Nobody learns if nothing is brok3n

So who is Alex Schaefer anyhow?

Alex C. Schaefer is a one-man army whom just this morning won "Coffee Drinker of the Week" at his local Pete's Coffee. In addition to enjoying free coffee, Alex C. Schaefer enjoys problem solving, computer programming, poetry, Portal for the Xbox 360, drinking Guinness until the bar closes, and talking in the third person.

Ah, so you got the Orange Box? Are there any other games on that which you like to play, or did you get the box for just Portal?

I bought it for Team Fortress, I stayed for Portal. Portal is just such a wacky way of thinking of physics - it's brilliant, and it's only flaw is that it's begging for more levels.

Are there any other kinds of games that you like, or is Portal it?

Halo 3 has been sitting on top of my 360, waiting be unwrapped, but Orange Box has too many goodies to play with. It's five games in one, it's such a ridiculously great package. I'm knee deep into Half-Life 2, been doing some Team Fortress 2 online, and I plan on unlocking all achievements on Portal. I'm also a big fan of PC gaming - Supreme Commander is such a great game, I only wish my PC wasn't breaking down from old age so I could start playing it again.

What is it about Portal that's so fascinating? For those who have never heard of it, can you explain some of the game play?

It's just gameplay and problem solving that I've never seen before, and it's just got my attention right now. I keep wanting to play it over and over again to try and shave a portal off my total-portal-count, or seeing how fast I can complete the puzzle. It's very much a thinking man's game.

So no Wii or Playstation 3? How about DS Lite?

I have my WII, but i find I don't play it nearly as much alone as I do when I have company over. I can play wii all night with a few people, but alone I am more likely to play Starfox on my DS Lite or Marvel Ultimate Alliance on my 360. The Wii is great, but without buds, I'm just not that into it.

I guess that's enough of the gaming tangent, let's get back to you telling us more about yourself

In all seriousness, I'm a young developer. I'm more or less home grown, having dropped out of Compsci after my fourth semester in favor of a degree in Creative Writing, and as such since the end of "Data Structures 419" I've found a happy medium between computers and writing (be it Obj-C or something in trochaic hexameter). The logical explanation for that gig is that I enjoy finding structure where there seems to be none, and well, it's lead me to some interesting places

I just have to ask, what sort of interesting places?

I like breaking rules, and finding structure where I hadn't thought it existed. When you find structure where one would not expect there to be structure you're instantly confronted with sets of decisions that someone made and a board is revealed of purpose and direction, and to play with those forces in such a way as to elaborate either has provided me with great clarity. What started this entire thought dichotomy, and subsequent leaving Compsci for the English department, was Oscar Wilde and the poem "The Harlot's House" - an example of trochaic hexameter, actually - it's very carefully crafted, and up unto it I hadn't given books or poetry much of a chance, because I didn't see how deep the rabbit hole goes kind of.

Good poetry, like computer programming, becomes about a structure, and that's how you rise above "amateur" and into something with substance and function - not to say that I'm not amateur, I am, but I've produced my best work while being consciously aware of the rules in a system. With Wilde, he exercised grace & meter in a way that each word pushed forward feeling, and that's hard to do in poetry - so much of it is variable meaning, but he had a unified message that was evident throughout the poem and ends on such a crashing note that you are forced feel something. It's this unified structure that, for me, becomes about finding something that someone never intended to do or something that people assumed is useless and pushing it into a different light to see where it takes the bigger meaning of what you're trying to do.

What kind of Mac do you develop on? Do you have any other macs, or have you had any others previously?

I was a terrible mac basher until a lovely little day in 2004 when the University of Colorado hired me on as a OS X administrator based on my Linux experience, and I suppose that would make my first mac a PowerMac Dual Proc 1.6ghz. From there I bought a black MacBook the day it came out, and now the rest is history. The platform is elegant and makes sense, and while I still keep windows around for some reasons - those reasons dwindle each day. I'm now anxiously waiting for the next MacBookPro update (be it six months or more) to upgrade.

I've quickly come to love Xcode and the entire OS X development environment. It's fun to do - and that's the important part to me. I've never had a real coding gig in my life, and so it's all exploration for me - and if I'm going to explore and wander around, I need to enjoy doing it. It's a beautiful system for old coders and young programmers alike, and I encourage anyone to look into the platform, the system, the api what have you. While .NET is an incredible improvement over Microsoft's old development environments, it's got nothing on objective C in terms of simplicity to slip into a program and sheer joy of production of the code itself.

