jackett: update to v0.8.490

[kaso17]

Motivation: This should fix various jackett specific issues
Linked issues: fixes #3015 #2986 #2881 #1995 Jackett/Jackett#1985 Jackett/Jackett#2217

Checklist

• Build rule all-supported completed successfully
• Package upgrade completed successfully (Only tested on DSM 6.1)
• New installation of package completed successfully (Only tested on DSM 6.1)

[m4tt075]

You probably want to increment the SPK_REV to 5, rather than decrement it to 3, or the package will not be proposed as update in the DSM package manager...

[kaso17]

done, the version is not considered by the package manager?

[ymartin59]

@kaso17 Thanks for your work. May you please rework package installation and start-stop process thanks to generic service support described at https://github.com/SynoCommunity/spksrc/wiki/Service-Support.
You may look at spk/demoservice demonstration package or to mosquitto PR #3049

[kaso17]

@ymartin59 unfortunately I've no longer access to a syno device so I can't test the requests changes.
I still would propose to merge this as it still significantly improves the situation.

[m4tt075]

@ymartin59 @kaso17 Happy to step in and handle the transition to generic service once TVH is done.

[kaso17]

@m4tt075 thank you.
Some notes:
${INSTALL_DIR}/share/${PACKAGE} must be writable by the jackett user (otherwise the build in update will fail)

If the user is changed as part of the legacy migration path make sure to chown the ${INSTALL_DIR}/var directory accordingly

The HOME environment variable must point to ${INSTALL_DIR}/var
The PATH environment variable must include the mono binary
The LD_LIBRARY_PATH environment variable should point to ${INSTALL_DIR}/lib

Jackett handles it's own log file generation including log rotation, so pointing the log to ${INSTALL_DIR}/var/.config/Jackett/log.txt is fine

As you can see in the PR I redirect the stderr output to ${HOME_DIR}/${PACKAGE}-stderr.log.

Under normal circumstances it should never write anything to that will but in case there's a low level mono crash it will contain the mono stack trace which might be helpfull for debugging.

The stdout output is sent to logger but that seems to equal /dev/null, it doesn't matter as it's all sent to the log.txt anyway.

[m4tt075]

@kaso17 OK, see #3070 for adaptations and let's continue the discussion there.

[m4tt075]

@kao17 Could you squash your Update and Makefile commits, please, and rebase your branch to get rid of the merge commit?

[kaso17]

@m4tt075 I don't see any conflicts.
I've no access to my dev machine at this time, will take me ~5 days to squash them. Can't you squash them as part of the github PR "Squash and merge" feature?

[ymartin59]

From my point of view, problem is that PR contains "obsolete" changes on makefile (mono version), installer.sh and dsm-control.sh... So I will probably discard it, to be replaced by a squashed commit of #3070, but no worry, credits for your work should remain in GitHub commit/history.

[kaso17]

Feel free to close it.
I don't care about credits.
I just hope once the official packages are released the number of Syno/DSM 6.1 related duplicate issues go down.

[ymartin59]

Included in #3070. Many thanks for your work !