tuesday, february 4th 2003, 8pm

i dropped the password protection. all sourcecode and documentation is now freely availably on the 'docs' page linked above. any comments are appreciated and should be directed to the entire team (email addresses are available on the 'team' page). as stated below, i didn't get around to cleaning up anything, so the code is not in a finalized state. also, i tried compiling using visual studio .NET today, and it works after setting the additional include path to ..\library\dx7asdk\DXF\include, and the additional library path to ..\library\dx7asdk\DXF\lib.

so go check it out! :) grab the game (5mb, including all binaries, libraries and source code) or the initial architecture (90kb, game architecture framework). email any questions to [andy at nealen dot net].

 
wednesday, may 29th 2002, 1pm

the prototype has been finalized end of march, and now i finally got around to posting the prototype package on the docs page. it's still password protected, as i'd really like to polish it some more before the final release. am not sure though if that will ever happen, so it's probably more likely that i'll simply drop the password protection and write up some disclaimer on the status of the prototype. the zipped package contains all source code, documentation and executables which were generated during the design and implementation phases of drug runner's prototype. here two small screenshots.

when testing, please keep in mind that this is a PROTOTYPE! the front end is practically non-existent and the game must be restarted after a 'game-over'. but i'm quite pleased with the quality of the prototype, given the very tight schedule we stuck to. some work well done. we shall see if it ever gets touched again in the future.

 
wednesday, january 30th 2002, 5pm

added jimmy's cvs documentation to the docs page.

 
wednesday, january 23rd 2002, 7pm

everybody on track with the roadmap?

i worked ahead a bit (due to assignments i have to work on in the coming week), so as i have not been added to our unix group yet (i have an undergrad account though) i have posted my current game architecture code on the docs page. this code comes with a VC++ 6.0 workspace, and a linux makefile. i just tried it, and it compiles and runs just fine on redhat and win2k. i'm guessing you are all familiar with make, but just in case not, simply type 'make' in the directory you unpack the gzipped tar to and it should produce an executable named drugrunner. i am still using the bouncing spheres dummy model, but have added the type database, and what is even more important, i have now implemented the cleanup routines, which are called when the model database is asked to destroy a model. this was a bit of a tedious task due to my lack of familiarity with the STL, but i think i am up to par now. i am also currently working on getting the sgi STL to work under win2k, so we can use the nice hashtable included within.

our first milestone is close (28th, monday), so i would strongly suggest a meeting as soon as possible. i will not be available all day this friday, but could meet at any time thursday, saturday or sunday. furthermore, i'd suggest you all browse through parts of my architecture code so we can discuss it during the meeting. i think this will be very, very important for further communication, so please take the time. i will gladly answer any questions and am open to any form of constructive criticism.

one one last note: i want to implement a framerate counter to the rendering part, but it is not clear to me if anyone is currently working on the font engine, i couldn't find a clear assignment in the roadmap. i will gladly take that part over for the time being (as the basic architecture is almost in place), but would also like to avoid redundancies. so please let me know, if anybody is working on that by the weekend.

other than that, hope everyone is making progress on their respective parts!

 
friday, january 11th 2002, 2pm

the new and updated game roadmap is now online, including the timetable for the specific tasks and the person responsible for them. this is (again) work-in progress and can be modified. i already realized, that the camera model will be needed much earlier in development and that it would make more sense to leave that to the rendering/modelling team (jimmy/cynthia). The AI/gamelogic team should also freely distribute the workload (serif/alan/connie). i'll be placed somewhere inbetween, i guess. and i also have to re-evaluate our progress before every milestone and update the roadmap accordingly.

i'd suggest bi-weekly meetings with the whole team, scheduled in such a way that we meet about 2-3 days before every milestone and (at least) once inbetween two milestones. our first milestone is january 28th, which doesn't leave us much time to get the initial manifestation going. so except for some smaller, group related meetings i think we should wait till the end of the month to review our progress. perhaps around the 24th. just a suggestion.

so good luck to all of you and let's get this thing going!

 
thursday, january 10th 2002, 3pm

posted the first version of the game roadmap on the docs page. this is very much work-in-progress and must be updated after our meeting today so it can be forwarded to david forsey.

