How many years will take Linus&co to develop Solaris contractfs+project

Posted Aug 4, 2014 14:37 UTC (Mon) by kloczek (guest, #6391)
Parent article: The unified control group hierarchy in 3.16

Anyone knows how long it may take?
Why Linux still is suffering on NIH (Not Invented Here) syndrome?
Why something so simple like managing tasks and processes must be driven by yet-another-stupid-fs?
Why no one from Linux developers is able to sit down study existing implementation of solutions of some problems, after this develop on first step consistent base API with plan how to extend base functionalities, and after this stick to agreed/approved plan?
Why .. ?
Why .. ?
.
.

Ten years after developing DTrace on Solaris most of the time spend on SystemTap, LTT, LTTng and many other attempts can be put in garbage and now more people on Linux is using DTrace delivered by commercial company.

Why Linux developers are trying again and again repeating the same errors and expecting that at some point it will Work(tm)?

How many years will take Linus&co to develop Solaris contractfs+project

Posted Aug 5, 2014 23:37 UTC (Tue) by nix (subscriber, #2304) [Link]

now more people on Linux is using DTrace delivered by commercial company

Well, I'd be very interested to hear where you got this information from. I'm one of the DTrace for Linux developers, and, y'know, I don't have that information. Possibly my bosses have it, but if so they haven't told me. To be honest I have no idea how anyone could know this sort of thing without horrendously invasive spying on users, or wildly unreliable usage surveys which have as far as know not been conducted.

(But maybe you mean some other Linux DTrace developed by a commercial company? Or perhaps you mean not 'more people than use SystemTap / perf / something else' but rather 'more people than used to use it', which is trivially true if it is used by anyone at all, since it has not always existed.)