Vacuum

Paul Harvey (as played by Rich Hall), 1987

Paul Harvey passed away today at the age of 90.

Rich Hall played him on Saturday Night Live - a clip I can't find right now - and also in this 1987 Showtime special I found on Youtube.

Time Magazine's obituary talks a bit about the Rich Hall piece on SNL:

A salesman for himself and his vision of the American dream, he was also a master peddler of many products, whose makers were as loyal to Harvey as his listeners. A skit from the 1984-85 season of Saturday Night Live had Harvey (played by Rich Hall) compulsively peppering his news items with sponsor names. The man remained unapologetic. "Some days," he told Larry King in 1988, "the best news in the broadcast is the commercial. You can keep your natural teeth all your natural life! There is a glove that doesn't wear out! There is a car battery that keeps its promises! That's good news! And I would use those things on the air if they were not in the body of the commercial." The finest huckster is the one who's sold himself on the product first.

I'll keep looking.

Ann Arbor Gold gift certificates turn to lead; CertifiChecks of Dayton, OH in Chapter 7 bankruptcy

First noted on Twitter:

      

CertifiChecks, administer of the Ann Arbor Gold program, has gone out of business

which leads to this Ann Arbor Chamber of Commerce news:

The Chamber was recently informed that CertifiChecks, the company that administered the Ann Arbor Gold program, has gone out of business. The statement below was available on their website. Calls to the CertifiChecks management team have not been returned. 

CertifiChecks will no longer administer Ann Arbor Gold. 

CertifiChecks gift certificate sales for Ann Arbor Gold have been halted. 

Do not deposit Ann Arbor Gold gift certificates in the bank as they will be returned. 

Merchants: Please remove all Ann Arbor Gold door and register decals. 

Ann Arbor Merchants affected by this are asked to contact Chamber president Jesse Bernstein for assistance.

CertifiChecks is based in Dayton, Ohio; here's what the news report from writer Ben Sutherly in Dayton says:

The company said it has halted gift certificate sales and its administration of gift certificate programs, effective Thursday, Feb. 26. The company said it is filing for Chapter 7 bankruptcy in federal court in Dayton.

A 2002 story in the Dayton Business Journal gives some sense for their scope of operations:

CertifiChecks, founded in 1999 by Steve Buchholz, serves nearly 40,000 associations, chambers of commerce and mall or shopping centers nationwide.

By 2004, when Certifichecks had signed up a network of Christian-owned hotels, they had grown substantially, as noted in the Dayton Business Journal:

CertifiChecks, founded locally by Steve Buchholz in 1999, now has about 72,000 restaurant associations, chambers of commerce and bed and breakfast associations as customers nationwide.

News is starting to trickle in about the impact of this. From Muskegon, MI, Chad Lerch of the Muskegon Chronicle writes in the story Muskegon Chamber fumes as widespread CertifiChecks gift checks bounce

MUSKEGON COUNTY -- An Ohio-based company that offers gift checks for retailers abruptly closed Thursday, leaving West Michigan chambers of commerce, merchants and shoppers fuming and searching for answers. Cindy Larsen, president of the Muskegon Area Chamber of Commerce, said up to $150,000 worth of gift checks for "Muskegon Money" could still be unspent. If that's the case, a myriad of shoppers are holding checks that are worthless -- at least for now. "We're freaking out," Larsen said. "We're freaking out because we're worried that someone is going to get stuck."

More will certainly come out of this story - if Certifichecks really has this many subscribers, gift cards and gift certificates are going to start bouncing all over the country.

Frozen blueberry price report + Eastern Washington Blueberry Workshop March 2009

As ever I am tracking the blueberry markets in anticipation of summer.

This from the Capital Press: Blueberry stocks in cold storage outpace past years; Berry dealers see bright side in large stocks of fruit

Frozen blueberries are selling at a steady clip, but carryover stocks from 2008's crop are still huge, according to industry experts.

