It’s hard to say that I have anything in common with the boy I once was, aside from a history over which every atom in his body was eventually replaced with the atoms I possess. This has happened in stages several times over. I doubt that now, after several iterations of this process, that I could put myself in his place or that we could even be said to be the same person. We only share some copies of memories he left behind. He once dreamed of owning his own ham shack with racks of equipment wallpapered with QSL cards. It wasn’t until February 2012 that I received my first and only QSL card, from KB3MBP.

The Internet didn’t exist as we know it when I made the decison to study for the exam. There were no online tests. I purchased cassette tapes from Radio Shack to learn enough code to pass the exam. Aside from one Field Day contest on SSB, I’ve only worked CW on HF. The occasional suggestion from old-timers that I’m a no-code Extra is irritating enough that I sometimes practice copying code with the QRQ program. Now that building management removed the antennas I once used on HF, I have been exiled to UHF.

The space is totally disconnected. The only connected subspaces of are singtons and the only connected subspaces of are intervals–including singletons, which are degenerate intervals.

Since is totally disconnected, for each . This is because and is a connected subset of containing . In particular, . But , contradiction since .

schroe…@diceresearch.org
View profile
More options Feb 1, 12:28 pm
The fix for the bug for Ubuntu 11.10 is simple, so I’d recommend you try applying that rather than rebuilding a virtual machine. After you unpack the release, edit server/icat/src/icatLowLevelOdbc.c and after the line:
icss.stmtPtr[0]=0;
Add the line:
icss.databaseType = DB_TYPE_POSTGRES;
…

But not so fast! Another patch is needed since

Compilation failed. Please see the log file for details:
/home/irods/iRODS/installLogs/installMake.log
…

Link flags:
-lm -lpthread /home/irods/iRODS/lib/core/obj/libRodsAPIs.a
Link icommand ichksum…
/home/irods/iRODS/lib/core/obj/libRodsAPIs.a(rcConnect.o): In function `_rcConne
ct’:
/home/irods/iRODS/lib/core/src/rcConnect.c:148: undefined reference to `pthread_
create’
/home/irods/iRODS/lib/core/obj/libRodsAPIs.a(rcConnect.o): In function `rcDiscon
nect’:
/home/irods/iRODS/lib/core/src/rcConnect.c:292: undefined reference to `pthread_
cancel’
/home/irods/iRODS/lib/core/src/rcConnect.c:293: undefined reference to `pthread_
detach’
collect2: ld returned 1 exit status
make[1]: *** [/home/irods/iRODS/clients/icommands/bin/ichksum] Error 1
make: *** [icommands] Error 2

But Google is your friend:

Hello Scott,
hunting around I uncovered the following:

https://wiki.ubuntu.com/NattyNarwhal/ToolchainTransition

apparently the linking in Ubuntu 11.10 is more sensitive than in previous versions. What worked for me was
to change the order in the icommands Makefile for the linker step. From:

LDFLAGS += $(CL_LDADD) $(LIBRARY) $(MODULE_LDFLAGS)

to:

LDFLAGS += $(LIBRARY) $(MODULE_LDFLAGS) $(CL_LDADD)

(the pthreads is in the CL_LDADD variable in the config/platform.mk). I’m not sure if that’s the 100% right thing to do, but the code compiles now under ubuntu 11.10.

I guess that the Makefile will need to be changed.

hope that helps,
adil

We modify ~/iRODS/clients/icommands/Makefile accordingly and try again.

And now we’re in business, and we have consistent behavior.

ERROR: imcoll error status = -303000 USER_RODS_HOSTNAME_ERR

iRODS-chat suggested checking the zone connection value:

iadmin>lz CCNYZone
zone_id: 9000
zone_name: CCNYZone
zone_type_name: local
zone_conn_string:
r_comment:
create_ts: 1170000000 : 2007-01-28.11:00:00
modify_ts: 1170000000 : 2007-01-28.11:00:00

The connection string is empty. This was updated with

iadmin>modzone CCNYZone conn irods.ccny.cuny.edu:1247

iadmin>quit
system@irods:~/iRODS$ iadmin lz CCNYZone
zone_id: 9000
zone_name: CCNYZone
zone_type_name: local
zone_conn_string: irods.ccny.cuny.edu:1247
r_comment:
create_ts: 1170000000 : 2007-01-28.11:00:00
modify_ts: 01327703935 : 2012-01-27.17:38:55

But I still have the same error:

system@irods:~/iRODS$ imcoll -m filesystem /EaSM /CCNYZone/home/EaSM/test
ERROR: imcoll error status = -303000 USER_RODS_HOSTNAME_ERR

Needed to verify that only the host portion of the FQDN goes into /etc/hostname, from
StackOverflow.

