Nick Bradbury

Dvorak is wrong about Internet Explorer

In his post The Great Microsoft Blunder, John Dvorak claims that Microsoft made a huge mistake by linking Internet Explorer to Windows. While I agree with much of what he says (Microsoft has taken an enormous number of hits due to Internet Explorer), as a Windows developer, I have to respectfully disagree with his conclusion:

"Microsoft should pull the browser out of the OS and discontinue all IE development immediately...People will not stop buying Microsoft Windows if there is no built-in browser."

Now, I'll be the first to admit that the way Microsoft embedded the browser into Windows was sloppy. I still wince when I think about the support problems I had with HomeSite that were caused by new versions of Internet Explorer updating important system files such as comctl32.dll. And don't even get me started on the anti-competitive ways in which Microsoft negotiated OEM deals regarding Internet Explorer, or the poor design decisions that have caused so many of the security problems that Windows users face. So, if we're talking about how the browser was embedded in Windows, I agree that Microsoft blundered. But if we're talking about whether embedding a browser in an OS is good idea, I say that it is (and Apple says it is, too).

What Dvorak ignores is the huge number of Windows applications that have benefited from the ability to embed a web browser. Microsoft has done a great job making it easy for developers to host Internet Explorer in their software, and this has been a good thing for customers. Think of all the software that relies on an embedded IE - not just commercial web authoring tools, feed readers, email clients, etc., but also the thousands of in-house applications that need to display web pages. This isn't a minor point: millions of people rely on software that requires an embedded web browser, and in this regard, these people benefit from having the browser included in their OS.

(And yes, I know that theoretically you can embed the Mozilla engine in Windows software. The problem is, embedding Mozilla is a pain in the ass. I speak from experience: TopStyle provides the option to view your work in either an embedded IE or Mozilla (or both at the same time), and despite the fact that the people who use the embedded Mozilla are in the minority, we've had far more support problems with it than we've had with IE.)

If Microsoft hadn't included Internet Explorer with Windows, developers such as myself would have had a much tougher time providing customers an integrated browsing experience. And, of course, given that Microsoft has bundled IE with Windows for such a long time, there's no way they could stop this practice without breaking scores of applications that depend on it.

Dealing with Connectivity Issues in Desktop Applications

One of the more frustrating challenges when designing a desktop application that connects to the Internet is figuring out how to deal with connectivity issues caused by firewalls, proxy servers and server outages. Firewalls can be particularly hard to deal with because:

• Almost everyone is using one
• It's hard to tell when your application is being blocked by a firewall
• Most firewalls are too techie for new users to configure

A common problem that we deal with is customers who upgrade to the latest version of FeedDemon, then contact support asking why the new version can't connect to the Internet. In virtually every case, the customer is using a firewall that permitted the previous version to connect, but it silently blocks the new version since the EXE has changed. We've seen this happen with several different software-based firewalls, and each time the solution has been to remove FeedDemon from the firewall's list of permitted applications so that it prompts the user whether to permit FeedDemon to connect the next time it's run.

And as we discovered last week, when your application relies on a server-side API, it has to be able to deal with the server being unavailable without significantly impacting the customer. This was something FeedDemon 2.0 failed to do, and I have to take the blame for this. Because of my poor design, synchronized feeds couldn't be updated while our server was down, and to make matters worse, FeedDemon kept displaying a "synchronization service unavailable" message every time it tried to connect - so not only could you not get new content, but you were also bombarded with error messages you could do nothing about.

