[EXT] TeX Live Installation failed at the very end

Johannes Hielscher jhielscher at posteo.de
Sun Mar 20 14:47:42 CET 2022

Am Sun, 20 Mar 2022 10:40:48 +0000
schrieb "Philip Taylor (Royal Holloway)" <P.Taylor at Rhul.Ac.Uk>:

> On 20/03/2022 05:20, Norbert Preining wrote:
> Hi Phil,
> Analogous problem reported 28/02/2022, 10:44; zero response
> received.  Earlier report appended.
> Yes, because it is Windows, connection problems are impossible to
> debug over the line, and above all not reproducible.
> Please explain "because it is Windows".   What, architecturally
> speaking, makes Windows "connection problems impossible to debug over
> the line", what operating system(s) do allow one to successfully
> "debug connection problems over the line", and what is the
> architectural difference between the two ?
> Feel free to throw your own infinite expertise into the ring and try
> to solve the problems, that would be *greatly* appreciated.
> At this point :
> [1] TeXLive::TLUtils::check_file_and_remove:   removing
> C:\Users\PHILIP~1\AppData\Local\Temp\RPAddWJRpP\nKUKULiiZu/spectralsequences.tar.xz,
> but saving copy in C:\Users\PHILIP~1\AppData\Local\Temp\csOK_gsvQ6
> [2] TLPDB::_install_data: downloading did not succeed
> (check_file_and_remove failed) for
> ftp://ftp.math.utah.edu/pub/tlpretest/archive/spectralsequences.tar.xz
> Installation failed. the diagnostics are confusing; line 1 says that
> it is attempting to remove
> C:\Users\PHILIP~1\AppData\Local\Temp\RPAddWJRpP\nKUKULiiZu/spectralsequences.tar.xz
> but saving a copy in C:\Users\PHILIP~1\AppData\Local\Temp\csOK_gsvQ6,
> whilst line 2 says that "downloading did not succeed" and goes on to
> say that "check_file_and_remove failed".  The transcript makes no
> mention as to what error(s) occurred during the attempt to
> "check_file_and_remove" — are these errors logged, and if so,
> where ?  And since "check_file_and_remove" would appear to be
> operating solely on local files, where do "connection problems" come
> into the matter ?
> Finally I repeat my original question: "If I "Re-run the installer
> [to] try to restart the installation", will it re-try just the failed
> part or will it need to download all 4313 packages again ?"

There are several ways to find out. You could have a look into the
sources and their inline documentation [0], and figure out what is
supposed to happen. You could also just try it.

Chances are, even if someone in your situation retries the installation
(by themselves, or tlmgr's internal logic), things will randomly break
again for the same rogue reasons why they have broken before, i. e.
simple reïteration would just increment your frustration, without
gaining any more knowledge about the actual cause of the unpleasant
As long as nobody understands (and has a good fix for) the download/
installation failure of unsuspicious packages, is there a point of just
hoping that things sort out by themselves?

Have you considered Norbert's suggestion to go the Divide&Conquer route
(installing scheme-minimal and then grow the installation step by
This is a good way to obtain a usable TL installation in an unreliable
environment, for pretesting TeX & friends (the “payload” of TL).
It is a very respectable motivation to also assist in the improvement of
tlmgr/install-tl infrastructure, yet this is a distinct topic.
Contributions are very welcome to improve this situation. There has
been no thorough discussion what behaviour and output verbosity to
expect, and in my impression it hasn't yet become obvious how much
tlmgr has improved lately at this front.
(Note that your issue on 2022-02-28 is not connected specifically to
TL2022 (= is not a regression introduced by TL22 pretest); Alexis'
report was with the TL2021 installer).



> --
> ** Phil.
> This email, its contents and any attachments are intended solely for
> the addressee and may contain confidential information. In certain
> circumstances, it may also be subject to legal privilege. Any
> unauthorised use, disclosure, or copying is not permitted. If you
> have received this email in error, please notify us and immediately
> and permanently delete it. Any views or opinions expressed in
> personal emails are solely those of the author and do not necessarily
> represent those of Royal Holloway, University of London. It is your
> responsibility to ensure that this email and any attachments are
> virus free.

More information about the tex-live mailing list.