Make changes back into generation framework for FireFox extension

[ATather]

Some changes I've made to generation scripts to provide end-to-end updates for firefox.

Installer scripts include commented out section for installation of mozilla signed .xpi file.

Have used wojwal's rearrangement of installer in-case we need the Win64 attribute setting.

[taxilian]

Hi; I'm a little confused, I see a couple of very minor changes which are mostly fixing typos (though I need to look closer to see why you're changing the output path of the msi file, that doesn't make sense to me) but I don't see the other changes you're talking about. Are you sure you pushed all your changes?

[ATather]

Hi, There are no intentional changes of output path :)

The changes are mainly minor, mostly in the current wix template script which doesn't compile in two places at the moment and some tidy ups.

This is from a test of using py.gen -> prep -> build on a clean workspace.

Bug 1: The line Source="$(var.BINSRC)\${PLUGIN_CRX_NATIVEHOST_NAME}_chrome.template is trying to pick up the template from say the MinSizeRel folder, but there is no copy command to put it in there, as the prep script puts it into the root. Hence the change:
Source="$(var.BINSRC)\..\${PLUGIN_CRX_NATIVEHOST_NAME}_chrome.template"

Bug 2: We also have two entries with Id="NativeHostManifestRegEntry" causing duplicate failure.

The latter section is a place holder to uncomment when their mozilla signed .xpi is available for self distribution
Added a placeholder to digitally sign the NativeMessageHost.exe
Made the name of the NativeMessageHost configurable via $FBSTRING_HostFileName
The rest are just tweaks and typos, nothing scarey.

Hope this helps,

Andy

[taxilian]

Odd; I see other changes now, but I didn't then, and github doesn't show that you've changed the pull request. Either I'm even crazier than previously suspected or github is acting weird.

Could you rebase on the current 2.0 branch, fix conflicts, and force push? a few of the issues you "fixed" actually remove fixes from changes I've made recently.

If you look at the changes in this PR you'll see that you have changed the commented-out output path of the xpi to "${PLUGIN_PATH}/${PLUGIN_NAME}.msi" instead of "${FB_BIN_DIR}/${PLUGIN_NAME}/${CMAKE_CFG_INTDIR}/${PLUGIN_NAME}.msi", which seems like it should not have been done. BINSRC isn't where those files are, they should be (if I remember) in VARSRC and that change has been made in the 2.0 branch as part of adding firefox support. Bug 2 is also fixed in the 2.0 branch and I believe that your fix is missing a reg key that is needed, so resolving that would be good.

It could be there are typos in the template files as well.

Thanks for responding quickly! I'm working on this stuff right now, which means I can respond to PR comments in a timely manner at the moment; that isn't always true.

[ATather]

Found the correct variable for the template TPLSRC, thanks.

Only updated the signing links to .msi/.exe using ${PLUGIN_PATH}/${PLUGIN_NAME} as we no longer have a ${FB_BIN_DIR} folder. Ran a test and everything is signed correctly.

Think it would be wise to merge the
#185
first (when everyone is happy) then I can re-merge with this.

[taxilian]

See code review comments. Finally figured out how to use the tools to ask intelligent questions :-P

[ATather]

Not seeing any inline comments, or notification for them.

[ATather]

Test in-line comment.