Sales seem to have picked up speed late last year, with warehouse stocks falling by 4 percent in October, 6 percent in November and 7 percent in December 2008, according to the USDA's National Agricultural Statistics Service.

At 153.7 million pounds, year-end frozen blueberry inventories in 2008 were nonetheless 35 percent higher than in 2007 and 55 percent higher than in 2006.

"The movement has been decent. It's just a lot of fruit that is still sitting in the freezer," said small fruits analyst Rod Cook of Ag-View Consulting.

Cook notes prices for berries are down since harvest time:

"Sometimes it's as big a challenge to get fruit into a freezer as anything else," said Cook. "Some have been so full they haven't been capable of taking additional fruit."

The strong supply has depressed frozen blueberry prices, which now range from $0.80 to $1.05 per pound, he said. "The price has definitely retreated from harvest time, 20 cents or more."

You can hear more on the Oregon blueberry market prospects for 2009 at the Eastern Washington Blueberry Workshop March 12, 2009 in Prosser, WA, free to the first 80 advance registrants.

ps. how do you get the title "Blueberry marketing expert"? sign me up!

On wandering into someone else's wiki

Every once in a while you find a smallish wiki that looks like it's healthy and has interesting material.  These wikis are bigger than the single person brain-dump, and smaller than Wikipedia, and less abandoned than most temporary bouts of enthusiasm that bring people into wiki-land.

Inevitably when I find one of these I have the urge to edit and organize just long enough to figure out what kind of stuff is in it.  Sometimes the wiki authors so much assume that everyone knows what some key terms or concepts are, and so those things never get their own pages until you wander in and link them.  Other times the wiki authorship collectively is so interesting in making links to things that they never manage to get around to putting pages at the other end of the links.  Doing a round of edits to random pages to try to improve them just enough to make sense of them is a way to figure out whether you want to wrap your brain more about the topic.

Wikipedia, in its timeless charmless categorical imperative way, calls the sort of people who wander into a wiki and start tidying things up as WikiGnomes.  This is a passive-aggressive way of saying "who the hell are you and why are you editing all of these pages and not adding anything to them".   If someone else calls you that, step back and head back to your blog.  The other term of art for this is "Wiki gardening", which implies some low-level drone job pulling weeds and fixing other people's punctuation, which also completely misses the point.

Sometimes, in order to make your own sense of something, you have to change it and see if it changes back after you touch it.

(Written on the occasion of wandering into the Code4Lib wiki, making a dozen changes, and stepping back.)

Borrowed nostalgia for the unremembered 90s at Slate

Slate's Farhad Manjoo wonders what the fuss was all about of the Internet of 1996:

I started thinking about the Web of yesteryear after I got an e-mail from an idly curious Slatecolleague: What did people do online back whenSlate launched, he wondered? After plunging into the Internet Archive and talking to several people who were watching the Web closely back then, I've got an answer: not very much.

Once you get past the "oh ha ha dancing cats" attitude of the article, you discover that the preserved online record of the World Wide Web circa 1996 is incomplete (at least as seen via the Internet Archive) and there's really not that much to explain why people were interested in it.  That's true especially if you constrain your view only to things that you can find now that were web based:

Some of Yahoo's 1996-era front pages have been saved in the Internet Archive. What's interesting about them is what they lack. First, no e-mail: The first webmail site, Hotmail,launched in July of 1996. There was no instant-messaging software; the first big IM client, ICQ, hit the Web early in 1997.

This would be sad if it were true, but it's not true.

Email has been around for a long time (since 1975 or so), and in 1996 companies like Netcom were selling dial-up Internet connections just as fast as they could install modems to serve the demand.  Netcom's NetCruiser software was state of the art, and it had wide support for the popular stuff of the day:

NetCruiser for the Mac will include support for GIF and JPEG image formats and drag-and-drop support. NetCruiser also includes access to the Web, email, Usenet newsgroups, Internet Relay Chat (IRC), Gopher, ftp, and Telnet.

