Weekly progress: fiddling around with Dr. Destructo.

So, Return of Dr. Destructo has been in development for three years already. I first started coding sky drawing routine somewhere around August of 2012. I’m still missing a few pieces of the new art, and mobile versions are not ready, and the new UI broke some screens, because a lot of stuff was hard-coded to save time.

So, for the last week I got back to writing code for this game and fixing things. There isn’t much to report just yet, but you can check out an updated version of High Score screen:

Updated High Score screenNotice that it uses the new font, Catatan Perjalanan, which replaced PressStart2P. Both fonts are free to use, but PressStart2P didn’t go well with the new UI, so I ditched it. Which led to problems, since it was a fixed-width font, and Catatan Perjalanan isn’t one, so now I have to fix all screens to make sure they doesn’t look like crap.

Also, I do not think I ever put any images with the new UI in the old blog, so take a look at what a beauty this game has become in the last two years! ๐Ÿ™‚

Briefing screen
The new briefing screen. It also has some animation, but you’ll have to wait for the full game release to see it ๐Ÿ™‚ I think the new character portraits are spiffy!
First level
All new artwork. Check out the new UI in the bottom part of the screen. I think it’s neat!

Now, about this post’s title ๐Ÿ™‚ I don’t know if I can keep up WEEKLY updates, since there may be just not enough meat to put into them: none of my current projects show any rapid progress. But maybe I’ll try to make it, and maybe this move will motivate me to work more on my pet projects, despite the lingering heat that settled over Moscow. So, till the next time, folks!

 

 

 

I'm a video games programmer from Moscow, Russia. I have worked in game industry since 2008 on MMOs and mobile games, but I really want to make niche old-school RPGs instead :) In my free time, I like to play rock'n'roll on my guitar and travel.

11 thoughts on “Weekly progress: fiddling around with Dr. Destructo.

  1. I read the tale about how your work went down the toilet. Quite unfortunate. However, I happen to have a copy of your open_horizon and the qdf_tool if you want them. And, if you still play Ace Combat GFWL or Steam, just send my your username.

    1. I should clarify that it’s only blog posts and screenshots that are lost. Source code and binary versions of qdf_tool and OpenHorizon remain. Unfortunately, Razgriz (author of OpenHorizon) does not want to work on this project any longer. I guess I should put up a post with links to qdf_tool and OpenHorizon archives, so even if old links are broken people still could find them…

        1. He has hoped to get a few more contributors to join his effort, because it was a huge undertaking, but the response on Reddit and elsewhere was somewhat lacking.

          He also hit upon a few hard problems (for example, it proved to be very hard to locate the description of clouds used in levels), but having nobody to discuss them with, he got stuck, and finally frustrated with the project.

          I’m pretty sure he could resume it if other people offered their help, but so far, nodoby had.

  2. To be frank, Open Horizon wasn’t published that well. It took me several days to find something involving Assault Horizon extraction/development, before I came across your site as well as the Open Horizon links to GitHub. From what I could find, anything that had to do with Assault Horizon’s source code and/or elements was (and still is) sorely lacking. Another point to make is that Reddit isn’t the only (and neither is it the best) site to have a serious conversation about playing around with a video game’s files. Nor is it the best place to gauge users due to the wild scale of opinion on such sites. To that effect, there are other sites. For example, there’s a whole Assault Horizon community on Steam that has a higher and more concentrated number of the games fans and developers who would be interested in the work.

    Neither Razgriz or others are at fault, of course, but I’m sure there would be a few more people in the world that would know about Open Horizon if there was better marketing across the board. It would have helped to say the least (and can help if he’s interested in resurrecting the project).

    Another thing I want to point out is that when it comes to Open Horizon’s documentation, the amount present is insufficient to warrant serious development without a lot of frustration involved. It took me some time to figure out a few things in Razgriz’s code until I played with it. However, there are plenty of coders that don’t have enough time on their hands and would get easily frustrated with a screw-with-things-until-it-works approach. Just my own two cents in this regard.

    Of course, if he’s willing to restart work on Open Horizon, just tell me to send me an email of what he wants done.

    1. I agree with you on the “marketing” angle, but neither of us really knows how to do that, we are not members of big communities and do not know how to promote our works.

      I’ll relay your offer to him, and we will see. I would be glad to see this work restarted, even if I don’t play Ace Combat at all.

  3. Maybe, I didn’t clarify my point well or get to it succintly. But what I meant was that information about Open Horizon could just be posted as a thread or two on more communities. When it comes to promotion, just do what you already do on your website. An Apple commercial isn’t a necessity to explain to the world that a coder can do more with Assault Horizon than previously thought. Just explanations of how the code was acquired (disassembly, hex editors, etc.), what was tried, what failed, what’s still needs work, etc. And good documentation when you what your doing works.

    Furthermore if you want to take a look at what Steam Communities have to offer for Ace Combat, you can take a look:

    http://steamcommunity.com/app/228400

    1. Did you see Razgriz’s reply? He also tried to contact you by e-mail, but received no response.

  4. My approximate ToDo is
    – near clouds
    – far flat clouds
    – accurate shaders and materials params (reverse hlsl assembly via Intel GPA)
    – bake plane textures (because each camo layer is a separate texture) and make my own xml plane assets (what textures to use, camo colors)
    – effects (engine, trails)
    – ui (at least hud, possible with own xml configs)
    Then all code should be a bit refactored to make something like an AH graphical engine lib, and continue with ai, gameplay, more precise physics, etc

    I donโ€™t use cpu code disassembly, only PIX and Intel GPA to reverse graphics.
    To understand how configs affects game I modify values in files via my qdf_tool, run game and see what happens.

    I can easily write code and reverse graphics things, but kinda stuck with investigating assets data, so the biggest help one can provide is to reverse assets. I think itโ€™s not necessary to find out how all assets work, itโ€™s sometimes faster to write some small configs instead.

Leave a Reply

Your email address will not be published. Required fields are marked *