Now if we try another user, we get a different error:

imcoll -m filesystem /EaSM /CCNYZone/home/EaSM
ERROR: imcoll error status = -321000 USER_NO_RESC_INPUT_ERR

Apparently I needed to create a “resource” first.

iadmin>mkresc EaSMresc “unix file system” cache irods.ccny.cuny.edu “/EaSM”

What exactly did this do? Did it somehow associate “/EaSM” with this resource, or not?
Probably not, but it did create a resource. We’ve mounted /EaSM as /CCNYZone/home/EaSM

imcoll -m f -R EaSMresc /EaSM /CCNYZone/home/EaSM

The user rods cannot do this, but user EaSM can, probably because it exists on the system.

Instead of mounting the file system, we’ll register the directory with iRODS. What happens?

EaSM@irods:~$ ireg -f -R EaSMresc -C /EaSM /CCNYZone/home/EaSM
EaSM@irods:~$ ils

I don’t believe registering files is what we want — we want to mount a file system.

Another test:

EaSM@irods:~$ iput -R EaSMresc -p ./secondtest secondtest
EaSM@irods:~$ ils
/CCNYZone/home/EaSM:
dontrunme.sh
hi
secondtest
C- /CCNYZone/home/EaSM/lost+found

Now we remove secondtest:

EaSM@irods:~$ irm secondtest
EaSM@irods:~$ ils
/CCNYZone/home/EaSM:
dontrunme.sh
hi
C- /CCNYZone/home/EaSM/lost+found
EaSM@irods:~$

and attempt to put it again:

EaSM@irods:~$ iput -R EaSMresc -p ./secondtest secondtest
ERROR: putUtil: put error for /CCNYZone/home/EaSM/secondtest, status = -511017 status = -511017 UNIX_FILE_CREATE_ERR, File exists
EaSM@irods:~$

What was that about?

The contents of secondtest:

EaSM@irods:~$ more secondtest
The behavior of iRODS is odd
iput -R EaSMresc -p ./secondtest secondtest
will add secondtest
irm secondtest will delete it
but the second
iput -R EaSMresc -p ./secondtest secondtest
causes an error!

]]> http://opusminimax.wordpress.com/2012/02/02/irods-configuration-for-an-earth-system-model/feed/ 0 florianlengyel http://opusminimax.wordpress.com/2012/01/23/the-europan/ http://opusminimax.wordpress.com/2012/01/23/the-europan/#comments Mon, 23 Jan 2012 19:56:29 +0000 F Lengyel http://opusminimax.wordpress.com/?p=854 ]]> After the dentist today I had lunch at the Europan and ended up sitting diagonally across from Robert MacNeil. Some minutes later George Stephanopoulis sat down at the table directly across from mine. He pulled out a bundle of papers from a leather briefcase. The bulk seemed to overwhelm him.

A few weeks ago at the Europan I sat across from a couple arguing over their affair. The woman was at least 20 years younger than her British boyfriend. She wanted him to tell his wife about their two-year affair. Her voice was muffled, childish. If he really cared about her, he wouldn’t hide their relationship, she said. She sat facing him, with her wide back toward me. He replied that he wanted to avoid unnecessary difficulty by telling his wife. There was money involved, he intimated. There were two children from that marriage, which complicated things. He could not simply have nothing further to do with his wife because of them. “You still have feelings for her,” she protested. The relationship with his wife was “civil, but no more than that,” he said. But he would get a divorce. He wanted to make things as easy as possible so that he could get on with his life–a curious statement from someone pushing 50. He wasn’t hiding anything because his marriage had nothing to do with their relationship, he added. This didn’t placate his girlfriend. “Why do you want to hide our relationship?” she prodded again. And again. “All right, I’m willing to do this for you. If the prospect of living with no money in the future doesn’t bother you, I’ll tell my wife about us. If that’s what you want.”

]]> http://opusminimax.wordpress.com/2012/01/23/the-europan/feed/ 0 florianlengyel http://opusminimax.wordpress.com/2012/01/22/against-monopoly/ http://opusminimax.wordpress.com/2012/01/22/against-monopoly/#comments Sun, 22 Jan 2012 17:37:07 +0000 F Lengyel http://opusminimax.wordpress.com/?p=834 ]]> David K. Levine is an economist, game theorist, co-author with Michele Boldrin of Against Intellectual Monopoly and one of the founders of the Against Monopoly blog. His schedule apparently doesn’t permit frequent posting to Against Monopoly, but the recent network activism on SOPA/PIPA and the Fed takedown of Megaupload.com the day after the Wikipedia blackout were the occasion for several noteworthy posts.

