Hi, For a while now I have been writing a buissnes plan and I would like help with some facts concerning salarys. What is the avarge salary for someone that has worked 2, 5 or 7 years in the buissnes. It would be great if you could give me some numbers on every position in the development team. Regards

Kent Bengtsson
bengtsson.sweden@telia.com

You can go to monster.com to fiddle with their salary thingy. For programmers it''s between 36,000 to 130,000 dollars a year. It''s about 60,000 to 130,000 dollars a year in the big cities while in smaller ones it''s about $40,000 to $60,000 a year. It all depends on cost of living as you can imagine the salary would be high in silicon valley but low in Yakima, Washington This salary range is the same for system programmers, application programmers and website or database programmers.

Good luck!

Jerry

my homepage
E-Mail: BlueOrbSoftware@mailcity.com

It is my view that game programmers are also often willing to settle for lower base salaries (than other software developers) with the promise of royalties later.

This is basically what I was told by a well-known game industry recruiter and a number of studios back in March/April when I was looking to change jobs. I have over a decade of experience in 3D graphics, C++, etc., and my current salary is $90K. The best offer from a game company I got was $80K and some stock options. One studio near LA wanted to offer $65K plus promise of royalties if they ever made money, which when translated into North Carolina dollars (where I am now) would be the equivalent of about $53K (cost of living, when renting an apartment, near LA, is 20% more than NC, so I just took 20% away from the $65K to get the "equivalent" to compare with my current salary.) A studio in Texas was indicating the offer might be $70K.



Graham Rhodes
Senior Scientist
Applied Research Associates, Inc.
email: grhodes@sed.ara.com

To grhodes_at_work,

What was your hardest problem you had to solve in your code? I am wondering how experienced you have to be (years don''t tell me anything) to be earning $90K/year. I mean, are you some kind of genius that can code circles around graphic programmers. Compared to Carmack, Sweeney, Abrash and other well knowns how would you compare to them? I just want to find out how much farther I have to be in terms of coding to be considered on equal footing with industry coders before I start looking for an employer or become self-employed. I really liked your honest answer in the above post otherwise I wouldn''t ask.

Thank you.

Jerry

my homepage
E-Mail: BlueOrbSoftware@mailcity.com

quote: Original post by JD
What was your hardest problem you had to solve in your code?

This is a longer answer than I thought I'd give. You don't have to read it all if you don't want to!

Whoa. That's a hard question to answer quickly, as I've had my hands in quite a lot of different things. Perhaps I'll just list a few of the key challenges. I started working in 3D around 1987---building an interactive mesh generator for airplane aerodynamics analysis while working on my Master's degree (I wrote a paper in 1991 that mentions this work), but the work that I consider challenging didn't start til the 90's.

- 1991-1992 Developed a full software 3D graphics pipeline that supported Gouraud shading, Phong lighting, palette-based data contours with palette animation, double-buffering, software-based 16-bit Z-buffer, an interactive 3D cursor/cutting plane, in a custom windows environment----under DOS (386 extended DOS). We used Z-buffer tricks to implement the cutting plane in near-real time on a 486/33Mhz machine with 8MB RAM. Ported part of it to OpenGL under Win32s in late 1992, early 1993. (We used an early implementation of OpenGL on Win32s from Pellucid, later distributed by Mediavision.) I am co-author on a paper from 1992-1993 that describes this 3D data visualization software, which we at Model Engineering called "Reveal". The compiled application fit on a 1.44MB floppy disk. We actually had a polygon offset feature, which you may know only appeared in OpenGL with version 1.1.

- 1993-1994 development of a 3D graphics package, based loosely on the pipeline from above, that we used to display pressure contours on airplane fuselages during simulated bomb blasts from terrorist attacks. Two issues existed. The display of pressure data that covered a variable portion of the mesh from time-step to time-step and dealing with something like 10,000 time steps of data. A tweaked polygon rasterizer handled the variable pressure coverage issue.

- 1996-1998 development of an interactive simulation framework architecture and implementation of the framework for inclusion in the WorldBook Multimedia Encyclopedia on CD-ROM. The framework includes a mini custom 3D pipeline and scene graph framework used to display the 3 degree-of-freedom flight of a small glider and the 6 degree-of-freedom motion of a frisbee (with gyroscopic effects). The 3D pipeline is integrated seamlessly with the primary, sprite-based 2D engine, and has support for simple projected shadows. 14 simulations using this framework are in the 1998 through the current version of the Worldbook encyclopedia. One challenge was supporting *all* versions of Windows, INCLUDING Windows 3.1. We weren't able to use Direct3D or OpenGL or any other DirectDraw or WinG features due to strict requirements of the overall WB application. And it does work. We had some hellacious palette issues that resulted in the executive producer personally flying down to our office with beta CD's in hand, sleeping in the office as I worked on the problem, and nearly resulted in the sims being dropped from the 1999 version of the encyclopedia until I solved the problem at literally the last minute. (My name is in the credits for this one, yeah!)

