Skip to main content.
Features: Calendar | Interviews | || Login

Interview with brian d foy

Besides being the publisher of The Perl Review, brian has written several Perl modules, writes a regular column for The Perl Journal, and is a partner in Stonehenge Consulting Services where he teaches and consults on Perl. He was the lead author for Learning Perl, 4th Edition which O'Reilly Media published this summer.

brian d foy (as interviewer): You're interviewing yourself. Are you really, really vain, have an insufferably huge ego, or just need attention?

brian d foy (as interviewee): Do I get to choose all of the above?

Actually, I needed an interview for September. I started doing interviews a couple of months ago and I've been able to get one every month. I ended up pretty busy with a National Guard mission to Egypt in September and immediately went on a GeekCruise after that. Basically, I was out of the country for five weeks without reliable internet access. It's actually the end of October right now, but I know the webmaster so I'll get him to pretend we published this is September.

Aside from that, I've had this idea of self-interview since I read Glenn Gould's interview of himself. His philosophy of interviews diverges from what most people do; he thinks the most enlightening answers come from questions completely unrelated to the interviewee's expert area. His interview with the great conductor Leopold Stokowski went over interplanetary travel, for instance.

Along with that, I was impressed by a comment that former Secretary of Defense Robert McNamara made to some questions Errol Morris posed to him in Fog of War. Secretary McNamara said you should always answer the question you wished you were asked instead of the one you were asked. I'm not sure how that will work out here since I get to choose both the questions and answers.

The Perl Review What's the question that most people ask you?

brian: I'm not sure if I can answer that in polite company, so I'll give you the second most frequent question. I suppose I should start a FAQ about myself, but that might be pushing this self-interview thing a bit too far. I'm really not that in love with myself because I think I'm an idiot. Anything I do I know I can do better. If you think I'm some great Perl hacker, just look at the CPAN Tester FAIL reports for any of my modules. If you think I'm a good publisher, look at how boring this magazine looked before Eric Maki came along.

The Perl Review You're answering the question you wished you were asked, I think.

brian: Ah, right. Writing is an interesting thing. Weird stuff comes out when you just let it flow. I've often read that a novelist sometimes loses controls of their characters that way. Still, I'm doing that again.

Most frequently people ask me about my name. I write it in all lowercase. That could be another sign of narcissism, but it was really more pragmatic than that. I needed people to remember my name. Before I made my mark on the Perl community, I was just another person doodling with Perl.

Some people use nicknames or handles for various online channels. Part of that has to do with the naming restrictions for logins, and some of it is historical. For instance, my PANIX and Pair account names are both "comdog". I never choose that. When I first signed up for a web hosting account, my one-man contracting company was called "Computer Dog Consulting". I needed a name for the incorporation forms and choose the object sitting on my monitor: a little stuffed dog. The web hoster condensed that to "comdog". I don't really like it that much, but most anywhere I go, 'brian' is already taken. If a guy was born in the US in 1970, their name is probably Brian.

I don't like people hiding behind nicks though. Just having a nick doesn't hide identity (everyone knows "merlyn" is Randal Schwartz, for instance), but I'd rather just put my name out there. I think people should attach their name to anything they want to create. There is the odd case of hiding from oppressive governments or litigious churches, but otherwise, I think people should take not only the credit but also the responsibility for what they do. The best way to do that is to attach my name to things.

That's getting away from the question though. I think I have a pretty simple name. My last name is "foy" and it rhymes with the other words spelled like it. It's not Hietaniemi, for example. People still mangle it's pronunciation. Worse than that, they come up with all sorts of interesting spellings. It's so simple it's hard, apparently. I decided to make it easier by specifying rules for my name. Along with that, I made it all lowercase thinking people would have to break out of their normal thinking and perhaps get it right.

That didn't quite work out. For a while I said my name, including the middle name "d", so people started coming up with annoying permutations of "defoy". I guess that makes sense, so I just made the "d" silent.

