Archive for the ‘Computing’ Category

Testing Der Mundo

Monday, September 23rd, 2013

This blog and the website of which it is a part are in English, which is a minority language (as are all other languages in the world).

Various services are available that could translate it into other languages.

With this entry, I am testing one of those services. It is named “Der Mundo”.

Here’s how it works. I give you a special address for this website, and you visit the site using that address. Der Mundo intercepts your request and detects your preferred language on the basis of the information provided by your browser. Der Mundo then submits the page content to a translation service, gets the translated output, and sends it to you in lieu of the original page.

Does it work? Let’s try it. The special address is

One problem, though: If your favorite language is English, then nothing should happen, so, in that case, to test Der Mundo you should temporarily tell your browser or your operating system that your preferred language is something else.

Good riddance, Android USB Utilities

Sunday, August 19th, 2012

My switch from an iPhone to an Android smartphone was motivated in part by the fact that Android permits a device to act like a storage volume (e.g., disk drive). Until a month ago, that worked, but it was awkward, requiring several steps with a setting named “USB Utilities” and the USB cable. As of the current version of the Android operating system, however, just connecting the device to a qualifying computer with the USB cable is enough. The computer recognizes the device, as it would recognize a USB disk or thumb drive. If the computer is a Macintosh, the user must first download and install the Android File Transfer application, but that’s an easy one-time prerequisite.

Another part of the reorganization was to unify the file systems. So, when you plug your device in, you see more directories to read from or write to than was true before.

I mention all this here because I was initially mystified by the change and looked all over for an answer before ultimately telephoning the helpful service folks at CREDO Mobile. So the information could use more exposure, and this posting is one small attempt to provide that.

Charging money for standards

Sunday, August 5th, 2012

The conflict is obvious: Some organizations want everybody to adopt standards, because this makes those organizations more effective. Nonetheless, after collaborating to define and publish standards, such organizations agree to prohibit anybody from getting a copy without paying money for it.

Sure, if I can get you to comply with my standard and also to pay me money to discover what my standard is, why not? But the world isn’t so docile. If I charge you money to see my standard, you are likely to decline the offer and thus ignore and violate the standard.

17 years ago experts were complaining about this conflictual practice of computing-standard creators. And they were only echoing an argument that has been made in the courts since the 19th century: It is in the public interest for the law to be known by all, so that it can be obeyed by all; therefore, the authors of texts that are incorporated into laws may not prohibit, or charge money for, the duplication of those texts. See Veeck v. Southern Building Code Congress International, Inc., for a 2002 statement of that argument.

Today the International Organization for Standardization (ISO) continues to finesse this issue by claiming, with no substantiation, that it can (1) limit the distribution of its standards by charging money for copies and enforcing copyrights while (2) “making sure standards are implemented as widely as possible”.

Nonsense. ISO’s conduct fits a different model: It’s a consortium (or perhaps cartel) of members who derive competitive advantage from their compliance with a standard, and are therefore willing to pay the price of access to (and, indeed, participation in the creation of) that standard. You don’t directly comply with the standard; instead, you buy standard-compliant products from consortium members and thereby enjoy the benefits of interoperability among those members’ products. You don’t need or want to know what the standard is or how it works.

Nonetheless, groups without any obvious competitive purposes (such as linguists who promote uniform language-documentation practices) continue to seek ISO sponsorship of their standards. Doesn’t this status, which prohibits the open publication of the standards, do them more harm than good?

Installing Latest Perl on OS X

Saturday, March 5th, 2011

The first times that I tried to install the latest version of Perl (5.12.3) on OS X 10.6.6 (whose bundled Perl is 5.10.0), the installation failed.

1. Installation of the ActiveState binary would have been easy, but it stalled without explanation, failing to show any disks from which to choose an installation destination.

2. Installation from source code failed with an error in the execution of “make”, namely:

./miniperl -Ilib cpan/Archive-Extract/pm_to_blib MAKE=make LIBPERL_A=libperl.a
Can't locate unicore/ in @INC (@INC contains: cpan/Cwd lib .) at lib/ line 96.
make: *** [cpan/Archive-Extract/pm_to_blib] Error 2

I found an explanation of this error in only one place: a forum of the “BitchX” project. Because that’s somewhat obscure, I am repeating the explanation and cure here.

The error was (somehow) due to the fact that the environment variable PERL_UNICODE was set, with a value of 63.

To prevent the error from occurring, I issued the command “unset PERL_UNICODE” and then executed “make” again. This time “make” completed without error.