Monthly Archives: February 2009

Chisimba and the Unviersity of the Western Cape

I spent this week working on connecting the Chisimba framework to IMS SimpleLTI as the first real producer of IMS LTI. With help from Paul Scott and Charl van Niekerk – I made good progress. I have working prototype code showing Chismiba tools working in both Sakai and Moodle. Screen shots when my connectivity is better.
I will continue to evolve the code in the Chisimba SVN over the next few months to make it production ready – that will probably wait for Summer break – as teaching once again takes over my spare time.
Tomorrow I visit Stephen Marquard, David Horowitz, Caly Fenlason, and Anthony with at the University of Capetown to talk about Cloud Social and Sakai 3. Afterwards there will be pubbing at which I will be dropped off at the airport for the ultimate red-eye.
I leave Captetwn at 12:30 AM arriving a Amsterdam at about 9AM and then leaving for the US after a long layover and getting home about 9PM and 7 timezones later.

SI301 – Chapters 9 – 12 Suroweicki Notes

Chapter 9 (173)
Columbia – January 21, 2003
Challenger memos – one person wrote an ignored memo. A minority voice was drowned out.
“… not much we can do about it”. (174)
small group judgments are more volatile and extreme (176)
confirmation bias – we seek information that confirms that which we already believe (178)
small groups emphasize consensus over dissent
absence of debate and minority opinions
NASA is meritocratic – but also hierarchical (182)
In small group, leaders must take active role in insuring everyone speaks
Lack of *cognitive* diversity at NASA (183)
A minority viewpoint improves group deliberation even if it is wrong
Group polarization (189)
Social comparison – if the group moves right you do too to maintain your place in the group
Talkativeness makes one influential – not necessarily well liked – but listened to
The one who talks the most is viewed as right, although that isn’t always true
Chapter 10 – Companies (192)
Zara – Agile Clothing manufacturer form Spain – 10-15 days elapse from design to shelf
Less wasted markdowns
Links from stores to designers – quick market feedback – adjust manufacturing
Zara factories for materials and cutting
Small craft shops for the actual sewing – Galicia and Northern Portugal (194)
Coordination happens without control because of the market.
Why do corporations exist? (195)
Reduce transaction costs – bid out, track and coordinate, evaluate, pay
Also some things ate the “secret sauce” – Zara manufacturing plants were central to agility – it is not always about cost
Independent films – gather – work – scatter
Amazon.com – used books
Gangster examples (198)
Classic corporations (200)
Many layers power from the top down
Came from factory pattern where line workers were really dumb
Lip service for decentralized decision making (202)
Often – lots of talk and meetings were seen as evidence of distributed decision making
Structure stops disagreement – someone might be a future boss or enemy – play the game (205)
Chances of promotion improved if you did not tell the boss bad news – stay in good graces (205)
Lack of cognitive diversity of top managers (206)
The illusion of perfectibility (208)
Targets and bonuses (209) Pay people to lie!
Does the Deli owner get more money when they “exceed expectations”?
In top-down corporation – there is no incentive to reveal private information. Markets cause information to be revealed – when used in particular. WHich increases knowledge (209)
People with local knowledge are often best suited to making decisions (212)
Linking authority and responsibility – Chuck
Experiment – Problem solving with noise and a button to stop the noise.
CEO’s as superheros / saviors – common workers look to upper management as being the final word (217)
Gary Wendt – GE Capital – Conseco (218)
Is it luck? Travelers leave Chicago with only one road with a gas station. Was the person who picked that road smarter?
Best CEOs – recognize their own fallibility. Alfred Sloan GM – Jack Welch GE
They built an environment / ecosystem – and set values so Wisdom of Crowd could work (222)
Chapter 11: Markets (224)
Short Sellers – Hated!
They gain when everyone else is unhappy – so folks blame them
Blamed for the crash of 1929 – but it turned out to be the long buyers that whipped the prices up to the point where they were wrong and then it all popped
Only 2% of the trades are short (227)
Massive PR all aimed at the market going up up up (227)
JS thinks there needs to be more, not less, short selling to reveal information
The pain of loss is much greater than the joy of gain – so folks hold stocks – because if it goes back up – they avoid the pain of loss
Stock prices jump around far more than IEM or NFL betting – and the crowd rarely flips (236)
Chuck: Cascade effect of rare flips??
Stock market just keeps going – we never know if we “win” or “lose”
Long Term Capital Management (LTCM) (237)
Pick pairs of slow/stable investments that track each other
When one is slightly off – buy it and sell when things rejoin
Buy on margin to magnify gains
Assumption: When prices deviate – they come back “quickly”
Problem: If prices don’t come back – because of the margin time runs out qiuckly
Ultimately: Each market is too small – when LTCM wanted to sell things freaked out
Chuck: Sometimes you need to just switch strategy – “cascade”
Bowling (241)
Things went well – cascade growth – then a crash
Small localized bubbles happen all the time – folks fancy something
Bubbles don’t happen in a real economy (245)
When prices go up – usually demand does not go up
Coffee “shortage”
Australian hurricane wipes out banana trees
Stocks are expected to grow
People tend to pick the stocks that will grow more – not so much about intrinsic value – which feels best – like a horse race
Bubble of the 1990’s CNBC (252)
A stock price would be affected within 15 seconds of being mentioned on CNBC
Bubble creation is like riot creation (257) Needs a spark and then a bit of influential folks nearby – to make it seem like a big trend
Chapter 12: Democracy (259)
Voting affirms partisan preference rather than an attempt to affect the outcome of the election (284)
Conservatives without health insurance opposed national health care – liberals with health insurance supported health care (265)
Government is like a big cooperation problem
Democracy is often not getting what you want but trusting that things will be generally OK