- late 1996 through present - lead architect and developer of the Next Generation Revolutionary Analysis and Design Environment (NextGRADE) graphical user interface in support of NASA's Intelligent Synthesis Environment (ISE) program (www.ise.nasa.gov). A couple of graphics-related challenging tasks on this project include:

* Development of a fast and robust computational geometry algorithm to compute the intersection of an arbitrary number of curves in 3-space----that's 2 *or more* curves that simultaneously overlap arbitrarily along a common region----and extraction/discovery of the topology of the intersection.

* Development of Linux and IRIX-based code to animate the nonlinear vibration of space structures due to radiation loading. The issue here is being able to deal with motion database files of around 300MB. The motion has to be streamed from disk at a sustained 96 frames per second (weird frame rate of one of the large immersive displays we use.) This code uses the IRIS Performer library.

quote: Original post by JD
Compared to Carmack, Sweeney, Abrash and other well knowns how would you compare to them?

Gosh, I have no idea. I've never met them personally. I've seen some of Abrash's and Carmack's code, and it looks clean enough. I will say that I used to follow Abrash's articles in DDJ and he has had an influence on some of the work that I did in the early 1990's. Carmack and Abrash both, Sweeney and others too no doubt, are damned good at developing optimized code that gives the best real-time performance. And no doubt they're able to use the latest techniques with somewhat less spin-up time than me. (Although I am very quick.) Until recently, my work has been focused mostly in the engineering world, and real-time was never that important. (I am now doing game-related work and real-time is a big focus.) I would say that my knowledge of 3D and geometry in general is very, very comprehensive, perhaps more comprehensive than these guys. I'm not sure they'd have a clue, for example, about the use of Christoffel symbols in mapping between generalized, curvilinear, and cartesian coordinate systems. They may not know the difference between covariant and contravariant tangent vectors. Then again, they haven't needed to. They are kings in the game development world and I don't compare at all (yet!).

In a way, I'm envious of these guys. They are my heroes of a sort. I was developing computer games on my VIC-20 and Atari 400 back in the early 1980's. Why couldn't I have gotten the idea *then* that I could have a career making games? Its only now, when *everyone* wants a career making games, that I've decided I *will* do it.

quote:
I just want to find out how much farther I have to be in terms of coding to be considered on equal footing with industry coders before I start looking for an employer or become self-employed. I really liked your honest answer in the above post otherwise I wouldn't ask.

I don't mind answering, as long as you don't mind my rambling, . I don't know if I'm paid too much or too little. I do know that my company recently issued salary corrections. In general, we were paying developers too little. We were having problems hiring and retaining new developers. A colleague with similar general programming skills (C++, with a knowledge of formal object-oriented techniques, etc.) but no 3D was able to demand 6 figures when he moved to Seattle recently. Not in the game industry. I'm aware of a small development shop in Ohio that develops an advanced (but not well known) CAD package and pays its developers roughly 30 percent less than the local market average (straight from the mouths of developers there). Yet no one seems to leave them. The promise of fame and fortune in years go come.

It seems to me that the best experience you can have to get into the game industry is to have already worked on a completed game. Catch 22? Maybe. But maybe you can achieve this just by giving up your nights and weekends for a few months. My many years of experience made it possible for me to get interviews and offers, but now I'm lucky enough that my company thinks highly enough of me to take a chance, and they're paying me to explore the uncharted waters of the game industry.

Graham Rhodes
Senior Scientist
Applied Research Associates, Inc.
email: grhodes@sed.ara.com

Edited by - grhodes_at_work on August 21, 2000 3:17:04 PM

>> This is a longer answer than I thought I''d give. You don''t have to read it all if you don''t want to! <<

I want to... I want to... I want to After I reread my message I was afraid that I might sound to you like an idiot and you would not want to share your experiences with me. Boy was I wrong. This is the exact info I was looking for because I program alone and really don''t know what other people do in their daily jobs. By asking around I get a better perspective of what it takes to become a good programmer so for that I thank you very much

>> 1996-1998 development of an interactive simulation framework architecture and implementation of the framework for inclusion in the WorldBook Multimedia Encyclopedia on CD-ROM. The framework includes a mini custom 3D pipeline and scene graph framework used to display the 3 degree-of-freedom... The 3D pipeline is integrated seamlessly with the primary, sprite-based 2D engine, and has support for simple projected shadows. <<

I wondered about this myself when playing Hasbro Falcon 4. They probably used the same technique in their 2D mission planner/bomb loading screen where you could manipulate the f16 in 3D. How interesting.

>> 1991-1992 Developed a full software 3D graphics pipeline... <<

I like the flexibility of a software engine but like many others I work with Direct3D api. In a way I wish I had time to bury myself into software engine as I feel many algos can be useful to learn and experience writing them rather than read them from a book and dismiss them because the 3D api already does them.