Because of this experience, I've been working feverishly on changing how FeedDemon handles the rare situation that our server is down. I don't think I've ever shared a design document before, but since so many customers were (rightfully) unhappy with how FeedDemon dealt with last week's outage, I'd like to share the details of what steps are being taken to improve this situation in the next build of FeedDemon:

Connectivity Improvements in FeedDemon 2.0.0.22 (DRAFT)

In a nutshell, FeedDemon will attempt to detect when a connection failure is due to a problem with our server, and when this happens it will provide a simple way to temporarily disable synchronization so that synchronized feeds are retrieved from their source rather than through NewsGator. This will make updating feeds quite a bit slower (since synched feeds can be updated much faster than non-synched feeds), but it does at least enable reading new content regardless of the state of our server. While synchronization is disabled, FeedDemon will cache items you read so that their read state can be synchronized with NewsGator once synchronization is re-enabled.

This design document - which is still in the draft stage - also covers how FeedDemon will better handle connectivity issues caused by firewalls and proxy servers.

NewsGator Outage

Yesterday NewsGator suffered an outage which lasted several hours, and as frustrating as this was for us, I know it was even more frustrating for customers who rely on our service to keep up with their feeds.

A service interruption is never a good thing, but I was pleased to see how NewsGator CTO Greg Reinacker handled the situation. I've seen some companies suffer a similar outage yet say nothing about it. Rather than take that route, Greg posted details of the outage, and he also had this to say:

"We screwed up. To our customers who were affected - you have my personal apology, and my assurance that we're doing everything we can to ensure this doesn't happen again."

I'd like to add my personal apology to FeedDemon customers who experienced the outage. In addition to the work our server team is performing to ensure this doesn't happen again, we're also working on ways to lessen the impact on our client products if/when connectivity issues rear their ugly head.

Goober Quail and the Unfortunate Toenail

Even though we've been married for over a decade, my wife Carolyn still tells me stories about her childhood that I never heard before. Carolyn grew up in a small town in Tennessee which had its fair share of unusual characters - including, as I recently discovered, her grandfather.

My wife fondly remembers her grandfather as a grouchy old fellow who looked like Archie Bunker with less hair. Despite his grouchiness, he retained a soft spot for two things in his life: his granddaughter, and his pet quail (yes, I said quail, as in the plump little bird that some people consider food).

The quail was affectionately named Goober, and it followed its master around the house, happily chirping while it hopped to and fro. Goober was a good quail, and by all accounts he lived a happy life...until his unusual and unexpected demise.

You see, one night Carolyn's grandfather started clipping his toenails on the arm of his sofa, with the ever-faithful Goober at his feet. Thinking it was a tasty treat, Goober grabbed a toenail clipping and hopped away with it in his mouth. He tried valiantly to swallow it, but sadly, it was too large for the poor bird, and it got lodged in his throat. Before anyone could figure out how to save the struggling quail, Goober choked to death on the toenail, an unfortunate victim of hygiene.

After Carolyn told me this story, I sat back and marveled at the fact that I ever attempted a career in cartooning. How on earth could my silly little cartoons have ever competed with the real world, where cranky curmudgeons can own cute quails named "Goober" that meet their untimely end from a toenail? The real world may be more tragic than my comic strip, but it's also a hell of a lot funnier.

15 Pounds to Go

Last week I announced that I was going to lose 20 pounds of blubber. This morning I weighed in and found that I'd dropped five pounds already - a nice start! Of course, the first week of eating healthy usually results in losing more weight than subsequent weeks, so I'm not about to get my hopes up that I'll continue to lose at this rate. But if I can drop two pounds a week for 10 weeks, I'll be happy.

Overall this first week of weight loss has been fairly easy, with the notable exception of yesterday - Easter Sunday. It was a challenge getting through the day without giving in to temptation and gorging on my kids' Easter candy, but I survived by remembering that I'd blogged about my planned weight loss. It's easy to cheat on your diet when you're the only one who knows that you're dieting, but it's not so easy when you've told a bunch of people.

So by reading this, you're actually helping me stick to my goals. And the nice part for you is that you don't have to put up with me during meals, where I guarantee that I'd lustfully eye your food or give you that preachy "I can't believe you're eating that crap" look despite having eaten buckets of that same food before starting my diet.

