Artel DeVries

Twitter Update

2009-06-14T22:53:00.001-07:00

Not all of my prophecies come true, but I think this one might be coming to fruition.

Link: Who Killed Twitter

-Artel

The Need to Integrate Parallel Programming into .NET 4.0

2009-06-07T16:40:00.000-07:00

I watched a great video on InfoQ today where Eric Nelson, who is working .NET 4.0 at Microsoft articulated on the Parallel Framework that will be shipping with said .NET release. In this video, he explained the merits of including Parallel framework as an API rather than including it as a straight-up language extension. The video can be found here.

I would have preferred that Microsoft included the Parallel extensions as part of the language as opposed to just making an API for it. This would have sent the message that parallel development is a more integral part of .NET 4.0 as opposed to some fringe feature that is cool but not necessary.

With the language extension, people would see the reserved words on Chapter 1 of their new .NET books and want to learn about parallel programming at an earlier stage of .NET 4.0 development. They would then hopefully incorporate this into their software more.
Like Eric said, CPU designers are now building horizontally with more cores as opposed to building vertically with faster CPU speeds. As we move forward in the next few years, parallel development is going to be an essential part of utilizing the plethora of cores that are coming standard on end-user machines.

Constructive feedback is always welcome.

-Artel

New Gig & Other Points of Discussion

2009-05-23T00:19:00.000-07:00

The job market for Developers in DC is quite different than the Buffalo market. In three short days, I had scheduled over 10 interviews, and had about as many tech screens. If any decent developer out there is reading this, come to Washington -- you will find work in no time.

When I decided I was leaving BearingPoint (Tuesday), I got an offer from the company I would eventually accept on that Friday. Overall I am pretty excited. The working conditions are better: I get an office instead of a cube, newer hardware, and more freedom to test out new technologies. On top of that, the money is a bit better. So, come to the Washington Metro area if you want to find a great development gig!

So you want to come to DC to find a development job, here is my advice -- take it or leave it.

1. Schedule as many interviews as possible. It turns out that most people ask the same questions for the technical part of the interview. Half the time I could tell what the next question was going to be based upon the question they were currently asking me.

2. Go to programminginterview.com and try a bunch of the questions to get your brain ready for some good problem solving exercises.

3. Besides the standard list of .NET developer questions (which I think all originated from here at one point), expect to do a programming exercise or two. If they don't ask this, I would be alarmed. All good developers should be able to write good code. If candidates don't have to do this in an interview, how do you know you will be working with other good developers? I don't think you can know that. It doesn't have to be a long question... just long enough enough to prove the person is smart and gets things done.

4. Don't be a shitty developer. That seems like trite advice, but if you are a person that cares about programming and you work to hone your craft, people conducting interviews will admire that, a lot (even if you can't write Objective-C and Lisp in your sleep).

5. Shut up about the economy. You are a Software Developer, your skills are in demand. Stay focused and you will do fine. If you were a philosophy major, or worse yet, an Art-History major, I would tell you it's a good time to panic.

For now, I am glad I am done with the whole interviewing process and can't wait to start working at American Systems.

-Artel

Left 4 Dead and Tech-Leads

2009-05-22T23:39:00.001-07:00

If any of you out there have not been playing l4d, I highly recommend you get on and play. It is made by Valve -- the same people that develop Half-Life and Counter-Strike. If you are reading this, you're probably a developer or some other type of nerd. A sweeping generalization allows me to conclude you're a gamer as well (Go buy l4d.)

The objective here is simple: kill mutant zombies and stay alive as long as possible with three other players in an online co-op mode. It turns out, the better the teamwork skills of the group are, the longer the team lasts before blood-thirsty zombies ravage your sweet, warm corpses.

If the team stays alive for 4 minutes, everyone gets a bronze medal, 7 minutes gets you silver, and 10 for gold. One of my philosophies in life is to never argue with results unless of course it involves compromises your ethics. Since, this is a video game, ethics be damned. If somebody has really good time a certain map, I will always take their advice and sit in a certain place, use a certain weapon, etc. But some of people just don't listen. When I get good time on a map I will tell people how I got it, so they can rack up good times as well. But some bull-headed people will try things their own way and they die, fast, every, time. So I just throw my arms up, get off my work-station, and walk the dog. What else can I do?

After experiences like this, I think I know experienced lead-developers feel when more junior programmers try to try something their own way and disregard the wisdom of the more senior folk on a project.

I still have some time before I fill the shoes of the tech-lead, but now I think I'm just a bit more prepared for what is out there. I'll be on the lookout for these types people, and hopefully I won't allow them on the team I am leading. Murphy's law tells me that won't happen, so I'm braced for the worst.

