@ebo Thank you for your testing.

@ikloecker indeed we try only for 5 seconds:

I think that the card reader is not connected and there is no Scardsvr at this time.
And the card reader connection to USB port results invoking Scardsvr. Then, "SCD SERIALNO --all" gets success.

For T6567 I changed the way that Kleopatra runs "gpgconf --launch gpg-agent". This change is not yet in Eva's test build. It seems my change is not good because running "gpgconf --launch gpg-agent" timed out after 5 seconds in 3 of 3 tests starting Kleopatra after a reboot of the VM. To check if "gpgconf --launch gpg-agent" really takes that long I measured the time in PowerShell after another reboot of the VM. The result is shocking.

Please note that a card insertion to a card reader and a card reader connection to PC are different things.
It may cause different results.

ebo: Thank you for your testing.

I found an issue in libassuan, which might be related possibly, it is described in T7399.
When the response is not consumed like the case in T7399, if repeatedly, it may fill up the buffer eventually.

This was fixed in gpgme and is therefore automatically in VSD 3.3

I believe this is a case of non-consumption of client. on Gpg4win-Beta-75 + updated GnuPG.
Setup: I had two cards connected, one Yubikey and one Netkey3.0 card. I rebooted windows and started Kleopatra. Nothing else.

This is included in test installers since some time already.

This change is also used for VSD 3.3

The same CSS has worked for years. Suddenly, after an update Firefox started to misbehave. I don't know if it's actually a bug in Firefox or in fontconfig or something else, but it's certainly not a problem of the CSS.

This website uses this CSS rule for body:

Works, you now get an error message that the card can not be found.

I put "scd" tag and let me claim this ticket.

The symptom of this bug was:

• there are multiple waiters for COND.
• COND is fired by npth_cond_broadcast, all waiters should be waken up, but only one wakes up by the old code of 1.7.
• other waiters keep waiting forever.

After I fixed the problem, I realized that the description of this ticket was not accurate, so, modified.

That's a bug in Firefox. Chromium doesn't exhibit the problem. I have problems with fonts in Firefox on many different websites.

This seems to be QWizard-specific behavior. One more reason to port away from that

FWIW, we should eventually get rid of the pipe + socket style connection model. It is just to complex with no real benefit.

After fixing two bugs, I changed the title to express the scope of this ticket.

Fixed in 1.51, by introducing gpgrt_spawn_actions_set_env_rev, which assumes utf-8 encoding.

Fixed in 1.51.

Fixed in 1.51.

Gpg4win-Beta-75, works:

For Beta-75 it looks similar judging from my first tries.

Fixed for all branches.

I managed to get the same "loading certificate" message several times in a row on this test instance by stopping and starting Kleopatra in a row twice. After removing the Signature Card 2.0 this did not happen again in 5-6 tries, although I collected 2 lingering listing processes again (not both started on the same startup). Even import of a X.509 certificate worked.

Next I managed to have one gpg and one gpgsm process each left over from the last execution of Kleopatra.
After starting Kleopatra new anyway, again "loading certificate cache" and an additional pair of gpg and gpgsm listing processes start.

Had a occurrence of the never ending "loading certificate cache" issue again.
There was a leftover gpgsm process from the previous tests (although Kleopatra warned when I closed it, that processes still running in the background were there and would be aborted).

I cannot reproduce the crash anymore. I guess this was fixed with the fix of T7372: Kleopatra: Crash when unplugging smartcard while operation is in progress. I'll close this ticket as duplicate of T7372.

@ebo I suspect that we want to fix this crash also for VSD 3.3 (if it is still reproducible). I found this ticket by accident while searching for READKEY.

This works with gpg4win-beta-70.

I found a problem of possible duplicate registration of another APP, due to no serialization for CARD access.

The resource leak was fixed in: rG40707c8bff49: agent: Fix resource leak for PRIMARY_CTX.

Thanks.

If 7z is used to create a tarball that tarball is then 7z compressed. At least this is how I understand the case.
When gpgtar now tries to extract the file it sees a 7z file and thus emits the octal number warnings because it assumes a tarball (after decryption by gpg).

This problem was also reported at https://bugs.kde.org/show_bug.cgi?id=479567#c1

Fixed and backported for VSD 3.3

While reviewing this task I noticed that I wrote adding a -p option. This is non-sense, because -p is to preserve permissions at extract time; this is unrelated to the last modification time. Standard tar extract files and set the modification to the one given in the tarball - unless you use -m to use the current time. Thus this task is actually a bug and not a feature request. For backward compatibility this will be done only for gnupg26 for now.

This ticket is obsolete, we did some work on gpgtar since then. Should the issue still occur, please reopen or create a new ticket

@ebo Thank you for your continuous testing.

Unfortunately, this seems not to have ended the sporadic hangs.
I just saw a hanging initial keylisting with gpg4win-beta-70 which has gpg 2.4.6

My fault: All my test had the relax flag set which is so common to me that I did not thought about this. So you fix is for the case that no flag has been set for a fingerprint.

(Please don't set a milestone tag for a fix to an already released version - we use the milestones to track done tickets). Use instead the branch specific tag so it ends up on the workboard.

@ikloecker : Thanks for investigating. Please note that gpg-agent is incompatible wrt LISTTRUSTED (2.2 vs 2.4). So, No data callback in IPC maybe expected with gpg-agent 2.4.

The last two usages of KIconLoader have been remove in kleopatra master. (libkleo was already good.)

Sorry, I've pasted the wrong link, I wanted to paste this one: https://lists.gnupg.org/mailman/listinfo/gnupg-users

Why would I turn to the Windows mailing list when I am a Linux user?

I removed a duplicated comment above.
Please do not duplicate information (no top posting) and keep your descriptions short and to the point.

"BTW, GnuPG 2.3.4 is a very old version."

In the story of my life, you are a mythological figure.