtnerdyeah so the import out will be much easier, and you can almost do it with unspent_outs call
vtnerd*export I think is what normal people say
Jaqueeyeah. i'm almost done with that
Jaqueei'm trying to make wallet2 fully compatible so i can switch between light wallet and using a normal node
vtnerdwell I meant but the export from the backend side, but ok interesting that you've already started the wallet side
vtnerdyup, understood
zueorehow monero is configured to only cpu/gpu mining?
Jaqueeso, to summarize. if i create the mymonero wallet first i can already switch between full and light wallet mode. This import call i'm suggesting also makes it possible to switch from an old full wallet never initialized on mymonero.
vtnerdok
Sources and notes
- Meeting log: Overview and Logs for the Dev Meeting Held on 2017-05-21, 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.