So... how cool of an interview would it be to play l4d as part of said interview? If people don't listen and die quickly, I'm pretty sure I would be scared to have them on my team. And... nothing beats getting paid to kill zombies.

This post just goes to show that video games are full of valuable life lessons.

You want to lead a team? Check out this book, its a fun read with lots of great advice! Managing Humans

-Artel

MVC and (no) CheckBoxLists

2009-05-16T18:01:00.000-07:00

What is up with there being no CheckBoxLists in MVC?

I shouldn't have to iterate through ListItems.

I've seen a few work arounds, but come on!

-Artel

Overdue Update

2009-04-30T17:21:00.000-07:00

I waited for Blockbuster to go belly up so I could have a glorious "I told you so" moment, but that never happened. Ohh well. I swallow my pride and forge ahead.

The buzz on the street for quite some time now has been that Software Developers have to step up their game and create some highly efficient software that takes advantage of multi-core processing, saves helpless puppies from raging infernos, and walks on water with confident grace. I equate that to inventing the paint brush, oil paint, and canvas -- then getting all pissed off because nobody has been inspired to paint the Mona Lisa or Starry Night (yet).

That is not to say that developers shouldn't spend time writing software that can be executed concurrently, but there's always just so many things that have to happen sequentially. Things like Parallel Extensions can help improve upon this, but software is hard to write and it is going to take time before we see awesome leaps.

In any regard, it doesn't really matter. Swine Flu is going to kill us all. Panic. PANIC!!!

More to come.

-Artel

Twitter Will Die... Here is why

2009-03-17T12:29:00.000-07:00

http://www.twine.com/item/123c9051b-g8/can-twitter-survive-what-is-about-to-happen-to-it

I couldn't agree more with this. Twitter will go the way of the dinosaur unless some filtering routine is implemented.

So... great, another spam filtering service to deal with.

Are tweets worth managing another spam filter... not for me.

-Artel

So long overpriced rentals

2009-03-04T21:01:00.000-08:00

So Blockbuster is most likely going the way of the Dinosaur. I would bet my non-existent first born son that Chapter 11 is less than a month away for this dying juggernaut. Is anyone surprised?

With a Business Model like Circuit City's -- offering something which online providers have done a much better job perfecting -- death is inevitable.

Like Circuit City set their prices higher and offered FAR LESS selection than Amazon.com, Newegg.com, and TigerDirect.com -- Blockbuster ain't goin' win much business offering rentals at 1/3 the normal retail price of a movie at Walmart. Their only saving grace was to offer an online service that copied Netflix's. But... they just don't stream nearly as much content as Netflix, they have higher priced plans, and have to pay to keep retail shops open.

The only competitive advantage they had was their retail stores. Then came Redbox and $1/night rentals. Score.

Consumer Reports also agrees on the relative shittiness of Blockbuster:

Consumer Reports’ survey respondents found Netflix to be the top choice, rating it higher than Blockbuster’s competing Total Access service. With monthly payment plans starting at $5, Netflix was among the standouts for price; it was top-rated for selection and service. Readers reported that orders were filled promptly and with very few errors. The company offers more than 100,000 titles, including 1,000 Blu-ray titles. Mail delivery is supplemented by a streaming service that can deliver videos to a computer or Netflix-enabled TV or Blu-ray player with an Ethernet port.

What the @#$# does this have to do with Software Development?

Are you thinking of starting an mISV? If you are, don't insult your (potential) customers' intelligence by assuming they can't do simple arithmetic. Price your Software fairly. (Yes, I know what you are thinking... this applies to all businesses.)

$800 for a new 42" TV at Newegg < $1200 for the same TV at CircuitCity. (I will also add that in the last days of the Circuit City liquidation, their prices are STILL marked up.)

Just like $1.00 for a movie at Redbox < $5.00 for the same one at Blockbuster.

I know this a shallow review of how Circuit City and Blockbuster dug their own graves, but these giants fell partially because they didn't think people had the common sense to shop comparatively. Don't you be so foolish.

I initially planned on delivering my trite advice in a much shorter post. So it goes.

-Artel

Best of the Best

2009-03-04T20:05:00.000-08:00

Here is a question I like to ask interview candidates for .NET programming positions
"If you could design an ASP.NET application any way you like, how would you do it.".

It is a really open ended question that lets me know

How people would apply lots of cool and new technologies
Whether the candidate is eager to try new and unproven technologies
(Or whether they are just not interested in discovering new things on their own at all)

However, I have never thought of how I would answer this question myself.

I know there plenty of different situations that would require the use of different technologies, but for the purpose of this post, I am developing an ASP.NET 3.5 Web Application.

