I noticed that facebook and gmail have really odd #id and .class names.
Do they obfuscate them on purpose? or does their IDE does it? It doesn't seem logical to have such unreadable names for development
for example - this is gmail's refresh button. It would be reasonable to have id/class as "refresh"
<div class="G-Ni J-J5-Ji" style="">
<div class="T-I J-J5-Ji nu T-I-ax7 L3" act="20" role="button" tabindex="0" style="-webkit-user-select: none;" data-tooltip="Refresh" aria-label="Refresh">
<div class="asa">
<span class="J-J5-Ji ask"> </span><div class="asf T-I-J3 J-J5-Ji"></div></div></div></div>
and facebook's post button for status update
<button class="_42ft _42fu _11b selected _42g-" type="submit">Post</button>
on performance - does have shorter name really impact load times? it would seem it is (for a large page such as facebook or gmail) a couple of 100kb more, which with today broadband line is negligible for the time needed
on exception - twitter and pinterest have readable names
See Question&Answers more detail:os