If Slate can't remember what the net was like back when it was doubling every six months, that's a reflection of the poor physical record that the virtual world has left behind.

More discussion at Techmeme

Trenary Outhouse Races 2009 - "Haul Your Arse to the Sweetest 16th Ever" - Feb 28, 2009

In Trenary, MI each year, when it's cold and snowy, they hold a race to see who can push an outhouse on skis (suitably decorated) the fastest down Main Street.

The official web site has many more details.

Past year's coverage:

2008: Trenary Outhouse Classic - Races show Yooper spirit from the Mining Journal's Winter Visitor's Guide has an interview with athlete Tom Felty

Tom said he built the outhouse in his yard back home and described it as a “conversation piece.”
Some of the outhouses built by competitors over the years were simple, stripped-down units built solely for the race. Others were quite elaborate and could be used for their original purpose.
“There ain’t nothing like this in Indy,” Tom Feltny said. “It’s something different. The first year I was here for it, I didn’t race. The next year I decided I better do it.”

Northern Michigan University's Military Science Department ROTC unit participated:

NMU ROTC participated in the 15th Annual Trenary Outhouse Race this Sat. 24 Feb. Members from the Wildcat Battalion had to created a wooden outhouse to compete in this local event. Their creation won first place in the design category. However, like the Jamacian Bobsled Team in the 1988 Winter Olympics, they sadly did not place as well as they planned or hoped in the actual race.

You do have to see it for yourself.

How to get invited to conferences

Brian Marick on How to get invited to conferences:

Someone asked me for advice on how to get invited to speak at conferences.  Key advice:

• become a valuable participant in a niche field,
• grow along with the niche,
• become a good speaker.

  There's more, of course; there's always more; it's a good discussion of how he found a growing niche and helped popularize it in part by viewing things from a novel perspective.

That discussion pointed to this essay by Phil Agre on How To Be A Leader In Your Field (2005), aimed at grad students.  He has a six part recipe (more pointless numbered lists; but you understand that as a grounding principle) on how to do it.  Starkly abstracted:

1. Pick an issue. You need an issue that the profession as a whole is not really thinking about, but which is going to be the center of attention in five years. 

2. Having chosen your issue, start a project to study it. 
3. Find relevant people and talk to them. First do your library work so you know any conventional wisdom that's out there. Then talk to some working professionals who are facing the issue, especially if they have spoken publicly about an aspect of it. 
4. Pull together what you've heard. Nobody is expecting you to solve the problems. 
5. Circulate the result. Send copies to the people who helped you. Call it a draft or interim report if you want. Give credit to the people whose ideas you've written down. Then follow up. 
6. Build on your work. Get invited to speak at meetings. Correspond with people who have contacted you after reading your work.   

Note that Agre's list has five steps before the "speak at meetings", and that Marick's list only has  two; you can chalk up most of that to the need in grad school to engage in more ritual behavior vs. the practitioner who is just looking to talk about something novel to the people who want to hear something novel.

--

I write about this stuff all the time, because in part I think about the tag line this blog has now ("I am a leader by default, because nature abhors a vacuum") and how it reflects on the leadership role of seeing something not yet there and working at it until it is real enough for others to take as their own.  Hard to say that this is an easy path to do - if you are too scatterbrained you have dozens of projects that are not quite full formed and you never get to the expert level - but there is something to be said for the notion that this is something which you or anyone can do by simple application of diligence and leverage in some small gap in the structure of how the world currently is.

Vacuum cleaners in the 1920s

why writing this?  a query came in for it, in part because my "Detroit in the 1920s" post is popular, and of course the title of the blog.  So here you go.

The Vintage Vacuum Cleaner Museum has a story with pictures of the Sears Energex, sold by Sears in the 1920s.  

