[solved] gitlab, no anonymous access?

seth · 2023-06-11 06:33:15

You could scrape the webinterface

curl -s "https://archlinux.org/packages/extra/x86_64/alsa-card-profiles/" | sed '/<a.*>Source Files<\/a>/!d;s/.*packaging\/packages\/\([^"]*\)" title=.*/\1/g'

constantchangesforwhat · 2023-06-11 17:51:22

Perfect! Thank you!!

constantchangesforwhat · 2023-06-13 18:34:31

Although I have started to discover where the base-package appears to be non-existent, for example:

https://archlinux.org/packages/extra/any/edk2-ovmf/

and

https://archlinux.org/packages/extra/x8 … tnet-host/

when clicking the base package link in either of those two it takes you to a page that has multiple packages titled "Split Package Details", unlike the "alsa-card-profiles" where you can click on exactly the same reference point for "Base Package" and the link correctly goes to the "pipewire" package.

Any thoughts or suggestions on how to obtain the base package in these situations?

Scimmia · 2023-06-13 18:40:30

Because those don't have a package with the same name as the pkgbase.

What are you actually looking for?

constantchangesforwhat · 2023-06-13 18:54:08

To get the same result as I would/have with the "alsa-card-profiles" scenario where, when I look it up it provides the "pipewire" base package for example in this case and it actually exists.

I am unable to see why this would be different from package to package as it was not an issue prior to the "asp" deprecation.

Scimmia · 2023-06-13 18:58:27

Looking it up isn't the actual goal here, what is? Nothing you're talking about makes any sense, we need to get to the root of the xyproblem.

constantchangesforwhat · 2023-06-14 03:32:11

I want to do this to compile on my own successfully which I have been doing for the last 8 years until the "asp" program got deprecated and now I need to adjust what I created to handle the new changes to continue compiling as I have.

Scimmia · 2023-06-14 03:57:19

You're just looking for what the git repo is called? That's no different than how it was before, you don't have to click on anything, what's listed as the "Base Package" is what you use.

Grogan · 2023-06-14 21:04:20

Trying to use --universe sure is proving to be a pain in the ass, even with the --sleep parameter I added for myself.

==> WARNING: Skip cloning cppcheck: Directory exists
==> Collecting packager identity from makepkg.conf
  -> name    : undefined
  -> email   : undefined
  -> gpg-key : undefined
  -> protocol: https
==> Configuring cppcheck
==> WARNING: configured remote origin may not exist, run:
  -> pkgctl repo create cppcheck
==> WARNING: Skip cloning cppcodec: Directory exists
==> Collecting packager identity from makepkg.conf
  -> name    : undefined
  -> email   : undefined
  -> gpg-key : undefined
  -> protocol: https
==> Configuring cppcodec
==> WARNING: configured remote origin may not exist, run:
  -> pkgctl repo create cppcodec
==> Cloning cppgir ...
Cloning into 'cppgir'...
remote: Retry later
fatal: unable to access 'https://gitlab.archlinux.org/archlinux/packaging/packages/cppgir.git/': The requested URL returned error: 429
==> ERROR: failed to clone cppgir

I'm just trying to use this to pull in new packages since last time I checked them all out (I'm updating them individually with git pull before use then copying them for editing/building). pkgctl runs through and skips the existing directories as expected (quickly, there's no sleep in the condition that skips existing) and it will run through that, but as soon as it runs into one it has to clone, the above happens immediately, regardless of the amount of sleep (even 20s). It fails on the first clone operation in that circumstance and I don't understand why. I mean, if I start with an empty directory 5 seconds of sleep is still plenty to clone every package (I don't want to do that a second time, it's wasteful for me and gitlab, I just tested for a while and ctrl-c'd it).

I don't understand how this circumstance is any different from last time, picking up where I left off after failure.

I'd sure like a better way to get the distro sources.

Last edited by Grogan (2023-06-14 21:19:00)

bulletmark · 2023-06-14 22:53:53

People, when you paste command output please include the actual command (and arguments of course) in the output. It's only 1 more line!

