| *** d34dh0r5- is now known as d34dh0r53 | 14:00 | |
| rosmaita | fungi: around? | 14:41 |
|---|---|---|
| rosmaita | we have a cinder patch proposed by "OpenRefactory Research": https://review.opendev.org/c/openstack/cinder/+/869285 | 14:42 |
| rosmaita | do we need to have a "Signed-off-by: <actual author>" in the commit message, or is it OK for the actual author to be unknown? | 14:43 |
| dansmith | I assume someone had to sign the CLA right? | 14:44 |
| rosmaita | that's what i'm assuming too | 14:44 |
| rosmaita | but we don't know if the actual author using the account has signed? | 14:45 |
| dansmith | I mean do we ever? | 14:45 |
| dansmith | anyway, I'm not the expert here, just providing peanut gallery commentary while you wait for fungi :) | 14:46 |
| fungi | yeah, i mean, i can ask a lawyer, but dansmith's right we never really know for sure | 14:46 |
| rosmaita | well, there would be a level of misrepresentation if i didn't sign, and then used dan's account to commit stuff | 14:47 |
| rosmaita | but here it seems like we know even less than usual | 14:47 |
| fungi | there's something similar that goes on with ericsson commits through their "nordix" proxy infrastructure | 14:47 |
| fungi | for some reason it's company policy that they're not allowed to push code directly upstream, so they commit to a local "nordix" gerrit which then has a periodic job that pushes a change into our gerrit | 14:48 |
| rosmaita | if we allow this, i'm OK with that ... just wanted to make sure I'm not doing something non-standard if we approve this patch | 14:48 |
| fungi | i'll bring it up with jbryce when he's around, but i wouldn't hold up a legitimate (and trivial, if subtle) bug fix over it | 14:49 |
| fungi | even not being a lawyer, i really don't see how replacing "is not" with "!=" is something their employer could try to make a patent claim on later | 14:50 |
| dansmith | never underestimate a patent troll, but yeah, probably not a concern here :) | 14:50 |
| rosmaita | ok, cool, thanks | 14:51 |
| fungi | someone has agreed to the icla there, at any rate. i'll ask around and find out if we have a ccla for that organization | 14:51 |
| dansmith | https://en.wikipedia.org/wiki/Double-click#Patent | 14:52 |
| fungi | touché | 15:00 |
| rosmaita | fungi: should i hold approval on the patch until you've heard back? | 15:01 |
| fungi | also this openrefactory is rather strange... https://www.openrefactory.com/ using ai to find (and fix?) bugs | 15:02 |
| rosmaita | and not doing a good job at it, judging from that bug | 15:02 |
| rosmaita | or maybe their plan is to use AI to introduce new, more subtle bugs | 15:03 |
| fungi | yeah, the concern for this may run deeper. still, if it's a useful change then i wouldn't want to hold it up. if it's wasting your and other reviewers time then it probably merits some outreach from the foundation | 15:03 |
| rosmaita | you will have to ask jbryce what our position is on whether an AI can sign a contributor agreement | 15:03 |
| rosmaita | not sure about it being a useful change, i think it's an edge case that hasn't been encountered in real life | 15:05 |
| rosmaita | I'm going to put a -W on the patch until I hear back | 15:15 |
| fungi | yeah, for the most part bugs found through automated scanners are of dubious value | 15:16 |
| dansmith | holy hell, AI signing contracts, we're all screwed | 15:18 |
| fungi | odds are ai is already writing contracts | 15:19 |
| dansmith | well, lawyers have already been caught offering AI-sourced legal research in actual court, which turned out to be hallucinations | 15:19 |
| * dansmith has his bunker well-stocked already | 15:19 | |
| dansmith | this is the case I heard about recently: https://www.bbc.com/news/world-us-canada-657 | 15:21 |
| fungi | yeah, that court case was hilarious | 15:21 |
| dansmith | hilariously scary | 15:22 |
| fungi | he assumed the ai was correct | 15:22 |
| dansmith | yeah, although his actual argument was that he thought it was just a "really good search engine" | 15:22 |
| dansmith | as we know, humans will be the weak link in the AI wars | 15:23 |
| fungi | so he previously used google search to write his briefs | 15:23 |
| dansmith | wait until chatgpt can register domain names and synthesize news sites to support its hallucinations... | 15:23 |
| fungi | can you be sure it isn't already? | 15:23 |
| fungi | i think i saw this movie | 15:23 |
| dansmith | haha | 15:23 |
| *** melwitt is now known as jgwentworth | 18:26 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!