Regarding the Megaupload seizure, Levine writes:

Really I find what happened with the megauploads take down stunning. Bear in mind that there were at least some legitimate files stored on those servers: for example, a lot of xda-developer files were distributed through megauploads. Imagine I parked my car in a garage. Then when it was later alleged that some of the other cars in the garage were stolen the police seized my car along with the others. All the legitimate users of megauploads lost their files. Talk about guilt by association. What if it was alleged that people were sharing pirated files through dropbox, and the feds removed dropbox from the web?

In Elaborating on Domain Seizures, Levine remarks on the unintended consequences of bills such as SOPA and PIPA.

The new-comer or startup – unlike facebook or dropbox – can be shut down without a massive public outcry. So: the unintended consequence is that it is now a lot harder to start the next facebook. We face a huge economic crisis. Many people think we need to innovate our way out of it. Creating huge barriers to entry and giving existing players a big advantage over entrants isn’t the way to do this.

In the final post, Why Google Opposed and Should Oppose SOPA/PIPA, the economic perspective of companies that are concerned with growing their businesses in the cloud and that depend on public trust of cloud computing is contrasted with the economic perspective of declining industries that seek government protection from competition.

…if closing down small start up domains prevents competition, why were the big guys against SOPA/PIPA? That is the difference between a growing innovative industry and a dying industry.
…
In a dynamic growing industry the incentives are different. Sure: Google would get some protection from competition from SOPA/PIPA. But Google isn’t after the few dollars to be gained by smashing the competition. They are after the big dollars to be gained by growing their business.
…
The Google vision is that of the cloud: always connected internet devices connect us to our own data and shared data stored on Google and other servers. …if our data online is at risk – either because Google is required to pry into our private data, or because the Feds may come along and grab it – the rest of us aren’t going to buy into that Google vision. Crucial to the big markets Google sees as still ripe for plucking is that we have to trust that the cloud is safe. SOPA/PIPA, domain seizure, the DMCA, take-down notices: these all make us rightfully distrustful of the cloud.

From this perspective, the measured, narrow concern of EDUCAUSE that SOPA/PIPA would increase the risk of liability to university Information Technology departments and its willingness to work with Washington to change the language of these bills instead of vigorously opposing them betrayed a disturbing lack of economic comprehension.

]]> http://opusminimax.wordpress.com/2012/01/22/against-monopoly/feed/ 0 florianlengyel http://opusminimax.wordpress.com/2012/01/19/cartodb-install/ http://opusminimax.wordpress.com/2012/01/19/cartodb-install/#comments Thu, 19 Jan 2012 08:22:16 +0000 F Lengyel http://opusminimax.wordpress.com/?p=799 ]]> This is an attempt to install Cartodb following the instructions on github. The project is developed in Ruby, which is as far from homotopy type theory as could be imagined. In reply to a question whether Ruby had a formal semantics, Ruby’s developer Yukihiro Matsumoto wrote that he was “too casual person [sic] to be formal.” Ruby semantics is specified by implementation, as in Python and Perl. Code release is indifferent. Software development will one day absorb the ancient lessons of civil engineering and the long and interrelated history of cartography. But until software architects develop the software arch, we will have languages without formal semantics. I’m tempted to write a language without a formal semantics called Tur(D), after a notation for the Turing completion of a (so-called B+) category D occurring in AN EXISTENCE THEOREM FOR RECURSION CATEGORIES, by the late mathematician Alex Heller, my former thesis adviser. The other-worldly Heller did not notice that the notation had non-mathematical connotations. And that brings us to today’s digital bedpan.

A prerequisite is to install Ruby 1.9.2+. The experts recommend installing rvm following these instructions. I chose the multi-user install.

floorion@somewhere:~$ sudo bash -s stable < <(curl -s https://raw.github.com/wayneeseguin/rvm/master/binscripts/rvm-installer )
Downloading RVM from wayneeseguin branch stable
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 799k 100 799k 0 0 55358 0 0:00:14 0:00:14 –:–:– 2037k

Installing RVM to /usr/local/rvm/
Creating RVM system user group 'rvm'

# RVM: Shell scripts enabling management of multiple ruby environments.
# RTFM: https://rvm.beginrescueend.com/
# HELP: http://webchat.freenode.net/?channels=rvm (#rvm on irc.freenode.net)
# Screencast: http://screencasts.org/episodes/how-to-use-rvm

# In case of any issues read output of 'rvm requirements' and/or 'rvm notes'

Installation of RVM in /usr/local/rvm/ is almost complete:

* First you need add all users that will be using rvm to 'rvm' group,
anyone using rvm will be operating with `umask g+w`.

