| 15 Nov 2018 |
 _hc | that's tricky though | 11:35:33 |
 mimi89999 | Hello | 11:36:15 |
 Coffee | That's pretty easy if you make a sub-site for every language, and go with full-manual translation. | 11:36:29 |
| Coffee | Although the problem then becomes how to prevent different language sites from diverging too much. | 11:37:04 |
| Coffee | Hey mimi89999 o/ | 11:37:29 |
| _hc | this approach seems appropriate https://forestry.io/blog/creating-a-multilingual-blog-with-jekyll/ | 11:37:51 |
| Coffee | Added to notes. | 11:39:32 |
| _hc | I have my report ready when we're ready to move on to the next topic | 11:40:06 |
| Coffee | Nothing further on this topic from me. | 11:41:19 |
| _hc | ok, here goes! | 11:41:30 |
| _hc | we were offered two more servers from GCC cfarm. | 11:41:31 |
| _hc | something like 32 thread (16 core) with 144GB RAM and 3TB disk | 11:41:50 |
| _hc | unfortunately not the most secure hosting, so maybe not for buildservers. that's still an open question | 11:42:16 |
| _hc | for release buildservers that is | 11:42:26 |
| _hc | once reproducible builds are widespread, then it is clearly yes | 11:42:37 |
| _hc | we already have the whole buildserver stack running on one | 11:42:53 |
| _hc | I've set that up as a gitlab runner | 11:43:04 |
| _hc | so there are now finally real, full builds for fdroiddata merge requests | 11:43:19 |
| _hc | the hard part is that the user who submitted the merge request must register the runner for it run on their merge requests | 11:43:46 |
| _hc | see admin#106 for how to get access for yourselves | 11:44:05 |
![@freenode_[gibot]:matrix.org](./avatar/%5Bgibot%5D) [gibot] | [admin] #106: get yourself a gitlab-ci runner for running `fdro… - https://gitlab.com/fdroid/admin/issues/106 | 11:44:07 |
| _hc | this also includes using LibScout to scan the final APK for tracker libs | 11:44:59 |
| _hc | see fdroiddata!3936 for an example | 11:45:23 |
| [gibot] | [data] !3936: WIP: use LibScout to output app libraries via git… - https://gitlab.com/fdroid/fdroiddata/merge_requests/3936 | 11:45:24 |
| _hc | I want to try running this on RFP also, using fdroid import then running a build | 11:45:50 |
| _hc | and a LibScout scan | 11:45:54 |
| _hc | also, Since we're exploring using machine learning for detecting trackers, we could easily need a giant machine just for that. | 11:46:14 |
| _hc | but that hardware is open for us to decide what to do with | 11:46:34 |
| _hc | we could allocate a full buildserver instance to each contributor | 11:46:51 |
| _hc | or shared instances | 11:46:57 |
