[Noisebridge-discuss] an interesting potentiometer failure mode (tin whiskers!)

D J Capelis mail at capelis.dj
Sat Dec 22 13:07:55 UTC 2012


On Sat, Dec 22, 2012 at 3:47 AM, Taylor Alexander <tlalexander at gmail.com>wrote:

> It just struck me as silly noise. Should we just copy the daily contents
> of Hacker News to this list? A large<http://www.nasa.gov/mission_pages/cassini/whycassini/cassini20121218.html>
> percentage<http://arstechnica.com/tech-policy/2012/12/behind-closed-doors-at-the-uns-attempted-takeover-of-the-internet/>
> of<http://www.dslreports.com/shownews/Its-Clear-Verizon-Is-Blocking-Google-Wallet-AntiCompetitively-122513>
> the<http://www.techdirt.com/articles/20121220/02365821447/intellectual-ventures-dont-mind-our-2000-shell-companies-thats-totally-normal.shtml>
> articles<http://torrentfreak.com/pirate-bay-censorship-backfires-as-new-proxies-bloom-121222/>
> posted <http://enpundit.com/a-5-lamp-powered-solely-by-gravity/> there<http://fernstrategy.com/2012/12/21/the-end-of-x86-an-update/>
> daily<https://takingsenseaway.wordpress.com/2012/12/20/letter-from-a-passenger-how-do-tsa-employees-feel-about-working-for-a-despised-agency/>are highly relevant to readers of the list. But of course it would be silly
> of me to do that. And likewise I thought it was silly for someone to just
> post an article here without much explanation as to why (excluding "it's
> interesting", because I read 20 articles a day you'd all find interesting,
> but I don't post them here).
>
> So in futile hope of discouraging silly conduct, I asked if perhaps there
> was more of an explanation (which, by the way, was not rhetorical). Of
> course I know that I can't fix people who are "wrong on the internet" by
> suggesting they stop being wrong, and that saying nothing may have been
> better, but you know how it's best to just ignore bad drivers when they cut
> you off? Sometimes you've still gotta give them the finger, even if it's
> not going to make a damn difference.
>
1) Tin whiskers are fascinating, in part because they happen on their own,
aren't something we have much control over, are somewhat mysterious and
totally mess up circuits in a way that's really incredibly frustrating.
2) Sensor networks in cars are increasingly relevant.
3) Faults in circuits that cause unintended acceleration is a failure mode
that speaks to a lot of people.
4) You seem to take this all very seriously.  Perhaps you shouldn't.
5) Speaking of silly conduct that seems futile to try discouraging anymore,
do you always top-post?

If you find 20 articles every day that have everything in 1-3, by all means
post it here I guess if they speak to you.  But if Susan wants to post an
article now and again that really speaks to her or her interests, I don't
really get why you feel you'd need to discourage that.  I haven't noticed a
huge pattern of bunches of links posted to this list all the time without
context, so one wonders which windmill you're tilting away at.

And sure, there could have been more context, if that's really your
objection.  But if that's your objection maybe instead of coming off as
"How about you shut that silly noise up?" you could try a tact of "Maybe
you could disseminate this information better by providing a better
overview of what makes it fascinating and amazing so that the content that
spoke to you can be seen by others the way you see it?" which might be a
useful thing to say.

Sometimes one of the better things I read is a reminder to read something I
intended to read but didn't.  (Or to re-read something, as recently
happened with that soon to be classic paper on exploiting computers with
lightbulbs[1] which it turns out is just as good on the second reading.)
One of the reason I even bother paying attention to some communities is in
part so I can hear how different pieces of knowledge speak to different
people.

In the meantime, here's the research paper that recently came out tin
whisker growth:
http://www.osti.gov/bridge/purl.cover.jsp?purl=/1013351-BXJitV/1013351.pdf

Perhaps someone on this list would be interested in digging into it and
trying to sum up whether we have any hope of finding solutions that are
lead free anytime soon.  With scale going the way it is, the problem
certainly isn't sucking any less.  I know my understanding of the topic
matter isn't what I'd like to be.  Maybe someone else wants to drop some
knowledge and take a crack at summing up where we are now on the ever
expanding tin whisker battlefront?  How is that war going anyway?

~DJ

[1] For those who haven't seen it or want to read it again, google: "Using
Memory Errors to Attack a Virtual Machine"  (I assume it must have been
posted here already at various points.)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.noisebridge.net/pipermail/noisebridge-discuss/attachments/20121222/13942f78/attachment-0003.html>


More information about the Noisebridge-discuss mailing list