Hidden for Gpg4win-4.3.0-beta571, too

The state of the brain is:

Fixes are already in GnuPG 2.4.4 and can't be easily tested. Thus closing also for gnupg24

Closing because we believe things are fixed and our test suite confirms that. Feel free to -reopen in case your own file does not import with 2.4.4.

• To configure a keyserver none I have now T6950: Kleopatra: Usability improvements for directory services configuration
• For tarball naming I created T6952: Gpg4win build system: Include commit hash in tarballs from gen-tarball.sh
• For the about dialog I have T6953: Kleopatra: show commit id in about dialog

In T6708#181592, @werner wrote:

I would also suggest that we show the git last git commit in Kleo's About dialog. That makes it far easier to see what we are testing. The Kleo version numbers are a bit arbitrary.

I would also suggest that we show the git last git commit in Kleo's About dialog. That makes it far easier to see what we are testing. The Kleo version numbers are a bit arbitrary.

Sorry, it was my fault building the test installer.

To be clear: This ticket is only about GnuPG (more precisely dirmngr) and the changes are included in VSD and Gpg4win.

Hi, ebo I would still think this is resolved. Because it was never meant that the user manually enters the value of "none" because there is no hint for the user that "none" is a reserved word. It should either be administratively configured which does not make much sense for Gpg4win or provided by the distribution. If left empty the default of GnuPG should be used. If we really want users to deactivate keyserver access by using "none" in the dirmngr.conf a much better solution would be a checkbox for this. In that case I would open a new issue.

The fix was not included in the Testbuid...

Does not work in Gpg4win-4.2.1-beta178

works in Gpg4win-4.2.1-beta178

Thank you very much on behalf of our S/MIME users. This also makes it easier for us in the frontend to show a consistent UI.

See also T6465

Testing in 2.4 will not be easy because it requires code modification just for testing. However, de-vs is not supported by 2.4 and the greater plan is to get 2.6 approved for de-vs.

Further investigation showed that this was due to a bogus key creating during I wrote the code.

works, the secret part is now imported, too, tested with VS-Desktop-3.1.90.258-Beta

For the VSD branch it works, VS-Desktop-3.1.90.258-Beta

Would love to test this, but I can't seem to compile this project, getting stuck at The system does not provide a working iconv function. Is there a Fedora based dockerfile or equivalent where I could build it? Here is the reference Fedora source. I have tried to hack it and build from a gitarchive, but I am still encountering issues No rule to make target 'audit-events.h', needed by 'all'. Stop.

Now fixed in 2.2 and 2.4 (commits rG08f0b9ea2e955209d467f1ff624bf7abd10ae7ac and rG7661d2fbc6eb533016df63a86ec3e35bf00cfb1f). See also T6752

According to Werner this should work.

That output was also misleading,. that was from before I added the ignore-crl-extension in there. I was confused because I still got the error:

So dirmngr already has that option.

yes, fixed

I think this was fixed with the fix for https://dev.gnupg.org/T6534

With VS-Desktop-3.1.90.246-Beta I can not import the secret part of the edward.tester@demo.gnupg.com.p12 Testkey (ECC brainpool).
I do not see any error message.

Okay, I found and fixed the import problem in 2.4 and will backport this to 2.2

With Gpg4win-4.2.1-beta31 I can no longer import the secret part of the edward.tester@demo.gnupg.com.p12 Testkey. Error is "Invalid object".

This has been well tested during development and is thus ready for a release.

We have no dedicated error to tell that the verification failed due to an non-compliant algorithm. Thus we return invalid public key algorithms as best approximation. You could use --override-compliance-check, though. We discussed things thing once at the Gutenbergweg.

Mmh, ok this does not seem like a regression, at least if I go back to one of my oldest appimages with 3.1.21 I still get ERRSIG.

Thanks for the pointer! I'll see how I can do what ecdh_param_str_from_pk does in gpgme.

The relevant commit is rGc03ba92576e34f791430ab1c68814ff16c81407b

We had to add the parameters because some keys don't use the default paramters PGP and gpg have used since the introduction of ECC 12 years ago. So yes, we could fallback to the standard parameters but it would bet better if Kleopatra could extract them from the public key (maybe via a GPGME helper).

The relevant logs are

2023-07-27 12:08:01 scdaemon[28156] opgp: ecdh parameters missing
2023-07-27 12:08:01 scdaemon[28156] operation writekey result: Invalid value

I can't find a missing forward port; need to debug this issue with gpg4win 4.2.0

yeah, sorry, didn't test different key types yesterday.
NIST encryption keys do not work either, so only RSA encryption keys can be moved with Kleopatra to a smart card in gpg4win 4.2.0.
I can confirm that authentication keys work.

In T6379#172803, @ebo wrote:

Noticed in gpg4win 4.2.0-beta373:

For Brainpool and ed/cv25519 keys it is not possible to move a subkey to a smart card with Kleopatra. The error message is "invalid value".
Moving the main key works, though. The command line works for all keys types, of course.

Noticed in gpg4win 4.2.0-beta373:

Same for the backport to 2.2 which uses the same test suite.

This has long been implemented due to the backport of the P12 parser and the recent rewrite of it.

I'm reopening this. Its probably not a regression but I was sure that we had progress for large files fixed in the past.

Yeah no progress for files larger then 32 bit o.O... But this used to work 😭

On 64 bit linux this works btw. so I think it comes down to the difference between 32 bit off_t and 64 bit off_t

Yeah, its the ugly off_t again. I am just testing how this works with single files above that threshold we worked quite a bit on this back in the days https://dev.gnupg.org/T2368

Yeah, probably a Windows/MinGW 32-bit problem. GpgME::Data does

off_t size = seek(0, SEEK_END);
seek(0, SEEK_SET);
std::string sizestr = std::to_string(size);
// Ignore errors as this is optional
gpgme_data_set_flag(d->data, "size-hint", sizestr.c_str());

Probably some issue with large files / integer overflow. I am testing on Windows with 32 bit.