Overview and Logs for the Dev Meeting Held on 2019-03-10

Geplaatst door: el00ruobuob / rehrar

Logs

<hyc> ding
<rbrunner> dong
<rehrar> oyo
<moneromooo> The witch.
<rehrar> 1. Greetings
<el00ruobuob_[m]> Hi guys!
<ArticMine> hi
<endogenic> sup y'all
<rbrunner> Hi
<ErCiccione[m]> Hi
<rehrar> 2. Fork Discussions
<rehrar> Is there anything that anyone present would like to discuss regarding the previous fork? We'll get to next point release items soon.
<hyc> looks like all quiet
<rehrar> it all went smooth as a baby's skin
<rehrar> Yay. It looks like this meeting will be short indeed.
<rbrunner> Just the first block was a real cliffhanger, right?
<el00ruobuob_[m]> it's been a long first block, besides that all's fine
<ErCiccione[m]> I take advantage of the quietness to say that i hope we will introduce a strict(er) release cycle for next release
<sarang> At some point the goal is to move to a longer cycle, no?
<hyc> maybe we should focus on making reproducible builds work
<ErCiccione[m]> i recently took a look at how nextcloud does, they release every 4 months, so it's somehow usable for us
<ErCiccione[m]> hyc: solid +1 for that
<ErCiccione[m]> sarang: i hope so, afaik we never really had a proper discussion about that
<sarang> As long as PoW tweaks are in the cards, I presume we'll stick to 6 months though
<WoomyZoomy> Why the use of "goto" in bulletproof_PROVE instead of more common flow controls?
<sarang> that's a question for moneromooo
<moneromooo> Ah, simplicity.
<rehrar> 3. Upcoming point release discussions
<ErCiccione[m]> sarang: sure, i was just pointing out that the timing is similar, so we could look at their system for inspiration. I can't find the page right now, but it was very solid, with planty of time for development, freeze and RC releases
<moneromooo> They presumably don't have to have misc other software update in sync.
<WoomyZoomy> Thanks
<rehrar> I know I had some questions, and the community at large is a little confused what is coming in the next point release, and if we have a timeline for that or not
<moneromooo> Everything that's on master, probably all/most of what's currnetly PRed and that's not super old.
<moneromooo> Except maybe the 128 bit diff change.
<hyc> * current master crashes
<moneromooo> Well, nohting says we can't add *more* fixes :)
<ErCiccione[m]> (nextcloud's release schedule: https://docs.nextcloud.com/server/stable/admin_manual/release_schedule.html)
<rehrar> because I know some features people are excited about (like Trezor, no?) are coming out with that?
<rehrar> we just going with "soon"?
<sarang> Better faster stronger bulletproofs :D
<el00ruobuob_[m]> not too soon, we need a proper code freeze for this point release
<ErCiccione[m]> I would love to just have that week of freeze for the GUI this time
<el00ruobuob_[m]> yeah me too
<moneromooo> Did the GUI not choose to do a full release already ?
<rbrunner> Should be easier now, with the threat of ASIC domination gone
<hyc> yeah, it would be simpler if it was only the monero project. but we have to coordinate with exchanges/mining pools/3rd party miner developers/ etc
<rbrunner> I think so, GUI was fully released
<ErCiccione[m]> moneromooo: yes, but would be good to have always that week for every release. I really need it to sync translations
<sarang> Is the intention to have new output selection for the point release?
<moneromooo> If you and suraeNoether are happy with the details, sure.
<sarang> It would be nice to have a rough deadline for freeze set
<sarang> (and we can discuss output selection at tomorrow's MRL meeting 17:00 UTC)
<rehrar> unfortunately hard to do without core?
<rbrunner> Maybe also depends on when master will be able to work with the 3 Monero nets again, because basically testing can start only then
<rehrar> Alright, was there any other meeting items that someone wanted to discuss?
<rehrar> now that we at the other side?
<rbrunner> The second hardfork will hardly give any problems, being rather trivial, right?
<moneromooo> Likely.
<sarang> If this is a good avenue for it, I'd like any opinions on output selection since it is likely to change at the next point release (non consensus, of course)
<sarang> We've discussed in MRL at length
<pigeons> oh the point release will be a hardfork also?
<rbrunner> No, I mean the second hardfork in what, 2 hours or so?
<pigeons> when was the first?
<moneromooo> Yesterday.
<rbrunner> er .. yesterday
<el00ruobuob_[m]> most likely 22h rbrunner
<pigeons> ok i see i thought i was reading things wrong
<rbrunner> Ah, ok, 2 days
<hyc> more than 2 days, at the current rate
<ArticMine> 0.9 days based upon 2 min blocks
<ArticMine> We are more like 15 in blocks now
<ArticMine> 15min
<rbrunner> But with nicely falling difficulty
<moneromooo> If someone can get the master daemon to crash in debug mode with log level 2 and a core, that'd be great :)
<rehrar> does anyone have any comments for sarang?
<moneromooo> Because mine ain't.
<el00ruobuob_[m]> is it the output selection about coinbase transactions highlight by sgp_ ?
<hyc> took quite a while for my debug build to crash but it eventually did. before I removed the ssl patches
<sarang> el00ruobuob_[m]: it will help to address that, but in a more general sense
<moneromooo> Did you have a better stack trace, and logs showing what was being done in other threads ?
<sarang> where output selection is poorly weighted by block "density"
<hyc> moneromooo: I gave you the stack trace, didn't check other threads unfortunately. I can try spinning that up again.
<el00ruobuob_[m]> i have to confess i don't understand very well this subject sarang
<moneromooo> Logs will help here.
<rehrar> sarang, maybe discuss tomorrow at the MRL meeting
<moneromooo> If another thread deleted some object, it might not show up on the stack trace if it's already done.
<rehrar> I think we can end here and business as usual can commence if that's alright with everyone.
<rehrar> Next weeks is community meeting, and we want to be desynced from them, so next meeting two weeks?
<hyc> rehrar: sounds good
<rehrar> thanks for coming, and everyone was here with the new time (stupid DST)
<rbrunner> Over here in Europe we will soon stop that DST nonsense
<el00ruobuob_[m]> thank you all!
<rbrunner> Maybe last change next year
<rbrunner> Likewise!


Post tags : Dev Diaries, Monero Core, Cryptography