Skip to main content
GameDev.net gamedev.net
🔒 Locked

We have some options - what do you want to do?

Started by khawk Jul 16, 2003 at 9:12 AM 76 replies 7.2k views
Original Post
khawk
After thinking about this contest and all the feedback that I have received, I have determined that we, collectively as a group here, have a few options. Option 1: We start the contest with the arena executable (almost) as-is with a few defect fixes. Advantages:
  • The contest starts within the next 2 weeks.
  • The core of the arena works well. It's just the presentation that needs work. Disadvantages:
  • The arena executable may not work properly for everyone.
  • The defects everyone has been seeing with the download, for the most part, will remain until a later date.
  • Everyone accepts the interface as-is.
    Option 2: I rewrite parts of the arena executable to allow for some of these other features that have been mentioned. Advantages:
  • The arena executable will (hopefully) work well on everyone's machine.
  • All defects have been fixed due to a redesign.
  • Some of the features that have been requested are implemented, or at the minimum, easily added due to the redesign. Disadvantages:
  • The contest will not start anytime soon. It could be another 2 months or more (keep in mind I only worked on this for about 3-5 hours per week for 2 months.. that trend would have to continue).
  • Potential for new defects due to redesign.
    Option 3: We start a contest without prizes (for fun) using the current arena executable - a sort of trial run. During the lifetime of this contest, the arena executable and the interface is redesigned and reimplemented to fix defects and include improvements/features based on feedback. Advantages:
  • We have a contest.
  • We get immediate feedback on: the arena executable, the arena interface, and the contest format.
  • Everyone gets a chance to give their input to make a better product (product being the contest).
  • The first official contest would be (in theory) better due to a low defect yield and a reworked/rehashed/reimplemented arena with improvements.
  • Everyone tests the current arena for the contest, as well as the new arena for feedback. Disadvantages:
  • We do not have prizes.
  • The real contest does not start for another couple of months - at least until this first one is over and the new arena version is ready.
    If anyone has any other advantages/disadvantages for each of these, I'd like to see them. Basically, I realized that my attempt at quickly developing the arena executable might not result in the best and most fair contest possible. And given that I cannot stand having people use my software if it is buggy, I would rather improve it if that is what everyone would rather have. Another reason is that this is a completely new contest for GameDev.net, and while I do not expect to get it right this first time, I would like to make it as good as possible. I leave this up to you, however. If Option 2 or 3 are decided, then just remember that I will not be working on this fulltime, and I will not be spending a couple hours every night trying to complete it. I have a life that I want to keep, a book that I need to get busy on, and everyday work. However, I would spend a few hours per week, and I would be willing to make this contest better if that is the consensus I received from those interested in participating. Also, feature creep will be kept at a minimum. If a new version is to be created, it will have a defined set of improvements, and only those improvements will be included. Anything else will have to wait until after the "official" contest. ***EDIT: To clarify a point made in a reply below, these features have already been determined for the next version of the arena executable. While people may request many features, some extraneous at this point, I already have a list of improvements and fixes. So by saying that feature creep will be kept at a minimum, I mean that my list of improvements is already defined for the next version, and new requests during that version's development will not be implemented. Anyway, as always, provide comments, suggestions, ideas, your name and credit card number, etc. and let me know as a collective whole what you want to do based on the advantages and disadvantages of each option. *** This post has been edited. Please review it for any changes since your last visit. ***
  • Admin for GameDev.net.
    SilentReaper
    SilentReaper
    My opinion is for #3 or a variant thereof. Start the for fun contest now -- it will maximize our amount of time we have to develop for the competitive contest. For those that feel like they would participate in the for fun contest and it would be unfair for both contests to not run on the same code, how about starting a new for fun contest once the competitive contest starts? And I think all of us certainly understand time constraints of real life. It seems to me this would take some of the stress off of you and give us something to play with in the meantime.
    Ysaneya
    Ysaneya
    I vote for Option1. That would allow you to work on Option2 for a new contest next year.