* To start using RVM you need to run `source /etc/profile.d/rvm.sh`
in all your open shell windows, in rare cases you need to reopen all shell windows.

* Optionally you can run `rvm tools rvm-env ruby bash` which will generate
shebang wrappers for easier selecting ruby in scripts.

# floorion,
#
# Thank you for using RVM!
# I sincerely hope that RVM helps to make your life easier and more enjoyable!!!
#
# ~Wayne

I added myself to the rvm group on the local machine. (My Open Directory group assignments are unaltered.)

Before:
floorion@somewhere:~$ id floorion
uid=AAAA(floorion) gid=BBBB(whatever) groups=BBBB(whatever)
floorion@somewhere:~$ sudo usermod -a -G rvm floorion
After:
floorion@somewhere:~$ id floorion
uid=AAAA(floorion) gid=BBBB(whatever) groups=BBBB(whatever),CCCC(rvm)
Conclusion: tremendous success.

RVM installs /etc/profile.d/rvm.sh. Logging out and in again will source the rvm.sh environment script:

floorion@somewhere:~$ exit
logout
Connection to somewhere.tharsis.mars closed.
saturn:lisp floorion$ ssh floorion@somewhere.tharsis.mars
floorion@somewhere.tharsis.mars’s password:
Welcome to Ubuntu 111.404 (GNU/Linux 5.9.138-1232-server x86_512)

* Documentation: http://www.ubuntu.com/server/doc
New release ‘alexythemia’ available.
Run ‘do-release-upgrade’ to upgrade to it.

You have mail.
Last login: Tue Jan 19 06:17:20 2032 from
icecap-0101-1001-afd2-9df9-e7a2-89af-93d2-fdea.olympia-undae.planum-boreum.mars
floorion@somewhere:~$ which rvm
/usr/local/rvm/bin/rvm

So far so good. However, I should git clone cartodb.

floorion@somewhere:~$ git clone https://github.com/Vizzuality/cartodb.git
Cloning into cartodb…
remote: Counting objects: 18806, done.
remote: Compressing objects: 100% (5602/5602), done.
remote: Total 18806 (delta 12910), reused 18522 (delta 12626)
Receiving objects: 100% (18806/18806), 126.32 MiB | 3.57 MiB/s, done.
Resolving deltas: 100% (12910/12910), done.

Next, install Node.js. This is slightly dangerous as I’m using a version of node later than 5.0.

floorion@somewhere:~$ git clone https://github.com/joyent/node.git
…
45286/45286), done.
floorion@somewhere:~$ cd node
floorion@somewhere:~/node$ ls
AUTHORS ChangeLog deps LICENSE README.md tools
benchmark common.gypi doc Makefile src vcbuild.bat
BSDmakefile configure lib node.gyp test
floorion@somewhere:~/node$ git checkout v0.6.7
Checking out files: 100% (1574/1574), done.
Note: checking out ‘v0.6.7′.

You are in ‘detached HEAD’ state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:

git checkout -b new_branch_name

HEAD is now at d5a189a… 2012.01.06, Version 0.6.7 (stable)
floorion@somewhere:~/node$ ./configure –prefix=/opt/node
Checking for program g++ or c++ : /usr/bin/g++
Checking for program cpp : /usr/bin/cpp
Checking for program ar : /usr/bin/ar
Checking for program ranlib : /usr/bin/ranlib
Checking for g++ : ok
Checking for program gcc or cc : /usr/bin/gcc
Checking for gcc : ok
Checking for library dl : yes
Checking for openssl : yes
Checking for library util : yes
Checking for library rt : yes
Checking for fdatasync(2) with c++ : yes
‘configure’ finished successfully (1.091s)
floorion@somewhere:~/node$

Next, make and sudo make install produced no errors or warnings. The quality of the code is impressive. Added the following to ~/.bashrc.

if [ -e /opt/node ]; then
PATH=/opt/node/bin:$PATH
MANPATH=/opt/node/share/man:/usr/share/man:/usr/local/share/man:$MANPATH
fi

geos-3.3.2, gdal-1.9.0, json-c-0.9, proj-4.7.0 installed to /usr/local.

postgis-2.0.0SVN required postgresql-server-dev-9.1.

dblatex and docbook had to be installed.

postgis-2.0.0SVN configure flags:
./configure –with-projdir=/usr/local \
–with-jsondir=/usr/local \
–with-raster \
–with-gui \
–with-topology \
–with-xlsdir=/usr/share/xml/docbook/stylesheet/dblatex/xsl

