But we shouldn't have to duplicate code that is already in a library we're using. But since upstream is brilliant fail in that regard, I'll let it slide.
This however, does introduce a regression. We must only add the bookmark once, when the first successful token retrieval and association is completed, and store a boolean somewhere to remind ourselves that we've done so, to avoid adding the bookmark at every log-in, if the user has decided to remove it, because they don't actually want the bookmark.
But we shouldn't have to duplicate code that is already in a library we're using. But since upstream is brilliant fail in that regard, I'll let it slide.
This however, does introduce a regression. We must only add the bookmark once, when the first successful token retrieval and association is completed, and store a boolean somewhere to remind ourselves that we've done so, to avoid adding the bookmark at every log-in, if the user has decided to remove it, because they don't actually want the bookmark.