Searching for the debug log file

[BeyondTheJob]

When I start Tine, I get the following error message:
“Some page files couldn't be read. The rest of the graph is available; the file list is in the debug log.”

Please excuse my probably silly question about where the debug log file is located. But unfortunately, I haven't been able to find the file yet.

Among other places, I’ve looked here: ~/.local/share/page.tine.Tine/, /var/log/, ~/.cache/, ...

Thank you very much for your response

[martinkoutecky]

@BeyondTheJob not a silly question: the message is wrong. The debug log only exists when Tine is started with debug logging on, so by default the file list goes nowhere. That's our bug, and I'm changing the message so it shows the unreadable files directly.

Until then, to get the list:

TINE_DEBUG=1 tine                 # or: tine --debug
TINE_DEBUG=1 ./Tine-*.AppImage    # if you use the AppImage

Tine then writes /tmp/tine-debug.log (it shows the path in a notice at startup). Open your graph, wait for the "Some page files couldn't be read" message, and search the log for unreadable-files: the line lists the affected files, with paths relative to your graph folder.

If you post which files they are and what is unusual about them (a very large file, an odd encoding, a name with unusual characters, a sync conflict copy), without their content, we can check why Tine couldn't read them.

Claude (working on Martin's behalf)

[martinkoutecky]

@BeyondTheJob The message I mentioned is fixed in Tine Beta 0.8.0-beta.6. The "Some page files couldn't be read" notice now names the files themselves (the first five, then "and N more") instead of pointing to a debug log that is off by default. Copy on the notice copies the full list, with paths relative to your graph folder. You no longer need TINE_DEBUG=1 for this.

Checked on Linux with a page containing an invalid byte; not tried on other systems. If you post which files are listed and what is unusual about them (a very large file, an odd encoding, an unusual name, a sync conflict copy), without their content, we can look at why Tine could not read them.

Claude (working on Martin's behalf)

[BeyondTheJob]

Thank you so much for the quick response to my question, for fixing the error, and for the great program in general!

I frequently create and edit the md files using a text editor. In 13 of the files, I accidentally used the WINDOWS-1252 encoding instead of utf8. The list of files allowed me to search specifically for the error. Without the list of files, I would never have found the error.

[BeyondTheJob]

I'm currently wondering if it makes sense to include the reason why a file couldn't be loaded right after the file name. That way, you wouldn't have to search for it at all. At the same time, the bug report would be significantly longer.