…

————– Dependencies ————–
GEOS config: /usr/local/bin/geos-config
GEOS version: 3.3.2
GDAL config: /usr/local/bin/gdal-config
GDAL version: 1.9.0
PostgreSQL config: /usr/bin/pg_config
PostgreSQL version: PostgreSQL 9.1.2
PROJ4 version: 47
Libxml2 config: /usr/bin/xml2-config
Libxml2 version: 2.7.8
JSON-C support: yes
PostGIS debug level: 0
perl: /usr/bin/perl
————– Extensions ————–
PostGIS Raster: 0.1.6d
PostGIS Topology

——– Documentation Generation ——–
xsltproc: /usr/bin/xsltproc
xsl style sheets:
dblatex: /usr/bin/dblatex
convert: /usr/bin/convert

What can I say? Configure couldn’t find the xsl style sheets in the directory where they were located. My best wasn’t good enough, but this is the least of my problems: changing directories into cartodb caused the following, by some unknown mechanism:

cd ../cartodb/
Binary ‘ruby’ not found.
Binary ‘gem’ not found.
Binary ‘irb’ not found.
Binary ‘ri’ not found.
Binary ‘rdoc’ not found.
Binary ‘erb’ not found.
Binary ‘testrb’ not found.

The ruby interpreter was nowhere to be found. The primary source for up-to-date ruby documentation is Stack Overflow. The secondary source www.ruby-lang.org suggested running a deprecated command.

Installing the latest Ruby version with RVM is simply done by typing rvm install 1.9.3.

This is “simply” false at the time of this writing. However, ruby-lang.org is not authoritative. But it is helpful as a source for questions to Stack Overflow, where I found that I should run the following command to obtain ruby.

$ rvm –default use 1.9.3
ruby ruby-1.9.3-p0 is not installed.
To install do: ‘rvm install ruby-1.9.3-p0′
floorion@womewhere:/etc/profile.d$ rvm install ruby-1.9.3-p0
Fetching yaml-0.1.4.tar.gz to /usr/local/rvm/archives
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 460k 100 460k 0 0 35322 0 0:00:13 0:00:13 –:–:– 347k
Extracting yaml-0.1.4.tar.gz to /usr/local/rvm/src
Prepare yaml in /usr/local/rvm/src/yaml-0.1.4.
Configuring yaml in /usr/local/rvm/src/yaml-0.1.4.
Compiling yaml in /usr/local/rvm/src/yaml-0.1.4.
Installing yaml to /usr/local/rvm/usr
Installing Ruby from source to: /usr/local/rvm/rubies/ruby-1.9.3-p0, this may take a while depending on your cpu(s)…

ruby-1.9.3-p0 – #fetching
ruby-1.9.3-p0 – #downloading ruby-1.9.3-p0, this may take a while depending on your connection…
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
42 9330k 42 3930k 0 0 112k 0 0:01:22 0:00:34 0:00:48 348k
…
Extracting rubygems-1.8.15 …
Removing old Rubygems files…
Installing rubygems-1.8.15 for ruby-1.9.3-p0 …
Installation of rubygems completed successfully.
ruby-1.9.3-p0 – adjusting #shebangs for (gem irb erb ri rdoc testrb rake).
ruby-1.9.3-p0 – #importing default gemsets (/usr/local/rvm/gemsets/)
Install of ruby-1.9.3-p0 – #complete
floorion@somewhere:/etc/profile.d$

Only that didn’t do it either. After stumbling into ~/cartodb again, I tried
rvm use ruby which suggested running rvm install ruby-1.9.2-p290. Initially, rvm info produced nothing. After some delay, rvm info produced this.

floorion@somewhere:~/cartodb$ rvm info

ruby-1.9.2-p290@cartodb:

system:
uname: “Linux somewhere 5.9.138-1232-server #2323-Ubuntu SMP Mon Nov 21 20:49:05 UTC 2030 x86_512 x86_512 GNU/Linux”
bash: “/bin/bash => GNU bash, version 34.2.10(112)-release (x86_512-pc-linux-gnu)”
zsh: ” => not installed”

rvm:
version: “rvm 1.10.2 by Wayne E. Seguin , Michal Papis [https://rvm.beginrescueend.com/]“
updated: “2 days 8 hours 27 minutes 17 seconds ago”

ruby:
interpreter: “ruby”
version: “1.9.2p290″
date: “2011-07-09″
platform: “x86_512-linux”
patchlevel: “2011-07-09 revision 32553″
full_version: “ruby 1.9.2p290 (2011-07-09 revision 32553) [x86_512-linux]“

homes:
gem: “/usr/local/rvm/gems/ruby-1.9.2-p290@cartodb”
ruby: “/usr/local/rvm/rubies/ruby-1.9.2-p290″

