<HTML><HEAD></HEAD>
<BODY dir=ltr>
<DIV dir=ltr>
<DIV style="FONT-SIZE: 12pt; FONT-FAMILY: 'Calibri'; COLOR: #000000">
<DIV>Sorry, forgot to copy Server-devel. I was hoping to gather some
statistics along the lines of what James Cameron suggested in the url below
regarding real world experience with wifi devices that work well in a
schoolserver plus XO setting. Please contribute anecdotes and any hard
measurements.</DIV>
<DIV
style="FONT-SIZE: small; FONT-FAMILY: 'Calibri'; FONT-WEIGHT: normal; COLOR: #000000; FONT-STYLE: normal; TEXT-DECORATION: none; DISPLAY: inline">
<DIV style="FONT: 10pt tahoma">
<DIV> </DIV>
<DIV style="BACKGROUND: #f5f5f5">
<DIV style="font-color: black"><B>From:</B> <A title=timmoody@sympatico.ca
href="mailto:timmoody@sympatico.ca">Tim Moody</A> </DIV>
<DIV><B>Sent:</B> Thursday, June 27, 2013 3:56 PM</DIV>
<DIV><B>To:</B> <A title=quozl@laptop.org href="mailto:quozl@laptop.org">James
Cameron</A> </DIV>
<DIV><B>Subject:</B> Re: [Server-devel] Testing and comparing wireless
setups.</DIV></DIV></DIV>
<DIV> </DIV></DIV>
<DIV
style="FONT-SIZE: small; FONT-FAMILY: 'Calibri'; FONT-WEIGHT: normal; COLOR: #000000; FONT-STYLE: normal; TEXT-DECORATION: none; DISPLAY: inline">
<DIV dir=ltr>
<DIV style="FONT-SIZE: 12pt; FONT-FAMILY: 'Calibri'; COLOR: #000000">
<DIV>Took a stab a creating a home for the information you suggest.</DIV>
<DIV> </DIV>
<DIV><A
href="http://wiki.laptop.org/go/User:Holt/XS_Community_Edition/Wifi_Experience">http://wiki.laptop.org/go/User:Holt/XS_Community_Edition/Wifi_Experience</A></DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV>-----Original Message----- </DIV>
<DIV>From: James Cameron </DIV>
<DIV>Sent: Monday, June 24, 2013 7:24 PM </DIV>
<DIV>To: David Farning </DIV>
<DIV>Cc: server-devel </DIV>
<DIV>Subject: Re: [Server-devel] Testing and comparing wireless setups. </DIV>
<DIV> </DIV>
<DIV>[Warning: this author is on coffee for the first time in months]</DIV>
<DIV> </DIV>
<DIV>On Mon, Jun 24, 2013 at 07:14:11AM -0500, David Farning wrote:</DIV>
<DIV>> As part of the XSCE, we would like to start offering wireless
setup</DIV>
<DIV>> guidelines.</DIV>
<DIV>> </DIV>
<DIV>> Large deployments might be able to afford a site assessment.
Small</DIV>
<DIV>> deployments are left scratching their heads and making decision
on</DIV>
<DIV>> anecdotal references. This results in a bottleneck. Smart and</DIV>
<DIV>> motivated people are spending time, early in a deployments</DIV>
<DIV>> lifecycle, on wireless which means other critical task are left</DIV>
<DIV>> undone.</DIV>
<DIV>> </DIV>
<DIV>> This topic is challenging because it is complex. Every workload
is</DIV>
<DIV>> different and every electromagnetic environment is different.</DIV>
<DIV>> </DIV>
<DIV>> My idea is start very simply by assigning a numerical score for</DIV>
<DIV>> various wireless devices based on simple criteria like
throughput,</DIV>
<DIV>> connection count, overheating, and range. My _very_ naive
experiences</DIV>
<DIV>> are that among consumer grade devices:</DIV>
<DIV>> 1. Some devices can handle more connections than others before
they</DIV>
<DIV>> start dropping connections.</DIV>
<DIV>> 2. Some devices can handle higher throughput - Several kids
watching</DIV>
<DIV>> youtube and me doing a download.</DIV>
<DIV>> 3. Some devices overheat and reset more frequently than others.</DIV>
<DIV>> 4. Some devices have better range than others.</DIV>
<DIV> </DIV>
<DIV>I think this is overly simplistic, yet I know a simple heuristic is</DIV>
<DIV>what is needed. So I suggest coming at it from a different
angle.</DIV>
<DIV> </DIV>
<DIV>> Does this information seem valuable to deployments? Does the
general</DIV>
<DIV>> approach seem sane?</DIV>
<DIV> </DIV>
<DIV>The stack is deep, so deep that anecdote can be inaccurate and</DIV>
<DIV>misleading. The phase space has a large number of dimensions.
It may</DIV>
<DIV>be better to accumulate test reports so that people can form their
own</DIV>
<DIV>opinions. The test report should include:</DIV>
<DIV> </DIV>
<DIV>- the wireless access point manufacturer, model, serial number, and</DIV>
<DIV> firmware version,</DIV>
<DIV> </DIV>
<DIV>- the XSCE version,</DIV>
<DIV> </DIV>
<DIV>- the XO OS version, model, and wireless card,</DIV>
<DIV> </DIV>
<DIV>- the measured capability of the internet service, in terms of
latency</DIV>
<DIV> and bandwidth, measured with ping and wget,</DIV>
<DIV> </DIV>
<DIV>- the number of access points deployed adjacent to the XSCE,</DIV>
<DIV> </DIV>
<DIV>- the number of XO users active during the test,</DIV>
<DIV> </DIV>
<DIV>- individual user XO performance observations, in terms of latency,</DIV>
<DIV> bandwidth, and packet loss, such as ping, wget, and curl POST,</DIV>
<DIV> rolled up into a total performance score in the range 1 to 10.</DIV>
<DIV> </DIV>
<DIV>Then, abstract from the collection of reports a list of access
points,</DIV>
<DIV>user counts, and total performance score. Link each line to the</DIV>
<DIV>actual report.</DIV>
<DIV> </DIV>
<DIV>This ensures that the claims the community make about the access</DIV>
<DIV>points can be substantiated ... which benefits the community, the</DIV>
<DIV>deployers, and the manufacturers of the devices.</DIV>
<DIV> </DIV>
<DIV>(With enough data, of the order of ten or so reports, the workload
and</DIV>
<DIV>radiofrequency environment aspects can be reduced in importance.
I</DIV>
<DIV>think those aspects are better handled by site design guidelines
based</DIV>
<DIV>on what we find is common to all working deployments.)</DIV>
<DIV> </DIV>
<DIV>-- </DIV>
<DIV>James Cameron</DIV>
<DIV>http://quozl.linux.org.au/</DIV>
<DIV>_______________________________________________</DIV>
<DIV>Server-devel mailing list</DIV>
<DIV>Server-devel@lists.laptop.org</DIV>
<DIV>http://lists.laptop.org/listinfo/server-devel</DIV></DIV></DIV></DIV></DIV></DIV></BODY></HTML>