| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
| |
vapi-popup wasn't even included in background.html.
Also, it seems vapi-contextmenu is useless, because vapi-tabs attempts
to register it without creating it first (making registration a
failure, naturally.)
Since it never worked since eMatrix 1.0.0 and it doesn't seem to add
anything useful, let's remove it from the XPI.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
| |
I can't really find a reason why the returned value is preferred over
the normal importing process.
Additionally, there's a good chance importing Services.jsm can be done
only once at the start of everything, instead of binding each object
to a separate closure.
|
| |
|
|
|
| |
Ideally HttpObserver should be its own module, but right now it
depends on too much "context" to be independent from vapi-background.
|
| |
|
|
| |
Also fix a typo in Makefile.
|
| |
|
|
|
|
| |
While they are technically informative, the splitting makes things
easier to follow already (somewhat) and there's not really a need to
list each global variable (there aren't many anyway.)
|
| | |
|
| | |
|
| |
|
|
| |
The same as CallbackWrapper.
|
| |
|
|
|
|
| |
In theory this way it can be used anywhere else if needed, but right
now it's simply because it has to be placed somewhere and a module
seems the best place for an object definition (constructor, etc.)
|
| |
|
|
| |
Also reorganize the Makefile a bit.
|
| |
|
|
|
|
| |
Also remove optional cachestorage script. It doesn't exists and the
cachedstorage, whatever it is, is defined as an alias for the normal
storage system.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
Exported stuff to handle the splitting.
|
| |
|
|
|
| |
Since things have been split, some of these have to be exposed to the
rest of the world.
|
| |
|
|
| |
That file is too large, let's split it up.
|
| |
|
|
|
|
|
| |
It generates an error in the browser console and it's rather annoying.
There is still one error being generated, but it's unclear from where
it comes from.
Not yet tested in Basilisk.
|
| | |
|
| | |
|
| |
|
|
|
| |
The cache should be available, either read from the database or as an
empty object, before any asset is fetched.
|
| |
|
|
| |
Happens when third-party assets can't be downloaded from the web.
|
| | |
|
| | |
|
| |
|
|
|
|
| |
uBlock Origin also uses "extension-data" as a directory, so moving the
whole thing is a bad move.
Instead, simply move the ematrix.sqlite database from one place to another.
|
| | |
|
| |
|
|
| |
It has come to this in the end.
|
| |
|
|
|
|
|
| |
The caller calling this function with topic equal to
"before-asset-updated" expects a non-false value, but is never
returned (a return with no value is equal to "return undefined" or
something like that.)
|
| |
|
|
| |
It's not actually needed anymore
|
| | |
|
| |
|
|
|
| |
Presumably, no one is using the old format anymore (especially since
eMatrix is a different extension than uMatrix/uBlock...)
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
Tests showed that it works without any noticeable problems.
As such, it was deemed better to throw it in the wild and see what
happens.
If there are any problems, this commit can simply be reverted.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
| |
the `forceLegacyToolbarButton' preference is useless because on Pale
Moon the button is always added with the "legacy" method, while on
Basilisk it's always added with CustomizableUI, so there is nothing to
"force" here.
Also add a check that was forgotten in the previous commit.
|
| |
|
|
|
|
|
| |
Unlike checking different versions of Firefox, the features that
Basilisk has over Pale Moon (and viceversa) are better
known and just checking the application ID is enough to understand how
to add the button to the toolbar.
|
| | |
|