The New Cleanness is a portion of Simply Switch On!  A virtual museum of small early electrical appliances , which has good coverage of appliances from the UK.  Its account of the 1920 "Air-Way Sanitary System" says 

40-page instruction manual illustrates a huge range of domestic cleaning tasks tackled including the removal of cockroaches and drying children's hair."

Some books on the topic to round out the search for now:

Chasing Dirt: The American Pursuit of Cleanliness

By Suellen Hoy

Edition: reprint, illustrated

Published by Oxford University Press US, 1996

ISBN 0195111281

In Chasing Dirt, Suellen Hoy provides a colorful history of this remarkable transformation from "dreadfully dirty" to "cleaner than clean," ranging from the pre-Civil War era to the 1950s, when American's obsession with cleanliness reached its peak.

Never Done: A History of American Housework

By Susan Strasser

Edition: illustrated

Published by Macmillan, 2000

ISBN 0805067744,

Finally back in print, with a new Preface by the author, this lively, authoritative, and pathbreaking study considers the history of material advances and domestic service, the "women's separate sphere," and the respective influences of advertising, home economics, and women's entry into the workforce. Never Done begins by describing the household chores of nineteenth-century America: cooking at fireplaces and on cast-iron stoves, laundry done with boilers and flatirons, endless water-hauling and fire-tending, and so on. Strasser goes on to explain and explore how industrialization transformed the nature of women's work. Easing some tasks and eliminating others, new commercial processes inexorably altered women's daily lives and relationships—with each other and with those they served.

There's some additional line of thinking I'd like to go down - a patent search - here's some from the 1920s that show up:

1699598 - Objects of the invention are:—to equalize the pressure of air passing through all points of the air inlet orifice of a vacuum cleaner; 10 to enable a surface to be efficiently cleaned in corners with a minimum of change of position of the machine; to prevent the curling upward of the margins of rugs, loosely laid carpets,"or other fabrics by the suction 15 when a vacuum cleaner is moved across their edges; to simplify the assembling and disassembling of the operating mechanism and otherwise to increase the efficiency, improve the operation and simplify the arrangement 20 of parts of traveling vacuum cleaners.

aha, much much more from the Vacuum Dealers Trade Association - A Vacuum and Sewing Hall of Fame!.. stopping now.

2,000+ posts and ten years later, what have I learned from blogging?

Typepad says that I have 2,000 posts in this weblog.  Some of them, to be sure, are drafts that were abandoned, and some are scheduled in the future, but that's as good a round number as I need to reflect on blogging (the practice and the habit).

I've been blogging since April 1999; here's the first month or so of posts.    They were edited oh so patiently with vi, and at the time the genre was new enough that I could say with confidence "Huh, this daily weblog stuff is too hard. " and have that be a novel observation.    Jorn Barger at the time noted in his Robot Wisdom blog

        "New weblog from Ed Vielmetti, who's been on the Net forever."  

Ten years later I'm still at it.  I suppose that I might have just enough experience in blogging that even I would think of myself as an expert at it, for what that's worth.  (As Forbes once said about my bright ideas, "Too bad everyone else had the same idea.") And like any good blogger, I have a nice pat seven item list to share with you - because people who read blogs love pointless numbered lists.

1.  Repetition is the soul of the net.  If I've told you this once, I've said it a thousand times.  Every year, regular as clockwork, there is an audience for people reading about where to pick blueberries, how to get election returns, who makes the best paczki, what to do on your birthday or your kids birthday.  You get a free pass to repeat your good content over and over again annually, do it.

2.  Don't think too much.  Think more, perhaps, than you would in writing a 140 character twitter message; think twice before saying something that you'd regret before a judge or in print.  But don't think more than twice, and be prepared to post something before it's absolutely polished and done.  Sometimes the people who read what you are writing will fill in the blanks for you.

3.  Omit needless words.  Your screen is small, and the day is short.  Don't ramble on and on - fill the text input box until it's full, and hit post.

