Gregory's Server@pieselpriemel @ProPublica The API is by far the easiest way to do this, and the problem is one of scale. The way you suggest might work if we knew exactly when thousands of accounts tweeted and could afford to engage in massive numbers of requests and if the service was reliable. Believe me, I've explored other options.
@derekwillis @ProPublica Fair point. wget has to call every single tweet constantly to see if it gets deleted. Let's hope twitter fix their new methode, should be in their own intresst that fake screenshots can be identified. But I have little to no hope for that platform.