[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: dmesg data avilable to others?

Others may find me out in left field on this, but I think that's entirely
too complex and a case of overcomputing for this data.

I am willing to bet that most (to nearly all) the queries of this would be
of the form "does this xxyy card work", which means that a simple
grep through the data is going to tell you a lot.  Further staring at
the particular dmesg(s) that had your match would be useful--likely
the simplest way would be to have each dmesg in its own file.  If
the comment of 1 - 5% is true, then there are depressingly few of
them to look at anyway.

One form of organization might be to sort them via their CPU class
I suppose, in that 486 boxes and Vaxen are less likely to have USB
ports than Pentium boxes.

Just having the raw data would be good for starters.

--STeve Andre'

At 06:15 PM 9/20/00 -0500, Matthew V. Sendak wrote:
Alex Charyna wrote:
> How would you arrange the data in database?
> the entire dmesg on it's own, sorted by processor?
> line by line, by retail model number....
> how is it currently managed?

I've seen no evidence that it is at all (I've been looking).  If there
were the resources, it would be incredibly convenient to have not only
dmesg data available, but also a way to format particularities with
particular hardware (perhaps organized by machine?).

Something on the order of:

Machine Name:
Memory (w/ vendor and speed, if known):
Cards (w/ vendor and model, if known):
Random Peripherals (w/ vendor and model, if known):
Weirdness that had to be performed to get everything to behave:
Known software that acts wonky:
Packages Installed:

If something like this could be setup, most of the folks I know would

I'm willing to help any way that I can.  Any other volunteers?


| Matthew V. Sendak                       m_(_at_)_procyon_(_dot_)_com |
| "The first thing we do, let's kill all the lawyers." /h