Hoo boy, almost let an entire month go by without posting!
I'm editing a new book: a collection of the best software essays published anywhere -- on the web or in print -- during 2004. Please nominate essays by posting a new topic here. Please include the title, the author's name, and a URL where the essay can be found. If you've read one of the essays mentioned and want to comment on whether it belongs in the book, please reply to the topic nominating it with your comments. Full names and email addresses are required to nominate or comment. Writers are encouraged to nominate their own works... don't be embarassed! The deadline for nominations is November 20th. The best articles will be published in a book by Apress (assuming we can secure the rights).
In other news...
The whole company (and four outside consultants, phew) has been concentrating on trying to get FogBugz 4.0 out the door. Exciting stuff there: this release, I think, will really show the benefits of our policy of listening to our customers, not our competitors. If all goes well we'll be pushing out the first beta in a couple of weeks.
The goofy name FogBugz is an historical accident which I rather regret, but the brand name is so well known by now that we're not going to change it. We have subtly reduced the number of capital letters (FogBUGZ became FogBugz) which makes it a little bit less embarassing, but I think for now we're stuck with the stupid name. I have actually gotten email from people who say they can't buy FogBugz because the name doesn't sound "professional." Sorry! Get over it, or you may have to use the professionally-named "Bugzilla" forever!
It's That Time of Year Again
Once again my company Fog Creek Software is looking for A Few Good College Students to work as interns next summer. We've doubled the size of the program, from two interns to four, and we're looking for both interns in Software Development, for CS types, and Software Marketing, for MBA types.
The application deadline is February 1st.
Ipswitch's IMail customers seem to be in open revolt over a new pricing scheme. IMail is a popular email server for Windows-based servers. Ipswitch is a normally smart company in Massachusetts. This story will be interesting to follow for anyone making pricing decisions about software and trying to move upmarket.
Disclosure: We use IMail for email at Fog Creek Software. The only reason we were using IMail is for historical reasons, when we only wanted to have one public-facing server, and it had to be Windows so it could run FogBugz. Now that FogBugz is available for Linux, and, more importantly, we can afford to run as many servers as we want, IMail is a lot less compelling than Postfix or qmail, which are free.
Now, on to Ipswitch's new price strategy. Clearly, management was drooling over the Exchange/Notes market and wondered why Microsoft was getting so much more money for Exchange than they were getting for IMail, and they probably misunderstood the reason. What they were thinking was, "oh, Exchange has all these other collaboration features besides email. Let's add some collaboration features and then we can charge as much as Exchange. In fact if we even go halfway towards the price of Exchange and be an even better value than Exchange. Mo' money, mo' money, mo' money!"
First mistake: they underestimated how much more Microsoft can charge for the same junk simply because Microsoft is the market leader. In MBA terms, Microsoft can capture a larger portion of the consumer surplus because they're the true market leader, a "safe" choice, and they have a worldwide sales force about 200 times larger than the entire staff of Ipswitch. These are people who spend every minute of every day explaining to customers why Exchange is great even if it isn't. Second, Ipswitch misunderstood that Microsoft is in the market for Large Corporate Email Systems and Ipswitch is in the market for small-medium ISPs and smaller businesses without dedicated IT staffs. These are two different markets with very different willingnesses to pay.
The second mistake was a fundamental pricing blunder of misframing the value equation. The price of software has very little to do with cost, and everything to do with perceived value. The classic book about pricing is Nagle and Holden's The Strategy and Tactics of Pricing: A Guide to Profitable Decision Making, a book which is offered in hardback and paperback, with the paperback sold at a higher price than the hardback in a brilliant pricing strategy (Thomas Nagle practices what he preaches). And one interesting point in this book is that customers perception of value is always based on comparisons to some baseline. IPSwitch was hoping their ISP customers, who have never remotely considered using Exchange due to the extremely high per-user cost, would compare their new price to Exchange, but of course, what they really compared it to was the price of the old version of their product, with which it compared very unfavorably.
By the way, if you're wondering what a good example of this strategy is, look at the four editions of Microsoft Visual Studio. There's a super expensive "Enterprise Architect" edition at the top of the line that hardly anyone ever buys; it's only there to make the other prices look reasonable by comparison. Consumers don't really know how to judge the value of software so well, so they grope around for nearby products to use as baselines. Pricing 101 says you give them something really expensive to think about so your product seems cheaper by comparison and thus a better value.
If Ipswitch really wanted to compete against Exchange and Notes, first of all, they shouldn't be competing on price in that market, because IPSwitch does not have any inherent price advantage. In fact Microsoft probably has the price advantage in that market because they can distribute development costs over so many more customers and they already have a sunk cost of a huge marketing force. Never compete on price when you don't have a price advantage. If you're the small company, you don't have the price advantage.
And secondly, Ipswitch shouldn't be abandoning their existing market to chase after an elusive new market in which they're going head-on against Microsoft and IBM (Lotus Notes). They can't win that way. There's a Will and Grace episode that proves it.
My sources inform me that this was a decision made by the MBAs in the company over the strong objections of the developers. It's ironic that the MBAs don't even get the basic MBA stuff right; this is entirely a bad move for reasons that business school graduates should be able to figure out without understanding the technology at all.
Gary Cornell: “While the reason we published the essays collected in "Joel on Software" is obvious to anyone who has read anything Joel has written, the cover of the book has given rise to more than a few questions!”
I've gotten lots of great nominations for essays that deserve a place in the next book, tentatively titled “// comment”. It's going to be your book: the 30 best essays about computer software over the last year... the kinds of things that Joel on Software readers want to read, edited and introduced by me. Check out the list of nominations, vote for the ones you like, and add your own nominations.
Pure CSS Slideshows
CSS expert Eric Meyer came up with a pretty awesome slide show system called S5:
Even cooler, CityDesk expert Brian Cantoni created a very slick CityDesk template for S5 so you don't even have to know HTML. CityDesk is free for sites with less than 50 files which should be enough for most presentations, so this is a pretty cool free, standards-compliant improvement on PowerPoint that lets you put your presentations on the web.
I am completely at wit's end debugging this strange (apparent) IIS bug...
... Update: Fixed! So far, at least. Since it's one of those intermittent things I can't be 100% sure yet, but it looks like Tim Chaffee got it right:
"I would look at the MTS/COM+ package that IIS creates when you choose HIGH isolation. What user is that configured to use? It may be IWAM And you may need to switch it to the user you need."
That seems to be it. If you look in the Component Services control panel, it is indeed the case that the COM+ application that IIS created has its identity set to the IWAM_Machinename account. Changing that fixed the problem. I've moved the whole story here.
Thanks to Sean Timm, Yakov Shafranovich, Simon Fell, Mike Surel, Sasha Do, Mike Openshaw, Marcus Tucker, Max Skibinsky, Scott Durow, John Christensen, Tim Chaffee, Steven Afdahl, Nick Parker, John Waterson, Beau Hartshorne, Olivier Dagenais, Gil Bahat, Robert MacLean, Nicole Calinoiu, Scott Wisniewski, Matthew Randle, Lars Bergstrom, Sikko2Go, Glenn Deschen, Brendan Tompkins, Bill Loytty, Sean Wheeler, and Justin Bowler for their helpful and intelligent suggestions.
1114 posts over 16 years. Everything I’ve ever published is right here.
There’s a software company in New York City dedicated to doing things the right way and proving that it can be done profitably and successfully.