RobbieAllen.com > Blog

Are you familiar with the “Work Less” mantra that has gained popularity in recent years from the people at 37Signals and from books like 4 Hour Work Week? It is the idea that by working fewer hours in a week, you’ll be more productive while you are working. Employees at 37Signals only work 32 hours a week (8 hours x 4 days). They take Friday off because “no real work gets done on a Friday afternoon” according to Jason Fried, 37Signals CEO. I’m a big fan of 37Signals, have read Getting Real, and even attended one of their workshops in Chicago. Likewise, I think 4 Hour Work Week and Tim Ferriss have some interesting ideas. However, on this topic, I don’t agree.

Why does Work Less work for them? Because they don’t love their work. It seems as if “work” is a dirty word. It is something they have to do and therefore want to minimize it as much as possible. Like most people, they work to live, not live to work.

If you love your work, 32 hours a week will not come close to squelching the insatiable desire to do it. Maybe the Work Less people find their lives fulfilling and don’t mind trying to minimize work as much as possible so they can do other stuff, but if not, they should instead try to find work they can’t live without. This idea may sound very odd to some people, but that is only because you’ve come to associate doing “work” as something that is not enjoyable, rewarding or downright fun.

When I was writing books from 2001-2006, I was immersed in it. During that time I liked to say that I wrote books to “become an expert not because I was an expert”. I wrote or co-wrote 10 books in 5 years and edited a dozen others (in addition to my “day-job”). People are often surprised by that fact, but I was completely enamored with writing at the time. There was no better way to learn a topic and I thoroughly enjoyed the book publishing process (warts and all). So much so that I became an editor for a time with O’Reilly and the Pragmatic Bookshelf. But going into my 6th year, I grew tired of it and didn’t think I had much more to contribute.

Now, I’m working on StatSheet.com. It is the most fun I’ve ever had coding going back to 1995. I’ve gone from nothing to a website that has over 1.5 million pages of dynamically generated sports stats covering college and pro football and basketball in a little over a year. And I just released a new site based on that work called StatFix.com.

I couldn’t possibly write the books I did or make as much progress on StatSheet has I have by “working less”. I’ve made as much progress as I have because of the fact that I enjoy “working more”. I “work” 7 days a week, 365 days a year. I work after my day job, late at night, early morning, on weekends, and any free chance I get.

One of my favorite books is On Writing by Stephen King. I’ve listened to it probably a dozen times. He has a similar work ethic. “I write 365 days including Christmas, workaholic dweeb or not.” He said he feels the most unease at moments when he isn’t working on a book at all. That’s what I’m talking about here. When I’m working on something I love, the feeling drags me out of bed early every morning and makes me stay up past midnight just about every night.

For the people that only work 4 days a week or try to minimize your work time as much as possible, what do you do all the other time? I listen to ~25 books a year (I’m a big audiobook fan), I don’t like TV much, do my weekly chores around the house, and I play basketball at the YMCA every chance I get. I have a supportive wife that I’ve known since high school and a 17 month old baby that enjoy immensely, and despite how much I work, I bet I spend more time with them than 90% of men spend with their families (I only go into an office 3 days a week). My wife and I travelled a ton before the baby and don’t have much a desire to do so now.

I’m sure I could fill my time with something else, but fortunately I don’t need to.

One of the few shows I watch on TV is Grey’s Anatomy. The scripts are usually entertaining and the cast does a good job. However, over the past couple of years they have steadily been getting more graphic with the episodes. It has gotten so bad that tonight they showed a guy’s scalp getting cut back and pulled over his face. It was completely disgusting. It is not uncommon for episodes to show blood squirting everywhere, things sticking out of limbs, or other very visually graphic scenes.

My question: Is this really necessary? Are people really tuning in to watch that? Can’t people just watch Discovery Health or something to get gore? I can understand some basic things. I can handle open heart surgeries, deep gashes, and the occasional impalement, but not scalps being pulled back.

The people are the show. The hospital is just the backdrop. Adding all the gore just takes away from one of the few shows on TV worth watching.

I continue to get a few emails asking what I’ve done since I reported problems with my web hosting provider Joyent. After quite a bit of research, I narrowed down my options to Engine Yard and SliceHost. Both companies have pretty good reputations in the Rails community. Engine Yard is a high end hosting provider and their price reflects it. They provide hands on support including help with initial configuration and installation in addition to ongoing support. If money was not an issue, I would have definitely gone with EY, but it is. SliceHost on the other hand offers more of a DIY model. They are similar to Joyent in that regard, but they are 1) less expensive, 2) offer a variety of UNIX platforms to choose from (Joyent uses Solaris) and 3) have much better documentation.

