01:11:32 Still RIP Windows CI :( 01:13:15 https://github.com/monero-project/monero/pull/6582 fixes it 01:13:34 in combination with https://github.com/monero-project/monero/pull/6531 01:25:42 merge merge merge 03:05:46 UkoeHB_: as moneromooo pointed out, there's no explicit undo code needed because aborting the DB transaction automatically discards everything 03:06:12 that is the point of DB transactions, they either completely commit or they completely rollback 03:18:26 ok interesting 03:25:20 My initial though was `BlockchainLMDB::add_spent_key` is adding key images to the database for each input in each transaction in a given block, so if two transactions in a block have the same key image then the second one will trigger the `KEY_IMAGE_EXISTS` exception. Would it be more accurate to say the key images are 'tentatively' added to the database, and then later on committed/rolled back depending on if any 03:25:20 exceptions are thrown? 03:33:47 Or maybe duplicates-within-a-block are caught somewhere else.. this stuff is difficult to track down. 09:30:37 re-re-re-tag? 09:30:38 :-P 10:29:14 your trolling is gaslighting me fluffypony, are we about to re-tag? :P 10:29:48 I think 6582 / 6531 are needed for the windows build 10:31:20 does msys need boost 1.73, or why is that a blocker? 11:00:06 TheCharlatan: I'm picking up stompies based on what sarang and selsta said 11:16:11 fluffypony: The 6531 branch equivalent is already merged 11:16:24 6582 only affects the 'Github builds' as far as I can see :-P 11:17:58 got it 11:20:54 Windows bins build fine in MSYS2. 11:22:02 so nothing in the way of release then? 11:32:07 MSYS2 has boost 1.73 now, so master build will probably fail until 6531 is merged. release-v0.16 and v0.16.0.0 should build fine. 11:44:14 TheCharlatan: Afaik, yes 11:44:18 selsta should be able to confirm :-P 11:46:33 Mmm I was about to move with signing and publish binaries. Should I hold on for now? 11:56:41 binaryFate: I think you can continue safely 11:58:24 TheCharlatan, do you want to merge the PRs? https://github.com/monero-project/gitian.sigs/pulls 12:17:11 binaryFate: Can you ping me btw once you are finished? 12:17:20 will do 12:52:11 sigs are merged now btw. 13:08:50 thanks 13:23:05 Do we have the release notes? I think would be better to post about the binaries when we have the changelog + contributors ready 13:24:01 actually, best thing would be to push the blog post and the new binaries at the same time 13:39:13 I think a reddit post would be fine as soon as the binaries are online 13:39:17 We can add both later via edit 13:41:50 binaries should be online in 1h or so 13:47:12 Please sanity check the hashes in this PR to website: https://github.com/monero-project/monero-site/pull/997 14:02:51 looks good to me binaryFate 14:04:59 gpg: Good signature from "binaryFate " [full] 14:05:45 ✔ 14:06:22 lgtm too 14:06:51 thanks, can you add a comment on the PR for the posterity? 14:09:16 yes 14:42:39 ErCiccione[m]: I have release notes 14:42:46 will send later 14:43:26 Ok. Thanks 14:47:54 CLI binaries for Nitrogen Nebula are online :) Seems good to me already, but fluffypony will flush cache to make sure 14:48:12 dEBRUYNE ^ 14:54:38 special thanks to everyone who posted their build results 14:55:16 binaryFate: Thanks 14:57:47 UkoeHB_: yes. In a transactional DB, everything is tentative until the DB txn commits. 14:58:58 release notes: https://paste.debian.net/hidden/d5bd6b70/ 14:59:07 please check for typos and if I forgot something 14:59:23 ErCiccione[m]: ^ 14:59:31 holy linebreaks batman 15:00:21 can you add the timelock verification to the ledger improvements selsta? 15:00:48 Has anyone manually tested all the links on the release notes to confirm they're correct? 15:01:25 TheCharlatan: done 15:01:44 I think the wording about checking hashes and signatures should be more direct. "You should check both that the listed hashes match the files you download, and that the signed list of hashes verifies using GPG or an equivalent tool." 15:02:19 we have used the same wording for all releases 15:02:27 But then the download server was compromised 15:02:30 I’m ok with changing it if you propose something better :P 15:03:30 "To ensure that the files you download are those originally posted by the maintainers, you should both check that the hashes of your files match those on the signed list, and that the signature on the list is valid." 15:03:43 thanks selsta. I'm gonna make the blog post 15:03:57 After all, some responses after the download server compromise happened were along the lines of "well, everyone should be checking hashes/signatures anyway" 15:06:41 like this? https://paste.debian.net/plainh/4d73b52e 15:07:57 Hooray! 15:08:29 Other thoughts on that language? 15:08:42 I think it's important to keep that as a strong recommendation, especially given the earlier compromise 16:35:27 blog post done: https://github.com/monero-project/monero-site/pull/1000 16:35:30 please revie 16:36:37 selsta i slightly edited it. You missed a ` at the end of your code block containing the hashes 16:37:08 ErCiccione[m]: it should be in the second version I sent 16:37:21 hmm or maybe nit 16:37:23 not* 16:38:04 lgtm 16:38:09 No, just checked 16:40:02 in the blog post i added links to the list of commits, to the verification guides and the source code 16:48:13 ErCiccione[m]: I will make a separate Reddit post OK? 16:48:22 Then we can do another post that links directly to the blog once it is merged 16:48:58 Sure :) 17:09:22 Oki, posted the thread on Reddit btw 17:19:32 yay 17:20:46 If somebody reviews the blog post we can have it up as well, if luigi and binaryfate can merge and update: https://github.com/monero-project/monero-site/pull/1000 17:20:58 I can review in a bit 17:21:16 it looks good to me, and I can merge/update when you want 17:22:11 L11 s/bulletproof/bulletproofs/, L86 s/te/the/ 17:22:39 Have we been capitalizing Bulletproofs elsewhere? 17:22:50 I normally capitalize it :-P 17:23:00 It was originally capitalized in its preprint and subsequent papers 17:23:11 but I don't know what's been done in other public-facing material 17:23:22 I'm probably super inconsistent when referring to it on IRC... 17:27:08 thanks iDunk fixed. I also capitalized Bulletproofs 17:30:31 If it's ok please approve on github 18:16:57 thanks hyc, it sounds like a robust approach (not that I know anything about these things)