Karelia Software

We've purposely avoided putting out a new beta for a few weeks so we can concentrate on bugfixes and improvements. There is still a lot to accomplish, but we have managed to conquer some important tasks, thanks to feedback of users like ... well, maybe you!

For instance, we have optimized the memory management of media (images and sound) to make Sandvox more responsive when dragging photos in. We've continued to improve the "iMedia" browser for better integration with iPhoto and the other iApps. We now will ask whether a user wants to install (or remove) our crash reporter (though it's highly recommended that people install it, at least during our beta-testing cycle). The HTML we generate now uses relative links, not absolute links, which will make a website easier to preview after exporting. We've improved the text encoding that is generated for non-English sites. And, of course, we've fixed lots of other big and little bugs along the way.

We should have our next beta out in a couple of weeks; certainly before the current beta expires. And we're planning on getting a new Internet Service Provider to handle the large load that we get when everybody tries to download 15-megabyte files at the same time!

Along with the rest of the Mac world, we saw the Macworld keynote, and naturally we were not too pleased to see how similar Apple's iWeb and its intended audience turns out to be to our vision of Sandvox.  Lightning has indeed struck twice.

Clearly, Mac users who just want to get their family photos and simple blog published on their .mac account are going to be using iWeb. There is no way that we can compete with what is essentially bundled software.

What Sandvox can offer is a compelling alternative to iWeb, just as Watson turned into an alternative to Sherlock 3; Path Finder is an alternative to Finder; NetNewsWire and a host of others are alternatives to Safari's RSS reader; and Adium can replace iChat. As each of these offer solutions to the limitations provided by Apple's software, so too will Sandvox.

Of course, with only a few hours having elapsed since the announcement, we are still honing these differences, but out of the gate, some of the biggest limitations of iWeb (.mac only, not pluggable, no ability for HTML content) can become strengths for Sandvox.

As we move forward past version 1.0, we will be able to further distinguish Sandvox from iWeb by focusing on features that our users demand that will never be a part of the iLife suite.

by Dan Wood

Astute Karelia observers and subscribers to our email list may have noticed the new logo we quietly unveiled at the beginning of the year. I had intended to write up the story behind the logo some time after Sandvox's release, but the rumors that started circulating on "the internets" yesterday prompted me to write it now. At the end, I have an announcement that will demonstrate how this logo applies to Karelia software today.

Ancient History

In early 2001, when the dot-com crash was just getting underway, I was laid off from my job, programming Java and Mac (OS 9) code at a Bear River Associates, a small company that, at the time, did Macintosh consulting and built websites for dot-com companies. Unemployed, I spent my spare time working on Cocoa programming for the fledgling Mac OS X, since that's what I really wanted to to do for a living. I wrote a small application, showed it off at Apple's WWDC that year, and licensed it to another company. (It was never published, unfortunately!) Mid-year, looking for a new project to start in between those rare job interviews, I got inspired by Apple's Sherlock 2, a desktop program wrapping web search functionality.  I came up with the idea of a toolkit of of interfaces to web services (as they came to be called): phone book lookup, auction tracking, weather, movies, TV schedules, and so forth.  I named the program Watson, inspired by Sherlock, meant to be its "companion." Toward the end of the year, I hadn't found a job yet — I had even interviewed at Apple, a horribly long commute for me, but they weren't interested. So I decided to finish up Watson, throw together a website (archived) and release it as Demo-ware to see if I could make some income to make ends meet. But by the end of 2001, I realized I was making enough money that I didn't need to look for a job anymore — I had made my own job.

I quickly formed a real company (I had chosen the name Karelia because I had the domain already, for my personal website), got some help in polishing up the User Interface and Website from Robb Beal (who later moved on to create the innovative Spring), and watched Watson take off. It seemed that the sky was the limit, until I was called in for a meeting with Apple's Phil Schiller. I listened to him tell me that Apple was going to announce Sherlock 3, and it was very similar to Watson. I watched a demo of their program: all but one of their modules connected to the same service that Watson did and looked almost the same. I was too stunned to be upset.

The Phone Call

I drove home, gradually realizing what I had just witnessed, and sent off an email to my contact at Apple Developer Relations expressing my unhappy sentiments. An hour later, Steve Jobs called me.

"Here's how I see it," Jobs said — I'm loosely paraphrasing. "You know those handcars, the little machines that people stand on and pump to move along on the train tracks? That's Karelia. Apple is the steam train that owns the tracks." So basically the message was: get out of the way, kid; this is our market. The conversation actually felt fairly productive, and I hung up with the impression that Apple would at least acknowledge their inspiration by Watson, and that they would "throw me a bone" — a little more compensation than offering me a job. (Although some people see an Apple job offer as the Holy Grail, I had no interest in working for the Sherlock team after what had happened; it didn't feel like any kind of compensation for the ideas they had borrowed.)

