~ Archive for metrics ~

Download size exponentiation


The size of downloads has been increasing at a record clip. 
Downloads have been growing in size since the inception of the
concept… today I direct your attention to Wikipedia and Wikimedia
downloads.  Unavailable as torrents, but rather only via http, the
full downloads at 2+GB each are unwieldy even for people who run
downloads over their broadband connections at night while they
sleep.  Is WP dump size growing faster than avergae pipe
throughput to homes and workplaces?  (yes)  What can be done
about this? 

How about… shipping hard drives to people who want them?  Guaranteed 5-day delivery; for a reasonable fee (perhaps $80 for a drive + shipping + overhead?)…

ETC and user-centered tools


Everyone seems to think that developing tools around people’s daily
lives, on cleverly-designed platforms, is the Answer to lots of things
– the next iPod/computer/phone, new PCs for people in China’s urban
households, etc.

It doesn’t sound terribly innovative to me; am I just a stick in the
mud?  How can anyone get excited about a PC-like platform when
there’s some real innovation being done for $100 PCs that torally
rethinks many layers in the development and distribution of
computing?  Not that I think the $100 PC is the be-all or end-all
of what target consumers really need…  I’m foolish enough to
think that most things that end-users really need doesn’t get developed
at all.  A completely silly suggestion, I know.

Blogging from the Emerging Tech conference @ MIT


I’m blogging from MIT’s Emerging Technology conference. Earlier today,
there were some great keynotes and a remarkable panel on innovation; a
full report on those to come. Up next: a panel on Nuclear-Power Comeback, featuring support from former opponent (and personal hero) Stewart Brand.

Incapacity in times of crisis


I am amazed by the number of people who think that a perfectly acceptable response to an emergency is disruptive, individual flight.  I can think of a number of positive responses to emergencies, but this is an entirely negative one. Roads jammed with uncoordinated traffic
and hotels overwhelmed in the absence of coordination; people
struggling alone to cope with traumatic decisions — what a gray joke.

A few positive alternatives:

  • Individually : stay and
    prepare when at all possible.  Start preparing at the first sign
    of possible trouble — at the neighborhood level — if you’re one of
    those people who thinks this is possible.  If you are trained for
    emergency response, make sure the local response offices know how to
    reach you.  The well-prepared New Orleans residents on high ground
    who insisted on staying long after the whole city was evacuated —
    there should be more such people, not fewer.  This requires education and preparation ahead of time; teaching citizens how to preserve themselves and their things through a tornado, hurricane, earthquake, flood, drought, heat wave, mud slide, or electrical/oil/water/food shortage
    Teaching citizens how to help their neighborhood in these events; what
    organizations to contact and how during the aftermath; how to identify
    and shelter affected survivors.  It would be worth a great deal
    for one family in each small neighborhood to be well and truly prepared
    to ride out a disaster.

    And this business of stores and people ‘running out’ of key supplies in
    the run-up to every disaster gets old fast.  In the first place,
    each neighborhood should maintain a decent supply of these
    staples.  In the second, if Wal*Mart can figure out how to alert
    their suppliers to up production every time there’s a sale, surely
    cities can find a way to alert the usual suspects every time there’s an
    impending disaster-alert.

  • Gathering together for shared action;
    by block, neighborhood, or district.  Thursday and Friday are days
    off?  Great.  Have a local neighborhood meeting Wednesday
    night to discuss plans and options.  Where are the nearby bunkers
    and reinforced buildings?  Where would there be food, water, and
    shade for a week-long holdout?  Where can people bring cars and
    belongings that need better protection from the elements than their own
    rooms afford?  Oh, you don’t have a way to contact everyone in the
    neighborhod on short notice… noone responsible for maintaining
    contact numbers for everyone and organizing such meetings?  Better get on that then.

  • Gathering together for shared flight.  Tell everyone to share vehicles; at least three to a car and six to a van.  Give direction,
    train citizens how to respond quickly and effectively.  Make
    contact with all neighbors; don’t bring more than two bags with you for
    safe-keeping — leave them with a protected depot, or secure them at
    home, depending on where you live.  Coordinate the use of large
    trucks, buses, and vans; reimburse owners for transporting
    people.  Promote central message-boards for ride-shares and shared
    floor-space in nearby cities.  Open nearby halls and other
    facilities for short-term emergency occupants.  Encourage people
    to stay as close-by as possible.  Expecting
    people to take refuge in hotels and find transport via rental cars and
    scheduled buslines in times of disaster is a disaster in itself.