Most of the rules evolved for usability and recognizability. At least its a conversation starter, which is always a good thing if you want to meet a lot of people. In the Perl community, having some weird name really isn't that weird. Just look at "chromatic". He's using a nick, but he uses it everywhere, including the covers of his books. When I was working as a physicist, I had to choose a form of my name and use only that in all of my publications to make it easy for people to find all of my work. I've just done a similar thing for the Perl community, but with a bit of geek twist.

The Perl Review If the most enlightening answers are the ones unrelated to the topic, what are you doing outside of Perl?

brian: I spend quite a bit of time on Perl things. After I finish whatever things I have to do with Stonehenge, I still maintain my Perl modules, the perlfaq, and a lot of other things I should pass on to other people.

Most of the non-Perl time seems to be taken by the military. I've been a military police in the Army National Guard for 16 years. I didn't really choose to be an MP, but that's the unit they had close to my university. I think everyone should perform some sort of government or civil service. Back then the National Guard, which is a state organization, was focussed on state emergencies. In California, the state where I started, earthquakes, forest fires, and flood were not rare, so there was a lot of that sort of work. The focus has shifted a lot since then, but I have a lot of experience I want to pass onto the next generation and I'm about five years away from a pension. The Guard just gave me a very large bonus to stay in until I finish 20 years, so that was hard to pass up.

Besides that, I'm always looking for time to scuba dive or ride a bike. Both of those are tough with all of the traveling I do with Stonehenge, though.

The Perl Review Are you traveling right now?

brian: Actually, yes. It's one of the reasons I'm interviewing myself. I'm headed to Tampa, Florida from Providence, Rhode Island so I can catch my flight from Tampa to Chicago. The airline industry is a big mess. Despite "Airbus" calling itself the bus of the air, you can't simply get on an airplane and go somewhere. They want you to buy a roundtrip so they lock in a sale on the return flight. That doesn't work very well for a traveler like me (or Randal, either). We sometimes stay on the road for several weeks as itinerant Perl trainers, so we want to go from one city to the next in a loop (or "open-jaw" as the airlines sometimes call them). To get reasonable tickets on a non-cattle-car airline (which are the ones who can deal with special issues much more easily and gracefully), we have to play their game. We end up flying a spoke pattern. For the past month, my plane flights look like I really like Tampa. In reality, that was the embarkation port for Perl Whirl 2005, and I left directly from there to go to San Jose. I then had to return to the starting point to finish the trip. From Tampa I started another trip to Providence and had to return to Tampa again. Finally, I fly out of Tampa (in about 6 hours, just enough time to claim the bags from this flight and recheck them) to finish the original trip.

Chicago -> Tampa

	Tampa -> San Jose
	San Jose -> Tampa

	Tampa -> Providence
	Providence -> Tampa

Tampa -> Chicago

The Perl Review Is this why things have been slow with The Perl Review lately?

brian: It certainly was a part of the reason. With some much travel its sometimes hard to get internet access. I know I can always go to a Starbucks, use a cellular modem, or use an EVDO card, but when you're on the road you're doing something, like teaching or consulting, in which case I can't just be on the net whenever I want. I do get web access at most places, but local firewalls don't always have the ssh port open. I can't read all of my mail, rsync things, or commit to source control in those cases. A lot of times, I'll be deep in the innards of a building (all the good spots are taken by real people), so I won't have the cell signal I need to for wireless access.

I did start The Perl Review thinking I could do it from anywhere. My wife is an opera singer so we travel quite a bit, including international travel. In those cases though, I have a decent hotel room where I can get high speed net access. That's worked out great. I have to figure out how I can still run the business of The Perl Review while being offline.

The Perl Review What would you need for that?

brian: As with all things connected to running a new business, I'm learning a lot as I go along. Most of the work happens the month before the next issue, but in the mean time we get new subscribers and we need to send those issues to people. That's the only real tangible thing we have to deal with. For short trips it doesn't matter that much because I'll be able to take care it in a couple of days. For much longer trips, I'll either have to carry copies with me (which I usually do anyway for user groups and classes) or get someone back home to do it. For international travel, someone in the US will have to do it since I don't want to pay international postage rates.

