Re: Disarchive update

[zimoun]

Hi,

On Thu, 14 Oct 2021 at 16:31, Ludovic Courtès <ludo@gnu.org> wrote:

>>   4. Add disarchive.guix.gnu.org to (guix download).
>
> Done:

[...]

> I think we’re making progress!  :-)

I added the SWH authentication token in patch#51216 [1].  Using a valid
TOKEN from an account of the Software Heritage Authentication service,
it reads something along these lines,

   GUIX_SWH_TOKEN=${TOKEN} guix lint -c archival

The token allows by default 1200 requests instead of 120.  The
interesting thing concerning the recent Disarchive additions are these
bits:

--8<---------------cut here---------------start------------->8---
Disarchive entry refers to non-existent SWH directory 
'aeae11cb3c33ab33374e222dc3bdf17039808a5b'
Disarchive entry refers to non-existent SWH directory 
'b25414c9864a270899ca1ff494e7ba4c437b166d'
Disarchive entry refers to non-existent SWH directory 
'128bbe76a82dd0b38b725565ed703a7148257ae0'
Disarchive entry refers to non-existent SWH directory 
'92625e2c6dbe3ad7c4f44a061ada24ce00637087'
Disarchive entry refers to non-existent SWH directory 
'6000a273dfff9de62725b53e41562fff711069c1'
Disarchive entry refers to non-existent SWH directory 
'c68ff8714c6fd360a38158f3d8f22e555c061452'
Disarchive entry refers to non-existent SWH directory 
'cb52aaa9500df2b674bf7922811deeea1b766139'
Disarchive entry refers to non-existent SWH directory 
'3e574043a04d77dd7231d23210547c4fe065a40c'
Disarchive entry refers to non-existent SWH directory 
'aa763150704fe06f34097b38e839409cee52366d'
Disarchive entry refers to non-existent SWH directory 
'127c0a03c7ccba74870aef7dac36019af35798cc'
Disarchive entry refers to non-existent SWH directory 
'd9745f29da983c6ad674871e68ac96362c4f11cc'
Disarchive entry refers to non-existent SWH directory 
'7d7ed9f88ee649a90493f54d3988a062c3ddeafb'
Disarchive entry refers to non-existent SWH directory 
'f5bd0fe7450175196c57d6f6d5aca8905393e814'
Disarchive entry refers to non-existent SWH directory 
'92bd3b93caa9a4b0840c70ddb96ac75b0684d7ec'
--8<---------------cut here---------------end--------------->8---

which needs investigations why the Disarchive database contains some
entries that SWH does not know.  It probably means that there is an
inconsistency from sources.json.

1: <http://issues.guix.gnu.org/issue/51216>
2: <https://archive.softwareheritage.org/oidc/login/>

Cheers,
simon