A quick note about using newer version of Ubuntu: From 8.10 (Intrepid Ibex) on, Ubuntu doesn’t come with a g77 package. g77 is not used for GEANT4 specifically, but with CERN’s ROOT, which is an amazing statistical analysis package. You won’t NEED this for GEANT4, but it allows you to do some cool stuff.  Anyhow, instead of g77, Intrepid (8.10) comes with ‘gfortran’ which is not the same code, it seems.  Now, there are two ways on handling it.  I, personally, advise sticking to 8.04 and the g77 it comes with.  However, you can use a newer version of Ubuntu (or its variants) and either use Hardy packages or install from source.  Either way, it’s nice to have ROOT installed for statistical analysis.

As a side note, I’m hoping to (eventually) post a complete walkthrough of the process I go through when compiling CLHEP, GEANT4, and ROOT on a 8.04 machine.  It’ll probably be a while before I manage to get it written up, but it sure would be nice…

After a bit of searching, though, I’ve found a nice updated program called Back in Time.  I really like a lot of the features it supports.  It’s written in python and uses rsync and diff to do its magic.  One of the nice features of Back in Time is that it only updates files that have been updated since the last backup.  This doesn’t seem like that huge of a thing, but it does it well.  The only problem I’ve found is that the it won’t accept symlinks that link to outside of its directories–you’ll have to manually include that directory to be included.  Other than that, it’s great!  Check it out.

"Can't open /usr/share/mime-info/air.keys" (error 2: no such file or directory)"
"Can't open /usr/share/mime-info/air.mime" (error 2: no such file or directory)"

Obviously, this wasn’t the worst thing that could possibly happen to me… but it was rather aggravating!  To solve this problem I went scrounging through the depths of the interblags and finally managed to fix it.  Lots of help comes from the Dusty Pixels blog, but I figured I would compile what worked for me here.

I will copy in these codes directly as I used them (and supress any output).  Please note that I do sudo into root for this fix just to make sure there are no permissions errors–if you wish to go the fully ‘approved’ method, just append sudo on every single command here.

nuke@nuke-laptop:~$ cd /opt/Adobe\ AIR/Versions/1.0/xdg-utils/

nuke@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils$ sudo su

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# xdg-mime uninstall --novendor /opt/Adobe\ AIR/Versions/1.0/support/AdobeAIR.xml

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# sudo nano ../support/AdobeAIR.xml

This step requires that you remove the three lines starting with <magic priority=”100″> and ending with </magic>.

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# xdg-mime install --novendor /opt/Adobe\ AIR/Versions/1.0/support/AdobeAIR.xml

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# sudo unlink /usr/share/mime-info/air.keys

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# sudo unlink /usr/share/mime-info/air.mime

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# sudo ln -s /opt/Adobe\ AIR/Versions/1.0/support/air.keys /usr/share/mime-info/air.keys

root@nuke-laptop:/opt/Adobe AIR/Versions/1.0/xdg-utils# sudo ln -s /opt/Adobe\ AIR/Versions/1.0/support/air.mime /usr/share/mime-info/air.mime

And there you go!  You shouldn’t have any more mime-type errors in Filezilla!

The biggest implication of Best Buy selling Ubuntu is just that–Best Buy is selling Ubuntu!  It is, as far as I know, either the second or third time a major retailer has marketed Linux (the others being Walmart’s $200 Linux PC and the Asus Eee PC).  Many people are hesitant to adopt anything that is free in the computing world due to the horrible experiences that most people have with free, spyware-infested software.  If the slightly computer-savvy average person is in the store and sees Ubuntu it should help them realize that it is a legitmate piece of software.  Once people start seeing Ubuntu in the store there will definitely be a large amount of random purchases of the software.  Also, on the bright side, a boxed copy of Ubuntu would make a much nicer present than CD-R with Sharpie on it!  Either way, the selling of Ubuntu, especially at the reasonable price of $20, is a good thing for adoption.

The biggest battle for Linux adoption is to get the software into the hands of new people.  But even after they have paid for it, many may not install their new, shrink-wrapped copy of Ubuntu once they actually try to install it.  Confusion over various things (Hey Jeb, what’s a partition?) that seem overly simple to experienced Linux users will be enough to offset users.  Luckily, the Geek Squad installation option is, ideally, great.  This is, of course, most likely the first and last time I will ever complement the Geek Squad as a whole, but I really think this is a good thing to have.  First of all, it makes it seem more ‘comparable’ with Windows’ high price tag.  $20 isn’t much, but when you add on a $130 installation charge it is more in the range of Windows (which is, after all, the main competition).  This also gives users a way to install Ubuntu without having to know anything about computers.  As long as the Geek Squad agent (and I use that word in the loosest term imaginable) is competent, this will definitely provide a decently painless way for people to obtain Ubuntu.

When someone is walking in the aisles of Best Buy and sees Ubuntu they will, most likely, be confused.  Even after reading the box they may still have questions and not buy it, and that’s perfectly fine.  At least the user will have been exposed to Ubuntu which is, in my opinion, a major victory.  If enough people start wondering about Ubuntu, they’ll start talking.  It’s not hard to imagine two co-workers chatting about this strange new Ubuntu thing they saw the last few times they were at Best Buy.  Retail Stores are, after all, about advertising products and making them look good enough that a person feels the need to purchase them.