Grogan · 2023-06-15 00:09:58

pkgctl repo clone --protocol=https -j 1 --sleep 5 --universe

The --sleep 5 in the command is only relevant to me, I added it to avoid the rate throttling. The same command I was using in posts above, to clone --universe

What I was more trying to show was that it was running through and skipping existing directories locally, but failing on the first actual clone operation for reasons I don't understand.

constantchangesforwhat · 2023-09-06 12:36:40

Has any one found any way to get past this issue? It is occurring with more packages and the apparent deflection to a direct answer is becoming more of a hindrance than a solution. For example if now you want to rebuild the openjdk kit (version irrelevant) you cannot do so because the constant username/password for gitlab issue is preventing this. Where as before this username/password issue was non-existent. The answers provided here seem to disregard that fact.

Lone_Wolf · 2023-09-06 12:52:14

While pkgctl repo clearly has issues keep in mind it's designed for use by archlinux developers and Package Maintainers (formerly known as TU) , not for users.

The proven method of manually retrieving a git clone some-url still works fine.

constantchangesforwhat · 2023-09-06 16:28:42

Ok this helps get things further along in my build script, especially regarding the aforementioned java. Thank you kind sir.

loqs · 2023-09-06 18:22:20

$ pkgctl repo clone --protocol=https java-openjdk
==> Cloning java-openjdk ...
Cloning into 'java-openjdk'...
remote: Enumerating objects: 214, done.
remote: Counting objects: 100% (7/7), done.
remote: Compressing objects: 100% (7/7), done.
remote: Total 214 (delta 2), reused 0 (delta 0), pack-reused 207
Receiving objects: 100% (214/214), 35.79 KiB | 916.00 KiB/s, done.
Resolving deltas: 100% (121/121), done.
==> Collecting packager identity from makepkg.conf
  -> name    : undefined
  -> email   : undefined
  -> gpg-key : undefined
  -> protocol: https
==> Configuring java-openjdk
$ pkgctl repo clone --protocol=https java8-openjdk
==> Cloning java8-openjdk ...
Cloning into 'java8-openjdk'...
remote: Enumerating objects: 290, done.
remote: Counting objects: 100% (4/4), done.
remote: Compressing objects: 100% (4/4), done.
remote: Total 290 (delta 0), reused 0 (delta 0), pack-reused 286
Receiving objects: 100% (290/290), 64.37 KiB | 1.69 MiB/s, done.
Resolving deltas: 100% (147/147), done.
==> Collecting packager identity from makepkg.conf
  -> name    : undefined
  -> email   : undefined
  -> gpg-key : undefined
  -> protocol: https
==> Configuring java8-openjdk

It does appear to be an issue with you using the wrong repo/pkgbase name that triggers the request for username/password.

Arch Linux

#26 2023-06-11 06:33:15

Re: [solved] gitlab, no anonymous access?

#27 2023-06-11 17:51:22

Re: [solved] gitlab, no anonymous access?

#28 2023-06-13 18:34:31

Re: [solved] gitlab, no anonymous access?

#29 2023-06-13 18:40:30

Re: [solved] gitlab, no anonymous access?

#30 2023-06-13 18:54:08

Re: [solved] gitlab, no anonymous access?

#31 2023-06-13 18:58:27

Re: [solved] gitlab, no anonymous access?

#32 2023-06-14 03:32:11

Re: [solved] gitlab, no anonymous access?

#33 2023-06-14 03:57:19

Re: [solved] gitlab, no anonymous access?

#34 2023-06-14 21:04:20

Re: [solved] gitlab, no anonymous access?

#35 2023-06-14 22:53:53

Re: [solved] gitlab, no anonymous access?

#36 2023-06-15 00:09:58

Re: [solved] gitlab, no anonymous access?

#37 2023-09-06 12:36:40

Re: [solved] gitlab, no anonymous access?

#38 2023-09-06 12:52:14

Re: [solved] gitlab, no anonymous access?

#39 2023-09-06 16:28:42

Re: [solved] gitlab, no anonymous access?

#40 2023-09-06 18:22:20

Re: [solved] gitlab, no anonymous access?

Board footer