Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.arch.embedded > #12743
| From | Don Y <this@isnotme.com> |
|---|---|
| Newsgroups | sci.electronics.design, comp.arch.embedded |
| Subject | Re: Cutting panels |
| Date | 2013-07-26 13:42 -0700 |
| Organization | Aioe.org NNTP Server |
| Message-ID | <ksun0e$p5a$1@speranza.aioe.org> (permalink) |
| References | (9 earlier) <b5aqkhFq6uqU1@mid.individual.net> <ksplpj$r4o$1@speranza.aioe.org> <b5flh4Fr90vU1@mid.individual.net> <ksufhc$4o7$1@speranza.aioe.org> <7x7ggdp2ef.fsf@ruckus.brouhaha.com> |
Cross-posted to 2 groups.
Hi Paul,
On 7/26/2013 12:06 PM, Paul Rubin wrote:
> Don Y <this@isnotme.com> writes:
>> So, you would consider putting "smarts" *at* the freezer and
>> sending *messages* to some "remote" location instead of running
>> thermocouple leads to that remote location?
>
> In networked systems I've generally found it preferable to put as much
> of the "smarts" as possible into one place, in order to keep the
> protocols simple. This is especially useful if the different nodes are
> programmed by separate people or teams, since complex protocols can
> drastically increase the amount of confusion, meetings, integration
> problems, etc that you end up dealing with. This would suggest making
> the thing at the freezer be a dumb computer peripheral controlled by a
> more complex host at the other end.
When I first created the irrigation controller, I took exactly
the *opposite* approach: smarts in the controller (SBC) and a
"dumb terminal" as the "remote computer".
I.e., the user interface (software) and application ran in
the SBC and just used the "remote terminal" for the visual
aspect of the user interface. This allowed the remote
computer to be powered down (since you rarely interact with
the irrigation controller!) nearly 100% of the time while
still having the functionality of the irrigation controller
operational. ("OK, there's no display. So what? I already
know what I have to do...") This was trivial and inexpensive:
serial ports are cheap and a character based interface easy
to implement -- even on a SLOW serial port! (think: curses).
The economics of today suggest using ethernet for connectivity
(or wireless, but that's another can of fish). Especially
if you want to do more than push a few characters down the
link (e.g., distributing multimedia; having nodes interact
with each other via RPC; RPC-over-SLIP?? ick! etc.)
But, the price of admission for ethernet usually comes with
a 32bit controller attached. (Are there any SoC's with
ethernet MACs that are smaller than 32b?) Protocol stacks
consume more resources than "serial port drivers" so these
devices tend to be well endowed (FLASH+RAM). So, even small
devices nowadays have more power and resources than, e.g.,
a PC/AT!
Therefore, why not *use* that power for more than just "remote
valve activation"? Why not dispatch some portion of the application
*to* this node in addition to the physical interface to the
field wiring interface? And, since each "remote device" needs
ethernet connectivity, you now have a bunch of devices that can
be slightly oversized for little/no cost and the extra capacity
applied to the application -- instead of just the "remote wiring".
If you have a framework that lets these resources be used AS IF
general purpose resources (i.e., MIPS and BYTES), then you can
run portions of the application wherever you have some surplus
capacity. Instead of running thousands of threads on *a*
processor, run dozens of threads on each of scores of processors!
If you open a many "consumer" devices, you;ll see that this is
how they have tended to partition the many "tasks" that the
product is responsible for performing during operation. E.g.,
a VCR might have four or five MCUs in it. An automobile might
have dozens (one *just* to control the power windows!). It's
easier than having a giant wiring harness come to a single
point for one, BIG computer! And, can enhance reliability as
the functions of each device are more definitely partitioned.
No way the power window controller is going to interfere
with the ignition timing! Even if it *isn't* running on a
processor with support for virtual memory! :>
> Running analog signals over long distances has its own problems but if
> you're ok with the wiring hassles etc., maybe that can simplify things
> even further.
