Jump to content
Posted Tue Feb 27, 2018 11:18 PM
Posted Wed Feb 28, 2018 12:53 AM
Heh, I'm looking for a chardef program now, but browsing and getting distracted almost as much. I know I have good graphics programs on the iron, but it's been put away and parked for a couple years. It has the PEB and usual compliment with speech. MacV9T9 has a pretty faithful speech emulation. That's how I knew the roll when I first tried it, but I don't really need to hear the sound effects after a few games.
Your AI is tough to beat but I have managed to win a few. Just to get some revenge for my many losses, I added code to change my zero rolls into a 5, but not the computer's , and it was still a battle! When I added an Accept At so I could enter my own choice for the ROLL variable, then I got the upper hand and sqaushed him like a bug!
Fun (and cheating) aside, my little hacks help me get a handle on the code. I still think the dice are weighted in favor of the TI, though I know that's not really the case. Maybe my hypothetical rewrite (it's been quite a few years since I've done much XB coding) will include a difficulty level, so dummies like me have a fighting chance once in awhile!
Posted Thu Mar 1, 2018 1:09 AM
On one of Tmop's disks, I found a Char Def program by MMG Software that is similar to one that I remember (unfortunately not the really good one I have in my floppy library), but screen text and menues were converted to Italian, so I converted it back to English so I can begin to create new graphics.
I want to display the board horizontally to use the extra space. 8 board squares works out nicely for the 28-character screen width, giving me 3x3 character spaces per square or a 48x48 pixel tile size for each square on the board, with room for a two-character-wide outside border. My goal would be to replicate the original board's detailed graphics within the two-color limitations of TI character color definitions and as large as possible to have more pixels and colors.
I found another pic of the board in the museum with a bit stronger color. I corrected the perspective and printed it out to fit a standard sheet. It looks almost 3D , printed and set up using vintage buttons for playing pieces and a single die to serve as temporary surrogate for the four pyramid dice. Rolling a 6 is not counted and rerolled, while 5 denotes the Zero: Lose a Turn. Odds of a zero might be higher with one die instead of four, but a test game against myself seemed to play about the same as the TI computerized version. One side gets the upper hand and dominant positions, but all it takes is a late-game upset (or two) in the roll of the die to determine the winner.
Maybe the game has lasted for 4000 years because it took that long for some poor soul to finally win!
Edited by Ed in SoDak, Thu Mar 1, 2018 1:55 AM.
Posted Thu Mar 1, 2018 7:35 AM
Nice I think a horizontal orientation would work even better and yield more natural play. Looking forward to see what you can come up with. If you need help understanding the AI engine, please let me know. You should be able to use it as is and just graft on new graphics. Even better if you can improve on it. If you check out the original post (link towards the top of this thread), I have the AI algorithm laid out. One of its weaknesses is that if there is no capture or home run available, and it has 3 or more tokens on the board, it will always move the one further away if possible, and sometimes this results in it putting that token in harm's way when it could just introduce a new piece. It should be feasible to add a check for this if you are up to it.
You could actually build your own board using a thick wood plank with cutouts for the board neck (simple rectangular cuts) and print the design on transparent adhesive sheets to stick on top and sides. If you want to be really ambitious, you could even paint the designs yourself!
Posted Thu Mar 1, 2018 10:50 AM
Linky to that auction? I just learned of this game through this thread, sounds like you have a long history with it.
Actually, making/faking the board is my plan. I was going to paste the printout on top, then apply a finish over it and the wood. We've done that before and it gives an antiquey look. Might hollow out a little cubby beneath to hold the pieces. It should be easy to cut pyramid dice in oak, but I need to know where the dots go. Sandy's quite the painter so taking that route gets her involved, always a good way to picque her interest.
I'd love to make one out of inlays like the original.
Thanks for the assist offer with the AI. Knowing in advance my abilities (or lack thereof) you can expect a knock on your digital door. Usually if I study the code and try some alterations, I can manage to get it to work. It's a slow process but I'll take a serious stab at it first.
Posted Thu Mar 1, 2018 2:46 PM
Yes, I am kind of obsessed with that game
If you do end up building a board, please share some pictures!
There is an artist on Ebay who is selling a gorgeous replica. What a beautiful display piece!
UR Replica.jpg 257.78KB 1 downloads
Here's the link to the Ravensburger board: eBay Auction -- Item Number: itm/Vintage-Game-of-UR-Otto-Maier-Ravensburg-Made-in-West-Germany-Age-10-99-Xllnt/152879863868?hash=item239859983c:g:Gq8AAOSwlzRaU6dV
Posted Fri Mar 2, 2018 2:32 AM
I see there's numerous UR games being sold on ebay. Imagine that! That fancy one from Russia looks really cool, but I see it's likely cast plaster and painted. For $425 with shipping, I'll pass, but the maker was very thorough with his replication, so it is probably worth it to someone.
Sandy & I played our first game today after work. She won of course, while I still had three tokens to remove, even though she was kind and didn't jump me several times she had the chance. The old die we used was all we had and it didn't roll well and hindered the game. We just went with the flow. It was enjoyable for both of us and worth playing again. So now I'll proceed with our own handmade board.
