DarkWyrm Interview.

Here it is finally. I know i said it would be during the last weekend, but some stuff got in the way so only now are we able to post the interview with DarkWyrm, uber-coder, aka Mr. App_Server. Like we say, better late than never. So without further ado, let's get this baby on the road. Note: This interview was conducted by e-mail.

TBJ: Why don't you tell us a bit about yourself for starters?

DarkWyrm: Well, I'm a white music teacher in my mid 20s, teaching at a private Christian school in Columbus, Ohio. I've always loved computers, and I tend to tinker with a little bit of everything. I guess you could say I'm kind of a Renaissance man in that sense. I have a beautiful little 8-month-old named Candace (a future OBOS user) and a wonderful wife of 4 1/2 years. Needless to say, free time is a premium for me.

TBJ: And what is your BeOS background?

DarkWyrm: I've been using BeOS since R5 was released. Somewhere around January 2000, I heard about a free operating system mentioned somewhere, checked it out, liked the screenshots, downloaded R5 PE when it was released, and the rest is history. I actually learned how to write code by cutting my teeth on Photon when it was mentioned on OSNews that following November. I released my work officially as BePhotoMagic the following January. Later on, I also helped Guillaume Maillard work on Eden, and when OpenBeOS was founded, I jumped on board.

TBJ: Will Proto #7 be REALLY the last one? :)

DarkWyrm: Definitely. The prototypes are essentially test programs for more complicated parts of the app_server. The Extreme Programming term would be a spike solution, but there is a difference - they're complicated enough that I try to make them at least sort of reusable instead of pure throwaway code. Once the font code is tested, the last of the complicated stuff should be out of the way, and the next code I write will be the real server. There's nothing really complicated left - just tying up little loose ends.

TBJ: You say in your site that the work for the "real" app_server might begin even before the end of the year. With all the "pre-work" done, the several protos etc, how long do you think it'll take till we have a working app_server?

DarkWyrm: To be perfectly honest, I'm not sure. I will start writing official app_server code when the specs for it are either done or 95-99% done. Quite a bit of the code in the last prototype will go into the server after modifying it to match the specs. In fact, parts of #7 have been updated to reflect the specs already. Having an app_server in even alpha won't guarantee that we will immediately have all the cool whistles and bells I'm building into it, though. There are key Interface Kit classes pretty much undone, like BView and BWindow. I've got some skeleton code implemented for the demo apps to go with the last couple prototypes, but nothing which will do a whole lot for a regular app. There's a job<->experience kind of problem for them - how to test the classes without a server, and how to test the server without classes. While I don't think it should take all that long to get something going, I guess my official answer will be As Soon As Humanly Possible.

TBJ: The font rendering screenshot provided is.. well... awesome. And it's just in proto phase (way too trekie...). Even though it doesn't look like it, is there any space for improvement with the real app_server?

DarkWyrm: Thank you, but I can't really take credit for the high quality work of the coders on the FreeType library, which does all the heavy lifting. Yes, there is room for improvement in the typography area for the server. Don't get me wrong, antialiased text is good, but what about limitations of the font API as it is? Checking the length of a string requires a trip to the server - just like a *lot* of other font calls - and, thus, it gets expensive if you want to do any serious text work. One thing going into R2 is full localization support. We're going to have to support vertical text display in the API if we want to support Chinese, for example.

TBJ: In your article "I want to see some skin!" (that could take us on a totally different interview), you wrote "A completely skinnable UI should be abandoned in favor of the simpler option of window decors and customizable system colors which all controls utilize, including window frames and borders.". Do you still maintain that opinion and if so, will it be for R2 and beyond only?

DarkWyrm: Yep, at least if I have a say in it. People get so frustrated with a computer because they can't get it to do what they want/need it to do. At work, I help the school with computer support on Windows machines, and there are days where I envy the Be Engineers in the Memorial Day videos and wish *I* could afford to throw something off a roof. Developers often just don't think of usability issues as is obvious with a great many applications on just about any platform out there. An operating system should place as few barriers for the user to get work done as possible. The interface has to have some consistency. I also recognize that people like to personalize their own computer with wallpapers, different colors, and such, but there needs to be a balance between the two. No matter what I think, someone is going to disagree with me. I want to set up something that the vast majority of us users can live with comfortably.

TBJ: Speaking of R2, which features would you like to see/implement for it, other than the above mentioned Skinning and the Color Management (with LCMS)?

DarkWyrm: I've got quite a list, but I'll name a few. I'd like to see a BBitmapUtils class (like what Michael mentioned in the newsletter recently) which operates directly on BBitmaps without bothering the server. I'd like to see some classes which would aid in working with .Net. While I can't stand Micro$oft, I think Web Applications will be important in future years. Support for multiple monitors on the desktop. More customizable GUI colors than what will be in R1.

TBJ: Last but certainly not least, tell us your feelings on OBOS and what you hope to see in store for it in the future (besides world domination of course).

DarkWyrm: What I would love to see is it entrench itself in education. We have so much to offer with making it easy to teach C++ and (eventually) Java, multimedia superiority, and stability. The fact that it's free, easy to use, and "just works" in almost all cases would be immensely helpful for teachers, students, and administrators, who often don't care what they use as long as they can do their work. While I doubt we'll ever seriously threaten Microsoft on the desktop, it would be interesting to see Steve Ballmer look at us like M$ looks at Linux.

TBJ: Thank you so much for your time, i'm sure all the TBJ readers will be most interested in reading what you have to say.