01:35:09 build hashes https://paste.debian.net/1221451/ 01:35:21 anyone else posted that I missed? 02:36:56 "it's... not. sorry" <- Nice one 08:18:41 no one else has gitian build hashes yet? 08:22:20 I'm beginning but it will certainly take me some hours 08:23:19 ok, just making sure I didn't miss a posting 09:06:02 hyc I tried to build yesterday but it errored out and I didn't have time to investigate 09:34:57 Does somebody know if the gitian buid should work on Debian GNU/Linux 10 (buster) ? 09:34:57 I ask since the doc says : Gitian host OS should be Ubuntu 18.04 "Bionic Beaver". 09:35:00 I guess Debian should be ok since Ubuntu is based on Debian, but I prefer asking. 09:44:18 Halver: maybe. should be ok. your best bet is to use dockrun.sh which isolates you from host OS incompatibilities 09:44:29 read the DOCKRUN.md 09:44:53 sech1: interesting. what's your build OS? 09:45:22 I use Ubuntu 18.04 VM 09:45:31 gitian builds worked there before 09:45:45 I used it to build p2pool-compatible binaries 09:45:46 strange yeah that should be perfect 09:46:15 Same VM that was successfully used to build 0.17.2.3 and hashes were correct 09:46:17 hyc: Thanks. I'll do. I'm just beginning the read. 09:46:18 First time I use Docker and 1st time I'll try a gitian build. Crossing fingers ! 09:46:27 I installed a bunch of stuff in that VM recently, maybe it broke things 09:50:11 I guess that's possible. that's why I wrote dockrun.sh; too much stuff on my host machine wasn't compatible any more 13:37:12 Starting Gitian builds now 13:46:58 Opened PR to add my new GPG key: https://github.com/monero-project/gitian.sigs/pull/157 13:50:30 Hmm, maybe I need to just add the new key instead of renaming old? https://github.com/monero-project/gitian.sigs/runs/4382112120?check_suite_focus=true selsta hyc 13:50:55 yes, add a new one 13:54:29 OK fixed: https://github.com/monero-project/gitian.sigs/pull/158 13:55:08 otherwise the old signatures can't be verified anymore 15:11:26 ... I don't exactly catch why gpg is needed in reproductible compilation 15:12:40 It's not needed for reproduction, it's needed for proving you didn't forge the hashes and fake it. 15:12:45 And that no one can pretend they were you verifying sigs. 15:20:30 Initial Linux sigs (unsigned):... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/7b47309defae595272da4ddda5c58c143ebcc7fa) 15:20:33 s/sigs/hashes/, s/// 15:20:57 use a pastebin... 15:22:36 hashes look good 15:22:52 full set has been posted in reddit https://old.reddit.com/r/Monero/comments/r60io7/cli_v01730_oxygen_orion_has_been_tagged/ 15:36:40 "use a pastebin..." <- Sorry, forget that doesn't come across well to IRC, and I will use a pastebin for the full hashes. 16:01:20 Hi, how can I compile monero-wallet-rpc without compiling other stuff like monerod? 16:06:11 make -C build/release/ wallet_rpc_server 16:07:27 change build/release to your build dir 16:26:24 "Starting Gitian builds now" <- Are all monero binaries built using deterministic building ? 16:35:27 "Are all monero binaries built..." <- They are all reproducible, yes. 16:45:09 Nice 16:58:31 https://github.com/monero-project/gitian.sigs/pull/161 16:58:40 Hashes: https://paste.sethforprivacy.com/?cb4552070ced5d8d#8YF2pefeknYTa8Bcqt8fgXp76cJLYxY9jDjphL7S1yWP 16:59:33 #158 has to be merged first for it to pass checks due to the new gpg key. 17:01:01 you can also add gpg key and sigs in one PR 17:07:58 True I can do that instead. 17:08:02 WIll close #158 17:10:34 Combined and passed in 161 🙂 20:05:20 My hashes: https://paste.debian.net/hidden/828b12fd/ 20:06:28 and https://github.com/monero-project/gitian.sigs/pull/162 20:07:04 hyc just removing all builder folders and starting from scratch helped with my error 22:47:27 An i understanding this right? You all submit a signed hash of the binaries to make sure no single person can sneak in malicious stuff? 22:48:00 prob 23:20:26 Technically it only proves that everybody is getting the same result when compiling the code 23:23:02 If anyone builds the code and gets a different result from the others, then there's an issue (either in that person's setup, or in the other people colluding to publish a malicious build) 23:27:00 And anyone "at home" can perform the same steps and verify that they're getting the same result too 23:30:34 Ok cool 23:34:42 Yup, essentially ensures that the builds of each release are the same as the source code that is publicly available. 23:34:42 All verifiers would have to collude to produce binaries that didn't match the source code, and still anyone could verify that the build was malicious/incorrect by running the verification themselves. 23:34:57 Makes it much harder to produce a malicious binary and pass it off as legitimate 🙂