How did you get involved in the ApolloIM project?

I started ApolloIM in late July as an exercise to get me back into Cocoa, and because there was allure in working with an undocumented API. It was a bit of a wild goose chase - hunting down how apple did certain things within this class-dump'd mess of an API that the hacking team for the iPhone rigged up.

The aim of the program was originally just to get AIM, and so I found an Objective C wrapper for Firetalk from a program called Boombot (long dead, I believe - http://boombot.sourceforge.net), and using that as a base customized it into a client. After it was clear that TOC support wasn't enough and that we needed something with Oscar for .Mac support, buddy icons, etc we switched to libpurple. It took us three weeks to even compile libpurple, and it would not have been possible if it wasn't for the husband and wife team of PHD's from MIT and Harvard (respectively) to get it done (Core, Pogue, thanks again) - reason being that the hacked-together compiler for the iPhone (while impressive in it's own right) just wasn't ready yet, but they found a way to fit the square peg through the round hole.

Anyway, I'm involved with ApolloIM because it's the big feature that most phones have that iPhone does not. Who knows why exactly Apple didn't include a MobileiChat, but really I saw it as something that could be done and so I ran with it.

You mention TOC being the reason you switched to libpurple due to needing OSCAR. What exactly are these things?

TOC is this basic protocol for AIM that AOL develop, from my understanding (or it could be conjecture), to create an easier to use OSCAR. The difference between TOC and OSCAR is that TOC is old school and Oscar has all the neat stuff like buddy icons, status messages without being away, Direct IM (picture sharing, etc). The move to libPurple was done to move towards that full IM client feeling - TOC works, but it's really not full figured, and Apollo wants to be a full featured mobile client. libPurple also allows us to move towards the other protocols - I have never had an MSN account, and now I do - the library itself is incredibly flexible and allows us to extend our base infrastructure into something that is so much more.

What are the benefits of using libpurple for the ApolloIM backend?

Multiple protocols, stable connections, and one sane way to modularize and move from there. We built our own libPurple/Cocoa wrapper (With some inspiration from Adium), and it's allowed us to move towards having the full client that we want. We're not anticipating video confrencing or anything, but we know libPurple is moving in that direction - and as long as we focus on abstracting libPurple from our code while be conscious of implementing it's feature, I find we work out a paradigm to focus on our client implementation while leaving the IM protocol development to Sean Egan & co ( as well as the Adium contributors) over working hard on libPurple.

How do you and your team communicate/work together, and what have you learned from this experience?

Instant Messaging has been the way to do it. It's really been the three of us, and lately more or less just myself. We're a very small project right now and so we don't have the need for a big wiki or a bugtraq quite yet - we're happy with Google Code's "issues" system.

Given what you know now, would you work on ApolloIM all over again? If so, what would you change?

I would have delayed our release of 1.0 by two or more weeks. One thing I've learned is that competition can make you stronger, but it can also make you a little too insistent on releasing a product that was completely ready. We're up to 1.0.2 now, and I feel that's the version we should have released as 1.0. I know I've learned my lesson - don't

Why is ApolloIM important?

ApolloIM is important... I guess it’s because it's the one glaring omission from the iPhone. The iPhone was really designed to be a platform for universal communication, and they forgot this one piece - Instant Messaging. Just as Adium fills the role of the missing as the client for the missing protocols from iChat, ApolloIM becomes important because everybody uses one IM protocol or another and without it, a communications platform is incomplete. I can't tell you how many times someone hasn't picked up their phone, or missed a txt message - but somehow, they still respond on GChat.

We've been in a lull at the moment, I've been settling in to a new job, and since the announcement of Apple starting to officially sanction third party applications - we've been tidying up code in anticipation of an official compiler. We're taking our time now - for a while there in September and October we were releasing once a week, and certainly lacked some testing in rushing things out the door. For now, we're going to polish and try and get the missing protocols that people have been asking for, and also try to figure out what exactly a “Gadu Gadu" is.

FYI, GaduGadu is an im protocol used heavily in Poland

I know what GaduGadu is - but man, is it a funny name, so funny that that I sometimes say outloud as a pickmeup when Coffee is out of range.