    Let''s be honnest: if you go for option2 now, people will never be happy. They''ll want more languages support, more portability (linux, 2D version), they''ll want more weapons, they''ll want bigger arenas with more complex obstacles, they''ll want 2v2, 3v3 (or more) matches, etc.. If you go for option2 now, i have the feeling this contest *might* never see the light of the day at all..

    I don''t like Option3 very much.

    Y.
    TheGhost
    TheGhost
    Option 1 wouldn''t be good for those who want to enter and have issues, so I think either options 2 or 3 are a must. I personally don''t care for prizes, and hence option 3 is most appealing to me. I think it would be fun to challenge others in the mean time. The only problem with this option is that it gives people time to build up experience and practically dominate any future contestants who know little to nothing about the system.

    Alternatively there is another option. You could make the program open source and let the public finish it up. Rather than actually developing the code you can just spend such time supervising the additions and modify them if people tend to go off track. Or perhaps deligate the activity to a localized group where things sould be more in control. In this case option 2 is possible with the change from X months to complete to X weeks to complete.

    In either scenario, option 2 would eventually be nice since I''m certain most people will want to make a ladder out of this game. It would be an effective way to attract new AI developers.
    Turt99
    Turt99
    I would say 3, that would allow anyone that wants to get going to do so but it wouldn''t stop others that maybe having a problem with some of the bugs..

    this would be a great way to test the system, you''d get all the the bugs worked out in the program and the whole contest format.. that way when the contest is started for prizes it would run smooth and there would be less chance of someone feeling cheated because of a flaw in the design of the program or the contest its self



    Please visit Turt99 Productions
    FOLLOW ME ON TWITTER OR MY BLOG
    okonomiyaki
    okonomiyaki
    I would say Option 1, except that you just work on the exe to work on most machines. Just work on the bugs that simply stop it from running on certain machine, and use that, and like someone else said, start developing a better one for next year''s (or whenever) arena contest.
    superpig
    superpig
    I vote #3. You never know what logistical problems you''re going to get until you actually send it out to a whole bunch of people; I''d rather that things be going horribly wrong while there *isn''t* a significant prize at stake. (Awarding a ''is teh win'' title next to the champion''s name would be small enough to not qualify as significant while still being cool).

    I''d also recommend you get a little help writing it. Getting other people to look at it will often help uncover bugs, and if you pick the right people then it''ll all be done faster. I volunteer to help out, if you''re interested - I''ve come up with this engine we might be able to use...

    Superpig
    - saving pigs from untimely fates, and when he''s not doing that, runs The Binary Refinery.
    Enginuity1 | Enginuity2 | Enginuity3
    Richard "Superpig" Fine - saving pigs from untimely fates - Microsoft DirectX MVP 2006/2007/2008/2009
    "Shaders are not meant to do everything. Of course you can try to use it for everything, but it's like playing football using cabbage." - MickeyMouse
    AtelierVisca
    AtelierVisca
    Option 3 ... that way people who aren''t familiar with AI get a chance to get acquainted with it and everyone gets a chance to see other tactics in action. Then, when the real contest starts, it will be a strong competition. It would, in a sense, be an extention of the initial idea of uploading "rough" bots half way through the contest ... instead, we''d be writing bots completely ... and then adapting to the new arena, opponent tactics, etc.
    James
    bastardos
    bastardos

    I like option #3. It would give beginners (like me) a chance to learn the basics, and (as we all know) the executable still needs some work.

    BtG
    .basprintf ( "And for the %dth time, I'm not American!", ++lIdiots );My homepage
    xEricx
    xEricx
    I''d say go with #1

    It worked well on all my PCs, which have different configs / OS...

    As for the problems with downloads... CDR are cheap as hell, and almost everyone has broadband connections nowadays... so I guess it would be pretty easy to ask a friend to download it for you or even get it from an internet cafe or something.

