12 Nov 2018 |
Mathijs | The easiest way I can think of would be using the legal system | 14:07:09 |
Mathijs | something like a notary | 14:08:02 |
Mathijs | but that increases the attack surface a lot as well of course | 14:08:23 |
Tovok7 | It would be nice if there would be a multi-party signature scheme where we can add/remove parties. | 14:08:33 |
Tovok7 | (that was one of the roadblocks in my decentralized repo brainstorming, not that the idea was not so much about decentralized distribution but rather building/signing of the repo) | 14:10:22 |
Mathijs | In reply to @freenode_Tovok7:matrix.org It would be nice if there would be a multi-party signature scheme where we can add/remove parties. that's not a thing in android, right? | 14:16:44 |
Tovok7 | that's independent of android. There are multi-party signature schemes, but I don't know enough about them to know if there are any that are feasible for our requirements. Also, that's an area you want to be absolutely sure that there's no flaws before using its output as a signing key | 14:18:08 |
Mathijs |
, that's an area you want to be absolutely sure that there's no flaws before using its output as a signing key
agreed
| 14:19:21 |
Mathijs | I'm no cryptography expert (obviously) so I won't be of much help there | 14:19:47 |
_hc | Tovok7: ever have a chance to try reprobuilds since the schweiger madness? | 14:25:21 |
_hc | I've since moved schweiger to stable internet, and also gotten my new connection stable, so I could move it back there | 14:25:42 |
Tovok7 | no, need to figure our what I can't connect via Tor | 14:25:44 |
_hc | plus there is the new gcc box | 14:25:47 |
Tovok7 | *out | 14:25:51 |
_hc | do you have the onion address for schweiger? | 14:26:16 |
_hc | its been stable for me for over a week | 14:26:33 |
Tovok7 | yup I have, but a different netcat than in your config | 14:27:07 |
_hc | then use the ssh only trick | 14:27:28 |
_hc | Tovok7: or I could set up your own complete buildserver instance on a gcc box | 14:29:48 |
Tovok7 | whatever server I can connect to, but atm the issue is more lack of time on my side | 14:42:02 |
_hc | Tovok7: you should try some things on the live infrastructure then, i.e. just commit to teh metadata file in fdroiddata | 15:13:52 |
Tovok7 | that's a nightmare for debugging and has huge rountrip times | 15:24:17 |
_hc | sure, but requires no setup time :) | 16:26:14 |
_hc | works well for very async | 16:26:24 |
_hc | e.g. only checking the job once a week | 16:26:31 |
Mathijs | Marcus (BubuIIC): the hard part about compiling riot android was the jitsi dependencies, not so much the matrix android sdk ones right? | 16:33:35 |
@bubuiic:matrix.org | matrix-sdk needs jitsi as well already | 16:44:16 |
Mathijs | oh, that's unfortunate | 16:47:01 |
@bubuiic:matrix.org | if you just want to test something you can use the precompiled versions in their repo | 16:47:43 |
Mathijs | nah | 16:47:52 |