I picked SliceHost and have been a happy customer so far.

My main problem with Joyent was with IO. There would be periods where the shared storage would slow to a complete crawl. And to make it worse, Joyent had difficulty in telling me the root cause (often pointing the finger at my app). Since I moved to SliceHost, I’ve had zero IO issues. In fact, I don’t have any issues to report. I’m running Ubuntu, mongrel, and nginx and my new layered architecture will allow me to scale easier in the future.

I’ve been very interested in Amazon’s cloud computing initiative. They’ve slowly been adding services to the point where some startups are able to use Amazon for storage, computing, and even database services. With StatSheet, I needed a simple content delivery network to host images, javascript, and stylesheets. S3 seemed like a good fit. I heard a few reports of S3 being unreliable when used as a CDN, but several bloggers I read seemed reasonably happy with them. I decided to take the plunge, change my some code, and start using S3 for static assets.

After a month I’ve decided to switch everything back to my own web servers. Why? Well of course there was the well publicized outage a couple of weeks ago, but the bigger problem for me is that periodically files would take an extremely long time to download. Occasionally I’d notice the site acting really sluggish. At least 3 times in the past month when that kind of slowness occurred, it was because an important file (like a javascript file needed for onLoad) is taking FOREVER to download. And by forever I mean on the order or 2-3 minutes! For just one file!

Below is a snap shot from Firebug that shows a 194KB javascript file being served from S3 taking 2.2 minutes to download! I could reload (with caching turned off) and it would take 2-3 minutes again. Each time I tried it.

Now eventually the file will start downloading fast again, but for an important file like that (or my main logo image), it dramatically alters the site’s appearance. Now I know there is a slight chance that it wasn’t S3 at all. Perhaps a network issue somewhere between me and S3. The catch is that all 3 times I’ve seen this issue happen, I can download the file with no problem directly from my web servers. I just don’t want to take the chance.

So I’m giving up on my S3 experiment for now. I’ll still use it for backups, perhaps, but as a CDN it just didn’t work well.

As I mentioned last year, I’ve stopped writing books and articles to pursue a new passion. Being the author or co-author for 10 books in 5 years can wear you out on writing ;-) However, an essay I wrote on blogging while at MIT has made it into a book that was just released. The title is Current Controversies: Blogs. The book is a collection of essays covering different aspects of blogging. I haven’t received my comp copy yet, but the table of contents looks pretty good.

For the first time in my life I’m starting to feel “old” :-) Having a child unlike anything else really makes you see how fast time flies.

But I wouldn’t trade this for anything:

That’s right. I’ve switched to a Mac. I bought a MacBook Pro a couple of years ago, but only used it sparingly. Now, I’ve turned in my Microsoft Club Card (btw, MS didn’t renew my MVP status this year and for good reason). There is no going back…at least for a while. There has been some initial discomfort during the move, but I’m getting used to OS X. The honeymoon period is already over though. I’ve experienced many application crashes. Firefox is still a pig. I’m going to give Safari a try. Because I’m still hooked on Outlook, I’ve been running VMWare Fusion. It is pretty good–especially Unity mode–but at times it slows down the system and doesn’t always suspend correctly.

Because I’ve been using Windows for so long as my primary desktop, it feels good to try something new. I have a theory that becoming too familiar with something is a bad thing. When you get too comfortable, you miss out on stuff. You never learn there is a better way to do something. Perhaps partly because I’ve written a bunch of books on Windows, I became too comfortable with it. There were no more surprises (besides the different ways it could crash on me).

Don’t get me wrong. I really really wish we didn’t even have to talk about which operating system I’m running. Modern desktop OSes are still too heavy handed. They should do all the heavy lifting behind the covers and get out of the way. I wish they were so dumb that no one needed to write books on them. But that’s not how it is today. Windows and OS X are big complicated pieces of software. With all the recent problems I’ve had with my desktop environment, I wonder if some higher power is pushing me to take on the cause of “thin” clients (for lack of a better term) because ultimately I still believe that’s where we are headed.

I’m both a BingoDisk (online storage) and Accelerator (virtual private host) customer with Joyent. I’ve only been marginally satisfied with BingoDisk to date (WebDAV access is flaky to say the least), and my Accelerators are decent but have experienced quite a few IO problems from being on shared storage.

