[mythtv-users] export/import of selected MythTV data

[Stephen Worthington]

On Sun, 02 Aug 2015 12:34:01 -0600, you wrote:

>I am preparing to upgrade my MythTV 0.21 FE/BE to new hardware and 
>software, put off by the need to overcome the 0.24-ish upgrading cutoff 
>for older versions.
>
>I did look into making a Ubuntu VM and upgrading my database to to the 
>intermediate version in order to upgrade it to the latest version but I 
>really do not have much in the way of content to carry over - I can convert 
>what I have on the old box (down to under 50 recordings now) to videos or 
>make a note to re-record them. I don't have anything invested in the other 
>media areas like MythMusic so it seems like building a fresh FE/BE is the 
>best route for me. 
>
>However, my largest "investments" that I'd like to maintain are
>
>  1 - recording rules
>
>  2 - what episodes I've seen already of repeating shows that are still
>      running.
>
>How could I best capture these two data sets and get them into a fresh 
>install of the latest MythTV when I make the switch?
>
>I could probably scrape my recording rules into some format that's easy to 
>re-enter by hand on the new box.
>
>But the previously-recorded episodes data - how could I capture and 
>transfer that? I don't want to have to pick "Never Record" for each of the 
>South Park episodes I've seen already that the new install won't know 
>about, for example.
>
>I'm comfortably-proficient in writing a script to do it, and I can use the 
>new box to test against. But with the greater knowledge of such things out 
>there, is it feasible? I looked at my 0.21 schema with phpMyAdmin and see 
>my "oldrecorded" table has a few *id fields that might need to be mapped 
>into new ids: chanid, seriesid, programid, findid, and recordid. A lot of 
>those can be culled because I don't have the recording rules (or channels) 
>that created them. And I can dry-run it all on the new box before making it 
>official.
>
>How feasible is this?

I would have thought that using a VM to upgrade the database would be
much easier than trying to manually upgrade just a couple of tables.
Upgrading the database is not insanely difficult or anything like
that, as long as you read the release notes for each of the versions
on the way between your current version and the latest.  And in a VM,
you can just roll back anything that does not work and try again, as
long as you have enough disk space to keep the previous version or are
using the option that VMs like VirtualBox have to only save the
changes to disk on shutdown, so you can abandon any changes that did
not work.

You only need to do two database upgrades to get from 0.21 to 0.27.