So you're hoping that the SDK that was mentioned is in fact a full on Cocoa/Carbon SDK with a compiler and the whole kit and caboodle? What if the SDK is in fact just support for dashboard widgets, or something else that isn't Cocoa/Carbon with a compiler?

Yeah, we're betting an awful lot on a real compiler, but frankly you're not supporting third party app developers unless you have a real SDK with a compiler and documentation. The way I see it, we're preparing for the future - if the future is on an unsupported platform, then we have to make ourselves flexible enough to accommodate changes to it; the switch from iPhone firmware 1.0.2 to 1.1.1 hurt everyone's development efforts and has been a real reboot for the entire development scene, but it's coming back with a vengeance right now. If the future is on a supported platform, then we have to be ready to make the switch from clever little hacks (such as us relying on the host file instead of doing a real DNS lookup due to Apple's DNS resolver on the iphone being wonky) and get ready to support real solutions. We're looking to be legitimate, and regardless of the direction, we are in a position now to prepare for that direction when it becomes clear.

Right now, the system that's already established is little more than widgets. I think with 10.5, and the iPhone os (which is very similar in file system layout and has a lot of the same API), you can implement a jailed region for third party applications that would avoid OS integration. Really, just give me a socket and 20k of r/w space for configuration files and Apollo should work fine. I am sympathetic to Apple securing the iPhone - it's a phone first and foremost, but it's also a waste of potential if you let it just sit idly by waiting for someone to call you. I've done coding for J2ME and have played with .NET for mobile platforms, and I know it's more than possible to isolate part of the phone for third party development.

Apple has a great relationship with Cocoa developers, and I am eager to see their infrastructure expand to include this, one more, OS X platform. They can do it, we can code for it, and at the end of the day there's hundreds of iPhone apps out there that are being actively developed anyway - so why not bring them into the fold and make it a secure environment?

What do you like to do for fun?

I like doing new things constantly - if I'm not changing and going in a different direction constantly, then I'm really not having that much fun. I started taking up white water rafting this previous summer - and that was pretty fun, and I'm hoping to branch into Skydiving or something similarly death defying by this Summer.

Lately I've been browsing various open-source projects looking for bugs to fix or a feature to add, and I'm hoping they'll accept them. I'm also working really hard to get the Vanilla Crazy Cake achievement in Portal.

Anything else, Chris? :)

Ah, yes, one last thing. What's your favorite kind of pie and/or cake?

Frankly, I believe the cupcake to be a modern marvel of snack delivery and consistent quality. With one fell swoop, some baker somewhere out there (I'd wager a Frenchman, but it could have been a Belgian Baker) created the first truly portable and satisfying dessert delivery system. Jacques, as I imagine he was named - because if it was truly Belgian it'd either take the form of a waffle or chocolate and while Cupcakes can take on either quality, it is not mutually exclusive - woke up one day in what I'm sure was an epiphany from god himself. This beloved edict of passion for baking and appreciation of the divine will inside Jacques could only be what produced this modern marvel of dessert specification.

Now, to further justify my general classification that the cupcake is my favorite cake - I am including this line graph demonstrating the growth in the tastacity / delicousosity of the cupcake over the last 100 years.

The cupcake starts off steady with it's invention in 1900 with a steady 5% popularity amongst elitist French (but, once again, possibly Belgian) bakers. This steady growth was interrupted in 1932 in part due to the Great Depression in the United States, where the cupcake had contributed a total of 15-20% of the popularity growth in 1910 and 1920. In 1940 the Cupcake Deliciousness index only grew by 10% due to the popular distraction of "staying alive in WWII" and thus the inevitable "taking for granted" occurred - however by 1945 the Cupcake skyrocketed back into the public's hearts, minds, and mouths.

The cupcake continued amazing growth until 1970 when "The Cupcake Man" serial murderer struck, sending baskets of cupcakes to unsuspecting childhood idols such as Captain Kangaroo and bozo the clown with Ricin laced carrot cheesecake cupcakes. It quickly recovered on New Years Day 1980 when pop-culture icon Punky Brewster was sighted downing a bottle of Jack Daniel's and washing it down with a Vanilla Frosting / Chocolate Cupcake.

In conclusion, the cupcake has withstood "the test of time" with it's magical innovations in portable deliciousness, and it's continued deliciousness growth in the future both excites and scares me - but mostly excites, because the cupcake is my favorite kind of cake.

Ah, interesting. What flavor cupcake is your favorite though?

Chocolate. :)