No matter if the address is an allowed one or not, or even if CTRL-I was pressed, EVOLUTION does not show the images in the message. Just shows a block with a question mark in the meadle. And this not happen to all messages.
Does anyone having the same problem? Any idea how to fix it?
Ubuntu 20.04.6 LTS (64-bits); Evolution 3.50.1 (by Flathub.org)
Hi,
I tried it, and the images are loaded fine after Ctrl+I here. The only
difference is the system, I tried with Fedora 39, but using the
Flathub’s Evolution of the same version as you.
I would try to run Evolution from a terminal:
flatpak run org.gnome.Evolution
to see whether it’ll print any runtime warning while trying to download
the images. You mentioned the images are not loaded only in some of the
messages, which makes me wonder whether those “broken messages” contain
valid URL-s to existing images, because Evolution cannot show images
which are not available on the addresses used in the mails. You can see
the references in the message source (Ctrl+U), in case the text/html
body is not encoded in base64 or similar encoding. Search for “<img” in
the HTML code, then the “src=” is the reference).
Bye,
Milan
Thank you for your message, Milan.
As I wrote, that is not with all messages that the images are not showen. I notice this with the messages the Bank sent. Most of the others, the images are showed because the address was allowed or after CTRL-I is pressed.
You mention about a “broken message” that I do not belive is what I have. One ot the Bank’s message does not show the images even after CTRL-I, but all the message could be seen in the web browser.
I think that this issue is related with Ubuntu updates. All this messages were seen in the past, but nowadays Ubuntu is asking for a keyring’s password, which was not asked a few weeks ago.
To try something, I removed all the addresses and sites that were allowed in Edit>Preferences>Mail Preferences>HTML Messages. And after this, the situation is repeated.
Even in a terminal run (flatpak run org.gnome.Evolution) the situation is the same. And I notice this error messages:
Gtk-Message: 22:57:31.801: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:31.803: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:32.520: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:32.522: Failed to load module “canberra-gtk-module”
libEGL warning: failed to get driver name for fd -1
libEGL warning: MESA-LOADER: failed to retrieve device information
libEGL warning: failed to get driver name for fd -1
Gtk-Message: 22:57:35.661: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:35.663: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:45.078: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:57:45.079: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:58:01.166: Failed to load module “canberra-gtk-module”
Gtk-Message: 22:58:01.177: Failed to load module “canberra-gtk-module”
(evolution.bin:28): GLib-GIO-WARNING **: 22:58:10.697: Your application did not unregister from D-Bus before destruction. Consider using g_application_run().
What about the above part, please? Examining the message source will avoid guesses.
As I wrote, “broken message” is not we have here. I checked all address shown after “<img” and all open fine in the browser.
(I’m writing here directly, because my mail reply had been discarded by Discourse…)
On Thu, 2023-11-23 at 07:08 +0100, EJSiqueiraJr via GNOME Discourse
wrote:
I checked all address shown after “<img” and all open fine in the
browser.
Hi,
supposing none of those are private images, even not addressing nor
referencing you, would it be possible to share the whole tag
here, thus I could test it, please? It’s possible WebKitGTK has some
odd problem to show the image, or the server doesn’t like download from
the Evolution, some such things maybe.
Bye,
Milan
Once again, thanks for your message, Milan.
When you wrote “share the whole tag” you means add the message? If so, I planned to attach the mbox file, but its type is not authorized. I planned to insert the text here, there are sending limitations for caracters number. I planned to broke the message into parts, there is a limitation in number of links in a message.
So, how can I send the message to you?
Hi,
supposing none of those are private images, even not addressing nor
referencing you, would it be possible to share the whole tag
here, thus I could test it, please? It’s possible WebKitGTK has some
odd problem to show the image, or the server doesn’t like download from
the Evolution, some such things maybe.
Bye,
Milan
Hi,
ah, I see the mail queue was slow, very slow, the message I sent finally arrived.
Anyway, by “share the whole tag” I meant to see the <img .....> tag with everything in it.
If you think the message does not contain anything private, you can mail it to me, mcrha (at) redhat dot com. Reference this thread, please, thus I know where it belongs and I do not overlook it in my spam folder.
Thanks and bye,
Milan
The message you sent me did not make it through the postman software here, it claims some problem with the message structure. Could you right click the test message and save it as .zip, then attach it to a new message and send it to me, please? I think this way the postman will not claim a problem. Ideally set the composer to the Plain Text, not HTML, format, thus the message structure is even simpler. Thanks in advance.
Hello.
I am not sure this message will reach you.
As I wrote before, I tried this on GNOME Discourse, but there are many
limitations.
