Messages in this thread | | | From | (david parsons) | Subject | Re: ARMS WAVING!!! Proposal to fix /proc dainbrammage. | Date | 26 Oct 1998 23:45:24 -0800 |
| |
In article <linux.kernel.199810270021.TAA23532@jupiter.cs.uml.edu>, Albert D. Cahalan <acahalan@cs.uml.edu> wrote: > >david parsons writes: >> Albert D. Cahalan <acahalan@cs.uml.edu> wrote: >>>Mark H. Wood writes: > >>>> (My take on this question is that programs should not be looking >>>> in /proc at all; everything shown there should be available in >>>> binary form via syscall, >>> >>> Oh yes! I really hate to parse /proc files. Parsing is not at all >>> reliable, considering unexpected format changes and spelling fixes. >> >> That's not the fault of the /proc filesystem, but of the people >> who promiscuously change the format of entries there without >> considering the consequences. > >It is the fault of the /proc filesystem for being so tempting to reformat. >Ugly files like /proc/*/stat are seldom damaged, while pretty files like >/proc/*/status are often broken.
Any interface is tempting to reformat. It's still bad software design, and it seems a really bad idea to strip kernel interfaces just because badly-trained programmers can tweak them.
>We wouldn't have so many problems if all /proc files were just a series >of space-delimited numbers without labels.
I don't know about that. I find it pretty easy to parse labelled items, but have been repeatedly bitten by code (procinfo, code in the procps suite) that makes assumptions about the format of the space-delimited numbers.
____ david parsons \bi/ And, of course, I (and any ps-style tools I write) \/ can easily read /proc/*/status without needing to know undocumented arcane details...
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |