On 04/10/2018 07:05 PM, Jonathon Fernyhough wrote:
> Much as I'm wary about raising this after last time... here goes. Really
> just putting this out there as a thing to consider if it's not already
> known about.
> With 3.3.13, upstream introduced support for a system-wide default
> toprc, /etc/topdefaultrc [1]. This is their "recommended" way of setting
> sensible/usable/accessible defaults rather than the "all or nothing"
> --disable-modern-top.
> I don't know what the specific implementation differences are between
> the two (e.g. whether topdefaultrc is compatible with/overrides
> --disable-modern-top) but it should provide an approach more "aligned to
> upstream". Coming up with a "sensible default", on the other hand, is
> likely to be more complicated...
> (I've gone for an htop-like scheme with a nice green 'column heads',
> idle tasks off (i), Zero-Suppress on (0), for Manjaro [2].)
> [1]
> https://gitlab.com/procps-ng/procps/commit/55a42ae040eaa19fd3089f56d98ccbde5a9abc3a
> [2]
> https://github.com/manjaro/packages-core/blob/master/procps-ng/topdefaultrc

I have boxes from 2 - 16 cores. In the past, you get old top on anything
4-cores or less. You get modern-top on anything with more than 4-cores --
that's just nuts. At first I didn't care for the new top, now I'm quite happy
with it, and in fact prefer it. Nothing more frustrating that what top you get
being "like a box of chocolates". Two xterms open, one to an 8-core box shows
new-top, one to a dual-core box show old-top -- that's nuts.

The PKGBUILD change for rebuilds is simple, but we ought to have one or the
other (modern preferred here, but I have nothing against someone else wanting
the old top view available) If a simple config can make both happy - that
seems like a no brainer. Otherwise, I'll keep using asp to grab the latest and
fix it myself by removing the --disable-modern-top build option.

       #       --disable-modern-top \

Consistency beats a 'box-of-chocolates' any day...

David C. Rankin, J.D.,P.E.

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to