added multi-softpatching support + OSD messages for patches (#9947)

[eadmaster]

followup

Moved into a separate branch as suggested.

[jdgleaver]

@eadmaster Thanks, that's much better!

@twinaphex I think we can merge this now

[jdgleaver]

Oh, hold on - Travis has revealed one mistake. The usage of basename() is incorrect - let me mark it in the code...

[jdgleaver]

@eadmaster Great, that's fine now. But it still needs a rebase to squash that second commit:

git checkout master
git pull upstream master
git checkout multi_patch
git rebase master
git rebase -i HEAD~2
<enter fixup for the 'fixed Travis error' commit>
git push --force origin multi_patch

I will indeed look at your other PR as soon as I get the chance :)

[eadmaster]

Can't you git merge+squash in this case?

https://stackoverflow.com/questions/5308816/how-to-use-git-merge-squash

[jdgleaver]

@eadmaster Okay...

You broke the PR again with your last force push. I waited a week to see if you would fix it, but you didn't.

Please run the following commands:

git checkout multi_patch
git rebase master
git push --force origin multi_patch

That should make it mergeable again.

[eadmaster]

sorry, i'm in a bad relationship with git atm, hopefully it'll get better with practice. :-)

[jdgleaver]

Ah, you'll get there in the end ;)

Git can be complicated, but for normal usage there are only really half a dozen commands to remember. It is indeed just a matter of practice - and also, probably a good idea to make a backup copy of the entire local repo before attempting rebases and squashes, just in case it all goes horribly wrong (that saved me from a couple of unfortunate merges back when I was starting out...).

Anyway - all good now, and glad to finally get this merged :)

[hizzlekizzle]

woooo! balloons and streamers, guys :D

[i30817]

@eadmaster @jdgleaver @hizzlekizzle this appears to show the notification of application of the patch like this:

game.ips
game.ips3
game.ips2
game.ips1

Is this on purpose? Isn't it more user friendly going up to make sure the patches apply in the natural order? Or is it just a bug from using a message model of notifications?

[jdgleaver]

Hmm... In the code, the patches are applied in exactly the order one would expect - so this seems to be a message queue issue...

[i30817]

Can a message push another message? That's how i worked around swingworker thread event bugs like that in the aeons past iirc. Though, better shut up because i can barely remember there was some kind of requirement on this to not deadlock.