In the end, we’ll all have to wait and see.  Maybe this will pay off like I (and so many others) hope it will.  Of course, it could also easily have the opposite effect.  If Best Buy doesn’t sell enough copies to justify keeping Ubuntu in stock, it will probably acquire an aversion to selling more Linux distros and Linux-based computers.  But for now, I’ll keep on hoping.

First of all, I must admit–the idea of making C code to run on the web is more appealing than Python to me.  Maybe I ‘get’ C better, maybe it’s the fact that Python makes me hate life, or maybe I am just partisan to good programming languages–who knows?  Either way, I just don’t get why people are always trying to move away from languages like PHP which were built to do this.  Do you understand?  PHP was made for the web.  C and Python were not (they appeared on the scene in 1972 and 1991, respectively).  Why do you insist on running those for the web?  Is it the same kind of mentality that drives people to run Linux on refridgerators? Do we just have to prove that we can run anything as a web language?  When do I get my Web-Fortran (and yes, I’m sure it exists somewhere)?

That all being said, this implementation seems quite interesting.  The programmer would write his C code and then compile it.  It would be compiled into a special format for a Virtual Machine.  This would then be converted into code for ActionScript, which is what Flash uses to code it.  This scares me on one hand–god knows Flash is whored out too much as it is–but it’s still very interesting theorhetically.  That being said, once I am forced to write some software in C to be used on the web I think I will go ahead and just stop programming for good.  I understand that the C code could quite possibly be faster than cooresponding PHP code… but the C code would have to be put into flash which completely kills any of that benefit for the end user.  Arrrgh.

First, in response to Dolinar’s article, I would like to make a few comments.  His very first sentence attempts to set the tone for what ends up becoming an article harassing Microsoft for not adopting more formats.  Dolinar wonders “Why doesn’t free trump expensive?“  I feel that this problem of non-adoption of F/OSS software (especially in lieu of expensive proprietary software) has two barriers to vanquish in order to gain the end goal of widespread adoption.  First, it has to be a better piece of software.  Secondly, it must have a serious driving force behind it.

For free software to win over expensive, commercial software it must be better.  The problem is that everyone defines the word ‘better’ differently.  For a lot of computer geeks, including me, better means something completely different than it would to a ‘normal’ person.  I love Linux because it allows me to hack at my entire computer system and customize whatever I feel like customizing.  At the same time, however, it allows me to mess up things that would stump a normal user.  When my Linux server wouldn’t initialize my X server session because the computer clock battery had died it stumped me for a while, but I enjoyed a challenge.  When my mom’s computer clock is off from real life, she has to call the nearest tech support person over to help her fix it.  The majority of the market for common, everyday software doesn’t necessarily care for more complex, hackable, or secure software.  They want something that is moderately stable and usable.  More specifically, they want something that they are familiar with it.  As a generation that has grown up with GUIs (though I ran DOS and was quite amazed by it at the time) we have always had some variation of Office, whether it be Lotus’ version or Microsoft’s version.  Later in life we were taught and tested on MS Office.  People who are older and, generally, less welcoming of new software are resisting more or less anything new that is not forced on them.

The problem that OpenOffice is encountering is not one limited to OpenOffice.  Hell, just look at Office 2007 and the new ‘Ribbon’ interface.  I, personally, love it–I feel that it makes things much easier to find and is more visually appealing.  The a large part of the workforce, however, has resisted the software.  Change is always bad.  Now consider for a second that, while people don’t like the new version of MS Office, they at least know what exactly MS Office is and how it (as a general rule) works.  Why would they install a completely new set of applications that they have to relearn?  Keep in mind that to get OpenOffice they have to visit a new website and download the installer.  If an office wants to do this they will have to install it on every single new workstation and then change settings that are needed to work with other offices that don’t use OpenOffice (such as changing the default document type).  While I was using OpenOffice Writer I don’t know how many times I have sent a file to someone, not thinking about the fact that the other person is basically guaranteed to not be running MS Word.

On the flip side, Firefox is one of the most popular open-source projects.  According to PC World, Firefox has a 19% market share and is predicted to hit 20% in July.  What made Firefox such a success?  I would argue that Firefox had a few things going for it.  First of all, it came from a decently developed background.  Firefox is the offspring of the Netscape browser, so a lot of the development team had a large amount of experience working on web browsers.  Secondly, it was better than the huge mess that was IE.  IE 6 was initially released in 2001.  The second of two major patches for IE 6 was released in August 2004.  Firefox 1.0 was released in November 2004.  At this point, IE 6 was a mess.  It didn’t have tabs, extensions, or anything close to a valid CSS intepreter.  Firefox, however, was new, fast, and didn’t have a lot of the problems that IE faced in being the major browser.  People started switching and word got around about a new browser that was better than IE 6.  Firefox then leveraged this excitement (along with a strong development team) to produce solid versions.

Firefox shows what can go right with a project.  If OpenOffice even had 15% of the market it would be viewed as a major success against Microsoft’s Office.  If OpenOffice really wants to win against MS Office, they need to go for visible successes and not hope that the format wars will push them to the front.