Topic: [ANN][MOTO] Motocoin - page 51. (Read 178256 times)

I would be willing to build the projects but I don't have the right dependencies installed. I need to install GLEW and glfb3 or something but I don't know how to get them.
I got them installed the first time I installed ubuntu but it didn't have enough space (only 25GB) so I had to make the partition bigger (125 GB) and reinstall it.
The "backup" of all of the packages I had installed didn't work because I think it was for a different version of Linux but the person didn't say.

Edit: Nevermind, I just installed glew and glfw-3.0.4 and now I can compile the game. I wonder if I can get any coins on the real net.

Edit 2: Wow, someone is still on the old client. When I started syncing it was out of 8001. That means the the first person I connected to was on the old client, right?

Edit 3:
I can successfully build the wallet! But it is with Qt creator and only for linux. Would I still be of help?

(EDIT: "A few days" might fall beyond the 104k block mark? )

How about we actually subscribe to the gitian philosophy and both make the builds? This way the community can know not to trust our builds if our signatures do not agree?    Ideally we should find more people willing to build, and have a larger group of signers.  (The more the merrier.)


I mean that we should move the coin/ground interaction and first frame death checks into the optional filters, and not do them at all, instead of have them done-but-ignored like originally, when verifying proof.  (From some quick testing, this also appears to speed up initial sync slightly.  Bonus!)

I will try again in a few days if you will not do this earlier.

This wasn't a bug, it was a feature and I knew about it. If you add mapfilter to the game will you then make all the maps that are considered bad by this mapfilter to be invalid? Of course no. This was just a simple mapfilter to exculde maps where coin is partially inside the ground. The solution would be simply to ignore this check during replay verificarion.

The original dev says that all premine (both their own holdings and what was "reserved" for bounties/promotions/etc) was already dumped into the market.  Some people dispute the possibility of this being true because one of the premine holdings addresses is still top of the "rich list" with 150k coins.

Personally, I don't really care what the truth of the matter is.  It would indeed be quite noble of the original devs to transition any funds remaining over to a new team, but I highly doubt that this would happen in any case.  (Has there been a coin yet where either the dev team was changed, or there was a community takeover, where the old devs did this sort of a hand off?  I'm not aware of any, but would be very interested to read about such a history.)

Does this imply that you still intend to make builds after all?  From the tone of your PM it sounded like you couldn't be bothered.

I had only been testing live net PoWs with the first 10k blocks, since I was mostly interested in verifying difficulty target.  You also could have just as well tested this before now, too.


Yes, it appears that in the original version the world generation result was checked in motogame when making maps, but not checked again when verifying a proof.  This was a latent bug that was exposed by making the world generation result checked when verifying proofs.  I did not change the "coin in ground check" itself, but I did make it so that this check is now performed both in map generation and proof validation.

The check doesn't actually test that the center of the coin is inside the ground, it tests if the center of the coin is within the wheel radius distance (why this?) from the coin.  In the first block it fails on, for example, only the top left edge of the coin intersects the ground.

We have a couple of options here.  

A) We could make the check conditional to the block height, so that these maps would invalidate proofs past the fork point.  I think this would not be good as it would make a lot of perfectly valid maps considered invalid.

B) We could change the result of the world generation from being a bool to being some enum value that indicates why the map was invalid, and keep the existing semantics where motogame disagrees with the network itself as to what constitutes a valid map.  This would probably be the easiest approach, but I'm not sure that the existing semantics are really sensible to begin with.  All components of the system should be in agreement as to what makes for proof validity/invalidity.

C) We could move this check outside of the "bare" map generation used by motocheck and into the map filters, so that proof validity definitions would remain consistent and miners would still have the option to use or not use this check in their filter selection.  This seems to be most sensible to me.

Is there another option that we should consider?

If the original Dev has given up, make sure he hands the pre-mine over to the new Dev.

the coin game looks like very interesting,lol

Seems good that I didn't build it. You changed rules for checking replays and didn't event test if old replays are still valid under these new rules. Issue is that replays where center of the coin is inside the ground are considered invalid in new version.

Oh, I thought there was a problem with the normal client. Well I'm stupid as I could have just looked at how many blocks we had. Wow, I even wrote 40000 of 100000 and you said it stopped on 24975.

Can you please compile the master branch for linux of windows (x64)? I know it is just a testnet type of thing but I want it anyway. Did you say it had a map filter? If it does, I really want it.

By 'standard client' what do you mean?  If you didn't compile it from the master branch on the github then it isn't really testing what we need to test.  We know the existing build is working fine already  

That is a real shame for this coin. Oh well.
It looks like you are our new dev! You seemed to be doing more work than the dev was doing anyway.

Ok clean syncing right now with standard client.

P.S. How come the progress bar doesn't work? That seems to happen to most other coins.

Edit: Currently 70000 of 100000 blocks.
Will update as I go.

EDIT: Nevermind, false alarm.  

EDIT2: Nevermind again, I think I was right the first time.  Master looks to have an issue right now.  Can someone else try a clean sync from genesis block with a build?  From here a new sync gets stuck after block 24975, failing the proof on 24976.

EDIT3: Some more details: it thinks the generated worlds are invalid because it thinks that the rider collides in the first frame.  I'm syncing from local nodes, I'm going to try starting a real sync with peers and see what happens, but it will take some time.

You should avoid running binaries from random strangers on forums.

That being said, WilliamLie has said in PM basically that he is giving up on building.  I'm not sure where that leaves us, now.  I'm assuming no-one went ahead with a block 100k fork, and 104k will be here soon enough. 

Can someone please compile a windows x64 mapfilter? I really want this to see if I can get anything. I think with a fall-through map, I can.

It would be too. I guess I just asked you to think even more. Well, probably just copy and paste, and wait a few minutes for your computer to compile the program but still.

I dunno it has had me thinking pretty hard lately

Remember when this was called "Proof of Thought"?
Fun times.

Apparently I've gotten a little better.  I timed a few runs in forfun and my best time was down to about 5 minutes on 60 second TT!  (not counting ~20 seconds of map filter run to make a particularly easy map.)

How long do you think it would take for a skilled player to finish a 9.35 TT run?  We can arbitrarily make N as large as we like, as long as miners agree to adopt the change.

I think that no human will be able to find and solve map in 9.35 or even faster.

No, but I am not a very good exemplar.  It takes me something like 20-30 minutes to complete a map in forfun mode at all, and a lot of that time is often spent repeatedly running out of time even at 60 seconds!  I don't think I've ever human mined a motocoin at all.  I've always been bad at bmx/moto games for some reason.