On The Margins

Tuesday March 20, 2007

[ Code ] John W. Backus

Today, The New York Times carries an obituary to John W. Backus, of the "Backus | Naur form" notation and the lead of the IBM team who brought us Fortran. Many a scientific computing wizard will today salute Mr. Backus for what he and his team accomplished.

While the need for new programming models was dire in the 1950s, a move by Backus to initiate an applied research program to invent a higher-level language led to a revolution in software. The first Fortran team worked on the language from 1953 to 1957. ("The first written reference to 'software' as a computer term, as something distinct from hardware, did not come until 1958," according to The NYT.)

In my experience with Fortran, I join many others who used this first-generation higher-level language to do useful things, including much scientific research.

I wrote my first toy computer program, which calculated the first 1000 primes, in Fortran. Later on, I wrote Fortran programs to calculate temperature profiles in three dimensional body embedded in a heated environment, to study dispersion and diffusion of particles in turbulent flows, to investigate the dynamics of particle-particle collisions and systems, and to perform direct numerical simulations of fluid flow and vortex-vortex interaction in an infinite body. In short, in the mid 1980s, I spent many hours doing scientific programming in Fortran. (Some of this work got its way into my masters dissertation and other to my Ph.D. dissertation. Much of it remained at the level of pure investigation and study.)

Note: For a modern progeny, see Fortress.

2007-03-20 08:59:43.0 -- ; Permalink ; Trackback.
code fortran programming software

Sunday February 04, 2007

[ Code ] Railing about Rails

While I may rail about Rails here, you could do something more useful by consulting  Peter Schow on how to get Rails 1.2 up and running on Solaris 10 with PostgreSQL, and by checking it out and sharing on Open Solaris. Alternatively, for another Web 2.0 development environment, consider Phobos.

2007-02-04 22:54:51.0 -- ; Permalink ; Trackback.
programming

Sunday January 21, 2007

[ Code ] Scrum and Sprints

Scrum is a light-weight agile software development process composed of Sprints, during which the next release of a system is being developed. To read about Scrum see ControlChaos and MountainGoatSoftware. (MountainGoatSoftware's Scrum page summarizes things quite nicely in the first paragraph on the top.)

2007-01-21 20:43:13.0 -- ; Permalink ; Trackback.
development process

Wednesday January 17, 2007

[ Code ] TDD

The cost savings and efficiencies of Test-Driven Development (TDD) should be clear not just to any hardware engineer but also to software and other engineers, system designers and developers. In fact, even when we imagine we are developing without a test, we implicitly have some kind of test in mind: We build systems in order to elicit specific responses to specific stimuli (tests?) upon those systems. Most software developers know how a good test can drive their work, particularly when the system is complex with multitude of constraints and also when there are volumes of existing tests to detect regressions.

Peter Marklunds describes TDD with Ruby in some detail.

2007-01-17 23:00:34.0 -- ; Permalink ; Trackback.
development ruby software tdd testing

Thursday December 21, 2006

[ Code ] Statistics on Open Source Projects

Now, we have places to go to for open source project statistics.

For example, see the Ohloh statistics for Apache / Derby.

2006-12-21 19:10:43.0 -- Comments [2] ; Permalink ; Trackback.
apache code open programming

Thursday December 07, 2006

[ Code ] Power Javascripting

jMaki means business with power Javascripting and more, including Phobos.

If you know your stuff, you'll check them out! 

2006-12-07 16:50:53.0 -- ; Permalink ; Trackback.
application java javascript phobos server web

Friday October 27, 2006

[ Code ] Innovation Communities and Open Source Software

For a brief but a very useful history of open source software, see chapter 7 ("Innovation Communities") of Eric von Hippel's Democratizing Innovation (also available online) under Creative Commons ("Attribution-NonCommercial-NoDerivs 2.0") license. Von Hippel underlines the user-based innovation communities and presents open source communities as a prime example.

2006-10-27 18:12:21.0 -- ; Permalink ; Trackback.

Monday September 04, 2006

[ Code ] Brevity

Having been raised on spinach, many may end up hating it.

Having been raised on FORTRAN and C, I find solace in brevity, constructs that do away with loops and object orientation. C can be very brief but I see nothing of object orientation in it. FORTRAN of my time had neither characteristics. C++ added the object orientation but only Smalltalk and then Java made it easy enough to use.

Recently, one of my office neighbor's asked me what I thought was distinguishing about Ruby. Without hesitation, I was surprised at my own quick response, given that I'm still learning about the language: "brevity." Of course, Ruby is not for everyone. One can enjoy it more with a combined understanding of object-oriented, functional, script and high-level programming, and an appreciation of how a compiler might have been written for it. A desire for brevity and a positive mental attitude towards it will also help.