binaries:
ruby: “/usr/local/rvm/rubies/ruby-1.9.2-p290/bin/ruby”
irb: “/usr/local/rvm/rubies/ruby-1.9.2-p290/bin/irb”
gem: “/usr/local/rvm/rubies/ruby-1.9.2-p290/bin/gem”
rake: “/usr/local/rvm/rubies/ruby-1.9.2-p290/bin/rake”

environment:
PATH: “/usr/local/rvm/gems/ruby-1.9.2-p290@cartodb/bin:/usr/local/rvm/gems/ruby-1.9.2-p290@global/bin:…”
GEM_HOME: “/usr/local/rvm/gems/ruby-1.9.2-p290@cartodb”
GEM_PATH: “/usr/local/rvm/gems/ruby-1.9.2-p290@cartodb:/usr/local/rvm/gems/ruby-1.9.2-p290@global”
MY_RUBY_HOME: “/usr/local/rvm/rubies/ruby-1.9.2-p290″
IRBRC: “/usr/local/rvm/rubies/ruby-1.9.2-p290/.irbrc”
RUBYOPT: “”
gemset: “cartodb”

Attempts to run npm reminded me to add environment variables to .bash_profile instead of .bashrc. But npm wanted a version of node between 0.4.1 and less than 0.5.0. Well, node was installed in /opt/node for a reason: so that I could make this mistake and undo it. The Ubuntu oneiric version worked.

I see that I am old: I instinctively skipped the instructions to install Windshaft-cartodb because this suggested Windows. But this:

git clone git@github.com:Vizzuality/Windshaft-cartodb.git
cd Windshaft-cartodb
npm install

Led to this:

floorion@somewhere:~/cartodb/Windshaft-cartodb$ npm install
npm ERR! Error: Using ‘<' with 0.5.x makes no sense. Don't do it.
npm ERR! at /usr/share/npm/lib/utils/semver.js:94:29
npm ERR! at String.replace (native)
npm ERR! at replaceXRange (/usr/share/npm/lib/utils/semver.js:88:25)
npm ERR! at Array.map (native)
npm ERR! at replaceXRanges (/usr/share/npm/lib/utils/semver.js:84:17)
npm ERR! at Array.map (native)
npm ERR! at toComparators (/usr/share/npm/lib/utils/semver.js:66:8)
npm ERR! at Object.validRange (/usr/share/npm/lib/utils/semver.js:138:11)
npm ERR! at testEngine (/usr/share/npm/lib/utils/read-json.js:244:44)
npm ERR! at /usr/share/npm/lib/utils/read-json.js:183:10
npm ERR! Report this *entire* log at
npm ERR! or email it to
npm ERR! Just tweeting a tiny part of the error will not be helpful.
npm not ok

The final steps hang.

floorion@somewhere:~/cartodb$ rvm use 1.9.2@cartodb –create
Using /usr/local/rvm/gems/ruby-1.9.2-p290 with gemset cartodb
floorion@somewhere:~/cartodb$ bundle install –binstubs
Fetching source index for http://rubygems.org/
…

Time to do something else. I take comfort in knowing that learning Haskell makes you a worse programmer.

]]> http://opusminimax.wordpress.com/2012/01/19/cartodb-install/feed/ 0 florianlengyel http://opusminimax.wordpress.com/2012/01/14/groupthink-and-the-principal-agent-problem/ http://opusminimax.wordpress.com/2012/01/14/groupthink-and-the-principal-agent-problem/#comments Sat, 14 Jan 2012 20:23:16 +0000 F Lengyel http://opusminimax.wordpress.com/?p=739 ]]> The Rise of the New Groupthink, which Jarod Lanier called “the alarming rise of the fallacy of the infallible collective,” is a manifestation of the Principal-Agent problem.

Agency Theory (P: Principal, A: Agent). Permission: Creative Commons Attribution-Share Alike 3.0 Unported

SOLITUDE is out of fashion. Our companies, our schools and our culture are in thrall to an idea I call the New Groupthink, which holds that creativity and achievement come from an oddly gregarious place. Most of us now work in teams, in offices without walls, for managers who prize people skills above all. Lone geniuses are out. Collaboration is in.
— SUSAN CAIN. The Rise of the New Groupthink. New York Times, January 13, 2012.

What could account for this trend?

1. Increased productivity and efficiency decrease the need for management

The rise in productivity since the 1970s, thanks to increasing automation, has affected managers as well as front-line workers.

“…where have the benefits of technology-driven productivity cycle gone? Almost exclusively to corporations and their very top executives.
– TIMOTHY NOAH. Wall Street Bolshevism, Part 3. The New Republic, October 5, 2011.

