Best Exchange Website

[tonyk]

I will not give my opinion on what to copy. mainly because I do not really like  the polo site, but most people praise it to no end. So I consider my personal opinions a bit off from the majority, in this area.

What do you not like about polo / like better elsewhere?  Perhaps we could combine multiple UI ideas and make something better.

I cannot see the orderbook and the graph or last sales at the same time...it takes scrawling. Maybe my small 15' screen is the issue.

[fav]

Yes lets get away from trying to reinvent everything, and instead try to take what is already proven to work, and adapt it into the blockchain.

But you don't want an exact clone. You start with a template based on whatever is the best out there but you have to add your own magic to it. If it's exact clone they'll use the original.

They all have strengths and weaknesses. Autosell for example is a strong feature, very popular.

I don't know how much work it is, but couldn't they just offer several skins to meet everyone's taste?

[tonyk]

I think the order matching rules are pretty good right now.  It seems to be standard (same as other exchanges).

The margin trading using the current orderbook to determine whether or not to execute sounds like the right way to go.  My question is how does it bootstrap?

If I lend funds to someone when there is no order book, then it is impossible to "call".

1.You cannot use polo-like call rules in a thin market. Which means the shorts must be called in relation to the feed

2.polo only puts  coins for margin trading if they have  solid order books for a period of time. - we can kind off copy that for the margined positions

3. Be very careful with the lending thing - you do not want to favour heavily the lending compared to shorting, as there is a chance if you do to dry up the shorters...i.e. move all of them to lending/borowing.

[Akado]

This pool should be done on Bitcointalk where we can get more people's opinions, specially from outsiders since we also want to attract new users. It will not be of much use here on our forums. Maybe even on reddit.

[luckybit]

Yes lets get away from trying to reinvent everything, and instead try to take what is already proven to work, and adapt it into the blockchain.

But you don't want an exact clone. You start with a template based on whatever is the best out there but you have to add your own magic to it. If it's exact clone they'll use the original.

They all have strengths and weaknesses. Autosell for example is a strong feature, very popular.

[luckybit]

Minebitshares and other miners can use the autosell feature. It's a very useful feature.

[Ander]

Yes lets get away from trying to reinvent everything, and instead try to take what is already proven to work, and adapt it into the blockchain.

[luckybit]

I keep telling people that Mintpal was the most successful altcoin exchange not because it had groundbreaking features or support, but because it had the cleanest interface of them all. It's a shame that nobody is using that UI anymore. Poloniex looks nice, but it became successful  because Mintpal disappeared.

https://web.archive.org/web/20140315230829/https://www.mintpal.com/

Here you have the complete front-end. The exchange came into life 1 year after Cryptsy and 1 month after Poloniex and managed to capture both user bases pretty fast.

BTC-E seems to be the most successful, but there are others.

Well, BTC-e was founded in 2011 and it also stopped being an altcoin exchange a long time ago as everyone is focused on BTC/LTC nowadays. BTC-e is to Bitcoin exchanges what Polo was to altcoin exchanges in 2014. Old and lacking of tools for serious trading. It is successful because it was founded many years ago and managed to build a solid user base without having to deal with a lot of competition along the way. With Mintpal, the story was different. A new altcoin platform that became 1st after a few weeks, stealing the clients from all the other exchanges. Reason ? The sleek, simple to use GUI.

LIST OF THE BEST AND WORST CRYPTOCURRENCY EXCHANGES in 2014
http://www.usacryptocoins.com/thecryptocurrencytimes/uncategorized/list-of-the-best-and-worst-cryptocurrency-exchanges/

Too bad i cannot find that screen with the top exchanges sorted by volume. Mintpal had so much volume in early 2014 that it was being compared with Bitstamp and Bitfinex.
________________________________________________________________________________________

Speaking about MetaTrader, i think it looks absolutely horrendous and its success had to do more with the fact that it was released in 2002-2005 and also that it had a great marketing campaign behind it. I remember that before adblock existed, i was receiving hundreds of pop-ups per day about the MetaTrader platform.
Another reason could be:

Why is MT4 monopolizing the Forex Biz?
http://www.forexpeacearmy.com/community/threads/why-is-mt4-monopolizing-the-forex-biz.23937/

________________________________________________________________________________________

There are some elements that you can take from Bitstamp and implement in graphene, but i'm afraid that "Tradeview" would be the last. It looks great, but it's clunky and it hangs often when you are dealing with great volume.
I agree with copying Poloniex too. Their last UI update made a night-and-day difference.

https://www.reddit.com/r/dogemining/comments/1xmddm/cryptsy_autosell_how_does_it_work/

Cryptsy was pretty good. One reason is miners could automatically have their mined funds deposit and sell. Copy that feature from Cryptsy at least.

I don't think any exchange should be copied 100%. Take the best features from each exchange and use science to figure out what those features are.

Bytemaster, try apply the design for experiments approach. DOE will allow you to figure out exactly the best features and cut away the rest. This requires a lot of feedback so maybe allow for feedback to be given directly from the interface itself so participants can make notes for developers.

[Ander]

1. Please have a 'novice' section so non-traders can buy/sell easily at the spot price.
2. Have this work on the iPhone, ideally with an app available in the AppStore
3. As clean and intuitive as possible

These are important suggestions as well! 

[Ander]

I will not give my opinion on what to copy. mainly because I do not really like  the polo site, but most people praise it to no end. So I consider my personal opinions a bit off from the majority, in this area.

What do you not like about polo / like better elsewhere?  Perhaps we could combine multiple UI ideas and make something better.