Lets start from the bottom and work up the stack.

SQL Server 2008 is going to be an awesome product when it matures. (I think it is awesome already!) It is also tons cheaper than Oracle 11g with a comparible feature set.

LLBLGen 2.6 is in my opinion the best ORM solution out there. The 2.6 release features an implementation of LINQ that seems to work quite well. I haven't been on top of the LINQ implementation for NHibernate, but you still need to configure XML mapping files. I prefer the feature rich designer and crazy amount of support that comes with LLBLGen.

For logging, I just stick with log4net. It's the de facto tool for logging, why deviate? (Sorry Enterprise Library!)

If the requirements dictate it, I am all for using some of the SP1 features of 3.5 like Dynamic Data and RESTful projections offered in WCF.

If the MVC framework is good enough for StackOverflow, then I guess I can't complain.

To tie all the .NET layers together, the Windsor Container, which is part of the Mono project, is a great tool for keeping your code decoupled.

I haven't found anything saying that Infragistics or Telerik have a control suite for MVC that is ready for beta. If they release this, I would love to use one of them with JQuery. Otherwise, I would ditch MVC and use the built in AJAX framework along with a control suite for ASP.NET forms built by Telerik all wrapped up using MVP.

Unit Testing can be all be done with the built in framework that ships with Visual Studio 2008. To isolate my tests, I would recommend some sort of Mocking Framework. I have using Rhino Mocks before, and it has a steep learning curve. If you are brave, you can try NMock. I have not used it, but ThoughtWorks has a propensity to put out great tools.

TFS 2008 does a great job of helping to manage projects and source code, no need to get creative here.

I know this can turn into an ugly religious argument, but that is what I would use. Feel free to try and change my mind! =)

-Artel

So you want to work for [awesome software company], but...

2009-02-19T21:14:00.000-08:00

...you just aren't that smart.

I used to think like that. But I realized that attitude would never get me anywhere. So if you just started your professional career as a Software Developer in the past few years, you have probably stumbled upon the works of a few awesome people. Specifically, I am thinking of Joel Spolsky, Steve Atwood, and Frans Bouma.

These guys hire great programmers, and they settle for nothing less. If you want to work for them, you almost have to be a gifted pupil -- having really high SAT scores, attending prestigious universities, or probably a combination of both. If you have those qualifications and are a passionate programmer, then you can a land a job at one of those places.

However, if you are like me, then you probably didn't get a degree in Computer Science from Stanford, you didn't have SAT scores in the 1400s, and you didn't graduate at the top of a class with several hundred (or thousand) people.

All hope is not lost (I think). There are several things you can do, and they all involve learning as much as you can, whenever you can. Check out this article http://www.ericsink.com/Career_Calculus.html. It talks about this much more elegantly than I ever could.

In order to learn, this is what I am doing:

Start a blog
...get your name out there, and build a reputation. Let everyone else realize that you are serious about writing good software, and that you enjoy doing it!

Do a side project
Like someone told me, test out new technology on your own. If you work for a big company, it will take a while before they adopt things like WPF, ASP.NET MVC Framework, or F#. Not only can you learn some awesome new things, but you can learn a lot about how to deal with customers and advertisers.

Read
Like the title of this article implies, there are a lot of people out there that are way smarter then us. Learn from them! Avoid mistakes they have made, and try to get some insight on what is new and promising in the world of .NET. You will be very surprised at the new things and people you discover while explorer good tech books and blogs.

Work Hard (and be patient)
Understanding things like how to prove a Turing Machine is the same thing as a PDA using a queue as opposed to a stack is hard! It takes a lot of work and and even more patience. When others settle for a good solution, keep at it -- and deliver the best solution.

Don't be afraid of new technologies
For me, any technology that wasn't Microsoft was hard to accept. When I started using a tool called LLBLGen to write Data Access code, I was more skeptical than I probably should have been. But I got over it, and now I love that tool!

Here is where this attitude has gotten me so far
I was selected as one of the 3 out of 20 so-odd Programmers from my AP Computer Science class to attend a Programming Competition at a local college.

I graduated top of my Computer Science class in college (there were about 13 of us).

I am working with a group of great people designing, developing, and supporting Software that is actually used by many people all around the country.

To anyone who thinks they can't program because they don't have the raw intelligence to do it... you can. Work hard and learn as much as you can.

If all of this helps you (or me) to get a job with Joel, then great! But as I was just reminded, there are probably a lot of unkown Joels out there to who would totally appreciate this list of things to become a great programmer.

...so keep on learning.

Hello World

2009-02-19T21:06:00.001-08:00