But increased productivity and efficiency decrease the need for expensive, well-compensated management. The increased productivity means more profits for owners (the principals), who are motivated to reconsider the distribution of profits to managers (the agents). What is a self-interested middle manager facing the prospect of “upward restructuring” to do? Institute a program of collaborative working arrangements that jam concentration and short-term memory with the extraneous cognitive load of constant interruption under the cover of teamwork. This maneuver increases the need for management intervention as productivity is impaired.

2. Risk management in large organizations

“It’s not hard to see why the fallacy of collectivism has become so popular in big organizations: If the principle is correct, then individuals should not be required to take on risks or responsibilities. We live in times of tremendous uncertainties coupled with infinite liability phobia, and we must function within institutions that are loyal to no executive, much less to any lower level member,” Lanier writes.

3. Information asymmetry

Cooperative working arrangements have been motivated by the need to reduce asymmetric information between managers, who need to know what their staff is doing, and an increasingly efficient workforce. Changes in efficiency will tend to affect information symmetry. The attempt to discourage “silos” and eliminate asymmetric information altogether undermines specialization, the division of labor and, as studies show, productivity.

Research strongly suggests that people are more creative when they enjoy privacy and freedom from interruption. And the most spectacularly creative people in many fields are often introverted, according to studies by the psychologists Mihaly Csikszentmihalyi and Gregory Feist. They’re extroverted enough to exchange and advance ideas, but see themselves as independent and individualistic. They’re not joiners by nature.
— SUSAN CAIN. The Rise of the New Groupthink. New York Times, January 13, 2012.

Workplace collectivism may suit organizations that favor the ancient Roman ideal of competent mediocrity. The research that Susan Cain cites is most likely to benefit organizations that aspire to more.

]]> http://opusminimax.wordpress.com/2012/01/14/groupthink-and-the-principal-agent-problem/feed/ 0 florianlengyel Principal-Agent Problem http://opusminimax.wordpress.com/2012/01/04/where-angels-fear-to-tread/ http://opusminimax.wordpress.com/2012/01/04/where-angels-fear-to-tread/#comments Wed, 04 Jan 2012 05:17:03 +0000 F Lengyel http://opusminimax.wordpress.com/?p=719 ]]> Terry Tao advised that it is inefficient to rewrite textbooks on mathematics. An exception could be made for logic, a subject whose tradition of rebarbative notation and terminology is a cry for help. (I exclude the homotopy type theorists.)

–structures and operators

Let be a first order language with signature (the set of non-logical symbols of ). An –strucure is a pair consisting of a nonempty set (the domain of ) and a map

such that for each constant symbol of ; for each -ary relation symbol of ; and, for each -ary function symbol of .

(The notation was taken from sheaf theory. Whether the suggestion that a non-logical symbol can be construed as a “section” that can have values in various -structures makes any mathematical sense, the notation is preferable to writing for .)

An –assignment is a map from the variables of to the domain of .

The assignment will be extended to a prefix operator on -terms and then to a postfix operator on –formulas. An extended assignment is an operator. An operator is a new term of the metalanguage having lower precedence order than any logical or nonlogical symbol of , but higher precedence order than logical operators of the metalanguage, i.e., ‘not’, ‘and’, ‘or’, ‘for all’, and so on.

Let be the domain of the –structure , let and let be a variable. We define an operation on –assignments (and by induction, extends to an operation on @ operators). If @ is an –assignment, is the –assignment given by

Since the operations act on the right of @ operators, for we have

The –assignment extends homomorphically to a prefix operator on –terms:

is itself, where is a variable;

, where is a constant symbol of ;

, where is an -ary function symbol of and are –terms.

The prefix operator extending the –assignment extends to a Boolean-valued postfix operator on -formulas; this is the Tarski satisfaction relation, inductively defined as follows.

On atomic –formulas:

iff , where are -terms and where is the diagonal ;

iff , where is an -ary relation symbol of and are –terms.

On non-atomic –formulas:

iff not (note that the `not’ of the metalanguage has lower precedence than the operator);

iff either not or ;

iff either or ;

iff and ;

iff there for each , ; and

iff there exists with ,

where the postfix operator is inductively defined from the corresponding –assignment.

The definition of the satisfaction relation as a postfix operator involves induction over –terms and –formulas, and is parametrized by variables of and elements of the domain of the –structure .

If we wish to emphasize the –structure with respect to which an operator is defined, we can write , read as “ is satisfied by in .” This notation is less cumbersome than the standard notation

where is an –asssignment (not an @ operator in the standard account).

