00:22 <schmars[m]> ah digitalocean now, ok

03:59 <mrkiko> hauke: I saw the GDSP support was backported. Just wanted ot say thank you !! I am grateful.

04:01 <mrkiko> ynezz: thanks!!

05:05 <owrt-images-builds> Build [#23](https://buildbot.openwrt.org/images/#/builders/55/builds/23) of `openwrt-24.10_bcm27xx/bcm2708` failed.

06:38 <schmars[m]> ynezz: i built new workers in a new location (on a home dsl line) and they look good so far, while the old workers are still failing. so, i'm gonna book this under "really fucked up mysterious glitches in the cosmos". we'll see if we can figure out the connectivity issues in these locations, nothing more for you to do, thanks for your help

07:34 <ynezz> schmars[m]: around midnight yesterday I've finished an IP based rate limiting on git.o.o where basically after 5r/s it will delay next requests up to 15r/s and then cut the IP with 429

07:37 <ynezz> schmars[m]: this is applied only on the "heavy" luci.git, packages.git, openwrt.git repos: `delaying request, excess: 14.669, by zone "limit_per_ip_heavy" .... "GET /project/luci.git/info/refs?service=git-upload-pack`

08:20 <totkeks> robimarko: unfortunately it is not a patch/email, its a PR on github https://github.com/openwrt/packages/pull/22167. I doubt I can edit their PR, only reviewers can, if they allowed it.

08:40 <owrt-images-builds> Build [#24](https://buildbot.openwrt.org/images/#/builders/119/builds/24) of `openwrt-24.10_ath79/nand` completed successfully.

09:28 <robimarko> totkeks: It doest matter, you just get the PR as a patch and apply it localy, do the changes and open a new PR

09:56 <totkeks> okay, thank you. I will try that. because that vim issue is very annoying :D

11:04 <owrt-images-builds> Build [#24](https://buildbot.openwrt.org/images/#/builders/111/builds/24) of `openwrt-24.10_lantiq/xrx200` failed.

12:21 <rua> hauke: Could https://openwrt.org/toh/linksys/ea4500_v3#upgrading_from_version_before_v24100 be added to https://openwrt.org/releases/24.10/notes-24.10.0-rc2#upgrading_to_2410 ?

13:28 <owrt-images-builds> Build [#29](https://buildbot.openwrt.org/images/#/builders/232/builds/29) of `main_at91/sam9x` completed successfully.

13:43 <owrt-images-builds> Build [#23](https://buildbot.openwrt.org/images/#/builders/187/builds/23) of `openwrt-24.10_ramips/mt7620` failed.

14:09 <efahl> does anyone know a way to scan all the dts files and grab out a value? I want to scan them all and make sure 'board_name' matches metadata/profiles.json

14:10 <efahl> motivation here: https://github.com/openwrt/openwrt/pull/17155 d1 target was missing proper names in profiles.json/metadata

15:07 <owrt-images-builds> Build [#29](https://buildbot.openwrt.org/images/#/builders/216/builds/29) of `main_mpc85xx/p1020` completed successfully.

15:59 <nick_> Did anyone try source tree override on current snapshot? seems not work anymore :/

17:09 <efahl> I used the git-src link in the recent rpcd PR I did, worked fine

17:15 <efahl> nick_ like this: ln -s /home/efahlgren/openwrt/rpcd/.git/ package/system/rpcd/git-src

17:16 <efahl> then edit the hash in the rpcd/Makefile to point my desired commit

18:14 <owrt-images-builds> Build [#28](https://buildbot.openwrt.org/images/#/builders/126/builds/28) of `main_bmips/bcm6362` failed.

19:23 <totkeks> Is there any tool used to autoformat / conform code on commit / PR? I just went through a Makefile in the packages repo and it uses sometimes spaces, sometimes tabs. Though not both together at the same time.

19:25 <efahl> yeah, we use 'ruff' for formatting the Python in the ASU server, but I've seen nothing in main for any formatting

19:26 <efahl> I just use "leave no tracks", make sure your changes are just like what you see in the bit you're editing

19:30 <hauke> mrkiko: We cherry picked all commits which were in main branch about 1 week ago and are not related to apk into the 24.10 branch

19:31 <totkeks> efahl, thanks. it just feels so weird, seeing all this. I'm used to other processes from my day jobs as a programmer.

19:34 <tmn505> totkeks: in Makefile You can use spaces for indetation except lines starting with shell command, those need to start with tabs

19:35 <totkeks> oh, so it is intentional. thanks!

19:36 <hauke> totkeks: You can write a github action which fixes this or warns people about such problems in PRs

19:36 <hauke> rua: I will have a look at this

19:37 <efahl> aren't there specific places were leading spaces (not tabs) are required? I'm thinking description text (or am I misremembering something?)

19:39 <tmn505> maybe You think about yaml. Makefie is good on both with exeption I mentioned

19:40 <totkeks> yeah yaml requires spaces. and a fixed number of those.

19:50 <totkeks> I guess I need to add some extensions to my VSCode for makefiles and such

20:10 <robimarko> Makefiles require a mix of tabs and spaces

20:10 <robimarko> Descriptions AFAIK require spaces

20:10 <robimarko> And then its the usual make logic where spaces get evaluated by shell

20:11 <robimarko> *I reversed it, tabs are shell

20:17 <totkeks> Mh, I guess I have to setup my environment to compile & run my changes before making a PR

20:18 <robimarko> I would not advise sending anything before testing first

22:36 <owrt-images-builds> Build [#29](https://buildbot.openwrt.org/images/#/builders/192/builds/29) of `main_gemini/generic` completed successfully.

23:20 <schmars[m]> ynezz: just fiy, our working theory about our buildbot connectivity stuffs is: it's black friday and ddos season, and we're victim to some upstream being very liberal with connection resets. our workers dont' actually have issues, they simply receive proper TCP RST in the middle of clone, and i assume these are not coming from g.o.o or fastly or github ;)