News for hackers. Stuff that matters.
Re:PDP-10/its: Incompatible Timesharing System by pjc50 on Sunday January 29, 2017 @01:25PM (#13515713)
What are/were the unique or distinctive features of ITS? It's mentioned in a couple of places in the "jargon file", but otherwise it's a part of computing history I'm not familiar with.
Re:PDP-10/its: Incompatible Timesharing System by panic on Sunday January 29, 2017 @02:02PM (#13515913)
The "PCLSRing" technique to essentially back out of an active syscall is worth studying: https://en.wikipedia.org/wiki/PCLSRing http://fare.tunes.org/tmp/emergent/pclsr.htm. It's much better than the UNIX approach of returning early with an error or incomplete data.
Re:PDP-10/its: Incompatible Timesharing System by waqf on Sunday January 29, 2017 @04:57PM (#13516963)
> much better than

Funny to say that, when the comparison between these two approaches was the subject of the original "Worse is Better" article: http://dreamsongs.com/RiseOfWorseIsBetter.html

Re:PDP-10/its: Incompatible Timesharing System by qwertyuiop924 on Sunday January 29, 2017 @03:33PM (#13516436)
Actually, as I understand it, the BSDs will now PCLSR some calls.

Might be wrong though, so don't quote me.

Re:PDP-10/its: Incompatible Timesharing System by SSLy on Sunday January 29, 2017 @03:57PM (#13516599)
The first article then goes on a synchronous vs async tangent. Is there some paper or textbook comparing the general syscall design between various OS families?