--- Log opened Fri Aug 09 00:00:35 2019
00:33 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has joined #rust-bitcoin
01:11 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has quit [Remote host closed the connection]
01:29 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has joined #rust-bitcoin
02:24 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has quit [Remote host closed the connection]
04:48 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has joined #rust-bitcoin
06:23 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-wgneqaurzjhyksyt] has joined #rust-bitcoin
07:27 < dongcarl> Taking another look at the unmerged PRs from last week today. andytoshi stevenroose Let me know if there are other things needing attention.
07:27 < real_or_random> hm travis consistently fails here on setting up the windows / rust stable build -.- https://travis-ci.org/rust-bitcoin/rust-secp256k1/builds/569555866?utm_source=github_status&utm_medium=notification
07:28 < dongcarl> real_or_random: Restarted, you said it was consistent?
07:28 < dongcarl> real_or_random: Across all rust versions or just stable?
07:29 < real_or_random> yeah, I restarted it a few times already
07:29 < real_or_random> the same step takes around 530s for beta and nightly, I guess it just takes > 600 s on stable
07:29 < real_or_random> but that step is not under our control
07:30 < dongcarl> Perhaps... rust-bitcoin should move to GitHub CI first :p
07:30 < dongcarl> They claim to have OSX and windows support already
07:30 < real_or_random> https://travis-ci.community/t/windows-builds-time-out-during-cache-restore/4642
07:33 < real_or_random> maybe disabling the cargo cache helps... hm
07:35 < real_or_random> ah it's possible to set a cache timeout. https://docs.travis-ci.com/user/caching/#setting-the-timeout but this seems unrelated because it defaults to 3 min and what we hit here is the 10 min limit
07:36 < dongcarl> smh... it seems like 10% of time is always spent on travis...
07:37 < dongcarl> Perhaps moneyball can help us get access to Github CI and test it out?
07:37 < dongcarl> in the mean time elichai2 try updating your .travis.yml to not cache, just to see if that's the problem
07:38 < real_or_random> do we need moneyball? it seems one can just "register" for it? or do you need approval
07:38 < elichai2> yeah the caching is so bad with the windows one
07:39 < elichai2> dongcarl: can we even register for the github one? I think it's still specific people only (i've registered for the beta a while ago and still haven't recieved anything)
07:39 < real_or_random> I nuked the cache for this PR for now... let's see
07:40 < elichai2> if anyone have a spare server I could set up a drone.io pretty fast. although this would be for linux only. maybe we should use azure/appveyor for the windows ones
07:40 < dongcarl> real_or_random: I've registered personally... But never got approved. :-/
07:41 < dongcarl> elichai2: I think GitHub CI > drone.io for this
07:41 < moneyball> I'm on an email thread with GitHub and several developers. I don't really want to be the gate here but am happy to connect a few more devs to GitHub for beta access.
07:42 < elichai2> dongcarl: as far as I know it's just github making Public Relations, I didn't see any actual CI yet
07:44 < andytoshi> dongcarl: everything tagged 0.19 is blocking 0.19
07:45 < andytoshi> but travis issues are fucking things up for rust-secp and rust-bitcoin
07:46 < dongcarl> moneyball: Understood, it'd be great be connected
07:51 < real_or_random> hm clearing the cache has helped this job to get across the cache setup step: https://travis-ci.org/rust-bitcoin/rust-secp256k1/jobs/569555871#L33
07:52 < dongcarl> real_or_random: good to know!
07:52 < dongcarl> Seems like #289 and #255 are blocked on stevenroose right now, looking at the rest...
07:53 < real_or_random> dongcarl: haha, I'm taking that back
07:54 < dongcarl> :-/
07:54 < real_or_random> it seems that the build was restarted after building up the cache? not sure if this is intended
08:27 < elichai2> I liked travis in the past and now I really hate that thing.