I figured this was the end for Watson, except that emails kept coming in from people who were beta-testing the new "Jaguar" version of Mac OS X that included Sherlock 3, encouraging me not to give up on Watson. It turns out that Sherlock 3 was slow and not much fun to use. Plus, I could keep adding new modules; Sherlock seemed stuck with its initial set.

I thought I had been swept off the tracks, but no! So I kept on with Watson. Toward the end of 2002, I got a few offers from several large companies that were interested in getting into the desktop Web Services game. Negotiations took almost a whole year, but I wound up selling the technology of Watson to Sun Microsystems, and quietly joined their team in late 2003 to create a sequel to Watson, in Java. I was excited to be working on a cross-platform application whose audience would dwarf Watson's.  And it was still going to look and work great on the Mac, my platform of choice.

Beginnings of Sandvox

But I'm getting ahead of myself.  Way back when I had thought Watson was on its last legs, long before I started with Sun, I had come up with some ideas for a really easy-to-use website builder.  I had envisioned an application that would allow me to make simple websites, weblogs, and photo albums.  The tools that were available at the time — and this still holds true — were just not easy or versatile enough. I was inspired by The CSS Zen Garden to build modern, beautiful, standards-compliant websites.  I pitched the idea to Terrence Talbot, who had contributed several tools to Watson, first as a third-party developer and then as a contractor; he expressed interest in working with me on the project. So we got started on the program in mid 2003; I spent some time building a rich-text-to-HTML converter (which we've abandoned now that Safari's WebKit allows you to edit rich text) and Terrence started on the application infrastructure. He continued to work on the program while I was working at Sun.

Well, the whole Watson port project eventually imploded at Sun; management had become interested in other shareholder-appealing opportunities. (Lesson learned: Do not sell a your Cocoa (née NeXTStep) application to Sun unless you are Jonathan Schwartz.) I was devastated to see Watson hung up to dry (though surprisingly, it is still quite functional; about half the tools still work fine), but relieved in a way — I did not enjoy writing Java "Swing" (user-interface) code, nor working for a big company. So in August 2004, I went back to work with Terrence on our web-building application. Yes, I still would rather create programs in Cocoa for the Mac than any other platform.

So, for over a year since I left Sun, Terrence and I have been hard at work on Sandvox. We showed some previews of it back in July 2005 at Apple's WWDC; we wanted to be able to talk about it so that we could get better help and feedback with what we were doing with Apple technologies such as WebKit and Core Data. (Plus, we thought we were a lot closer to release back then than we really were.) We went into beta testing in Fall 2005 and have been moving closer to release of the program.

Of course, since we started the project, I've always been looking over my shoulder at Apple. To switch metaphors for a moment, I've been worried about lightning striking twice in the same place. We even made fun of the idea, and commissioned a company logo paying tribute to the metaphor of two guys on a handcar. (It took 3 artists to finally get what we wanted!) But every recent Steve Jobs Keynote that Terrence and I have watched has been filled with apprehension: is Apple going to release a program that steals our thunder?  So far the answer has been no … until now.

Enter: iWeb

Yesterday, an Apple page describing their upcoming iLife '06 slipped out listing a single new application: iWeb.  Here we have been moving closer to release (ironically, not feeling like there would be any competition from Apple because I have been finding bugs in Apple's WebKit that implied that Apple couldn't possibly be doing any kind of Webkit-based website-editing application).

Since the "iWeb" page is still (as of this writing) in Google's cache of Apple's site, I know it's not a hoax. I wasn't able to get much information out of anybody I know at Apple (tight-lipped as they are!) but it's pretty clear that there is something brewing for next Tuesday's keynote.

At least this time, I'm not expecting that Apple's offering will be a clone of our application, but the timing is certainly unfortunate. Although at times I feel like we are Apple's unpaid "Research and Development" department, the truth is that Karelia's product ideas just happen to be mainstream, like Apple's. I won't be surprised if iWeb turns out to be as lame as Sherlock was, so we can get off the tracks, let Apple pass by, then get back on the tracks and back to work.

Being Nimble, Being Quick

With that single word, on that one leaked page, our plan changed yesterday. And this is where we come back to the meaning of our logo. Because we are a small company, we are nimble — much more so than a Fortune 500 company or a bowl of petunias. We were going to keep Sandvox under wraps (save for a few hundred beta testers) for a few more weeks, but now we've decided that it's going to be important for people to get a chance to see Sandvox this week, and compare it to Apple's iWeb — whatever it turns out to be. Thus, we will be releasing Sandvox as a Public Beta on Monday, the day before Apple announces iWeb. So sign up on our email list so you will be notified of its release, and give it a spin.  We'll keep chugging along....