Interarchy Interview: Peter N Lewis and Matthew Drayton

Gruber: As the PR announcing this deal says, this more or less amounts to an employee buyout of the product. I take it we can presume this is an amicable agreement?

Lewis: The agreement is entirely amicable, we see it as a great way to ensure Interarchy’s future.

Gruber: Peter, when did you start writing the first version of Anarchie?

Lewis: I started writing Anarchie in 1993 as an archie client (archie was a very primitive search engine for finding files available on public FTP archives) coupled together with an FTP file downloading engine to allow the user to perform a search and then directly download the file.

At the time I was working full time for supporting a small University department with 20 Macs, and writing Mac Internet applications after hours (I even had to pay them for the computer usage!).

Gruber: Why give it up now?

Lewis: I’ve enjoyed working on Interarchy, but I have been doing it for a dozen years now, and it is time to try something else. The problem with having a successful program like Interarchy is it essentially stops you from working on anything else that doesn’t succeed as well as it does.

Matthew has been an excellent employee, and I have had a great time working with him, but this gives him a chance to go out on his own, which is something he would have needed to do sooner or later anyway.

As for what that next thing might be, I have no specific plans at the moment. Our third child is due any day now, so that will keep me busy for a while and then I’ll see what I might like to do next. Perhaps a revision of Keyboard Maestro or a Mac OS X version of Greebles, or perhaps something entirely different.

Gruber: Good to know that Stairways Software will continue.

A decade ago, you really had a slew of small utilities under active development. Most of the individual networking utilities were folded together under the Interarchy umbrella, but things like Obi Wan simply faded away.

I’m not asking you for specific ideas on what you might do, but do you have a general sense of whether it’d be one big new thing, or a handful of smaller projects?

Lewis: Actually, I still use ObiWan (now called Wanobi) and ThufirCalc, but many of our early programs served to bring Unix Internet facilities to the Mac, but many of those facilities have since faded away and our programs with them.

Currently I doubt I’ll write a new big program, at least not in the short term. I’d like to write a game, but they generally take a large group of people these days. I might build a few web sites to try something a bit different.

Gruber: Matthew, when did you start working for Stairways on Interarchy?

Drayton: I started working for Stairways at the beginning of 2001; 19th of March to be precise. I’d just finished my Computer Science degree and desperately wanted to write Mac software. Stairways was the only Australian company I knew of that developed Mac software so I sent them an email. A couple of days later I had a job.

Gruber: Pretty cool. Are you doing Nolobe on your own?

Drayton: I’ve enlisted the help of my brother David. We have wanted to work together for a long time and now we have the opportunity to do so.

David’s current situation is not dissimilar to mine back in 2001. He has just finished a Computer Science degree and wants to write Mac software.

Gruber: Interarchy is pretty unique, architecturally. My understanding is that it’s still written mostly (or completely) in Pascal. Is that right?

Drayton: Yes. There are a few C and Objective-C files but the rest is Pascal.

Gruber: There certainly aren’t many modern Mac apps left that are written primarily in Pascal. For one thing, support for Pascal isn’t provided by default in Xcode. How much work is it just to maintain a working tool chain?

Lewis: It is actually not too bad. There are a group of us who maintain the Pascal “Universal Interfaces” which are just translations from the C headers. There are GNU Pascal people who work at maintaining GNU Pascal, so we don’t have to worry about that, although early on we worked with them to minimize the inter-operability problems when converting from CodeWarrior Pascal.

Now that it is all set up, it works very nicely. We work in BBEdit and hit a key to cause a make or compile, which builds the program, potentially including all the documentation, widgets, actions, command line tools, disk images, etc. BBEdit supports ctags, so there is good syntax coloring and quick navigation. It also supports subversion, so we can use all of those facilities. Debugging is a bit limited, but we use lots of assertions which detect most errors at their source, cutting down debugging time enormously.

Gruber: I can see two sides to taking over the reins of an app as long-standing and popular as Interarchy. On the one hand, it gives you the opportunity to start your own company with an enormous existing customer base.

But on the other hand, existing customers have, shall we say, strong feelings about how Interarchy should look and work. How much do you feel bound to Interarchy’s tradition?

Drayton: I have been working on Interarchy for a long time; nearly six years. I like to think that I have played a big part in creating this tradition. It is something that I am very proud of and hope to continue with Nolobe.

Gruber: One of the things I like best about Interarchy is that it works so much like the Finder, e.g. its list views offer hierarchy disclosure triangles. And in other ways — and I mean this as a deep compliment — it works like the Finder used to, getting certain things right that the OS X Finder never has. For example, when I open a folder into its own window in Interarchy, the window opens in the same size and location as the last time I opened that folder.

Have either of you ever considered working on a Finder alternative? Something that leverages the UI architecture of Interarchy but for local file browsing — more or less a competitor to Path Finder?

Lewis: Not really, no. Going head to head with Apple on what amounts to a straight user interface does not strike me as a good idea. Who knows when Apple might decide it is time for a major change in direction for the Finder, so after a lot of work it would be very easy to be left behind in the dust if Apple ever got serious. It would not surprise me to see a big change in the Finder experience for 10.5.

Gruber: Well, congratulations. I’m sure this has been an exciting day for both of you.

Matthew, if you accept as many of my feature requests as Peter did, I’m sure Nolobe is going to do just fine.

Lewis: Thanks John!

Drayton: Thank you.