4.  Leave comments.  Comments on other people's blogs will remind them of stuff that you have collected that's relevant; comments on your own blog will let you update things that are changing without going back and re-editing.  Every once in a while, you'll unearth someone who is a bona fide expert who reads the news clippings you pasted together and gets an insight.

5.  Don't feed the trolls.  (And there are so many of them.)  Squash idiot comments, commercial spam comments, irritating comments, stray trackbacks, and any other crud out of your blog as soon as you see it.  

6.  Did I mention pointless numbered lists?  Use them.  Everyone's attention span is short, and lists let people know just how far along they are before they can be done.  They let you jump from idea to idea without pesky transitions, and make it easy for someone else to cut and paste that one good idea hiding in your pile of dreck and take it for their own.

7.  Quote liberally (and credit liberally).  Half the battle in blogging is figuring out who knows more than you do about something; the other half the battle is figuring out who is uselessly ignorant but still writes a lot.  If you are good and thorough and relentless in clipping out good paragraphs from other people's work, and you are good and thorough and relentless in doing the searches that end up in results that aren't on everyone else's front page of Google, you can learn a lot and share a lot.

I've been averaging a post a day for a long time.  Some days more, some days less, and some blogs are mercifully left for dead from ideas that didn't pan out.  I wish I knew what it all meant.  In some ways it reminds me a lot of my short lived but illustrious high school forensics career of impromptu speaking - all it really shows is that you can be coherent on short notice about pretty much anything, and that you have enough diligence to keep being coherent for a long time.

If someone can tell me what career that qualifies me for, please let me know - I'm looking.

Optical Character Recognition with parallel computing?

Some references to collect here on the use of general-purpose graphic processors in OCR.  Nominally, you run some kind of MapReduce based algorithm with a GPGPU over some image, and out spits text much faster than any ordinary processor of any speed; of course, there are lots of details to attend to.

From Berkeley EECS: Algorithms and Frameworks for OCR and Content-based Image Retrieval

Jike Chong, Bryan Christopher Catanzaro, Narayanan Sundaram, Fares Hedayati and Kurt Keutzer

A new breed of general-purpose manycore computing platform is emerging. Exemplary examples include the Niagara from Sun Microsystems, the G80 from Nvidia, the Cell from IBM, and the up-coming Larrabee from Intel. These manycore processors each pack 8-32 relatively simple cores on a chip, capable of supporting up to 100s of threads, and boast tremendous potential peak single-chip performances up to the range of Tera-FLOPS. However, traditional algorithms and applications in many domains cannot take advantage of much of the parallelism provided by these platforms.

The ParLab at Berkeley was recently founded in part to help meet this acute need for novel algorithmic approaches to unleash the performance potentials of emerging manycore platforms for a wide range of application domains. It proposes to concentrate on analyzing the communication and computation patterns (or Dwarfs) of important classes of algorithms underlying modern application domains, and develop techniques to efficiently parallelize them for the general purpose manycore platforms.

We concentrate on the domain of image recognition and retrieval, leveraging the Intel PIRO content-based image retrieval framework as a motivating application. Specifically, we study the parallelization of classification algorithms for machine learning, and develop parallelization techniques to improve the performance of these algorithms and applications on the emerging manycore platforms.

Here's a discussion of an OCR system that uses the CUDA libraries called cocr

The cuda implementation allows for exceptionally fast image manipulation, cleaning and segmentation before being presented to the template based ocr system.  Traditional cpu based ocr systems are very slow, especially in the image rotation, cleaning etc departments and although cocr is by no means a complete ocr package it is orders of magnitude faster than the various cpu based ones I’ve tried. With a bit more work and an additional of a neural network ocr it could easily become a system able to do greater than realtime ocr’ing.  Neural networks are extremely well suited to GPU implementations due to their inherit parallelism. I will post more snippets and modules of the system here over the coming weeks.

I wasn't able to find any evidence that the ocropus OCR system that Google has used and released is using any parallel computing mechanisms, so this still looks like "gee, it should work for someone, if you can throw a PhD student or two at it."