06:46:57 <rbrunner> Just wondering: Is everything working as intended if the daemon (started without any special options) defaults to not accepting SSL connections from wallets
06:47:27 <rbrunner> and the CLI wallet in turn (again without any special options) defaults to *using* SSL, which results in no connection to the daemon?
06:47:39 <rbrunner> Just tried today with everything compiled from current master
06:49:01 <rbrunner> I seem to have that state currently on my Linux system
09:11:26 <blackpenguin> rbrunner: I thought default is accept SSL if client supports it (and create an ssl-cert on the fly)
10:52:28 <rbrunner> blackpenguin: I do wonder whether something is special with my system, that's why I hope for somebody in the know to tell how it definitely *should* be
11:02:30 <rbrunner> To check / reproduce, create a new wallet.
11:10:14 <moneromooo> Bug ? I have no scrollback here. What is it about ?
11:11:00 <rbrunner> I start the daemon without any special options, then the CLI wallet without special options, and the CLI wallet does not get a connection to the daemon
11:11:21 <moneromooo> Same host/VM (sorry) ?
11:11:26 <rbrunner> Reason: My daemon defaults to not using SSL, my CLI wallet wants to use SSL, the two do not find each other for creating a new wallet
11:11:30 <rbrunner> Yes, all local
11:12:04 <moneromooo> Does it work if you specifically set ssl to autodetect (which ought to be the default) ?
11:12:06 <rbrunner> It works as soon as I tell the CLI wallet to give up SSL
11:14:00 <rbrunner> --daemon-ssl disabled
11:14:44 <rbrunner> Yes, it shows "autodetect" as default in the help text: --daemon-ssl arg (=autodetect)
11:14:59 <rbrunner> No, with that it does not work on my particular Linux system
11:15:17 <rbrunner> If I try to create a new wallet, or restore one with seed
11:15:21 <moneromooo> Does it work with SSL set to enabled on both ?
11:15:36 <rbrunner> Did not yet try that combination
11:17:04 <moneromooo> I don't recall how the whitelisting works now, but also try looking for a "allow any cert" flag.
11:17:10 <rbrunner> Will try now
11:19:56 <rbrunner> Nope, daemon with SSL enabled and wallet started like that does not work: ./monero-wallet-cli --testnet --daemon-ssl enabled --daemon-ssl-allow-any-cert
11:20:49 <rbrunner> What I wonder the most with all that: That the default (as it seems "autodetect" on both sides, right?) does not work for me
11:23:10 <rbrunner> Somehow I can't believe that it should be that broken, but on the other hand no idea what should be special with my system, or what I could do wrong ...
11:24:28 <rbrunner> Maybe, because I self-compile, some library with a wrong version comes in that then has this strange result?
11:38:26 <moneromooo> Try allowing any cert on the daemon too.
11:39:15 <moneromooo> Could be, if they don't have matching ciphers. But seems unlikely unless you built on different machines.
12:25:07 <moneromooo> There's openssl s_client which can be used to probe a server (and possibly a s_server, let's check)
12:26:05 <moneromooo> There is. Then you can try s_client vs monerod, and monero-wallet-cli vs s_server, and see what it moans about.
12:26:27 <moneromooo> I do not remember how to use, but there are man pages for s_client(1) and s_server(1).
12:27:15 <moneromooo> IIRC monero allows only a very short number of suites.
12:51:34 <rbrunner> Thanks for the support moneromooo, but I think I put it away, as I tried now with official release software, and there it works.
14:34:10 <selsta> rbrunner: something that works in release and not master seems like a bug
14:35:07 <rbrunner> No, I think the problem is merely local on my machine somehow: My *self-compiled* binaries behave strangely, for whatever reason
14:35:44 <rbrunner> I noticed that I had the Linux headers as a "held back" package for a long time and finally updated properly, but the problem persists, so ...
14:36:02 <selsta> when did you start noticing this issue?
14:36:39 <rbrunner> A few weeks ago for the first time, then I forgot, and faced it again today.
14:37:11 <rbrunner> What's the most strange about this: Some wallets that I created earlier work. Is it stored in the wallet files whether to use SSL or not with the daemon?
14:48:27 <moneromooo> No.
14:55:17 <rbrunner> Then it's even stranger :)
14:57:41 <moneromooo> Do the logs tell you what it's doing on connection (on both client and server) ?
15:01:27 <rbrunner> Never checked so far. Do people think something interesting could possibly come out of a further inspection, beyond merely "rbrunner7's system is flaky, so there"? If yes, I can continue to investigate.
15:10:13 <moneromooo> We do not know whether it is interesting without debugging. It depends how far you want to go.
15:18:40 <one-horse-wagon[> rbrunner: I run Debian linux on my computers.  If you do also and if you like, tell me how you "self-compiled" the binaries and I'll try the same on my computers.  That way it will verify where the bug is at.
15:19:06 <selsta> .merge+ 8593 8594 8590 8580 8578 8571 8570 8569 8564 8543 8529 8527 8525 8517 8516 8355 8319
15:19:06 <xmr-pr> Added
15:21:31 <rbrunner> one-horse-wagon: I am also on Debian. I just check out the Monero source in Git and then "make" or "make debug", that's it. Absolutely nothing fancy.
15:21:57 <rbrunner> Dependencies installed a way ago already, never looked too closely whether they get properly updated automatically however.
15:23:24 <one-horse-wagon[> rbrunner: Not to insult you, please, but how long has it been since you did apt-get update and apt-get upgrade?
15:24:08 <rbrunner> One hour?
15:24:34 <rbrunner> Installed aptitude today to finally get rid of "held back" packages in a safe way
15:24:41 <one-horse-wagon[> rbrunner: LOL.  Sorry.
15:24:51 <rbrunner> In the hope that this brings the solution.
15:25:04 <moneromooo> I could matter if you built with some libs and are running with others.
15:25:18 <rbrunner> No problem, do ask, I have done absolutely dumb things in the past.
15:25:27 <moneromooo> Ideally sonames would find what;s needed but you never know.
15:25:49 <moneromooo> Like updates in /usr/share.
15:26:10 <moneromooo> Anyway, s_client/s_server and monero logs are the next things to try.
15:26:39 <rbrunner> Ok, maybe a start an attempt this Sunday. It's kind of interesting, the problem.
15:27:17 <rbrunner> But as I already mentioned, probably not urgent, because our release software does *not* exhibit that problem on my system.
15:31:13 <one-horse-wagon[> rbrunner: I'll try building the binaries this afternoon and let you know here, what happened.
15:32:04 <rbrunner> Splendid, a cliffhanger :)
18:09:59 <binaryFate> Binaries for v0.18.1.2 are now available at getmonero.org, great job everyone