Posts by Scott Fradkin
I’ve finally decided to go to Madison Barcamp. It should be a whole lot of fun. It’s taking place on March 3 and 4 at the Inn on the Park on the capitol square. Barcamps describe themselves as an un-conference. What it seems to me is a whole bunch of people hanging out and talking about a bunch of really fun technical things. The website is http://www.barcampmadison.com.
One of the rules about attending Barcamp is that you have to help out in some capacity. I’ve decided to give a presentation about Gentoo Linux.
I’m real interested in seeing some of the other presentations, too. It should be a good place to meet a bunch of people and have some fun learning about a lot of different things. Be sure to sign up soon.
Filed in: Team Member Blog
I’m still trying to figure out whether Google’s Testing on the Toilet blog will actually be helpful. It’s an interesting premise, though. According to their blog, they regularly write some stuff about various testing topics, then plaster their bathrooms with what they wrote. The first official testing entry includes a PDF link that will let you plaster your own bathroom.
This is pretty cool, though. There are all kinds of rumors about Google’s development process and how it’s extremely agile. I can’t imagine they’d create this blog if they weren’t big into testing.
I’ll certainly be checking it out regularly to see what kind of tidbits of information they’ve got. Here’s a handy link for your favorite feed viewer: Google Testing Atom Link
Filed in: Team Member Blog agility
I ran across a thread on Slashdot [it.slashdot.org] that linked to an article on the Register [theregister.co.uk] talking about Bill Joy, his impending retirement from Sun, and how he created the vi editor. It’s interesting that someone decided to post a thread about an article that was written almost three-and-a-half years ago. But, that’s not the point. It brought up the whole Emacs vs. vi argument.
It seems like this argument rears its ugly head every so often on Slashdot for some unknown reason. It’s well known in the UNIX and Linux circles that more often than not, developers/sys-admins/etc. either use Emacs or vi to get all their editing done in the terminal world. It’s a similar argument to whether you like the Xbox or PS2 better.
Which do I use? Well… if I had to choose, I’d choose Emacs. I tried… really, really tried to figure out vi when I was back in school, but to no avail. Even though people tried to tell me that vi was waaaay better than Emacs, I would have none of it. I turned to Emacs. For me, Emacs did the job better than vi. It just seemed more natural. I couldn’t get used to what I thought were incomprehensible control keys and no easy to find documentation with vi. So, my entire college career was Emacs-ful and I still maintain that bias.
Truth be told, though… At home, I don’t even use Emacs anymore. When I need to do some quick editing of a file from a terminal window I just use Nano. Now, Nano isn’t the greatest text editor in the world. It’s not even close. It’s a Pico clone [wikipedia], so you get all the un-robustness (is that a word?) of that type of editor. Why do I use it? It’s simple to use. There’s no hassle in learning all kind of key combinations. I just want to edit text quickly, and Nano fits the bill. This may be somewhat of a bias from using Gentoo since Nano is the default terminal editor that is installed, but it works for me.
This gets me thinking. What does make a good editing program? I use Nano from a terminal. For quick text editing in the GUI realm I usually use gedit [wikipedia] (think glorified version of Notepad). For programming, I use Eclipse. In the Windows world I’ll use TextPad [textpad.com].
For the most part, the editors I’ve chosen are fairly simple, uncluttered, and don’t require vast tomes of information to figure out how to use.
What it really comes down to, I think, is pure personal preference. The reason that there are so many different text editing programs available for so many different operating systems is because people’s preferences range so widely. Still, it seems that there’s a large contingent of developers and sys-admins that treat this as some sort of polarizing thing. Either you’re with us or you’re against us. I find the whole Emacs vs. vi debate to be amusing.
Can’t we all just get along?
Filed in: Team Member Blog linux
On January 22, 2007, I’ll be giving a presentation at the Madison Rails meeting. The meeting is at MATC downtown and the presentation should start somewhere around 6:15ish.
The title of the presentation is Web Development Framework Comparisions. During the presentation I’ll show the development of some administration pages for a simple store application using Java/J2EE, Ruby on Rails, and Grails. Full demo applications and presentation materials will be forthcoming.
For more information check out the Madison Rails website athttp://www.madisonrails.com.
Filed in: Team Member Blog ruby
I ran across a link to an article on Slashdot [Slashdot] that purports to be a list of search engine evolutions for 2007. What it amounts to is a press release by a company called Hakia trying to garner interest in their search engine product. The press release dredges up all kinds of search buzzwords such as “semantic analysis” and “fuzzy logic”. It is a beta product, but even so, if you take a look at the Slashdot comments, others have written that this company really doesn’t do a good job returning relevant results. I’m sure Hakia would add, “yet” to the end of that sentence.
This does bring up an interesting subject, though. What would make a search engine better? And, would it actually make any difference? Has Google done such a good job pulling all of us in that we won’t look elsewhere for our Interweb searching needs?
Google has some amazing resources at their disposal. They hire the best and brightest programmers, many of which are Phds that know academic programming. They have an extremely agile programming process in place. They seem to have endless money to throw around. Are they in it for the long haul? I have to believe so. I just can’t believe that they’re sitting back on their laurels and not trying to advance the state of searching.