>
Back to comp.arch.embedded | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Cutting panels Don Y <this@isnotme.com> - 2013-07-21 02:47 -0700
Re: Cutting panels John Fields <jfields@austininstruments.com> - 2013-07-21 08:05 -0500
Re: Cutting panels Roberto Waltman <usenet@rwaltman.com> - 2013-07-21 10:38 -0400
Re: Cutting panels "Carl Ijames" <carl.ijames@nonono.verizon.net> - 2013-07-21 12:36 -0400
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 09:48 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 13:29 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 14:10 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 16:23 -0700
Re: Cutting panels Grant Edwards <invalid@invalid.invalid> - 2013-07-21 21:04 +0000
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 15:07 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 16:35 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 17:11 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 16:27 -0700
Re: Cutting panels Robert Wessel <robertwessel2@yahoo.com> - 2013-07-23 21:50 -0500
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 21:24 -0700
Re: Cutting panels Jasen Betts <jasen@xnet.co.nz> - 2013-07-24 09:09 +0000
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-24 16:02 -0700
Re: Cutting panels hamilton <hamilton@nothere.com> - 2013-07-21 16:52 -0600
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 17:13 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 12:35 -0700
Re: Cutting panels amdx <nojunk@knology.net> - 2013-07-21 09:54 -0500
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 09:32 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 13:36 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-21 14:04 -0700
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-21 13:14 -0700
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-21 13:26 -0700
Re: Cutting panels [solved?] Joerg <invalid@invalid.invalid> - 2013-07-21 14:06 -0700
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-21 16:08 -0700
Re: Cutting panels [solved?] Rich Webb <webb.ra@example.net> - 2013-07-21 19:42 -0400
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-23 21:45 -0700
Re: Cutting panels [solved?] Jan Panteltje <pNaonStpealmtje@yahoo.com> - 2013-07-22 07:51 +0000
Re: Cutting panels [solved?] Charlie E. <edmondson@ieee.org> - 2013-07-23 05:57 -0700
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-23 14:46 -0700
Re: Cutting panels [solved?] Don Lancaster <don@tinaja.com> - 2013-07-23 16:41 -0700
Re: Cutting panels [solved?] Charlie E. <edmondson@ieee.org> - 2013-07-23 17:21 -0700
Re: Cutting panels [solved?] Jasen Betts <jasen@xnet.co.nz> - 2013-07-24 09:26 +0000
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-25 12:11 -0700
Re: Cutting panels [solved?] Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> - 2013-07-25 16:04 -0400
Re: Cutting panels [solved?] Don Y <this@isnotme.com> - 2013-07-25 13:03 -0700
Re: Cutting panels [solved?] Jasen Betts <jasen@xnet.co.nz> - 2013-07-23 13:40 +0000
Re: Cutting panels [solved?] Joerg <invalid@invalid.invalid> - 2013-07-23 07:38 -0700
Re: Cutting panels [solved?]/summary Don Y <this@isnotme.com> - 2013-07-29 15:18 -0700
Re: Cutting panels andrews@sdf.lonestar.org - 2013-07-21 12:45 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 19:17 -0700
Re: Cutting panels Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> - 2013-07-21 19:45 -0400
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 17:07 -0700
Re: Cutting panels Peter Bennett <peterbb@somewhere.invalid> - 2013-07-21 16:57 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 20:58 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-21 21:02 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-22 09:20 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 12:15 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-22 16:17 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 18:52 -0700
Re: Cutting panels Jasen Betts <jasen@xnet.co.nz> - 2013-07-23 14:00 +0000
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 15:13 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-23 07:09 -0700
Re: Cutting panels Charlie E. <edmondson@ieee.org> - 2013-07-23 10:03 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 21:59 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 15:54 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-24 13:09 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-24 15:51 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-26 09:13 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-26 11:35 -0700
Re: Cutting panels Paul Rubin <no.email@nospam.invalid> - 2013-07-26 12:06 -0700
Re: Cutting panels Jan Panteltje <panteltje@yahoo.com> - 2013-07-26 19:44 +0000
Re: Cutting panels Paul Rubin <no.email@nospam.invalid> - 2013-07-26 22:21 -0700
Re: Cutting panels Jan Panteltje <pNaonStpealmtje@yahoo.com> - 2013-07-27 08:12 +0000
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-26 13:42 -0700
Re: Cutting panels Paul Rubin <no.email@nospam.invalid> - 2013-07-26 18:50 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-26 22:49 -0700
Re: Cutting panels Paul Rubin <no.email@nospam.invalid> - 2013-07-27 00:53 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-27 10:16 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-26 23:52 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-27 17:36 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-28 09:22 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-28 12:31 -0700
Re: Cutting panels Paul Rubin <no.email@nospam.invalid> - 2013-07-28 13:48 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-28 14:20 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-28 16:35 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-28 17:39 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-28 19:38 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-29 11:15 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-29 16:30 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-29 17:38 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-29 21:45 -0700
Re: Cutting panels Joerg <invalid@invalid.invalid> - 2013-07-30 12:20 -0700
Re: Cutting panels Jeff Liebermann <jeffl@cruzio.com> - 2013-07-21 21:38 -0700
Re: Cutting panels Robert Wessel <robertwessel2@yahoo.com> - 2013-07-22 00:39 -0500
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 01:39 -0700
Re: Cutting panels Robert Wessel <robertwessel2@yahoo.com> - 2013-07-22 19:09 -0500
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 20:10 -0700
Re: Cutting panels Jasen Betts <jasen@xnet.co.nz> - 2013-07-23 14:07 +0000
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-22 18:33 +0100
Re: Cutting panels Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> - 2013-07-22 13:44 -0400
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-22 19:00 -0700
Re: Cutting panels Jan Panteltje <pNaonStpealmtje@yahoo.com> - 2013-07-23 10:09 +0000
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-23 11:45 +0100
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-23 13:46 +0100
Re: Cutting panels 0bama <0B@nocontrast.org> - 2013-07-23 13:06 +0000
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-23 22:19 +0100
Re: Cutting panels 0bama <0B@nocontrast.org> - 2013-07-24 08:05 +0000
Re: Cutting panels Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> - 2013-07-23 10:07 -0400
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-25 01:26 -0700
Re: Cutting panels Jasen Betts <jasen@xnet.co.nz> - 2013-07-23 14:18 +0000
Re: Cutting panels Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> - 2013-07-23 11:38 -0400
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-23 22:32 +0100
Re: Cutting panels Frank Miles <fpm@u.washington.edu> - 2013-07-23 21:37 +0000
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 16:16 -0700
Re: Cutting panels Theo Markettos <theom+news@chiark.greenend.org.uk> - 2013-07-24 10:37 +0100
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-25 01:27 -0700
Re: Cutting panels Gonadicus <gonadicus@no_spam.com> - 2013-07-23 12:46 +0800
Re: Cutting panels sms <scharf.steven@geemail.com> - 2013-07-23 04:58 -0700
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 16:13 -0700
Re: Cutting panels Jasen Betts <jasen@xnet.co.nz> - 2013-07-23 13:25 +0000
Re: Cutting panels Don Y <this@isnotme.com> - 2013-07-23 21:42 -0700
csiph-web