As of last Saturday, BingoDisk and Strongspace have been offline. This wouldn’t normally be an issue for me because I access my BingoDisk infrequently, but it just so happens that my Lenovo T60p was starting to act up. I could sense a crash was coming. Unfortunately I couldn’t upload my latest files and pictures to BingoDisk (my online backup location). I figured they would get it back up any day. Well yesterday my system crashed and I spent a whole day trying to salvage the files I could. I lost data. Of course I’m not blaming Joyent for my laptop crashing, but it really sucks that when I needed BingoDisk the most it wasn’t there for me.

It looks like the problems aren’t just limited to BingoDisk and Strongspace. While working on another issue with Joyent, I stumbled upon this:

Ouch, that means their public site (at least one of them) is also down. The reason I went to their website is because one of my Accelerators has been experiencing IO problems all day. I have two Accelerators, one for dev and one for prod. The problem is with my prod Accelerator. Joyent support suggested I optimize my app. Eh. The problem sprang up out of no where last night. iostat shows an obvious bottleneck on the storage side. I asked if this was related to all the data being restored from BingoDisk and the support guy said it was unrelated. The only option they gave me (besides optimizing my app) was to move me to another host and take 2 hours of downtime. Ugh.

To summarize:
1) BingoDisk down for almost a week
2) Accelerator experiencing IO problems and only solution is to move me to another web hosting provider.
3) Joyent’s own home page is down

Should I consider switching to another provider? Any recommendations?

Personally, 2007 will be a hard year to beat. I saw the birth of my daughter and have been enjoying her ever since. Professionally, it has been a mixed bag, with some things going well and other things not so well. Obviously the new baby had an impact on my priorities.

Below are my highlights (and lowlights) for 2007:

• My daughter was born :-)
• I got promoted at work
• For the first time in over 5 years, I did not write a new book
• I stopped editing books too! In fact, I don’t plan on doing anything related to writing or editing books for a while.
• I started a new pursuit
• For the first time in several years, I did not speak at or even attend a conference, go out of the country on business or pleasure, or play golf at least once a month.

Biggest highlight of the year? My pumpkin’s first Halloween:

About a month ago, I described how I was fed up with Firefox memory problems and instability so I decided to try out IE7. Well I’m unhappy to report that my experience with IE hasn’t been much better. Yes, it also bloats up to several hundred MBs and over a period of a few days my whole system starts to exhibit memory leak behavior. Part of the problem is that I’m running Windows Vista. I should have just stayed with XP. Perhaps I should “upgrade” back to XP. Or better yet, maybe I should put down my Lenovo and dust off my Mac.

It amazes me that in my 13 years of using the Windows operating system, stability is still a huge issue. Sure the UI has improved and the apps are more feature rich, but we’ve sacrified stability. I’m about fed up with it. The periodic need to reboot, memory leaks, SYSTEM process going to 100%, inability to handle multiple monitors well, video driver crashing after undocking, etc., etc.

All of this just furthers my belief that “thin web-based clients” are the future.

I stumbled upon an interesting post on Brazen Careerist titled “Twentysomething: Why I regret getting straight A’s in college”. I too was obsessed with getting straight A’s in college (undergraduate). I graduated with all A’s except 1 measly B. However, I don’t regret it. I don’t disagree with any of the main points made in the Brazen Careerist post, but I do disagree with the conclusions. Let me address each point and explain why.

1. No one has ever asked about my GPA.
Same here, but then again I didn’t do it so I could brag that I was a straight A student. In fact, I didn’t announce it to anyone. It shows up on my resume, but I don’t think it has done anything for me as far as being perceived as a “straight A student”.

2. I didn’t sleep
Me too! For a whole semester I only slept 3 hrs per day (using the Polyphasic method). But you know what? This instilled a great work ethic, which I carried over into my professional career (ie, cube sleeping). Not sleeping much doesn’t work for some people, but it does for me. Even now I need a max of 6 hours per night.

3. I’ve forgotten 95% of it.
Same for me. My brain is definitely wired to “use it or lose it.” That means I forget stuff relatively easily if I don’t recall it periodically. Most of the stuff I learned in college I haven’t had to recall since then, so it is lost in the nether regions of my brain. But is that really a reason not take classes (or read a book)? If you applied that same logic further, you should never take any classes (or read a book) because you’ll likely forget most of it. I believe learning (and brain development) is a cumulative process. The important point isn’t that I’ll remember some obscure detail, but connections are made in my brain that enable to me learn other things more quickly or to gain insights not possible without the prior learnings.

