when i try Server = https://mirrors.manjaro.org/testing/$repo/$arch , i only get 404 - i guess i’m understanding you wrong?
After this update in Xfce , all the application icons packs with name: Menda-Circle ; Menda-Square and Vertex-Maia are apparently non-existent …
All the Whisker menu are with the correct circle / square color but the detail is gone …
All the others ( Faenza ) are correct , all the non default installation system are all correct.
How can I reload the Menda-Circle icon pack?
yes you do, click that link and follow the instructions.
The “link” doesn’t contain any URL for me (checked in multiple browsers to be sure, but even the HTML doesn’t contain a link) - i kinda feel really dumb right now if that is some oversight on my side ^^
yeah, that’s pretty much what i did - i’m getting 404s though when doing pacman -Syy
is not the same with
https://mirrors.manjaro.org/repo/testing/$repo/$arch
1 Like
oohh, no i see it - and that way, it works 
Thanks!
For me, in Australia, https://mirrors.manjaro.org/repo gets 404.
https://mirrors.manjaro.org/ without ‘repo’ shows the message.
(Not a problem for me, as I use stable branch on a local mirror,
but thought you might want to know.)
You’re not supposed to access and browse the CDN from your web browser, that’s probably why.
1 Like
Sadly i am not allowed to post links…
Last known working versions are 5.14.2 and 5.13.15.
Here is the bugreport:
https://bugzilla.kernel.org/show_bug.cgi?id=214455
cdn77 is in australia
yes, as @omano written, we can’t list directory
but we can test one file
curl -IL "https://mirrors.manjaro.org/repo/testing/core/x86_64/core.db"
HTTP/2 200
date: Sat, 18 Sep 2021 06:34:00 GMT
content-type: application/octet-stream
content-length: 173248
last-modified: Thu, 16 Sep 2021 10:50:07 GMT
etag: "6143215f-2a4c0"
x-accel-expires: @1632983569
server: CDN77-Turbo
Other tool (in aur) for display real download speed
httpstat -L "https://mirrors.manjaro.org/repo/testing/core/x86_64/core.db"
HTTP/2.0 200 OK
Server: CDN77-Turbo
Accept-Ranges: bytes
Access-Control-Allow-Origin: *
Content-Length: 173248
Content-Type: application/octet-stream
Date: Sat, 18 Sep 2021 06:20:07 GMT
Etag: "6143215f-2a4c0"
Last-Modified: Thu, 16 Sep 2021 10:50:07 GMT
X-77-Cache: HIT
X-77-Nzt: AbldAhRO7WTvhAAAAA==
X-77-Nzt-Ray: mCtPyLp42qs=
X-77-Pop: parisFR
X-Accel-Expires: @1632982675
X-Age: 132
X-Cache: HIT
Body discarded
DNS Lookup TCP Connection TLS Handshake Server Processing Content Transfer
[ 66ms | 28ms | 142ms | 49ms | 169ms ]
| | | | |
namelookup:66ms | | | |
connect:95ms | | |
pretransfer:237ms | |
starttransfer:286ms |
total:456ms
2 Likes
For switching mirrors simple use the pacman-mirrors cli tool.
Command to execute for Global Mirror (CDN77)
sudo pacman-mirrors -c Global
Done 
But I noticed it’s not faster in download speed compared to local Mirrors (Germany) only the mirrors are synced more often.
sudo pacman-mirrors —geoip
3 Likes
Something still going on, there is not much server marked as ‘up to date’ here https://repo.manjaro.org/, and servers marked as such are not, for example manjaro.re is broken despite being regularly synched according to https://repo.manjaro.org/
//EDIT: even weirder, mirrors marked as out of date, are actually up to date, like all the *.garr.it mirror, all marked as out of date, but synched regularly, forcing these mirrors gives me the updates…
Whatever, nobody cares, switching to Manjaro CDN and ‘eff’ this problem. Looking forward to seeing this arise on Stable branch…
4 Likes
WARNING! Libre Office 7.2.1 is broken, because filtering doesn’t work correctly.
There is no way to uncheck or check selected items on the filter list in Calc. I downgraded to 7.2.0 and the problem is gone. Without properly working filters, LO is unusable for spreadsheet users.
If the same issue happens for more people, I would recommend withholding that update from stable branch.
Well, this patch slipped thru to many kernel series: [PATCH 5.14 286/334] net: qrtr: make checks in qrtr_endpoint_post() stricter - Greg Kroah-Hartman
- 5.14.4
- 5.13.17
So you may want to change to linux510 to have a working wifi. Check with Dan Carpenter from Oracle about this. Seems this needs to be backported to both series: kernel/git/stable/linux.git - Linux kernel stable tree
Kernel runs:
Did you report this upstream and linked the bug report?
I was about to write about the failing wifi for 5.13 [EOL] & 5.14 and to install 5.10 in order to test with that as well, thanks for the heads up 
Don’t know if it is related to this newest kernel, I would assume so, after the update I kept getting errors like this:
Sep 20 00:15:25 tmo systemd-journald[6206]: Missed 30 kernel messages
Sep 20 00:15:25 tmo kernel: pcieport 0000:00:1d.5: [ 0] RxErr (First)
Sep 20 00:15:25 tmo systemd-journald[6206]: Missed 29 kernel messages
Sep 20 00:15:25 tmo kernel: pcieport 0000:00:1d.5: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
Sep 20 00:15:25 tmo systemd-journald[6206]: Missed 24 kernel messages
Sep 20 00:15:25 tmo kernel: pcieport 0000:00:1d.5: device [8086:a11d] error status/mask=00000000/00002000
Sep 20 00:15:25 tmo systemd-journald[6206]: Missed 30 kernel messages
Sep 20 00:15:25 tmo kernel: pcieport 0000:00:1d.5: AER: Multiple Corrected error received: 0000:00:1d.5
device a11d == 00:1d.5 ==
$ lspci -v | grep 1d.5
00:1d.5 PCI bridge: Intel Corporation 100 Series/C230 Series Chipset Family PCI Express Root Port #14 (rev f1) (prog-if 00 [Normal decode])
had to add pci=noaer to the kernel line to get rid of the advanced error reporting.
Can you share how to reproduce this step by step?