Let's say hybrid? I use FB API for gathering some data to display, but I mostly use HTTPrequest to emulate FB functions.
You need to be more clear to these ppl, they aren't programmers.
Hao does something similar to what I do.
On my youtube bot what I do is that I emulate the login with a headless browser, then get the cookies and attach them to my http bot version of some other yt functions. I suppose is the same for fb, some things on facebook you really need to use a browser/headless browser because there is just too much javascript to deal with. However, on many other tasks its as simple as an api call or just a simple http emulation without browsers. So by having the cookies and attaching them to your http requests, you can actually make a hybrid bot.
For me, the youtube logging was absolutely necessary to make it browser login . Once it was completed i usually just took the cookies straight away from the SQLlite database that chrome or firefox use to store them, or I used phantomjs . Other things like scraping usernames, viewing channels, I didn't need any javascript just pure http requests.
Let me clarify, browsers are a pain in the ass . Same goes for phantomjs.
I actually using phantomjs on my current job, nothing blackhat related or botting, and I run that baby for hours . By hours I mean like 12-24 hours straight processing heavy javascript sites. Phantomjs, has some huge memory leaks that I literally always have to kill the process forecefully because it always steal all of my cpu cycles allowing my machine to do nothing.
That is why , I always try to go for http requests and avoid browsers a lot.Even though, social sites are becoming more and more javascript dependant that its getting harder to scrape them with just http requests. Not to mention footprnts you leave behind, by not usig a normal browser.