 
thurdsday, december 13th 2001, 9pm

something is not working with the messageboard, so i'll post this message on the news page. i just got done with my final term project and am completely exhausted. as i recall, the roadmap document is not due till the 7th, right? so i would suggest we meet a day or two before that for an extensive meeting during which we also write up the document. i guess the doc will not be more than a few pages long anyway.

i'm pretty sick of work now, so i hope you all understand that i would like to avoid more of it this year. we have surely all deserved a break. i'll be back from whistler on the 2nd of january, so i'd suggest meeting on friday the 4th.

cheers

 
monday, november 26th 2001, 6pm

i just found a dreadful error in my code (and therefore in the techdoc) in the most crucial part, the mainloop! i actually try to resolve collisions before moving the objects (duh). the correct mainloop code would be:

    GetInput();
    if (m_timer->HasTickPassed()) {
    // for each frame
        m_kernel->RunProcesses(m_timer->FrameRatio());  // move stuff
        m_colMgr->resolveCollisions();          // resolve collisions
        m_renderer->RenderScene();          // render stuff
    }
    
as soon as i get to it, i will update this in the prototype code and both techdoc versions on our webpage. sorry about that, but it was late :) with the code corrected, spheres do not accidentally pass through the bounding walls and also change state correctly. in the final code, all these states will be double buffered so the controllers behave consistently. (see pattern double buffered state on the game patterns page).

 
friday, november 23rd 2001, 3pm

it is DONE! added the finalized version of the technical design document to the docs page (as a word .doc and as an html version). This is the final version we are handing in today. Please note, that the html is simply dumped from msword and looks awful; good enough for browsing though.

 
thursday, november 22nd 2001, 2pm

added OBJ file specification to the links page. thanks jimmy!

 
monday, november 19th 2001, 9pm

added cynthias part of the tech doc to the docs page (a little late, sorry :)

we met today and discussed how we will go about writing up the remainders of the tech doc and that we will meet thursday evening at 6pm for a final merge and cleanup of the document. i'm pretty sure this will be an 'all nighter' (well, at least for jimmy and myself), so try to bring some time. on friday, it'll all be over :)

on the docs page you will also find the first basic engine prototype i have been working on for the past week. it would be cool, if all of you could familiarize yourself with the code and give me constructive criticism. the code is meant as a rough architectural guideline, so we can make the tech doc as consistent as possible. please take a (closer) look and let me know what you think. keep in mind, that the code contains a few nasty hacks (which i have pointed out) and is in no way stable. the only model type currently 'implemented' (if you can call it that) is the 'sphere'. all others are just dummies, introduced to show their (possible) placement in the overall architecture.

and now, i will sleep :) goodnight.

 
thursday, november 14th 2001, 2pm

i worked on some timing issues for the main gameloop last night and used some pretty cool c++ stuff i found on the web. it uses a high performance tick counter (microsecond resolution) for both win32 and linux platforms. from my notes i see that jimmy is responsible for this part, so we should team up on that one for the documentation. I also tested the general purpose hashtable (for buffers of memory) which i assume we will be using to organize the contents of each cell (i.e. each cell will have a hashtable of models, etc. associated with it) and it works fine. Dynamic memory allocation might be an issue, but i already found the tweakable spots of that container so we can grab a big chunk for each cell at initialization time. I got a cool matrix/vector library, which i tested, and it also works. It has nicely overloaded operators, but it doesn't support quaternion calculus, yet. And finally, I got a general purpose (templatized) version of the STL priorityqueue running, which has the advantage that you can update priorities and resort the element in the queue in logarithmic time (STL priority_queue doesn't support this without using some nifty STL algorithms).

the actual reason for this posting is that while researching all this stuff i stumbled over online versions of both 'design patterns' and 'effective c++', two of my absolute favorite books! model-view-controller architecture and the 'memory pooling' stuff i've been talking about all comes from just these two books. so check them out on the links page when you find the time. the 'memory pooling' is item 10 in the book 'effective c++', which describes how to write your own operators new and delete.

 
wednesday, november 14th 2001, 4pm

added connies techdoc draft to the docs page.