Upgrading my 1and1 Hosting Package for www.dr-chuck.com

I have had free hosting from 1and1.com since 2003 based on a special deal they sent out through IEEE Computer Magazine called the “1&1 Professional Linux” offer. It was cool – 100% free, shell access, – all in all a lot of fun – it has allowed me to do lots of cool things (including this very blog) for the past six years.
Recently I noticed file uploads were failing. So I dug into it a bit and realized I had hit my disk Quota! My free quota was 500MB. So I guess it was time to get off the free thing and pay for my hosting (I already pay over $1200 per year to 1and1 for domain names – another $60.00 for hosting is barely noticeable).
So I was looking at the limits of the $5.00 per month account – the storage is 120GB and the transfer was 1200GB. If I paid $10.00 per month the storage went up to 250GB and the transfer went up to 2500GB per month.
I started looking at my monthly transfer for my current dr-chuck.com – my monthly transfer was 2700MB out of 5000MB. This kind of concerned me because my new account only had 2500GB per month and I was *already* at 2700MB per month! I figured that while I could easily afford the $10.00 for hosting, I really was uncomfortable paying for bandwidth. I get hit a lot by search engines – and I just did not want to be nailed with a $100.00 bill when my bandwidth limit was exceeded – and it was not even my fault!
I went back and forth – how could the free account from 2003 give me 5000MB transfer per month and the pay account from 2009 reduce that to 1200GB transfer per month! It seemed like my free account was just too good to be true with its awesome 5000MB per month transfer – perhaps this was a trick to get us off those free accounts and stick us with bandwidth bills! What is up with that! Good thing I caught these folks before I mistakenly upgraded!
It took about 20 minutes of going between screens at my ISP before I realized my mistaken logic – and went ahead and ordered the upgrade. Perhaps you can find the flaw is my thought process in less than 20 minutes :).

Ironic Call For Papers: International Symposium on Peer Reviewing: ISPR (http://www.ICTconfer.org/ispr)

I got this message on my iPhone so I read it sequentially I could tell right away that it was a call-for-papers spam message but the first paragraph grabbed my attention.

Subject: Invitation to a Symposium in Peer Reviewing
Only 8% members of the Scientific Research Society agreed that “peer review works well as it is.” (Chubin and Hackett, 1990; p.192). Horrobin concludes that peer review “is a non-validated charade whose processes generate results little better than does chance.” (Horrobin, 2001). This has been statistically proven and reported by an increasing number of journal editors. Since a growing number of studies conclude that peer review is flawed and ineffective as it is being implemented, why not apply scientific and engineering research and methods to the peer review process?

This sounds sweet – I am pretty critical of the notion of “peer-review” – I will blather on about the subject for a long time – particularly after a long day and a few beers at the pub. This was the perfect tease for me – I was hooked – I kept reading while pumping my gas as the snow was falling around me. I also was thinking – this is Orlando Florida – could be a family trip combined with business. Hmmm. Lets see how this works…

So I looked further and kept reading as the gas tank filled. But I stopped when I got to this bit:

All Submitted papers will be reviewed using a double-blind (at least three reviewers), non-blind, and participative peer review.

Hmmm – somehow this does not make sense – why not choose papers randomly? Since random is better than peer review according to the above references.

This reminds me of the feeling over the years watching endless bad PowerPoint presentations describing how PowerPoint is the *worst possible* approach to teaching and learning. Somehow irony hits me differently than others I guess. Ah well – I thought the call for papers was uproariously funny while pumping gas in the snow.

Fixing Bad Checksum in SVN Commit

I started getting this messaage

Sending        wiscrowd/mod/freerider/templates/index.htm
Transmitting file data .svn: Commit failed (details follow):
svn: Base checksum mismatch on '/trunk/contrib/drchuck/wiscrowd/mod/freerider/templates/index.htm':
expected:  1c2fe4a8f013f8af2729280d89567095
actual:  2ea55328bc430c0f6e738303e4393f3b
svn: Your commit message was left in a temporary file:
svn:    '/Users/csev/dev/cloudcollab/contrib/drchuck/svn-commit.4.tmp'

The solution was to make a copy of the file, delete the file from svn and then re-add it.

This loses history – but is quick and dirty when you are in a hurry.