Certainly, even as good as Google is most of the time, their searches could be better. Hakia is definitely on to something. Searching needs to branch away from using keywords and meta tags and try to make some sense out of page content in order to categorize it correctly. Being able to write your searches in the form of a real question is also a good move. I know that’s been tried before, but I think the problem is that it’s hard to pull out the real meaning of the question and apply it to a search when you’re searching with keywords.
One of these days, some company or individual will figure it out. How to parse page content and index it in such a way as to allow people to search with more than just keywords so that we can have a better search experience and retrieve better results.
Filed in: Team Member Blog
The One Laptop Per Child (OLPC) organization [website] is creating a laptop aimed at children that only costs 100 US dollars.
Wikipedia has a really nice entry about it [wikipedia].
This laptop is a really impressive feat of engineering as far as I’m concerned. The amount of stuff they’ve packed into this machine for 100 bucks (well… apparently 130 to 140 dollars to start out) is astounding.
It uses Linux as its OS, flash RAM rather than a hard disk, ad-hoc wireless mesh networking, a video camera, microphone, sound, USB, and a pretty decent LCD screen. It’s engineered to use 2 to 3 watts of power for long battery life. There’s even a separate power unit that will allow you to use hand power or foot power to recharge the battery.
The OLPC is interested in using this laptop as a modern learning tool for children living in some of the poorest places on this Earth.
There are a lot of interesting discussions that surround this project ranging from people who think that what the OLPC is doing is great, to people who believe that this is money better spent on the basics that these children need to live.
I think it would be kind of neat to be able to get one of these to see what it can do and what it can be made to do. Unfortunately, the OLPC website says that it won’t be available to the general public. Hopefully some of their ideas make their way into consumer models in the future.
Filed in: Team Member Blog
Sun announced earlier this week [announcement] that they would be open sourcing Java. Millions of Java programmers rejoiced.
What does Sun’s decision actually mean for the typical programmer? Probably not much.
Sun decided to release the JDK along with the Java compiler and the virtual machine under the GPL v2 license [GPL website]. What this means is that Sun grants a royalty free license to anybody to redistribute the pieces of the JDK along with the source code. Developers can also create new pieces of work based off the original code as long as they indicate the changes and the original license. This does not stop anyone from charging money, however. It is legitimate to charge for media distribution or warranty support.
What this may really do is kickstart more innovation with regards to the internals of the language implementation itself.
While in the past new JVMs have been created by third parties, Sun would expect those third parties to license the Java technology. Now, since all aspects of the JDK will be under the GPL, anyone will be able to create a new JVM implementation, a new compiler implementation, or even a brand new implementation of the language itself.
I expect the real benefit will come in the form of JVMs for operating systems other than the standard UNIX/Linux, Windows and MacOS. Hobbyists will be able to tweak things and redistribute them to others. Academia could use the source code as a teaching tool.
This announcement is also nice for various Linux distributions and individuals who don’t like to use products that are not licensed with one of the open source licenses like the GPL.
So, at the end of the day this doesn’t really mean much for a regular user, or a non-interested developer. For those who like royalty-free software and the ability to peek under the covers, this is a wonderful opportunity to have some fun.
Filed in: Team Member Blog java
I’m a huge fan of Linux. I’ve been using a Linux desktop as my primary OS for about three years now.
Call me crazy. Call me a glutton for punishment. I enjoy it.
I realized a few years back that there just wasn’t a need to use Windows anymore. Everything that I really use on a computer has a Linux equivalent (well, almost… but there are ways around that). So, I made the big switch.
I started out using one of the older versions of Mandrake (now Mandriva if I remember correctly). Unfortunately, Mandrake had a tendency to corrupt itself on my computer, so after the third or fourth time of starting for scratch, I decided to look at a different distro. One of my brother in-laws who also uses Linux as his desktop OS had tried out a distro called Gentoo. I checked it out and have been using it ever since.
Gentoo is an interesting distro with an interesting philosophy: everything should be built from source on your machine. The upside is that everything runs fast. The downside is that it can sometimes take a long time to compile packages depending upon your processor speed.
The package manager, Portage, is top notch. It does a great job at managing dependencies and performing the builds through a glorified scripting system called ebuilds. Gentoo can be kind of heady, though. There is a ton of documentation and a really great community surrounding it, but it’s still not as simple as other distros to get things to work. You’re routinely editing scripts and config files to make things work.
I don’t mind it. I’ve currently got three machines running Gentoo (soon to be four). It runs really well on low-memory and low-processing speed machines.
So, what’s the point? The point is that as a Java developer there’s no need to be tied down to the Microsoft world view. Java works just as well on Linux as on Windows. Eclipse runs great on Linux. (If you really want to use it, WSAD is available also)
Developing on a Linux box is so much snappier than on a Windows box. (Disclaimer: I do not run a full Desktop Environment such as Gnome or KDE)
The nice thing is that if you absolutely need Windows for something, you can always run it inside VMware. VMware Server is now a free product. It works spectacularly well.
Give Linux a shot. The more people try it out, the better the OS and the various desktop environments will get. This is critical to give Microsoft and Apple a bit of competition in their arenas.
Gentoo: http://www.gentoo.org
VMware:http://www.vmware.com
Filed in: Team Member Blog linux