08:28 < real_or_random> well they said windows builds are in beta
08:29 < elichai2> but not just them. lately it's been timing out on github in the linux builds too
08:30 < andytoshi> in fairness, it's pretty generous of them to let us run fuzztests in free CI
08:30 < andytoshi> but yeah, it's really frustrating
08:33 < real_or_random> I'm working on a PR that disables some of the cache. this should hopefully speed up the windows builds a lot (and slow down linux builds just a little)
08:34 < real_or_random> s/some/some parts
08:48 < elichai2> real_or_random: doing some tests of my own to see if it's worth it: https://ci.appveyor.com/project/elichai/rust-secp256k1/builds/26592410
09:02 < real_or_random> I'm really wondering what they're doing. restoring the cache takes 20s on linux and > 600s on windows
09:04 < elichai2> not sure. but splitting the work between travis->linux appveyor->windows should also give us faster CIs (they'll do it in parallel to each other)
09:04 < elichai2> real_or_random: I even manage to catch the correct OS on the windows weird env vars https://ci.appveyor.com/project/elichai/rust-secp256k1/builds/26592675/job/n5gvqg0853455eb2#L90 :)
09:05 < real_or_random> this could help too: https://travis-ci.community/t/current-known-issues-please-read-this-before-posting-a-new-topic/264/15
09:37 < real_or_random> when I clear a branch cache, it falls back on the master cache *shrug*
09:44  * elichai2 sigh
09:44 < elichai2> I'm trying to fix emscripten to also pass tests on windows, hope to open a PR for appveyor yml soon
09:45 -!- dpc [dpcmatrixo@gateway/shell/matrix.org/x-empbwlmqcegqsmla] has quit [Remote host closed the connection]
09:50 < real_or_random> okay this looks reasonable now: https://travis-ci.org/rust-bitcoin/rust-secp256k1/builds/569916636 I'll try to restart it to see if the cache will kill it again
09:53 < elichai2> real_or_random: does the windows tests take only 7 minutes *without* cache?
09:53 < real_or_random> yes, indeed. but I also disabled windows defender (see link above)
09:53 < elichai2> lol
09:54 < real_or_random> not sure if this has any impact. if this works here, I'll test
09:54 < elichai2> that's awesome
09:54 < real_or_random> but apparently unpacking the cache with defender on is slower than installing rust and all deps from scratch ...
09:54 < elichai2> appveyor without cache, the stable tests take upwards of 10 minutes. but they include emscripten too. i'll see soon how much the beta/nightly ones take
09:55 -!- dpc [dpcmatrixo@gateway/shell/matrix.org/x-zezvfqtzphmqizux] has joined #rust-bitcoin
09:55 < real_or_random> interestingly, rust stable on linux took much longer than the other linux builds
09:56 < real_or_random> whereas on windows all jobs take all around the same time, no matter what the rust version is
09:56 < elichai2> yeah on windows travis doesn't do emscripten
09:56 < elichai2> only on linux stable
09:59 < real_or_random> oh sure
09:59 < elichai2> and I think travis is only 64bit machines
10:55 < elichai2> I'm starting to give up on emscripten in windows. For some reason it doesn't find it in path and it's so hard to debug something with no way to reproduce it locally(no windows machine)
11:07 < real_or_random> https://github.com/rust-bitcoin/rust-secp256k1/pull/147
11:18 -!- reallll [~belcher@unaffiliated/belcher] has joined #rust-bitcoin
11:22 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 246 seconds]
11:23 -!- reallll is now known as belcher
11:50 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has quit [Read error: Connection reset by peer]
11:51 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has joined #rust-bitcoin
12:27 -!- DeanWeen [~dean@gateway/tor-sasl/deanguss] has quit [Remote host closed the connection]
12:27 -!- DeanWeen [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
13:15 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has quit [Remote host closed the connection]
13:15 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has joined #rust-bitcoin
13:15 -!- mauz555 [~mauz555@2a01:e35:8ab1:dea0:b523:9b26:7d33:44f5] has quit [Read error: Connection reset by peer]
13:26 < elichai2> ok appveyor is just stupid. https://ci.appveyor.com/project/elichai/rust-secp256k1/builds/26597799/job/j50pac30jv7d860o idk what's going on
13:36 -!- DeanWeen [~dean@gateway/tor-sasl/deanguss] has quit [Ping timeout: 260 seconds]
16:31 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-wgneqaurzjhyksyt] has quit [Quit: Connection closed for inactivity]
18:08 -!- TamasBlummer1 [~Thunderbi@p200300DD6712643901BB4490E7A0D008.dip0.t-ipconnect.de] has joined #rust-bitcoin
18:10 -!- TamasBlummer [~Thunderbi@p200300DD67126430586CA50BF0656CBC.dip0.t-ipconnect.de] has quit [Ping timeout: 264 seconds]
18:10 -!- TamasBlummer1 is now known as TamasBlummer
18:21 -!- sanket1729 [~sanket172@72.36.89.11] has joined #rust-bitcoin
21:00 -!- DeanWeen [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
--- Log closed Sat Aug 10 00:00:35 2019