Born to Code, Part II

At the end of part I of this post, I had given up computers to try my hand at cartooning. A short version of the rest of the story may be found in this post, but here's the long version...

I survived high school by doodling during many of my classes, and I always wanted to create my own comic strip. So I finally gave it a try. My daily comic strip Dexter appeared in the student newspaper at the University of Tennessee from 1989-1991, and I had a blast doing it. The strip won several awards during its short run, so I had high hopes of turning it into a career.

As luck would have it, my first job out of college was as an editorial cartoonist for a local newspaper. It didn't last long (the paper went belly-up), but it did enable me to hone my skills a bit. At the same time, I submitted "Dexter" to various comic strip syndicates, hoping it would catch the eye of an editor.

It didn't.

After months of submitting my work, I received nothing more than a pile of rejection notices. Realizing that it might be a while before my comic strip career took off, I found a temporary job to pay the bills. And that was a lucky move for me, since the temp job had me surrounded by "modern" PCs - and this was 1993, right when the web started to take off. It struck me that the rise of the web meant that I didn't really need a publisher for my work. Why submit my comics to syndicates when I could publish them myself?

I decided to jump back into the computing world to figure this web thing out. Most nights I stayed after work to teach myself how to use a PC, and I remember convincing my supervisor to upgrade from DOS to Windows 3.1 so that I could experiment with Windows programming. I also remember that during the day, I had to use a really badly-designed data entry program - it was an awful, painful, incredibly user-unfriendly piece of software, the kind that you write not to help people but to punish them. To this very day I'm fueled by the memory of that exceptionally bad software and the effect that it had on those forced to use it.

Anyway, I learned enough HTML to create a bare-bones site for "Dexter," and I still remember the thrill of seeing the site in Mosaic for the first time. Cliché as it may sound, that was a life-altering event for me. Self-publishing - which used to be an expensive, time-consuming process - was now cheap and easy.

Every week I'd post a new cartoon (actually, they weren't really new - they were the same ones I created in college), and the strip gained a decent following. But a few months into it, I started running into problems maintaining the site. Notepad just didn't cut it, and none of the HTML editors I tried did what I needed, so I decided to write an HTML editor that did what I wanted it to.

Now, I'd like to report that I immediately created a useful tool, but the truth is that my first attempt was pretty lame - it was an ugly, slow application that was only a cut above Notepad. But I kept at it, and soon felt confident enough in my programming skills to leave my temp job and seek a full-time programming position. I'll skip the details of the various projects I worked on, but suffice to say, I didn't really enjoy them (think "Dilbert" without the humor). It bugged me how I'd be handed a list of requirements for a specific application, yet I never got to talk with the people who would actually use the finished software - I knew my work didn't meet the needs of the people using it, which bothered the hell out of me. I decided that if I was going to be a career programmer, I should at least like my work, and I should feel that I was spending my time doing something useful.

With that in the background of my thoughts, I started over with my HTML editor project, rewriting it in Delphi to make it faster and more compact. One night, I decided on a whim to make my HTML editor available for download from my comic strip site, and I let a couple of shareware sites know about it before going to sleep. When I awoke the next morning, I was shocked to find that over a thousand people had downloaded it.

The HTML editor, of course, was HomeSite - and it turned out to be far more popular than my comic strips. Almost immediately, HomeSite seemed to be everywhere. I was completely unprepared for the success of HomeSite, and I remember it being an exciting but very sleep-deprived time for me. I had a full-time programming job and a comic strip site to maintain, and suddenly I was receiving a couple hundred emails a day about a program I'd written for myself.

I couldn't work and support HomeSite for free, so I decided to create a for-pay version of HomeSite, hoping I could use it as a way out of my day job. My memory is a little hazy about the numbers, but I think it sold something like 200 copies the first day it went on sale - more than enough to justify quitting my job and jumping head-first into the world of shareware.

