2012/11/12 Asmus Freytag <[email protected]> > Khaled, don't mind Philippe - his experience is a bit on the > "theoretical" end. >
Exactly the opposite, it was just targetting the common "practical" cases encountered most frequently by most users, ignoring the individual needs of Khaled in specific situations (and notably the situations that programmers need to take into account for building a usable GUI) : Most users won't care about this technical effort, that YOU (a programmer of GUI interfaces will encounter), they just have expectations for the text to be entered easilyin the softwares they use, for their own languages in their normal orthographic rules where "latinARABIC" and "ARABIClatin" cases are effectively very infrequent — this is asserted in IDNA for example). They will just want the unnecessary complication of their GUI interface to occur when they are NOT trying to enter these very infrequent cases in very specific situations that they certainly don't use every day ; and for the rest they will assume immediately that weaker-direction characters will loose in favor of stronger-direction characters to determine the behavior that the editor will perform when they type something : two positions for the caret creates an unnecessary and perturbating interface that does not help reading (you would need to focus on two eye-catching positions, which you can't perform instantly, ONLY if this is really necessary, but NOT everytime because users will start ignoring constantly this information: as a general rule, displaying too much unnecessary information kills the information).
