Community Directory


Go back

User Activity

Forum Posts

  • Replies: 271
  • Topics: 24
  • Likes Received: 178
  • Re: Exchanging bitcoin to Fiat and back.

    @TNT3

    The local trader function in mycelium is an open trading book in which you can list buy and sell orders. It does not match orders, instead it stores geolocation information on orders and you can start an encryped chat with the other party and reach an agreement on when and where you would like to meet to complete the order. It has a rating system and at least where I live it is the easiest way to obtain bitcoin.

    I encourage you to read more about it, it's very well made.

  • Re: Token Quantity

    I personally am very fond of the values 100 and 1. These are intuitive for all cases where you need to represent percentages.

    Obviously they'd make horrible units in some or most other cases, but this goes to show you that previously suggested numbers are just that: suggestions.

  • Re: Exchanging bitcoin to Fiat and back.

    I can strongly recommend using the local trader function in the Mycelium android wallet. It works wonderfully!

  • Re: It's Expensive Using Bitcoin!

    Regardless of how we solve this particular issue for this particular usecase, if we still have a high base fee in the bottom of the structure and scale up, then this will never work out.

    So how do we get the base cost lower? Even if there was a way to group up all the transactions in to a single one, the sheer data size of the addresses alone is an issue.

    Storing the blockchain is rather cheap today even though it's soon 20gb in size (harddrives are cheap). Broadcasting the transactions and blocks today is relatively cheap (bitcoin volume still fits nicely in most consumer internet connections).

    Should the bitcoin volume go up much faster than the development of faster internet and larger harddrives though, then fees is supposed to be the filtering mechanism between the transactions that are valued and useful to people and the transactions that are (subjectively) less valued and useful.

    Could it be that we are seeing the limits of the current architecture for bitcoin by pushing its boundaries with counterparty?

    Here's a crazy idea that may or may not have merit (I dont know how counterparty transactions look like):

    If the counterparty transactions are sent and handled in clear-text, compress them. If the order of the components inside the transactions are not important, randomly re-organize them to achieve better compression ratios. (trial and error procedure, costs CPU power but could save storage and network utilization)

  • Re: It's Expensive Using Bitcoin!

    One option to reduce cost (and security) would be to apply the counterparty protocol on top of another coins blockchain where the transaction costs are lower.

    Also I'm not sure fee's are guaranteed to rise with the new fee model. And even if it does, whenever we get a solution to the costs of block distribution and allow larger blocks the fees should go down again.

  • Re: It's Expensive Using Bitcoin!

    I believe this problem will soon manifest itself in a different manner, as bitcoin core will release an updated client with a dynamic fee model that might make the problem become irrelevant, just as well as it could make the problem become very urgent.

  • Re: BTCBLUE.XCP Marketplace Credits

    I should've read things properly before making an order: "Note: You need to include your CounterParty address in the comments at checkout for a smooth process."

    Also, I'm fairy sure I put in 5 BTC.BLUE tokens to buy at $1 each, but looking at the screenshots of the payment process I paid $10. Might be a bug, might be something else I didn't care to read carefully enough >_<

    Either way, you already have my address and if not you can ask me for it :)

  • Re: BTCBLUE.XCP Marketplace Credits

    I got mine just minutes after you sent it. Though it feels odd since you also gave me some tokens just before. Anyway, I'm off to buy some more tokens to show some support.

  • Re: Game Development.XCP Coins?

    One of the ideas I'm very fond of is a more realistic approach to "jobs", where the character, when logging out, uses his time to earn in-game money, thus incentivizing people to spend some time offline every now and then. If you assume this to be the main method of creating a token-supply and balance it so that it has a declining return based on amount of time spent offline per calendar day or other time frame then you would have the basis for casual gameplay.

    If one wants intense gameplay, that theory does not hold up, however.

  • Re: Game Development.XCP Coins?

    I am definately interested in having a token with a limited supply as the default in-game currency, even if only to see the interesting marketplaces for virtual time and items that would arise. I've already been down that rabbit hole when planning ideas like dynamical quest generation based on resource scarcity: Take for example a simple game which has wood, stone, iron and gold and you have a blacksmith selling basic items to the players. If this blacksmith has a current supply of the resources he uses to create the items players can buy, and only creates new items if his current items get sold, then he can in theory end up in a situation where his own resources gets depleted and he would then generate a quest to gather the resources he needs to create new items that re-stock the store.

    For the multiplayer auction house in a single-player roguelike I can see that it would work, but I don't think I would find it enjoyable personally. Mostly because you'd alienate other players and discard valuable social interactions by reducing them to "providers or random goods".

Pages 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

© Copyright 2013–2016 The LTB Network. All rights reserved .