[time-nuts] 5370 controller topics (was: processor boards available)

[Tom Van Baak]

> This is actually pretty interesting to me.  The questions of how to choose
> an RTOS and how to handle power management seem to be relevant engineering
> topics when discussing the 5370 processor boards.  It's not an academic
> question in the professional world, where certain instruments I won't name

Hi John,

It's very interesting to me too! And, yes, threads like this belong on time-nuts -- as long as the postings are constructive, educational, and friendly.

What happened in this case is that thread degraded to the point where people were arguing about batteries and sync'ing file systems. Then I start sending emails in private, and when that doesn't work I suggest the thread stays on-topic or ends, and when that doesn't work we turn on full moderation until the list quiets down.

This case is especially awkward when a hero (John Seamons) freely shares an incredibly cool project with the list and then instead of a gradual, constructive series of confirming measurements or accolades from group members, his design is quickly peppered with accusations about a minor potential problem by people that probably don't even own a 5370 or haven't ordered his kit yet. It's at this point I jump in an play referee or group nanny.

Our time-nuts list works really well when the postings are few and each one conveys either an on-topic question or a really interesting, well-researched, informative set of answers from people who share their hard-earned experience with the rest of us.

The list starts to fail, and actually repel good members, when people make cheap postings that don't reflect any research or lab experiments or measurement or life experiences. I would rather read one belated posting by someone who describes how they actually faced and solved a problem than dozens of live postings about what might be problems that someone else should have solved. The goal is to keep the group as practical, numerical, experimentation, and communal as possible.

The S/N ratio degrades when there are too many postings from the couch instead of the bench.

Anyway, if you or anyone else on the list has serious experience with modern embedded OS's for T&F instrumentation, please start a thread and share it with the group, or pass along white papers, product manuals, requirements, etc.

Thanks,
/tvb

p.s. My TSC 5110A has worked fine for a decade using freeBSD running off a CF card. It's the 5120A that always seems to have issues. On my bucket list is to capture all IDE traffic to/from the CF card on these units and see first-hand were the problem is.