 
wednesday, november 14th 2001, 12pm

added alan and serif's tech doc draft to the docs page. I will be posting all drafts you guys send me to the docs page, so we all have access to the other parts of the 'work in progress' to see if we are all heading in the same direction. so as soon as you have something worth posting, email it to me.

 
monday, november 05th 2001, 8pm

added a link to 'car physics for games' to the links page. this small tutorial already includes almost all we will need to simulate quasi-realistic 2D physics for the vehicle model including sideslip. there are some other interesting links at the bottom of the page as well.

i also added the 'email reply notification' option to the messageboard, so you'll be notified if someone answers to your posts.

 
sunday, november 04th 2001, 4pm

added second draft of the technical design document to the docs page.

 
monday, october 29th 2001, 9pm

added the first draft of technical design template and our meeting notes to the docs page. the current deal is, to get done whatever is possible until friday 1pm and email it to me. i will then merge the parts so we can pass it on to dave forsey for some necessary feedback. i'll also add a part to the introduction, describing our plan to build a miniature prototype of the game engine using the model-view-controller architecture and game design patterns early on in development (possibly mid december for approximately 2-3 days).

 
friday, october 26th 2001, 7pm

ok, just as a reminder (mainly for myself): our next meeting is on sunday, october 28th at 3pm. if i find the time, i'll write up an agenda so we can focus on the important stuff for the technical design doc. also added a link to the official 'grand theft auto 3' site to the links page.

 
wednesday, october 24th 2001, 5pm

just updated the links page, adding a link to 'design patterns for computer games'. a very good read based on the model-view-controller (MVC) architecture. might be good for the high level view of the overall design.

 
monday, october 15th 2001, 4pm

uploaded newest version of the 2nd design doc draft. also posted a message on the messageboard concerning a possible meeting this week.

 
sunday, october 7th 2001, 4pm

added some links to the page. also uploaded an updated version of the preliminary design doc. and finally, the html version of the first draft is also online. i am currently adding dave's comments, so expect that to be up soon too. the html of the design doc is generated by microsoft word and not hand optimised, so don't expect too much of the layout :-)

 
friday, october 5th 2001, 3.15pm

posted the final version of the first draft on the docs page. this is the version we will give the producers today for further review.

 
friday, october 5th 2001, 2.15pm

again, newer version of the preliminary design document online (120k). merged parts from jimmy, connie and myself.

 
friday, october 5th 2001, 1.30pm

newer version of the preliminary design document online.

 
thursday, october 4th 2001, 6pm

did it! ok, just posted my part of the preliminary design document. Please register, that this is a 'work in progress' and by no intentions final. I have been brainlessly hacking away at the keyboard just to get all the stuff out of my head that we had thrown up and discussed in the past few weeks. i'll be gone from here now, but hope to find all other missing parts in my mailbox tomorrow by 2pm so i can do a final merge and printout. i will not be able to print it out seven times like last week, but will bring three copies along to class, two for us, and one for the producers.

 
tuesday, october 2nd 2001, 4pm

i just linked to a new messageboard provider, as the missing threads really sucked bad. try it out and let me know what you think.

 
tuesday, october 2nd 2001, 5pm

connie mailed us all a suggestion for the game-logo and i like it. i think i'd change the font, but i like the dark, sinister style. also, i found a link to some screenshots of an xbox batman game which has quite a nice visual style. check it out here. so here's the logo:

looks not like all of us had time to test the forum. go ahead, i'd like to know if all of you are comfortable with the 'german' stuff in it. i will get to the gameplay part now. after that it's cartoon rendering and particle systems. i found some cool demos with sourcecode on gamedev.net (see links page) yesterday, so i have enough resources for my research.

 
monday, october 1st 2001, 8pm

ok, added the 'links' page. again, please bring on suggestions for good links and resources. the links currently on the page are the ones i mostly use for any graphics or game related research (aside of the obvious 'google' of course).

 
monday, october 1st 2001, 7pm

first redesign. got the forum and our document repository linked in the navigation bar. more to come as time permits. please give feedback, and let me know if i got any names or e-mail addresses wrong on the team page.