jasonpowers.net

The TIOBE index managing directory, Paul Jensen, was recently interviewed by Dr. Jobbs.  The question posed was regarding C and C++ losing ground in the top programming languages being used, in which he claimed they were simply because they did not have better memory management and automatic garbage collection like C# and Java.  

Just because C and C++ do not have such memory management does not mean that will no longer have a place in the industry.  C has been around for ages and is still being used to write new code.  Have you heard of LINUX?   The Linux kernel is written in C.  Does this mean Linux is losing ground also, because it will eventually lack developers that know C?   C++ code is found everywhere, including most of the computer games that are being enjoyed today.  Does this mean those games are losing ground as well and should be rewritten in Java or C#?    You also mentioned that Python replaced Perl on the index because Python was object oriented.  Your group needs to research that a bit more as Perl also has an object oriented structure now.

 While I like coding in C# and Java, I seriously doubt they will dominate the market completely.   Not only due to maintenance of legacy applications, but also because speed and overall performance is superior to automated memory management, especially when such memory management is still available in C and C++, if the programmer is able to implement it correctly.  You will need a bigger bullet than garbage collection to kill off these languages.  

 On a side note, are you sure that the level of web presence a programming language has on the web is truly indicative of its actual usage or value to the industry?   I wonder if I were to post thousands of videos on YouTube  showing how to write code in C++, if it would “regain” its position on your index?  Bah.

I recently modified the security scripts that are analyzing attacks against my home network and decided to see if any of the systems that were brute forcing me had websites.  I wrote a small port scanner in Python and let it rip… several had HTTPD running.   I browsed to one in particular, which happened to be a French shipping/transport company.   There were hundreds of order information PDFs and packing slips scattered throughout the site.  Complete with customer names, addresses, what they had ordered, etc.   If their customers new about their lack of security, I’m sure they would go out of business.

 Being the nice guy that I am… I didn’t retaliate with an attack of my own.  Because I realized that the owner’s of this website had no clue what was going on.   The site appeared to be an unauthorized web server that contained company information.  My guess is, someone in the shipping department thought they were technically inclined and wanted to track the shipments by uploading the slips to a centralized server, which appeared to have been bought and hosted outside of the company.   To bad the system administrators forgot to secure such important content.

 I discovered their official company website… and contacted the info and abuse staff to let them know that they may need to re-evaluate their security policies.  I wrote the message in English, but translated it into French as well.  For starters, they have business sensitive information in public view.  Not to mention the fact that this particular server is probably a nest from which attacks on other servers are staged.  

After reading a few industry reviews, I quickly realized that ASP.Net was the language of choice for most web applications.   For years, I have been favoring PHP due to its close similarities to C++.  However, outside of a few open source projects, PHP is not in very high demand, at least in my region.   I will still maintain my existing PHP applications, write the occasional PHP script to administer a few web sites, and of use it for a few projects at work (seeming as how we are a Solaris shop).   However, most of my web development effort will be migrated over to ASP.Net in order to user C#. 

 The first thing to resolve was the fact that my home web servers are running Apache over Linux.  Obviously ASP.Net favors IIS over Windows for the web server.  Being that I have had experience in writing portable applications between Windows and Linux via Mono… I decided to take a look at Mono to handle running ASP.Net over Linux.

 The solution was mod_mono available and documented here:  http://www.mono-project.com/Mod_mono.   Mod_Mono is an Apache 1.3/2.0/2.2 module that provides ASP.NET support for the Apache web server.  Stay tuned for a few web apps I have coming down the pipes…

The Apple iPhone Software Development Kit has been released!    I may switch gears a bit from mobile development on Palm TX over to the iPhone.  The SDK and related documents can be found here:  http://developer.apple.com/iphone/.   I’ll post my first iPhone program as soon as I get a chance to test it out.  

There is also a $100 million iFund for startups developing applications for the iPhone.

I have migrated my blog software into WordPress.   I selected a few blog entries to come along for the ride, you’ll see them below.  Updates to this blog will be made weekly, if not daily, so check back soon!

For a few years now, I have been wanting to build a Beowulf Cluster… however, the cost of such a project has just became within range with the prices of memory and processors steadily on the decline.

