Ok, good news first. The bid now rotates correctly.

The bad news, this update made the MP less stable. We struggled to get through a match since it kept timing out or freezing. 5.1 was far more stable, so not sure what changed to cause that.

Another question. Do you think it matters if some of us are using the HD app and some the regular?

Thanks for all your help. This has been a fun outlet during the quarentine.

Really? That's surprising. It's pretty simple to put it back how it was. Intuitively, this should have been better, but laws of unintended consequences rule some times.

5.1: The non-host phones can race ahead of the host once a hand is complete and begin the shuffle. This kind of behavior used to spell instability trouble.

5.2: Forces the non-host phones to wait for a signal from the host that all phones were ready to be shuffled.

So, it makes sense to me that 5.2 should be better, but I'll revert that bit of code tonigh

Alsot, it should not matter about HD and non HD.

Yeah, not sure why it was less stable with the update. I'd wait to see if anyone else has experience with the MP to see if it got better for them.

Been playing with 5.3. I'll upload it tomorrow.

I love 42. Is there an android version anymore? I'm wondering if this is the same one I used to play on the PC years and years ago?

Edit...I believe this is the same one I bought back in the 90s. I found it on amazon's app store for ANDROID here:
https://www.amazon.com/MaroonEntertainment-Texas-42/dp/B00EUY85B8

Definitely not the win42 version, but I'm flattered.

Did 5.3 get uploaded? Not seeing it. We are having pretty terrible problems staying connected in a MP game on 5.2.

It's in the review queue now. If it doesn't get approved soon, I can always make you a beta tester and get it to you sooner.

HD is approved and ready for sale. The iPhone version should be ready soon.

I've had you're game from when you first released it-just had my boys pick it up.

It's great for learning-playing versus AI or multiplayer when mom is at work and we can't sit at the table.

One observation-when bidding two marks for splash/plunge winning awards 3 marks instead of 2.

I think that's how it's supposed to work? Otherwise if you bid 3 marks off the bat, you are telegraphing that you are splashing before the bid is over?

We still are having trouble even starting a multiplayer game with just 2 people. We text the invite to each other, tap the invite, a message says "Retrieving" and opens Game Center. Then nothing happens. Any ideas?

We tried playing with 4 of us a couple nights ago. We were never able to make it past a couple of hands. One time it got stuck for one person and his screen would not update, then the next 2 games we started both froze on the 'synchronizing' message on the 2nd or 3rd hands.

Best advice I can give is to make sure everyone is on 5.3.

Until I can get in a game with some of you guys, I'll only be guessing at the problem.

So send me an email at maroonentertainment@gmail.com and we can try and work something out. Might be possible to use XCODE with a live app to see what's happening when it stalls.

Unfortunately, the answer has often been that the Game Center service blows. It drops connections out of the blue and doesn't offer the developer a way to bring the connection back seamlessly.

I thought the connection issues were just us.

No. It's always been systemic to a certain point from what I can tell.

I've added features in the background where it pings the server every two seconds or so just to try and maintain continuity, but I'm not sure how much it helps. It is weird though, I can usually play with my mom across country with no issues, but other times have issues with two devices in my hands. Even then, it would be sporadic. Issues one day, none the next.

I'm certain the only real way to fix it is to go to a third party server. It's just a lot of work to get that running...plus it isn't free. I've never heard complaints of MP from android users. Anyone have any experiences with it in the wild?

Fry, is it possible to pay you to create a private server for a small group of friends to use instead of Game Center? We have terrible problems using Game Center even after all your updates. Our group would be willing to pay extra to create something more usable.

Thanks and reach out to my username at Gmail if there's anything you can do.

Bump for Phillip to see.

No longer in the App Store?

I let it lapse. Just didn't feel right keeping it up there when I don't the ability to manage it like I used to. I recently started porting it all over to UNITY and updating the graphics, but haven't been able to touch it in a few months. My hope was that ChatGPT could be my personal developer, but it's honestly it's not there yet. Maybe v5.0 if it ever comes out can help me finish it.

Been a busy couple months. The code for single player has been almost completely ported over to C++. I gave up on UNITY as my old brain just refuses to learn it for some reason.

