I recorded this 23 minute podcast on October 31. I didn't publish it then, but I figured at some point I would. It's the story of how a product like Frontier comes into existence.I had done this before, in 2020, in an oral history I did for a book a friend was writing. This podcast is how I remember it in 2025. :-) If you want to hear how a complicated project comes together when you're developing as you're designing, which I always do -- this is for you. It takes a while to...
I recorded this 23 minute podcast on October 31.
I didn't publish it then, but I figured at some point I would.
It's the story of how a product like Frontier comes into existence.
I had done this before, in 2020, in an oral history I did for a book a friend was writing. This podcast is how I remember it in 2025. :-)
If you want to hear how a complicated project comes together when you're developing as you're designing, which I always do -- this is for you. It takes a while to get started, and then I talk fast, and use technical terms without explaining them. Sorry for all that.
I want this kind of story told, because the folklore about how software is built or even that software is built at all, by humans, is usually wrong. It's not about invention, it's about building a new machine out of mostly pre-existing parts. Note that in the story there are zero components in the mix that we had not already perfected and commercialized. Some of them came from other developers, but most of them were remixes of themes that had appeared in earlier stories, or maybe ones that had been considered for inclusion but ended up on the cutting room floor, as in a movie editing process.
The thing about Frontier is that it made it easy for us to iterate over blogging tools when the time came to work on those. Frontier was the ideal platform for that kind of work, it's why were able to move so quickly and try out lots of approaches. But our runtime was no competition for PHP or Python with SQL. Our database wasn't written to work at that scale, unfortunately -- or a lot more of the world we use today would still be running in our environment. But the ideas persist.
Interesting sidebar not mentioned in the podcast, when we did MORE which was a really popular product on the Mac platform of the mid-late 80s, we took everything we had and put it into the product. We didn't leave a single thing out. This was because we had a devteam that could do it, and we were fairly desperate as an ongoing business just before we shipped it (1986). Apple had to loan us $400K to get to shipping! Anyway -- it worked. And that's why we called it MORE, we had no idea which if any of the features would pull people in. Turned out it was the presentations.
Anyway -- glad to finally get this out there.
Happy Thanksgiving! :-)
View more