But despite being focused on HomeSite full-time, I still found it impossible to handle the amount of email I received. Email just didn't cut it for tech support, so I switched to web-based forums and soon a lively community of HTML coders sprung up. That, to me, was great - unlike my previous programming jobs, people who used my software were providing direct feedback and discussing features that would make their lives easier. Before long I was conducting "feature votes" where new features suggested by customers would be voted on by other forum members, and the results determined what to add to the next version of HomeSite. That experience taught me the importance of involving customers in the evolution of a product, and ever since then I've provided a public forum for those who use of my software.

It was still too much work, though. I was overwhelmed with the demand for HomeSite, and knew I'd need to either hire people or seek acquisition from a company that could support it. I still wanted to code full-time, so I took the latter route. The first suitor was Microsoft, who wanted to acquire HomeSite and have me work on their Visual InterDev project. My visit to Microsoft was a good one - I liked the people I met, and I liked how they worked in small teams despite being a huge company. I also liked Seattle, but my wife's father had become very ill, and we didn't want to move hundreds of miles away from him.

Before I could accept or decline Microsoft's offer, a Dot-Com startup named Allaire Corporation made a counter-offer which would enable me to work from home. I met with Jeremy Allaire, and soon afterwards I accepted their offer. HomeSite was re-branded as an Allaire product, and for the next 18 months I was an Allaire employee.

OK, I just took a breather and I'm now realizing how long this post is - talkative little geek, aren't I? I'd say this is a good place for me to stop, since the rest of the story is fairly recent history which I can write about when I'm older and wiser :)

PC Magazine Reviews FeedDemon 2.0

Wow - PC Magazine just reviewed FeedDemon 2.0, and they love it! The review calls FeedDemon 2.0 "the best desktop RSS aggregator that money can buy," and says it's "the most comprehensive, feature-rich, and intuitively organized RSS feed aggregator/reader for Windows."

Born to Code, Part I

In the comments to a previous post it was suggested that I blog about how I became a programmer. I'm not sure I can make that an interesting story, but if nothing else it might be a fun nostalgic trip.

My introduction to computers came when I was a 15-year-old extreme underachiever. I absolutely hated school, which I'm sure frustrated my parents to no end. One day, completely unannounced, my dad came home with a brand new TRS-80 Color Computer, hoping to spark some interest from his brainy-but-disenfranchised son.

And it worked - I was transformed, an insta-geek. I spent hours learning how to write silly little programs in BASIC, and within a year I had graduated to creating full-fledged games.

The first game I remember writing was called "Possum Run," a bloody takeoff on Frogger which was inspired by the squished opossums that decorated the roads near my Tennessee home. My friends thought it was a fun game, so I toned down the gore a bit and submitted the source code along with a short article to a magazine called Hot CoCo - and much to my surprise, they accepted it. I was 16 at the time, and I remember being thrilled at seeing my work published.

Of course, being a teenager, I also used my programming skills to play pranks. I had a lot of nerdy fun writing a fake "personality test" which I asked people to try. At first the program would compliment the user, telling them that they had desirable personality traits. But before long the program would start questioning their integrity, eventually degrading into hurling crude insults their way. I also wrote a war simulation which I used to trick a friend into believing that I'd hacked into a military computer and had accidentally launched a few missiles (let's just say that he wasn't the brightest of my friends).

I was published a few more times before graduating high school, but I had bigger plans. I decided I wanted to write games for a living, so I spent some time learning assembly language in order to make my games run as fast as the ones in the arcade. Assembler is an ugly and cryptic language, but I loved writing it - it felt like second nature to me. Although I no longer program in assembler, learning it way back then has proven enormously helpful throughout my programming career because it forced me to deal with how computers really think.

Anyway, at 18 I formed my first company, "D & N Software" (I was the N, my dad was the D). D & N Software sold two "professional" games of mine: a Moon Patrol clone called "Moon Runner," and an original action/adventure game called "Legend Quest." Despite receiving very nice reviews, they earned just a few hundred dollars - not enough to warrant the months of effort involved in their creation, and also not enough to enable dumping my aging TRS-80 and upgrading to the new world of PCs.