What I like is when the code speaks for itself, and that can be done in almost all programming languages.

2006-09-04 20:58:31.0 -- ; Permalink ; Trackback.

Friday August 25, 2006

[ Code ] Just for the Record

Just for the record, it seems that satellite news programming can be banned much more effectively in the U.S. than in any other place in the world. (I had written about this particular case earlier.)

2006-08-25 23:15:06.0 -- ; Permalink ; Trackback.

Saturday August 19, 2006

[ Code ] Another Law Blogger

While those who follow copyright issues and other legal topics surrounding new technologies are sure to run to Lawrence Lessig's blog and find a ton of relevant material to consume, another professorial law blog that might be worth following for those interested in patent issues might be that of Beth Noveck, professor of law at New York Law School.

Recently on her blog, she has been proposing wiki based commnity reviews of patent applications.

While this is an ineresting idea, it is not clear to me whether it will reduce the work of the patent office.  An immediate question: Will there be even more information to digest, leading to an information overload and to mis-information? 

It is highly likely professor Noveck has addressed this and similar questions in her writings under the topic of "patent and innovation."

Here is another question, can professor Noveck patent the idea of community patent review on wikis?

If not, who was the one who first proposed it or something like it? What evidence do we have? And what about the question of durability of paper compared to digital information, raised by others, like John Seely Brown and David Duguid in The Social Life of Information, not to mention myself, earlier here, on some other aspects of the paper "trail"?

And now that we're talking about patents and shameless plugs, here's one for me, which I just found when I typed my name and "patent" into Google. Search for my name in this Nr. 6052 edition of Patents and Design Journal from the U.K. It is good to know more about patent GB2391091.

Endnotes:
Lessig points to this style sheet for turning blog entries into letters to the Congress.

2006-08-19 22:38:35.0 -- ; Permalink ; Trackback.

Friday August 18, 2006

[ Code ] Test, Test and Test Again

I wrote about the upcoming test marathon for Apache Derby 10.2 here.

Good products always involve great testing efforts.

This goes back to the early years of the industry.

General Motors board, in the first couple of decades of the 20th centry, implemented a plan for its famous GM Testing Grounds because the copper-cooled engine produced less than satisfactory results and because the importance of quality control in gaining market share was very, very visible to GM leaders.

In software industry, tests have been a main-stay. Notice the ratio of testers to developers at Microsoft as outlined by Michael Cusumano in his books.

Quality and testing, while sometimes taken for child's play by some developers, involves a very high art that imposes discipline, good, repeatable results and acceptability constraints.

2006-08-18 12:52:44.0 -- ; Permalink ; Trackback.

Friday August 11, 2006

[ Code ] The Whole Point of the Dictionary

2006-08-11 16:36:44.0 -- ; Permalink ; Trackback.

Wednesday August 09, 2006

[ Code ] Open Source Licenses

Here's a listing of various open source licenses.
(Thanks go to a Sun colleague who pointed me to this resource.)

Networks, Code

2006-08-09 07:50:00.0 -- ; Permalink ; Trackback.

Thursday August 03, 2006

[ Code ] A Nice Little Introduction to Programming

Chris Pine's Learning to Program is a gentle start and a well-written guide on elementary concepts in programming.

Chris has picked Ruby for this book. In fact, he says he wrote the book because of Ruby.

2006-08-03 15:06:27.0 -- ; Permalink ; Trackback.

Tuesday July 25, 2006

[ Code ] Open Source Mantras

This week, sun.com is hosting open source mantras by Simon Phipps. The article is a great read and touches on all the essentials.

Simon will also be sharing his ideas at OSCON, Portland Oregon, Thursday, July 27.

For similar topics, you can also try to read Sun's white paper on open source licensing, my earlier note on the economics of open source and on the evolutionary advantage of open source.

2006-07-25 15:31:18.0 -- ; Permalink ; Trackback.

Sunday July 16, 2006

[ Code ] Equal, But Not Equal Before The Law

This vacation time has been very important to me because I have had the good fortune to visit my brother who lives in Turkey. This is the brother with whom I had been unable to visit since the summer of 2001, when my whole family last spent a month in Kalkan, Turkey, where he lives with his Turkish wife and daughter and owns and operates an art shop.

Now that I've started, let me tell you a bit more.

