| 13 Jun 2025 |
 K900 | @magic_rb | 15:19:01 |
| K900 | Because like, UPnP does not make sense behind CGNAT | 15:19:10 |
 magic_rb | yeah | 15:19:11 |
| magic_rb | public and static by contract | 15:19:15 |
| K900 | Then have fun I guess | 15:19:43 |
| K900 | I haven't even tried to set up UPnP because, well, CGNAT | 15:19:52 |
| magic_rb | yeah i dont want to | 15:21:19 |
| magic_rb | if it was up to me i wouldnt | 15:21:28 |
| K900 | I mean if they're using it for gaming, you can also probably just hard forward some UDP ports | 15:21:59 |
| magic_rb | "BuT I WaNT uPNpD" | 15:22:23 |
| magic_rb | * "BuT I WaNT uPNp" | 15:22:26 |
| magic_rb | 🤷♂️ | 15:22:30 |
| magic_rb | also the answer is no | 15:22:58 |
| magic_rb | ew ew ew | 15:27:12 |
| magic_rb | ew upnp | 15:27:18 |
| magic_rb | miniupnpd has to run as root because it wants to directly write into the fucking firewall | 15:27:33 |
| magic_rb | and there is no backend which would allow me to split the upnp implementation from the "editing the firewall" part | 15:28:04 |
| magic_rb | ew | 15:28:09 |
 emily | K900: I am deeply unhappy to report that that patch does, in fact, fix PPPoE negotiation | 15:46:54 |
| emily | and v6 | 15:47:15 |
| emily | Q: why is it sending all the same things as OpenWrt and getting different results?
A: IT'S NOT SENDING THEM | 15:47:35 |
| K900 | Uhh | 15:48:16 |
| K900 | OK that's interesting | 15:48:19 |
| K900 | OK I wonder if | 15:49:04 |
| K900 | Just that last header change | 15:49:09 |
| K900 | Is what does it | 15:49:12 |
| emily | by that patch | 15:49:34 |
| emily | I mean the frank-w tree one | 15:49:36 |
| emily | not the lkml one | 15:49:38 |
| emily | I just pulled down https://github.com/frank-w/BPI-Router-Linux/commit/1a6d1e12425127e5f9fe3012581d756500342856 | 15:49:44 |
