Cloudron Forum

Hello, first time here. I am hosting InvoiceNinja v5.7.6-C123 on a Cloudron, and <a> links in generated PDFs stopped working all of a sudden, meaning they still appear (blue and underlined) but are not clickable anymore.

• Unless mistaken, it doesn't seem to come from an IN update (last update in August, stopped working in September).
• the HTML sent to the PDF generator seems to be well-formed and does contain the proper <a> link.
• I tried two PDF generators (PDF_GENERATOR=snappdf and PDF_GENERATOR=hosted_ninja), slight output differences but same problem.

I posted on IN forums, no luck so far.
Thanks in advance.

Thanks you guys! Not sure if there's a "solved" button that I'm missing in here but problem and questions solved.

I support this.

To clarify, this is a UX bug and not a feature request, because this leads an operator to see "No new updates available" when clicking "Check for updates", while there are new updates, which is misleading.

Why yes, I have been asked for a PDF with a clickable link so I provided . Now for an unclickable link (which does sound more appropriate indeed) (same expiration date).

Would it be possible to have an automation of the upgrades until there are no more upgrades, then? For example a button "upgrade to latest" that would upgrade from 1.0 to 1.1, then without intervention from 1.1 to 1.2, etc.?

Reply by email does not work on my Discourse.
Discourse logs say multiple instances of:

Job exception: Timeout to open TCP connection to my.soletic.cloud:9595
Job exception: Timeout to open TCP connection to my.soletic.cloud:995

My mailbox looks good to me:

(I actually just changed the owner to the app, before that it was a user group, not sure if that matters)

And my Discourse mailbox settings as well:

When test-sending an email reply to a Discourse message notification, it seems to be received:

But the message does not reach the Discourse thread to which it replies.

Any help appreciated.

@jdaviescoates Yes these were on-the-fly translations. And it seems that you couldn't reproduce my problem. Thanks for checking.

@james I don't know but it was a bad example, I've got this error for 995 as well:

@james I didn't realize it was necessary, thanks.
I set it up on the 10th of march at 17:49. From this moment, the Discourse logs hasn't shown any TCP timeout on 995 anymore. But mails remain ineffective.
These errors remain afterwards:

I suspect the SMTPServerBusy reveals a new problem?

@james Ah yes sure.
POP3 polling enabled and :

Note: POP3 polling port was actually 9595 (and wasn't considered modified by discourse (no green dot)), so that explains the 9595 call (but why was there a 995 call too then?). I just changed it to 995 and tried again to send an email (and waited for the 5 min of polling period), but no luck.

Bumping this thread in case someone got an idea.

@nebulon FINALLY! the mail arrived. So, Discourse was not correctly configured, as for the POP3 host and password. Good to know, I'll document that internally.

Thanks a lot, I really appreciate the time and thouroughness!