Database Versioning and Sample Data

[lornajane]

Hi everyone

A couple of database things have come up this evening which I thought
you might either know about or help me on.

Firstly: I've got a pull request for a script that runs the database
versioned scripts automatically (https://github.com/joindin/joind.in/
pull/428) - didn't we have this already? I'm confused, can someone
please reply or comment on the pull request on how we do this?

Secondly: Joshua is working on a way of generating sample data that we
can use on our development copies - this sounds really interesting but
I guess not many people actively read the bugs list. He has code for
us to try out so go and have a play if you're interested (this has
been a big issue for us - thanks so much to Josh for pickin this up).
It's here: http://joindin.jira.com/browse/JOINDIN-101

Comments, suggestions, etc are all welcome

Lorna

PS Greetings to all our new members! Do feel free to post to this
mailing list if you have questions or comments :)

[lornajane]

I feel completely ridiculous talking to myself, so do feel free to
chime in at any point!

It turns out I was not imagining another script We have a script for
running database patches, which was recently updated by Paul Court
(recently = July). You can find it in /src/scripts/patchdb.sh. I
think this means that we don't need Marcus' script, but that we
definitely do need to document that this one is there. What do you
think?

Lorna

[Rob Allen]

Hi,

I've had a play with patchdb.sh and it works well, so no need for another one as far as I can see. patchdb.sh could do with some improvements to its documentation though as the options are a little opaque. (what's a "target" for instance?!)

We need to consolidate the set-up documentation though. There's two different files with info on setting up: README and doc/SETUP. I'll have a go at tidying up on Sunday.

I'll try and have a play with Joshua's data generator on Sunday too. I've had a look at the diff on https://github.com/jaytaph/joind.in/tree/JOINDIN-101 and at a casual glance, it looks good. I especially like that it creates genuine-looking data.

Hassle me Monday, so I can hang my head in shame if I forgot to do it!

Regards,

Rob...