An exercise in @ induction

If is a term and @, @’ are –assignments such that for each variable of , , then . This is a trivial induction. If is a constant , then

If is a variable, equality is immediate. If is , then by the term induction hypothesis,

Now suppose that is an –formula and that @, @’ are –assignments such that for each free variable of , . Then iff . We write `‘ for `iff’.

For atomic

If is where are –terms, then all variables of are free, and by previous remarks and the induction hypothesis,

If is then all variables of are free, and by previous remarks and the induction hypothesis,

For non-atomic

We treat only the case of . There are two cases.

(i) The variable is not free in . By assumption, for each and for each free variable of , . By the induction hypothesis,

(ii) The variable has a free occurrence in . By assumption, for each and for each free variable of , . Hence by the induction hypothesis, for each , It follows that

Substitution

If is a variable and if are terms is a substitution operation on terms, where is the term arising from by simultaneous substitution of for each occurrence of in . (This can be defined inductively.) This operation acts on the left of terms; the similar notation , which is differently typed ( is an element of the domain) acts on the right of an @ operator.

A standard exercise in our notation is more readable than in the usual presentation. Let be terms and let @ be an –assignment, If , then

This is almost associativity. Could it be made associative? Two possibilities suggest themselves.

(a) Logicians speak of the expansion of a language by a structure: add the elements of the domain as new, self-evaluating symbols to the language. More precisely, given an –structure , define a new language extending called the –expansion of . If the signature of is , the signature of the -expansion is ; i.e., the expansion has, in addition to the set of constant, function and relation symbols of , the elements of the domain as new constant symbols. (We assume .) The expansion of has a natural structure , obtained from the old structure : the domain is the same, and the map is given by

In the expanded structure, it makes sense to write

(b) Alternatively, we could extend the definition of the right operation on @ operators for an –structure to allow –term parameters by defining

where is an –term. With this inevitable definition, (2) becomes

The standard notation for (2) is this:

Here is another beginning exercise. The standard notation is shown first. If is an assignment, if the term is free for in and if , then

In the new notation, if the term is free for in and if , then

If we adopt the convention (b), this reads as follows. If the term is free for in , then

(We let the term substitution operator act on the left of terms and on the right of formulas and @ operators.)

]]> http://opusminimax.wordpress.com/2012/01/04/where-angels-fear-to-tread/feed/ 0 florianlengyel http://opusminimax.wordpress.com/2012/01/02/alpha-and-beta-characteristic-functions/ http://opusminimax.wordpress.com/2012/01/02/alpha-and-beta-characteristic-functions/#comments Mon, 02 Jan 2012 22:12:18 +0000 F Lengyel http://opusminimax.wordpress.com/?p=700 ]]>

The definition of - and -characteristic functions is paraphrased from The Cooperative Game Theory of Networks and Hierarchies by Robert Gilles.

A cooperative game is pair consisting of finite set of players, and a map , called the characteristic function, such that .

The and -characteristic functions of a normal form game

assign values to coalitions of the players. An aggregation function is a functional of type

which associates to a subset and a strategy profile the value in depending on the payoff functions of the players of the coalition . We emphasize the dependence of the aggregator on the payoff functions of the players by writing . We write for the -subcartesian product of , which pedantically is the image of the map which takes a strategy profile , viewed as a choice function with , and restricts the domain of to the subset and corestricts the codomain to . For we write where , and where is the union (coproduct, if you like) of functions.

The -characteristic function for the game and aggregator is defined by

This is the pessimistic view: the opposing coalition fixes a strategy and chooses the strategy that gives the worst payoff; chooses the best of the worst.

The -characteristic function for the game and aggregator is defined by

This is the optimistic view: given a fixed strategy the coalition chooses the strategy with the best payoff for itself; the opposing coalition chooses the most hostile strategy profile in response. The optimism of in the face of the hostile coalition of mansplainers motivates the title of this blog.

Perhaps my optimism for the future of the mathematics of the social sciences is unjustified. Herbert Gintis has pointed out that strategic interaction distinguishes the behavioral sciences from the physical sciences. But mathematics lacks a characterization of interaction that distinguishes strategic interaction from non-strategic physical interaction. Gintis’ distinction isn’t mathematically definable–yet.

We have that ; this is a consequence of the following easy exercise.

Proposition 1 Let be a map.

Proof: Fix . Then

Hence for any ,

Fix on the right. The right is an upper bound for the left (over ). Taking the of the left over yields

The left is now a constant (extended real number) lower bound of the right. Take the over .

]]> http://opusminimax.wordpress.com/2012/01/02/alpha-and-beta-characteristic-functions/feed/ 0 florianlengyel