That said, v3.0 will be completely compatible between android, apple, and even windows when I get around to the multiplayer side. I'll be asking for beta testers here soon!

Right now, I'm working on hunting down some of the dumb bugs/moves and basically re-writing whole portions of the bid and play code. That said, if you were Jay, what would you bid with this hand any why? Curious to get some early feedback.

I'd give it a go at 30 with 6 as trump.

Same. Lead 6-2 hoping to draw out the 6-6 so I can trump back into the lead later. But I might consider passing more hoping my partner would bid. That's a good support hand too with all that count.

So what's happening in this game is Jay is winning the bid with a 30 and making 6s trumps. Really close call on 5s as trumps though so when I build in some randomness, it may push it over the edge.

First move is a 6-2. Fry is playing the 1-6, me the 0-6, and Philip is winning with a 6-3. Unfortunately, my hand sucks for support.

I mostly agree with this move since it doesn't seem reasonable that Jay would play the 6-2 without owning the 6-4. Where I hesitate is whether I really came up with code smart enough for Fry to figure this out. So while I'm happy with Fry's move, I want to understand better how he's making it before I pat that code on the back

Philip wins two tricks here and starts with the 0-1. Jay trumps in with the 5-6 which blew my mind at first, but now I understand why he's doing it here and I'll have it fixed today. Basically, he knows the 66 is still out there and is afraid of losing the ten spot but isn't correctly accounting for whether Fry would even be able to play it. The code involved with deciding to Trump and what to Trump in with has gotten a little unwieldy over the years and needs a bit of an overhaul. Honestly though, I'm just happy having it up and running as it was quite a bit of work to get it going.

6 trump was my first thought, as well, but I think I would go 5 trump. I would want to guarantee I get one of the 10 counts. You can attempt to flush out the 6-6 with the 6-2, but if they have two 6's then you have to burn the 6-5, as well.

I would have passed if my partner bid or if I have no confidence in the dealers skill.

How about this one? Fry won the bid with 30 on aces. How annoyed would you be if he played the 44 instead of trumping in?

His only hope to win at this point is to trump in and hold off playing the 4-6 until the last round while hoping I play the 66 to avoid giving up the 55. I've been hesitant to add a ton of monte carlo simulations to that code. For one, it sometimes results in fairly non-human behavior, but more than anything, I have a concern that it will make the game unfairly hard to win against.

Can kinda see both ways on it. In a bad situation no matter what, you either give up the lead immediately and hope someone throws the 66 and 65 out or waste a trump and still give up the lead at the end.

If I couldn't see the other dominoes, knowing me I'd be more worried about the 66 being led so I would want to stay in the lead as long as possible in case someone sheds it, so I would play 12 then lead 44 and 13 afterward.

I think he has to trump here; he has a bad 64 lead eventually but letting opponent keep lead and just delays it and gives opponent the chance to pick up points in the meantime.

OP questioning the AI logic when he passed on bidding with the 5-5, 6-6, 0-0, 0-5, 0-4, and 0-3

Kampfers said:

---

OP questioning the AI logic when he passed on bidding with the 5-5, 6-6, 0-0, 0-5, 0-4, and 0-3

---

Yeah, that's an easy two mark bid.

Philip J Fry said:

---

How about this one? Fry won the bid with 30 on aces. How annoyed would you be if he played the 44 instead of trumping in?

His only hope to win at this point is to trump in and hold off playing the 4-6 until the last round while hoping I play the 66 to avoid giving up the 55. I've been hesitant to add a ton of monte carlo simulations to that code. For one, it sometimes results in fairly non-human behavior, but more than anything, I have a concern that it will make the game unfairly hard to win against.

---

Yeah, definitely playing trump, trump/44, 64. He's going to get one of the 10s. But, if I'm you and your partner I'm holding count until the last hand.

Is the code publicly available anywhere? I'm curious how you determine bids for the bots, as well as what they play on any given hand.

My code? No. Not something I would share freely.

I can say generally speaking, it uses a scoring system based with bonuses and penalties depending on what tiles they have, what AI (if you can call it that) setting they have, and what their partner bid. It does not use Monte Carlo's or anything. So it isn't perfect compared to running several thousand iterations to find the best move/bid. Sort of a conscience choice.