    For the compatibility issue... I guess only people with exotic videocards are experiencing problems?
    slepyii
    slepyii
    I would opt for Option 3. This would give you a chance to work out bugs that may only appear when someone is writting and running their bot in the arena. These wouldn''t necessaraly be crash bugs, but logic type bugs. This would also allow for the arena API to be tweaked if need be before the official contest begins.

    I would also agree with making the Arena core Open Source so that others can help develop the arena code, which should make it more stable in the long run. Of course you would be in charge of what was approved for changes, and what was not. I feel that making the arena core open source would not give anyone an advantage over anyone else as the competition is in writing the bot AI and not how well you can interface with the Arena.

    - Timothy
    Wandering Relic
    Wandering Relic
    Man, some of the replies so far are scaring me. I was really looking forward to entering a worth while contest. It is true that the current executable is buggy. I fixed the main slow-down/white screen problem by replacing the 11MB model with a more light-weight one, and shrinking the textures to reasonable sizes. A lot of systems just couldn''t handle the unoptimized resources.

    My main concern is that by having an unofficial contest or by releasing the source code, there no longer is any competition left among players for the official contest: everyone will have access to the resources necessary to make super bots.
    Naaga
    Naaga
    My vote goes to #3. Just think of it as a beta test tournament or something. There should be no prizes, not even a title next to the name or anything, as that would be unfair to those who can''t get it running right away. It''ll give you time to stabilize the exe, and it''ll give us time to get used to how the bots work, which will make the first official contest much more interesting and enjoyable. I also like the idea of people running there own unofficial tournaments. That could get some really good competition going.
    __________America seems to like crap because its what we make popular. - Goober King
    khawk
    quote:
    Original post by CodeDemon
    Man, some of the replies so far are scaring me. I was really looking forward to entering a worth while contest. It is true that the current executable is buggy. I fixed the main slow-down/white screen problem by replacing the 11MB model with a more light-weight one, and shrinking the textures to reasonable sizes. A lot of systems just couldn''t handle the unoptimized resources.

    My main concern is that by having an unofficial contest or by releasing the source code, there no longer is any competition left among players for the official contest: everyone will have access to the resources necessary to make super bots.


    I am actually looking into this (the changing of the model).. I have not tried another model yet, but if it makes the contest more accessible, we might go ahead with it while I rewrite what I would like to rewrite. The executable may still crash occassionally, but since that crash is primarily during shutdown, it may still be worth running the official contest now (option 1), or maybe even a mix of options 1 and 3, which was my original plan.

    As for open source, I do not want to do that. Not for a while anyway..
    Admin for GameDev.net.
    Wandering Relic
    Wandering Relic
    quote:
    Original post by Khawk
    quote:
    Original post by CodeDemon
    Man, some of the replies so far are scaring me. I was really looking forward to entering a worth while contest. It is true that the current executable is buggy. I fixed the main slow-down/white screen problem by replacing the 11MB model with a more light-weight one, and shrinking the textures to reasonable sizes. A lot of systems just couldn''t handle the unoptimized resources.

    My main concern is that by having an unofficial contest or by releasing the source code, there no longer is any competition left among players for the official contest: everyone will have access to the resources necessary to make super bots.


    I am actually looking into this (the changing of the model).. I have not tried another model yet, but if it makes the contest more accessible, we might go ahead with it while I rewrite what I would like to rewrite. The executable may still crash occassionally, but since that crash is primarily during shutdown, it may still be worth running the official contest now (option 1), or maybe even a mix of options 1 and 3, which was my original plan.

    As for open source, I do not want to do that. Not for a while anyway..



    Yeah, I don''t want to be pressuring you either, as I wouldn''t want you to neglect your most recent "find". I say that if you can fix it up so that it runs somewhat reasonably, within a couple more days, then run the contest; otherwise, go with 3 or a mix of 1 and 3.

    Topic Locked

    This topic has been locked by a moderator. New replies are not allowed.

    Sign in to reply to this topic.