For those of you who do not know what a Beowulf Cluster is… it is a collection of personal computers that are connected together to acheive high performance computing. Google it to see some pretty cool pictures (if you are into that sort of thing).

I do not have the funding of a enterprise company, a university, a hospital, a research center, or some government… so the cluster I build will be relatively small in comparison. My plan is to start off with 5 nodes, including the server… and add nodes as I streamline the process.

Primarily, I am doing this to mentally torture myself… really just to learn more about parallel programming. In the process, I may be able to offer researchers the ability to simulate chemical reactions, atmospheric modeling, protein folding for cancer research, physics simulations, etc… or in the end, I may just have one huge system to serve all of my favorite games 24/7.

I was starting off with setting up the master node with Fedora Core 8. Which was a mistake, because OSCAR does not support it yet. So, I can either drop back to an older version of Fedora, change Linux distributions to say… Redhat… or, download the source from their subversion directory and modify the code and recompile it… thereby creating support for Fedora 8 myself. The SVN content is still downloading… Pending how much time I anticipate wasting, I may just switch distros or replace OSCAR with a different clustering utility.

More later…

“C++0x is the planned new standard for the C++ programming language. It is intended to replace the existing C++ standard, ISO/IEC 14882, which was published in 1998 and updated in 2003. These predecessors are informally known as C++98 and C++03. The new standard will include several additions to the core language and will extend the C++ standard library, incorporating most of the C++ Technical Report 1 libraries - most likely with the exception of numerical libraries.”

http://en.wikipedia.org/wiki/C%2B%2B0x

I decided a while back to scale down my network and related equipment, to get rid of things that I have not used in a long time. Call it a technical spring cleaning of you want…

I went from about 21 computers in my home office down to 10. The other computers were either scrapped for parts, combined into a single computer and given away, or thrown to the garbage. I could not bare to part with my SGI workstations or the 72 port network hub, so I decided to store them in my attic… until I had a new project in mind that may require them.

I imagine that the number of remaining systems will be cut in half again soon. Although, I plan on making a dedicated computer room soon when I get the power requirements figured out, I will end stacking in the rack mounted servers. and take the system count down to two, while probably boosting up to 12 to 24 rack mounted servers. Depending on if I am still interested in the beowulf cluster and parallel processing projects at the time.

After having dug to a depth of 10 meters last year, Scottish scientists found traces of copper wire dating back 100 years and came to the conclusion that their ancestors already had a telephone network more than 100 years ago.

Not to be outdone by the Scots, in the weeks that followed, British scientists dug to a depth of 20 meters, and shortly after headlines in the UK newspapers read: “British archaeologists have found traces of 200-year-old copper wire and have concluded that their ancestors already had an advanced high-tech communications network 100 years earlier than the Scots.”

One week later, “The Advertiser”, a Lafayette, Louisiana, newspaper, reported the following: “After digging as deep as 30 meters in cane fields near New Iberia, Gaston Boudreaux, a self-taught archeologist, reported that he found absolutely nothing. Gaston has therefore concluded that 300 years ago Cajuns were
already using wireless.”

-Muhahaa.

I had a need for a list of zip codes and some other relavant information and a need to tinker with Python… So here is a small script that will go to a few zip code sites and save the codes and other info to individual text files for later parsing. I must say this was relatively quick and easy compared to what I would have had to write in C++ to do the same thing.

*NOTE: Don’t use this code to plague the sites, they normally shut you down around 130 hits any way. Although, I programmed a way around that, but due to those interested in DoS attacks, I will not post that here. If you have a need for the zip codes, I can make my zipcode database available to you.

import urllib2

zip=10000
while zip < 99999:
#address = “http://www.zip-codes.com/zip-code/” + str(zip) + “/zip-code-” + str(zip) + “.asp”
address = “http://zipinfo.com/cgi-local/zipsrch.exe?cnty=cnty&ac=ac&tz=tz&ll=ll&msa=msa&zip=” + str(zip) + “&Go=Go”;
website = urllib2.urlopen(address)
website_html = website.read()

print “Processing zipcode: ” + str(zip)

filename=”data/” + str(zip) + “.html”
myfile = open(filename,’w')
myfile.write(website_html)
myfile.close()
zip = zip + 1;