This message comes from a white listed address and even after CRTL-I most images
are not shown. And they all are shown if openned in a browser (“ver no
navegador”, on top).
Hope it helps looking for the issue.
Best regards,
(Attachment Voe mais longe com os descontos da Azul e Itaú Personnalité…mbox is missing)
Hi,
I need to see what the references to the images look like, which, I hope, will help me to reproduce the problem. It’s hard to know why the thing does not work when I cannot reproduce the problem locally.
Having the whole message would be an ideal state, but it can contain private and sensitive data, thus I agree it cannot be shared in public. You tried to send it to me, but the postman software claimed the message structure is broken. I hoped when you right-click the message in Evolution and pick “Save as mbox” (or “Save to file”) and then compress the created file (for example to .zip or any other common format supported under Linux), then it’ll workaround the problem with the broken message structure.
I do not speak the language the last sentence is written in, but I guess due to the “mbox is missing”, it’s from time when you try to send the message? Maybe name the file differently, not with its subject, say something like test.mbox or such. The spaces and non-ASCII letters in the name should not be a problem, but just in case I’d try it this way. And when you zip the file afterwards (zip -9 test.zip /tmp/test.mbox) and attach the test.zip, then it’ll be ideal. Otherwise verify the file is completely written to the file. I do not know how large the mail is, but if it has large attachments, then it can take some time to be saved into the file.
Thank you for the re-send of the message. It had been received fine this time and I can reproduce it too, many images (I’d say all) are not loaded. One of the failed image reference is https://www.itau.com.br/assets/dam/publisher/01_itau/47_sf_footers_padrao/2022/footer/04_foo_ico_wha_05.png . When I try to wget https://www.itau.com.br/assets/dam/publisher/01_itau/47_sf_footers_padrao/2022/footer/04_foo_ico_wha_05.png I get:
--2023-12-04 08:17:47-- https://www.itau.com.br/assets/dam/publisher/01_itau/47_sf_footers_padrao/2022/footer/04_foo_ico_wha_05.png
Resolving www.itau.com.br (www.itau.com.br)... 23.212.110.144, 23.212.110.201
Connecting to www.itau.com.br (www.itau.com.br)|23.212.110.144|:443... connected.
HTTP request sent, awaiting response... 403 Forbidden
2023-12-04 08:17:48 ERROR 403: Forbidden.
thus the server rejects download of the image. Interestingly, opening the image URL in a browser (in my case Firefox), it is loaded properly. From this I guess it’s a server configuration, preventing download of the images due to some missing identification of the downloader or such. The server admins may know better what they expect in the request headers to allow download of the images.
Dear Milan:
Thanks for your message, and glad that finally you got it.
So, you wrote that the problem is with server that probably expected some marker to permit that the message download the images. And to fix this, I must ask to server admins. A very interesting point, is that a few weeks ago, all were fine. All images were downloaded. What could have changed?
Anyway, I will ask the bank for a solution.
Best regards,
Hi,
I do not know what they changed, but they surely did change something. How they decide when they allow the download and when not I do not know either.
You can try the wget command I provided above. The URL does not contain any “shell-recognized” letters, thus I believe the address is used verbatim, when passing to the server. I mean, the address does not seem to be garbled due to characters, which are recognized and interpreted by the shell, because the example URL does not contain any such letter (like an ampersand, backslash, and so on).
I believe the wget command is a good test also for the server admins, they can try with it whether their server settings work as expected.
Bye,
Milan
Hello, Milan.
Once more, thanks for your attention.
I asked bank for an answer about these issues and included your previous
message.
Now, just waiting.
Best regards,
Hello, Milan.
This week I had a Bank’s answer, and as expected they did not did any change on their messages and all others can read all message without any error. This error is related just with me. And I have no ideia how to fix it. Any thought?
Bye.
Hi,
as far as wget fails, there is clearly something wrong with the server
settings. I do not know what precisely. My uneducated guess is that
they expect some headers in the request or certain connection settings,
and when it’s not there, they reject the file download.
The wget test is very simple to try, but I understand they are not
willing to do it, which is a pita.
I do not know what other apps they use for a reference, maybe Outlook.
How Outlook does the images download I do not know either, maybe they
use “internet explorer” under the hood, thus it pretends to be a web
browsers (or it uses “necessary” headers/connection settings), thus it
works for those users.
Trying to add workarounds into Evolution just because one server has
some odd setting/enforcement is something I’d like to avoid. I’m sorry.
Bye,
Milan
This topic was automatically closed 45 days after the last reply. New replies are no longer allowed.