  • Helpful city responses.  Recruit
    a few thousand short-term staff from the ranks of the trained
    citizens.  Don’t have enough of those whom you trust?  Start
    a national emergency reserve program asap.   Offer safe, guarded repositories
    for belongings.  Provide guards for such repositories, and for
    sensitive or priceless areas such as hospitals and museums and those
    reinforced hotels/halls being used as shelters.  Do not double-book these guards; this
    is a full-time job.  Are people starved for food or water? 
    Set up ration lines.  This is one of your primary duties while
    people remain in the area.  Are half-destroyed stores and
    pharmacies vulnerable to looting?  Gather key goods in an orderly
    fashion, to distribute or preserve them.  Are there armed people
    wandering the streets?  Give them something useful to do, a
    partner, and proper gear.  No spare gear for such
    situations?  Better get on that, then.

“As long as we can choose our networks…”


I suppose that should be refined to “as long as we can choose what routes our traffic
takes…” — that is, which peers, what types of lines and routers, perhaps even what
last-mile providers.    It should be possible to say “if
there’s no way to send the following content along routes I trust,
don’t send it.

You don’t have to be paranoid to want this.  You might distrust a
route because you expect it to attempt to reconstruct, alter, and
resend content; because you suspect it of not accepting content from
certain areas or sites, because you worry that it keeps track of what
you send when, without your permission…  You might not want to
send content through any router that doesn’t respect the “return
receipt” flag which sends back information on how your packets
travelled on their way to a destination. Or you might just not want to
support in any way certain traffic providers, explicitly asking to
patronize other providers whenever possible.

“I’ll take ‘Arlnet Secure Wifi’ from my
house to the Arlington Center hub, ‘Hub of the World’ or ‘OpenBelNet’
from there to the Harvard U. hub, ‘HU Internet2’ to the Cambridge
Internet2 backbone hub,  and any lines/routers run by WorldCom or
UUNet, or on the dynamic ‘Debian-Class1’ network list.”

“Oh, as for my other options and preferences:
  •   Latency : as long as total latency is under 3 seconds, stick to the above networks rather than leave them.
    I use Return-receipt packet delivery, which sends me back a packet for
    every packet I send out, announcing the route it’s taken, or that it’s
    been dropped.  This more than doubles my bandwidth bill, but has
    its advantages… feed the resulting datastream into a route-analyzer;
    and one can set up all sorts of useful triggers.  When my
    preferred routes are all down, I can opt to use secondary networks –
    either outside my normal prefs, or at a higher rate.  I can also
    tell my Net-enabled applications not to complete sensitive transactions
    if there have been any fishy routings in the past minute. (Besides,
    whenever my provider makes too many routing mistakes, I get a free
    month of service.)”

Implementing this would seem to take significantly more intelligent routers and middleware than currently exists.
For a great coverage of some of the topics brought up at the Web of Ideas discussion tonight, see Geoff Huston‘s killer essay on the finance of networks, with its diversity of options laid out in gory detail.

People Finder XML Spec 1.1


Standards are sexy.  Reuniting families is sexierPFIF is worth the time it takes to read it.
In use by the grassroots Katrina PeopleFinder project [Katrina help wiki | search for refugees here).

How to help [KATRINA]


Public lists of “ways to help” with Katrina relief : a
retrospective.   Below are a collection of links from the
past weeks, and some public timelines.  How to do better next
time?  Is a “Disaster 2.0” effort the answer?

Timelines: from TPM | from Wikipedia

On contingencies; Ray in Austin


Ray in Austin
is my favorite blogger at the moment.  He’s writing solely about NO; his anger is tangible and practical.  He
provides a recap of Walter Maestri’s work in predicting hurricane
damage and evangelizing for preparedness, apparently in vain.  An NPR story from last year describes how explicitly this very storm had been played out in the minds of people preparing for it.

Meanwhile, skilled volunteers are actively not being called in
Chains of command are still worrying about looking good to others,
while the “related deaths” toll is steadily soaring.  I’ve seen
this kind of careful negligence before, and cringe to observe it when
so much is at stake.  10,000 deaths doesn’t sound unlikely to me any more.  According to
some sources (the NYT?), we’re up to 250k refugees in Texas, far more
than the 100k I predicted last week.