On my TI graphics rewrite, I cobbled up a grid in a draw program that allows editing of every object and precise placement. Looks like it'll be a lot harder than I first imagined to do a decent clone of the original board. Just not enough pixels for the fine details, so I'll have to greatly simplify the symbols. Symbols are in a 3x3 character grid, allowing 24 pixels on a side. For the six symbols, this will take 54 characters with a few more for the borders and such, so it should be doable if I can get the look down.
Here's how the Rosette symbol looks overlaid by my character grid. Easy to see that with only two colors to work with, character-based graphics will look very crude without altering them to fit the relatively large TI pixels.
Oh, and I did finally get DOXbox to run on Sandy's Intel Mac. It runs the DOS version of V9T9 just fine. I was unable to get Wine to run Classic99 or install Java to use JS99er. Java installed ok, but kept reporting security issues when I tried to run stuff, I must be missing something. With a full Windows Install on the Mac, I can use Bootcamp to set it up, but I probably won't go that far. V9T9 while old at least runs on it, good enough for now.
Edited by Ed in SoDak, Fri Mar 2, 2018 2:37 AM.
Posted Fri Mar 2, 2018 10:22 AM
Oops, I meant the current version of V9T9, which does use Java.
Posted Mon Mar 5, 2018 12:01 AM
I made a bit of progress. Using the chardef program mentioned above, and a couple short XB routines I wrote to help plot it out onscreen, I worked out a very basic board layout that uses the full screen width. So far, I have only the "bullseye" symbol and a few simple diamond border characters, but it was enough to recreate the board's form. I'll flesh it out with more symbols to refine and complete it. For now, I just Photoshopped a little variation into the symbols to see how it looks. Not sure I like the divider border characters between the short single blocks, so my design may shrink 2 characters to be 26 wide instead of the full 28.
For the tabletop "real" version, I've made a set of 3-sided pyramid dice from two-part white plastic casting resin. That idea began when I decided to make a paper mockup to use as a crude guide for setting my chopsaw angles to cut them in oak, when I realized I'd just made a workable mold form to cast them instead. I printed out the layout on heavy cardstock and backed one side with wide packing tape so it would release. Folded them up, taped the seam and poured a mere 5ml of resin into each mold. They turned out fairly nice, considering, and kept that homemade look. I plan on spraying them in black and adding the white dots, but first I want to see how well they roll random results in a game or two.
I spared no expense, using only the very finest cardstock, buttons, plastic resin, tape, magic marker and laser toner I had laying around. Splurging on a piece of oak scrap, some glue and varnish will probably put my all-in cost at about a buck.
Posted Mon Mar 5, 2018 10:20 AM
Posted Mon Mar 5, 2018 11:12 PM
Thanks, you're being too kind! The 3x3 symbol gives a lot more graphic real estate than a single character. I need a better chardef program than I've found so far online. Looks like this Game of Ur will coax my ol' TI back out of hiding so I can boot up the chardef program I have on my floppies and transfer it. Fortunately I long ago indexed, numbered and cataloged all my 253 floppy disks to a single text file, so I'll able to find it quickly.
What's amazing to me is the ruthless AI you added to the program!
Posted Wed Mar 7, 2018 2:39 AM
I plod onwards with progress on several fronts.
I have a fairly simple rewrite of UR with a couple additions. I have found a single small bug in the version Vorticon released. Simply remove line 191 to fix the end of game "replay? Y/N" display. It functioned as written, but did not display.
I have kept the numeric score display I added early on, so sound is no longer needed to announce the roll of the dice with a small emulator display where the dice can't be easily seen. The speech commands could be removed if one does not happen to have a speech module or decent speech in their emulator, which should speed up gameplay.
I added a short intro screen with description of the keys used and a couple options. One option is a beginner's "cheat" that prevents UR's AI from jumping you and removing your token. It can still jump you, but it lands hidden beneath, and you keep your token! Hah! Take that, TI!
If your hardware or emulator can be accelerated, I added an option to add a delay to the player move code to slow down cursor movement. Set it to 0 for normal play or input a number that allows more exact movement of the cursor around the board.
I run MacV9T9 at 190% speed and 4.5 mHz clock for manageable but much zippier play, especially during sprite movement and other logic-intensive sections of the code.
Otherwise, it plays as before. I hope someone enjoys my minor enhancements!
I also unpacked and plugged in my Real Iron. Stock TI and basic PEB, three drives and SXB as my main cart. Dragged out all my floppies in search of my favorite CHARDEF program. Haven't located it yet. I guess I rearranged my oh-so-perfect disk indexing system, but it should in there somewhere!
Anyway, I'm having fun just revisiting things I've come across and it's good to see everything still works. I'm guessing some of these disks have not been mounted in 14 years or longer. Newest date was in 2014, some last-minute stuff I did before packing it away. Something's up with DSK2, but I have several backup drives if that's the problem. More likely it's due to flaky, jury-rigged wiring.
All part of my efforts towards adding my redesigned-gameboard-in-progress to a revised UR game and making it work!
Here's my rewritten UR4 program. In order to edit it, I was forced to resequence it, so line numbers are now very different from the original. Remove the ".txt" to reveal the V9T9 format program.
Edited by Ed in SoDak, Wed Mar 7, 2018 2:50 AM.
0 members, 0 guests, 0 anonymous users