I've resided in the U.S. since 1979, and my two brothers have settled in two other countries---one has settled in Germany since 1986. He is an architect and designer. The other has settled in Turkey since 1984 (or was it 1983?). He is a city planner, painter and store owner. Both of them left for these countries prior to their university education. Both have masters degrees in their fields of study. Both have been recognized by their peers as accomplished and creative. Both have abided by the laws, formed families and supported them. Both are my brothers. We share the same parents and grand parents. We were born in the same city, went to same high school and lived in the same homes.

However, they have not been treated equally before the law--or at least the U.S. law, if such a law exists--when it comes to their rights to visit me, and my rights to have them visit me in my place of residence.

This is a very simple, and one can argue, one of the most basic human rights, i.e. the right to have family members visit you where you live. (On one extreme, most prisoners in the U.S. have such visitation rights by law.) Anything less would be considered inhuman.

Yes, the case is simple and definitive.

My brother who lives in Germany can travel to the U.S. as he wishes and has done so in the past to visit me and my family.

My other brother, who lives in Turkey, has never been able to visit me in the U.S. If my memory serves me right, he has requested to be granted a visa on three occasions (all going back to seven years earlier or before), and on every one of these occasions his requests have been denied. We no longer trouble ourselves with these requests. We know the automatic answer already--denial--so, why bother brother?

Where is the justice in these denials, and why should my brother and I have no right to visit each other in my place of residence, and why should I have to write a weblog entry about this, and why should I have to bother to go to a state senator or a congressional representative in order to make an attempt to assert and bring this self-evident human right to stand against all odds and, most probably, as a special case?

(In closing, it may be worth noting that Ms. Dianne Feinstein, our esteemed senator from California, has been behind harsher measures designed to deny visas to people from Iran and several other Muslim countries. Her proposals have continued to baffle and puzzle me if not other people of similar heritage.)

2006-07-16 15:27:46.0 -- Comments [2] ; Permalink ; Trackback.

Sunday January 29, 2006

[ Code ] The Mind of the Maker

Frederick P. Brooks, Jr., in his essay The Mythical Man-Month, refers to Dorothy Sayers' The Mind of the Maker. Sayers divides creative activity into three stages: the idea, the implementation, the interaction.

The idea stage occurs outside of time and space. It represents an ideal of what is to be made. The implementation stage occurs in the confines of time and space and has to come to grips with the limitations of the medium used to realize an implementation. The interaction stage begins when that which is made arrives at the hands of its users.

Brooks notes that the medium of implementation for programming has proved relatively tractable compared to that for other creative work. This tractability, along with the positive, theoretical aspect of the idea stage, Brooks takes to be the sources of over-optimistic estimates in programming projects.

Brooks' other important insight has to do with the simple fact that not all "men" are equal and that not all tasks can be partitioned into perfectly parallel pieces whose accomplishment requires no communication among those who work in parallel. In fact, addition of resources to a project often pushes its delivery date farther back.

2006-01-29 18:01:16.0 -- Comments [1] ; Permalink ; Trackback.

Tuesday January 24, 2006

[ Code ] Gonzales vs. Google

FindLaw has posted U.S. Department of Justice motion against Google. (You might want to check FindLaw's Law and Internet pages.)

Among other things, Gonzales is asking Google to turn over "a multi-stage random sample of one million URL's," and "the text of each search string entered onto Google’s search engine over a one-week period (absent any information identifying the person who entered such query)."

The Internet, is the greatest distribution, copy and search machine in the world.

2006-01-24 08:56:34.0 -- ; Permalink ; Trackback.

Monday December 19, 2005

[ Code ] Open Source Jahrbuch

If you know German and want to read an European perspective on Open Source, you might consider turning to the Open Source Jahrbuch 2005. (This link comes courtesy of Craig Russell, a.k.a. the father of JDO, who also just recently led an Apache project of the same name out of incubation.)

The Jahrbuch, in its 498 page length (available for free in parts as well as in whole), touches on everything related to open source: case studies, technologies, economics, political and social aspects, open content and open innovation.

The chapter on economics, contains a paper on the use of open source as economic signal and a paper on standardization and open source.

2005-12-19 17:42:03.0 -- Comments [1] ; Permalink ; Trackback.

Sunday October 09, 2005

[ Code ] Good Time To Support Creative Commons

If there was a good time to support Creative Commons, it might be now.

Does this mean that even with all the wide use of Creative Commons licensing by the public, on the web and on many weblogs, actual financial support is required to grant non-profit status to Creative Commons dot Org?

2005-10-09 10:37:33.0 -- Comments [0] ; Permalink ; Trackback.