Meanwhile, the Army relief forces seem to have dived into NO from a standpoint of total war:

Combat operations are underway on the streets “to take
this city back” in the aftermath of Hurricane Katrina.
“This place is going to look like Little Somalia,” Brig. Gen. Gary
Jones, commander of the Louisiana National Guard’s Joint Task Force
told Army Times Friday as hundreds of armed troops under his charge
prepared to launch a massive citywide security mission from a staging
area outside the Louisiana Superdome. “We’re going to go out and take
this city back. This will be a combat operation to get this city under

… next up : you didn’t know what when???

Statements of unreliability, and earning trust


There are an increasing number of articles and works published whichrefer to Wikipedia as an implicitly reliable source — often ininappropriate contexts.  As its quality improves, Wikipedia
seemsto be shirking a certain quiet

to be modest; something which wasnot a problem back when none would
have mistaken it for a meticulouslyedited compilation. 

Example:  Ann Simmons, writing in the
LA Times on a matter of British peerage earlier this summer, used the
clause “according to Burke’s
and Wikipedia,”
a snippet which should immediately give one pause.  For one
thing,the two references have nothing in common.  It seems that aneditor tacked on the clause, “,
an online encyclopedia,”
in a vain effort at
clarification.  The full quote:

to Burke’s and Wikipedia, an online encyclopedia, Fredericksucceeded
his father, Robert Capell, the 10th Earl, who died in June.(The late
earl was a distant cousin of the 9th Lord Essex.)

The 11th Earl is a bachelor and has no children.
With no otherapparent successor in sight, Capell is the new heir to the earldom.
Hisaristocratic genealogy is documented in the 106th edition of “Burke’sPeerage & Baronetage.”

Please understand me; I will be the first to tell you that you can
articles and collections
on Wikipedia – including many
on peerage and
– which are among the
overviews in the Englishlanguage; if only you know where to look, and how to check the latest
revisions in each
article’s history.  

the process for checking information added to Burke’s and that
foradding information to Wikipedia are vastly dissimilar. 
TheWikipedia overview article on the Earl of
instance, continues to list no references, two months after theabove
(widely syndicated) article drew new attention to the wiki
articles on Frederick andRobert Capell. 

It is
embarrassing to imagine some newscasster, writer, lawyer,politician,
student, professor, or publicistciting a random article from Wikipedia,
on peerage or anything else,without somehow verifying
thatthe article had been carefullyresearched.  So what can be done?  Short of the
full-fledgeddrive for moderated or static views of the project, that is. 

What I would like to see is an internal quality review group that
issues regular recommendations
to the rest of the world.  At first these
recommendations would look like a brief whitelist of the categories and
subsubfields thatare really
top-notch and being monitored by a healthy community ofrespected
users.  As content improves, it would add various
hard metricsfor each of
various top-level categories — spot-check accuracy;vandalism
frequency/longevity; proportion/longevity of POV and otherdisputes;
rates of article creation, editing, and deletion; &c,

The recommendations could go out to educational, librarian, andresearch bodies –including
some of you reading this.   Theywould be prominently linked
to the sitewidedisclaimer[s].  The metrics would be available to
anyone asfeedback, including those working on relevant WikiProjects.
What do
youthink? (… read the full
A tip o’ the cursor to

(Update: quintupling of this post reverted.  Now how did that happen?  Rogue content editor alert…)

Goodness and trust in strange lands


Shako Mukulu is one of those people I think of every month, though we have not spoken since my last visit to his hometown of Kibwezi
over 5 years ago.  He taught me many things while I stayed with
him, off and on, for two months.  Most importantly, after “never blow your nose on anything but tissue paper (or you’ll get sick),” was his maxim not to mistrust anyone without very good reason.

I once came home after a day of packed matatu rides
with a neat rip through my pants pocket and missing a 1000 KSh note
that had been there the day before – my luxurious budget for last week
in the country.  I worried that it had been stolen;
Shako reproved me roundly, saying “don’t think such things if you don’t
know.”  I checked through all of my belongings, and found I had
turned my pockets out into a small bag without remembering it. 

Trust is a funny beast, but it is in many communities the right
default.  This is a complex topic, worthy of a few chapters of a
book, but of particular relevance to travellers in strange lands with
professional pickpockets. 
Differentiating between these pros and everyday people, and the milieus
each group prefers, is the difference between prudence and prejudice.

My mother left behind a makeup case – stolen?  perhaps.  I
left behind a stack of newspapers and a baseball cap (the Sox… you
had to ask?), and retrieved both of them once I found the right people
to ask.

Log in