@Pressedflowers
For this kind of large military border, without clear political affiliation, I generally elect to do something like this (al-Andalus/Asturias),the stripes being reversed comparatively to co-dominium and with red (de facto/fought over) borders.
And not changing the border between Byzantium and Arabs? That's an understable choice, to be honest. I wouldn't agree too much because I think it's representable, but it's more a matter of disagreement.
Euratlas...is more interesting as freely avaible sources goes, but isn't that to be trusted when it comes to geopolitical representation with several really weird choices popping up now and then. Nothing outrageous, again, and certainly better than the Website-That-Should-Not-Be-Named, but not to be trusted blindly.
Personally, I rely quite a bit on Autrement's atlases which provides some good cartographic and textual explanations (even if they don't focus on geopolitical mapping).
Ian Mladjov is another interesting source I agree.
Eventually, tough, nothing beats checking informations on history books, because everyone can overlook something, decide to make a choice you wouldn't or simply because nobody can be an expert on everything. For instance Mladjov's map have some representation I don't really think are fit (in the Alexander's map, not putting Armenia as distinct from the rest of surrounding regions, not distinguishing clients and conquest in Asia, etc. nothing really world-shattering of course)
The question on what is to be represented is another interesting debate : personally, I focus on a geopolitical representation, and trying to include primitive forms of state such as sophisticated chiefdoms.
