On Tue, 2011-12-06 at 01:47 -0800, Dale Miller wrote: > John McKown wrote: "With Z/OS, Windows, and Mac OSX, I don't even have > a change to understand some of the internals." > > Read "Mac OS X Internals" by Amit Singh- Addison-Wesley 2007. Also the > source code for Darwin (OS/X) is downloadable from Apple. > Not to defend Apple, but it is interesting to note that Apple refers > to internals as "implementation detail", and discourages attention to > internals by noting that implementation details are subject to change, > and that depending on implementation details in applications leads to > migration difficulties.
I don't like using undocumented internals. That's why I like GNU/Linux. OK, it's not really "documented". Unless you go with source==documentation. And talk about not guaranteed to stay consistent. Well the binary API to userspace is fairly consistent. And usually backward compatible. In z/OS, I stick with GUPI defined interfaces. It's much safer. Especially since my boss said he'd hang me out to dry if I ever make the system dependent on a non-GUPI interface. > > That said, any vendor whose software writes undocumented SMF records > is not playing cricket. If the SMF records are to be useful me, they > have to be documented. If they're not useful to me, I'll turn them > off, and they won't be useful to the vendor. > > Dale Miller > And, from what I've been told, if you run CA's CAS9 (RIM) (and who doesn't), it will front end IBM's SMF processing in such a way as to cause the records to be produced, regardless of anything you do. Of course, the actual writing of them may be suppressed. But not the production of the record. -- John McKown Maranatha! <>< ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