4. I didn’t have time for people.
Me either. But that has more to do with the fact that I’m by nature an introvert. Can you make a lot of friends and still make straight A’s? Absolutely! Study groups, clubs, intra-murals, etc. are all options. And you can still attend social events, but not every night. Would you want to do that anyway? Sounds like a pathway to alcoholism rather than making a bunch of friends.

5. Work experience is more valuable.
Work experience is critical. Your work experience will definitely provide more possibilities for career advancement than getting a 4.0 over a 3.0. That said, I don’t see the two as mutually exclusive. I did both. I had two internships at high tech companies while I was in college. I didn’t have to do any classwork during my internships and yet I still got some class credit. It was the best of both worlds and I highly recommend it to any college student.

What about Graduate School?
A big reason I wanted to do well in college was so I could attend MIT for grad school. It wasn’t until many years later that I realized that goal, but knowing what it took to push myself and do well in an academic setting was extremely helpful to perform well in a top school like MIT. (i’ll save the benefits of going to grad school for a later post.)

In summary, even with many years of experience under my belt and knowing what I know about the “real world”, I’d still recommend to an 18 year old Robbie to do his best in college. For type-A personalities like myself, it is very difficult to turn my passion for achievement off and on like the flip of a switch. I’m concerned that if 18 year old Robbie started intentionally slacking early in college, it would have continued into my professional career. I would have regretted not giving it my all.

I’ve been a long time Firefox fan. I’ve been using it since the early betas, but I’m so fed up with it now that I’m officially switching to IE 7 (and this despite the fact that I don’t like the IE 7 UI). The main reason for my dissatisfaction? 1) Memory consumption, 2) CPU utilization.

Microsoft gets a lot of flak for writing bloated software, but Firefox is the biggest consumer of memory on my system. A distant second is Outlook. It is not uncommon for Firefox to consume 500MB-700MB (Private Working Set) on my T60p (2GB) system. I may only have 4 or 5 Firefox windows open with 4 or 5 tabs open each, but Firefox will get so slow and unresponsive that I have to kill it. Fortunately, the “Restore Session” feature lets me restore my windows, but it won’t be too long before the memory usage goes back up and the CPU starts spiking.

Maybe I could blame this on my specific laptop, but I’ve had the exact same problems for a while now even on previous laptops. There has been quite a bit of press on the Firefox memory leaks and I know the Firefox team is working on it, but I just can’t deal with it anymore. Since much of my day is spent staring into a browser, I’d rather sacrifice a little usability (IE) than to sacrifice on performance (Firefox). Perhaps I’ll try it again after the next major release.

I’m a college basketball fanatic. I’ve been a North Carolina Tar Heels fan for as long as I can remember. Last year when I started thinking about interesting websites I could build, college basketball bubbled up as a potential topic. Currently, most of the big sports websites do a rather poor job of catering to fans that have a hunger for stats. Most existing sites use the same text-based, current season only stats as everyone else. We’re taking a fresh approach at StatSheet.com. Since I’m such a big college basketball fan (and blogger), I figured we’d develop a website that catered to my needs to start with.

We’re planning the first official release during Midnight Madness. This will include all the base features to view historical stats, compare players, coaches, teams, and conferences, graphical charts which can be embedded in your website, and much more. I’m announcing it now because we have enough completed to start getting feedback. Some pages will be rather empty, that’s ok, we’re still working on it. But we are interested in any general feedback about usability, functionality, and features you’d like to see added.

After the first release we plan a follow on release in time for the beginning of the season. This will include features that allow users of the site to contribute their insights and knowledge. That’s all I’ll say for now!

It has been a blast developing the site so far. Stay tuned for more!

The second edition of the successful Pragmatic Unit Testing book is heading to bookstores. This is the first paper book I edited for the Prags. Matt Hargett was responsible for updating the content in the second edition.

If you are interested in Unit Testing, especially using NUnit, this book will be extremely helpful. I’m not a C# person, but the discussion on Unit Testing practices along with the examples described in the book were informative.

One of the books I’m editing at the Pragmatic Bookshelf is Deploying Rails Applications. It just went beta this week so you can get your hands on an early version. It is a fantastic book and will prove very helpful to anyone that has to deploy Rails applications. I hadn’t made the leap to Capistrano yet, and this book guided the way. Good work Ezra and Bruce!