While all this was going on, I had also entered the University of Missouri-Rolla, hoping to receive a degree in computer science. Let's just say that I didn't do so well - I was more inspired by the parties than I was by the classes and assignments. Embarrassing as it is to admit, the university "asked me to consider another career" (that's right, I flunked out of computer school).

Truth is, I was losing interest in computers anyway, and strange as it may sound, my real love was cartooning. To make a long story short, I switched schools, changed my major to journalism, and became a cartoonist for the student newspaper. My last game - a platformer called "Penguin Icarus" - was published in 1990 by Rainbow Magazine (a popular Color Computer magazine at the time), and it was several years before I returned to programming.

And that seems like a good cutoff point for this post. I'm guessing some long-time readers already know how I got back into programming after an attempt at cartooning, but for those who are interested, I'll supply the rest of the story in part II.

Niall Goes to Microsoft

Niall Kennedy just announced that he's joining Microsoft and is tasked with helping build Microsoft's RSS platform. This is a good move for Niall, and an even better move for Microsoft. Niall has the ear of everyone in the syndication community - and he's a trusted participant within that community - so by hiring him, Microsoft's voice in the conversation has gained more authority. And I'm personally excited to see Niall make this move, since he's obviously someone who is motivated by the sheer joy of building something useful. Now he has a chance to tinker and geek-out on a much larger scale :)

As an aside, I've heard various comments from journalists and bloggers suggesting that by creating an RSS platform, Microsoft hopes to put RSS aggregators out of business. While I shared that concern when I first heard of Microsoft's plans, I don't believe this is true. If you look at their RSS API, it's clear that a major goal is to simplify development of RSS-enabled Windows applications, and I believe they want to entice developers to stick with Windows.

Had I not already written code for FeedDemon that provides the same functionality as Microsoft's API, I'd consider using it. As simple as RSS may be, creating a working RSS aggregator is far from simple - there's a lot more to it than just parsing and displaying XML. Of course, Microsoft's RSS platform isn't even at version 1.0 yet, and it is missing some important features, but I'm confident it will become much more useful a version or two down the road (the fact that IE7 relies on the RSS API without resorting to "undocumented" functions says to me that they've got to get it right).

But this doesn't mean I've got my head in the sand. It is scary for small developers who see bigger players like Microsoft, Apple, Google and Yahoo getting into their space - especially when some of those bigger players have a track record of "embrace and extend" (followed by "destroy") - and I fully expect Microsoft to further RSS-enable many of their applications. For some users, these applications will provide enough functionality to make third-party RSS aggregators unnecessary. At the same time, though, it will also broaden the audience for RSS, providing a larger base of customers for syndication companies that stay on their toes and remain innovative.

Fat Tuesday

Note to Chris Pirillo: you're not the only one who has gained weight. I usually keep myself fairly fit, but I used last year's surgery as an excuse to eat whatever I wanted (and boy, did I ever eat what I wanted).

Sooo...I sit here with twenty pounds (!) of extra blubber, and today is the day I start making it disappear. This isn't the first time I've lost that much weight, but I plan to make it my last. Truth is, as much as I enjoy bad food, life is simply better when I take care of myself - I feel better, I think better, I sleep better, and yes, sex is better. Trading all that for the short-term pleasure of pizza and beer (mmmm...pizza and beer) just isn't worth it.

So do I plan to do it by following some trendy diet? Nope. I'm just going to eat less, eat better and exercise more (that plan may not sell many books or videos, but it does have the benefit of actually working).

I'll blog about my progress along the way, but for now, I just wanted to publicly state my goal to force myself to stick to it - after all, I've just told a whole bunch of people that I'm going to lose weight, so it would be embarrassing to fail.