>> 1993-1994 development of a 3D graphics package, based loosely on the pipeline from above, that we used to display pressure contours on airplane fuselages during simulated bomb blasts from terrorist attacks. <<

This was probably a fun project. Might be quite useful in today''s terrain engines for games. Destructive Solid Geometry on a fly

>> Development of a fast and robust computational geometry algorithm to compute the intersection of an arbitrary number of curves in 3-space <<

My current headache Well, actually I''m working with lines and just finished union, intersection, two difference boolean operators working on any two 2D shapes (both clockwise or counterclockwise winding order)creating infinite or zero number of inside loops. Currently am working on triangulation function that will cut the final result into tris clockwise/CCW ordered vertices and give them to the 3D api. Fun stuff but almost impossible to debug complicated shapes due to time it would take me to step through many loops and keep track of the changing directions, etc. I certainly have a lot to learn and do before this becomes a decent CSG editor.

>> They are kings in the game development world and I don''t compare at all (yet!) <<

I don''t think I ever will but I can try

>> A colleague with similar general programming skills (C++, with a knowledge of formal object-oriented techniques, etc.) but no 3D was able to demand 6 figures when he moved to Seattle recently. <<

Oh, my... I wonder if he got the job. Kinda crazy because us graphic programmers need to know other disciplines like math, physics, artificial intelligence, etc. all subjects outside of computer domain and yet some of you earn less in games industry.

<< Not in the game industry. I''m aware of a small development shop in Ohio that develops an advanced (but not well known) CAD package and pays its developers roughly 30 percent less than the local market average (straight from the mouths of developers there) <<

This scares me. Not that I''m greedy but I know developing CAD is very hard work. I would like to create a game CAD or a CAD with measurements and have that be my daily job at some company but as you wrote, it seems that there might not be enough demand for those type of programmers, thus lower pay and or more competition from other programmers. I don''t know if I would be happy doing other type of development outside of graphics because I like to see my results graphically I''m having blast watching my shapes take form and knowing that I''ve done that myself with a little help from books of course

Speaking of books, are there any good CSG type books that take more than few pages talking about this topic? Recently I bought "Real-Time Rendering" by Tomas Moller and Eric Haines (pretty good but poor on CSG more like 3D pipeline and speed-up techniques) and "Mathematics for computer graphics applications" by M.E. Mortenson. Using Mortenson''s book I was able to do the simple CSG (only a few pages in his book) but I am hoping for more rigorous books. Any that you like and feel are worth of purchase?

Once again, I thank you for taking time of your busy schedule to answer me. I really appreciated it. I wish more people could be open minded like you are.

Jerry

my homepage
E-Mail: BlueOrbSoftware@mailcity.com

quote: Original post by JD
Speaking of books, are there any good CSG type books that take more than few pages talking about this topic? Recently I bought "Real-Time Rendering" by Tomas Moller and Eric Haines (pretty good but poor on CSG more like 3D pipeline and speed-up techniques) and "Mathematics for computer graphics applications" by M.E. Mortenson. Using Mortenson''s book I was able to do the simple CSG (only a few pages in his book) but I am hoping for more rigorous books. Any that you like and feel are worth of purchase?

I like Moller and Haines'' book, and their website as well. I attended part of a session at SIGGRAPH this year, and Haines (or was it Moller) was one of the presenters. Unfortunately, I had to go to a meeting after one hour. I''m faced now with educating a customer about just how powerful desktop graphics have become. And about the fact that laptop 3D graphics accelerators are actually available now. The customer has been brainwashed by a person who seems stuck in 1995.

Mortenson also has a book dedicated to geometry, called "Geometric Modeling: Second Edition". It has some discussion on CSG, but still not much. I have it on my bookshelf, but its not a reference I use often. Mortenson is associated with a company called AeroHydro, who have a software tool (MultiSurf) for designing boat hulls, e.g, for racing. It is a surface modeling package. They don''t seem to believe in CSG or implicit surfaces or anything like that. They use a method called relational geometry synthesis where every intersection and edge/corner is explicitly modeled.

There is a book called "CAD/CAM: Theory and Practice" by Ibrahim Zeid, published by McGraw-Hill in 1991 that goes into a *lot* more detail on CSG than Mortenson''s book. It also has a very good discussion of boundary rep modeling. In the most general case, you might want to be looking at non-manifold modeling, which is what all the fashionable solid modeling kernels (Parasolid, SHAPES, ACIS) are supporting nowadays.

I also like "Computational Geometry: Algorithms and Applications" by M. de Berg et al., published by Springer. There is a new edition out. It talks a bit about the map overlay algorithm, which is basically 2D boolean operations. Only a few pages and no 3D, but it is a good book on some of the more fundamental geometry problems.

quote: Original post by JD
Once again, I thank you for taking time of your busy schedule to answer me. I really appreciated it. I wish more people could be open minded like you are.

No problem! I would rather be open. To hide everything seems paranoid.



Graham Rhodes
Senior Scientist
Applied Research Associates, Inc.
email: grhodes@sed.ara.com