The Perl Review What's the most expensive part of publishing?

brian: Right now, the most expensive part is the actual printing. As we print higher volumes, I expect the price to stay the same because I also anticipate higher page counts and nicer paper. At the moment I'm paying just under $2 per copy. First class postage in the US is $1.06, and I charge a subscription at $16, or $4 per copy. Out of the that left over dollar are the credit card transaction fees and monthly fees. So far all of the authors have donated their work (so they also keep the rights), and that's pretty expensive for a lot of magazines. The trick is to sell a lot of copies.

I'm not publishing the magazine to make money. I'd make more working at Starbucks (or some coffee house with free wireless). I might be able to make some money with it later, but at the moment I like the learning the business and providing the magazine to the community for almost at cost.

The Perl Review What's the easiest way to reduce costs?

brian: The best way to reduce costs would be to have a web magazine. The Perl Journal went from a print magazine, to a magazine supplement Sysadmin, to a digital edition. Now it's only available digitally. CMP (the publisher) already has sunk costs in servers, hosting, and some of the staff, and they no longer need printing or postage costs.

I started The Perl Review the other way around. At first I supplied PDF editions only, but a lot of people wanted a print magazine. I certainly understand that since I hate reading things on the screen. I could print things out, but at 6 cents a page (the cost of my ink-jet printer), a 40 page magazine costs me $2.40. That's just $1.60 less than my per issue cost of the print magazine where I print it for you, slap a nice cover around it, and staple the pages together (most of the time in order).. It's half the size of self-printing because the magazine has stuff on both sides of the paper (unless you have a duplex printer). The cost for someone to print it themselves also involves the time they have to wait on their printer, dealing with loading paper, and anything else that's probably worth a $1.60 for me to do for them.

The more interesting question, and I guess the one I should have answered (according to McNamara), is how to increase profits. I can't do much to cut costs and keep the quality I want, but I can do things to make more money.

The first answer is volume. I pay fixed monthly fees for some services, so higher volumes amortize those over more transactions. I have to get five new subscribers every month to just cover those fees, or about 20 subscribers to cover those fees and the cost to deliver the magazine.

After that, we can do other things to continue to make money on the older issues, such as selling back issues, collecting the best articles into special editions, or other things that don't create a lot of extra labor.

The Perl Review You mention labor. What takes the most time?

brian: I think it's a time really. Acquiring and editing articles takes quite a bit of work, and it's not something that just anyone can do. An unknown person is going to convince a notable Perl person to write an article for free, and very few people have enough writing experience to make an article ready for publication.

Before you (that is, me) asked me (that is, you) about reducing costs and my answer was to publish to the web. There a magazine merely comprises a bunch of articles, but those articles don't have to live anywhere in particular. Anyone can publish just about anything and I could link to it. However, the article loses out on stuff that external review and editorial help can provide. Wikis seem to provide that technology, but the really good articles have a single person's vision behind them even if multiple people implement that version. For instance, Wikipedia, which is an amazing project, has a particular style that they enforce. Even though a lot of people can contribute, other people come along and enforce the rules. Even more come along and correct the mistakes.

Projects like Wikipedia divorce themselves from human stories though. It's all about facts. It specifically asks that people exclude commentary. They have to do that to get the community to work together. The same thing happens at any sort of wiki site. You either have to limit the content form and scope or the limit the contributors. I think it makes a lot more sense and gives us a lot more freedom simply to create our own magazine so we can do whatever we like. We can put an article about how much Perl OO is broken right next to one saying how amazingly flexible it is. We get to make that decision rather than mediating a wiki-war.

The Perl Review If you can't reduce costs, can you do less work?

brian: We're doing less work as we build up the software for subscriber management and other business like things. It took us a while to get a mini-dashboard going, but that was really the fun stuff. We wrote that mostly using CGI::Prototype, which is something I use for work, so it really wasn't that much of a hassle.

The most work is the actual content. We could reduce the work by simply lowering the standards or simply publishing the articles just as people send them too us, but it wouldn't be as good. We use a lot of first-time authors so there is extra work there. If we went with experienced authors who actively write and publish, we'd have to pay for their time. Most of that stuff can go right into the magazine without much editing. It costs more, though.

The Perl Review How about...

brian: No more questions, please. It's time to get back to work.