HI, guys, I have noticed that messages from Kevin Fenzi and Stephen Gallagher do not load in evolution. So I surmise that I am missing a credential source or something. Any guidelines... This seems to have happened when I upgraded from F-24 via DNF if that makes any difference.
I tried to reply to them privately, but evolution just hung up and eventually gave me a time out error.
Regards, Les H.
On 12/12/2016 12:10 PM, Howard Howell wrote:
HI, guys, I have noticed that messages from Kevin Fenzi and Stephen Gallagher do not load in evolution. So I surmise that I am missing a credential source or something. Any guidelines... This seems to have happened when I upgraded from F-24 via DNF if that makes any difference.
I tried to reply to them privately, but evolution just hung up and eventually gave me a time out error.
I suspect that evolution is having an issue with messages that are signed with GPG keys (which both Kevin and I do by default). This is a mechanism I use so that people can trust that what I sent actually came from me. (I'm not signing this single message so that you can see the reply).
Probably worth filing a bug against Evolution.
On Mon, 2016-12-12 at 12:13 -0500, Stephen Gallagher wrote:
I suspect that evolution is having an issue with messages that are signed with GPG keys (which both Kevin and I do by default). This is a mechanism I use so that people can trust that what I sent actually came from me. (I'm not signing this single message so that you can see the reply).
Probably worth filing a bug against Evolution.
That sounds highly plausible. evolution currently shells out to gpg which is pretty fragile, so it's not very surprising that some issue could cause it to hang forever. It needs to be rewritten to use GPGME.
Michael
On Mon, 2016-12-12 at 12:43 -0600, Michael Catanzaro wrote:
That sounds highly plausible. evolution currently shells out to gpg which is pretty fragile, so it's not very surprising that some issue could cause it to hang forever. It needs to be rewritten to use GPGME.
Hi, I looked on GPGME sometime in the beginning of the year and it's missing some features the evolution(-data-server) uses. I do not recaal what exactly, I'm sorry. It's still just a front end for the gpg/ggp2 binary, thus basically the same what the evolution-data-server does.
These things are run in a dedicated thread, thus they do not block the UI, only may result in an infinite wait for a response from the gpg/gpg2. It doesn't result in a timeout though. Being it about gpg/gpg2, I'd reference: https://bugzilla.gnome.org/show_bug.cgi?id=769204
Being it about WebKit2 usage as such (the 3.22.x uses WebKit2, while 3.20.x used WebKit1), I'd reference for example: https://bugzilla.redhat.com/show_bug.cgi?id=1398806#c2
Hope it helps, Milan
On Tue, 2016-12-13 at 09:44 +0100, Milan Crha wrote:
On Mon, 2016-12-12 at 12:43 -0600, Michael Catanzaro wrote:
That sounds highly plausible. evolution currently shells out to gpg which is pretty fragile, so it's not very surprising that some issue could cause it to hang forever. It needs to be rewritten to use GPGME.
Hi, I looked on GPGME sometime in the beginning of the year and it's missing some features the evolution(-data-server) uses. I do not recaal what exactly, I'm sorry. It's still just a front end for the gpg/ggp2 binary, thus basically the same what the evolution-data-server does.
These things are run in a dedicated thread, thus they do not block the UI, only may result in an infinite wait for a response from the gpg/gpg2. It doesn't result in a timeout though. Being it about gpg/gpg2, I'd reference: https://bugzilla.gnome.org/show_bug.cgi?id=769204
Being it about WebKit2 usage as such (the 3.22.x uses WebKit2, while 3.20.x used WebKit1), I'd reference for example: https://bugzilla.redhat.com/show_bug.cgi?id=1398806#c2
Hope it helps, Milan
The first one GNOME Bugzilla – Bug 769204 says: Status: RESOLVED NOTGNOME
The second one Bug 1398806 - Weird rendering for Evolution on Wayland doesn't seem to apply.
In the past when evolution blocked on gpg messages it was because I didn't have the correct url to the cert for the sig or encryption. could that still be the case here?
I have also noticed that now when I click reply, and then minimize the evolution window, the reply no longer is interactive. This is new behavior with F25 because I could do that with all prior versions of fedora. But I suspect that some interaction with wayland/evolution is the issue.
Regards, Les H
On Tue, 2016-12-13 at 10:08 -0800, Howard Howell wrote:
The first one GNOME Bugzilla – Bug 769204 says: Status: RESOLVED NOTGNOME
Hi, yes, that's correct, there was a problem in the gpg/gpg2, not in the evolution-data-server as such. Please read through it for some pointers into the gpg configuration.
The second one Bug 1398806 - Weird rendering for Evolution on Wayland doesn't seem to apply.
The exact comment suggests how to run the evolution, maybe just for testing. That problem comes so often that I wanted to suggest it here as well. It can be completely unrelated, of course.
I would need a backtrace of the evolution to see what it tries to do. That's out of scope of this mailing list, thus better to file a bug and we can investigate there.
In the past when evolution blocked on gpg messages it was because I didn't have the correct url to the cert for the sig or encryption. could that still be the case here?
Yes, it can. Just note that the "evolution stuck" is most likely a consequence of the gpg/gpg2 doing something (or waiting for something).
I have also noticed that now when I click reply, and then minimize the evolution window, the reply no longer is interactive. This is new behavior with F25 because I could do that with all prior versions of fedora. But I suspect that some interaction with wayland/evolution is the issue.
I do not think so. See https://bugzilla.redhat.com/show_bug.cgi?id=1401239 which, by the way, references the change I suggested with the above bug link, which you claimed as "not to apply". Who knows.
Bye, Milan