[Ander]

Since 80% of paid fees go to referral, I think 0.1 BTS for all is not feasible. Instead, we can decrease the difference by lowing the trading fee to 1 BTS (0.8 BTS for normal member vs. almost 0 BTS for lifetime member. Based on current fee system, it's 8 BTS vs. 0 BTS)

No, we should keep the trading fee for orders that actually execute higher.  (Or percentage based). 
There is a huge difference between the two.  Some people will place and remove tons of orders and most of them wont execute, for automated trading strategies and such.  We need these people and they provide liquidity and volume of trading.  But they will NOT come to bitshares if they have to pay a fee for cancelling orders, of any amount of real importance. 

Everyone expects to pay some fees when they actually have an order execute, so we can charge a decent amount for that.  But everyone expects to pay NOTHING for a cancelled order, because they pay nothing for it on all sites currently.  We have to get this as close to 0 as possible, and the only reason its not 0 is to prevent spam from eating all our TPS up and allowing an attack on the network for free.

Referrals will still be very awesome when you get 80% of someones trading fees, plus 80% of their small cancelled order fees.  The cancelled order fees should end up being a small portion of the total fees.

[tonyk]

Could someone create a poll (preferably on bitcointalk, but also here) that asks people to vote for the best exchange website. 

It would be great to have some real market research regarding which features people like, and which ones they do not.

We will be migrating BitShares GUI to copy as closely as possible the best interface out there.   We will also be copying the best fee system and best market rules.

Everyone seems to like poloniex margin trading, so we will adapt their rules unless Bitfinex is better.    Any help on identifying the best choices let us know.

We are going to stop trying to invent new market mechanics if at all possible and use what is proven to work.   So please help us identify places where our system deviates from the market-accepted solutions.

1.  Fee Systems
2.  Trading Interfaces
3.  Order Matching Rules

In other words, we want people to be able to use BitShares without having to learn anything new!

So we are giving up on the reinventing new and improved square wheels, already? 

Honestly though, this is great to hear! + 1, + 1


I will not give my opinion on what to copy. mainly because I do not really like  the polo site, but most people praise it to no end. So I consider my personal opinions a bit off from the majority, in this area.

[Ander]

I think the order matching rules are pretty good right now.  It seems to be standard (same as other exchanges).

Yes I think this is true.

The margin trading using the current orderbook to determine whether or not to execute sounds like the right way to go.  My question is how does it bootstrap?

If I lend funds to someone when there is no order book, then it is impossible to "call".

When you start out with an empty orderbook, no margin trading is allowed.  In order to place a margin order in the first place, the system checks to see that the orderbook could handle the new state (where you have a margin position), and checks that you have at least 40% collateral in that new state.  (Thats the amount for polo, 40% to start a position or add to it).  Then, if things go against you, you are allowed to go down to 20% margin before a margin call happens.  While in the 20-40% range, the interface gives you a warning that you are near a margin call and it does not allow you to add to your position, only to reduce it.


I am not sure exactly how the polo system works, but there might be an additional requirement:

If an order on the books would be cancelled, and this cancellation would leave the orderbook with not enough supply remaining, then instead of allowing the order to be cancelled, it instead executes the margin call FIRST, to ensure that the person can pay back what they owe, and then it cancels the rest of the order if needed. 

This is needed because otherwise I could set up the following scenario:


I have an order in to buy a large amount of asset at price 100.  My order is most of the orderbook.
People then took margin positions which were within collateral requirements, based on my order being in the order book.  That is, as long as they can sell into my order, they have sufficient collateral and will be able to pay back the loan.

But then I remove my big order!  There is very little remaining orders on the books, and everyone gets margin called and the system tries to sell into what is left, but they cannot get enough funds from it to pay back the people they owe.  This would be a failure of the system to protect lenders, it would leave the person who got margin called owing a bunch more BTS that they cannot pay, to other people.


Instead, at the point where I try to remove my big order, the system checks that this removal will cause a margin cascade, and instead of pulling the order, it executes the margin cascade into the order.  Then, if some of the order is left unfilled, it is allowed to be pulled.



If you use polo you will know that at times when trading gets crazy and a margin call happens, the system freezes up for a couple seconds.  This occurs because the system is forces to do a bunch of calculations and then execute automatic margin call trades, and then recompute again, and do more margin call trades, etc.    Only once this process is completed are people allowed to trade normally again.


This might be an issue with the blockchain times, witnesses computers needing to process all these calculations in time, etc.  I would imagine that this is a hard technical problem to do on a blockchain.  (It was probably hard enough already for polo to implement flawlessly...and it must be flawless).  So maybe because of this difficulty we are forced to use a simpler system?

[ValentinJesse]

I keep telling people that Mintpal was the most successful altcoin exchange not because it had groundbreaking features or support, but because it had the cleanest interface of them all. It's a shame that nobody is using that UI anymore. Poloniex looks nice, but it became successful  because Mintpal disappeared.

https://web.archive.org/web/20140315230829/https://www.mintpal.com/

Here you have the complete front-end. The exchange came into life 1 year after Cryptsy and 1 month after Poloniex and managed to capture both user bases pretty fast.

BTC-E seems to be the most successful, but there are others.

Well, BTC-e was founded in 2011 and it also stopped being an altcoin exchange a long time ago as everyone is focused on BTC/LTC nowadays. BTC-e is to Bitcoin exchanges what Polo was to altcoin exchanges in 2014. Old and lacking of tools for serious trading. It is successful because it was founded many years ago and managed to build a solid user base without having to deal with a lot of competition along the way. With Mintpal, the story was different. A new altcoin platform that became 1st after a few weeks, stealing the clients from all the other exchanges. Reason ? The sleek, simple to use GUI.

LIST OF THE BEST AND WORST CRYPTOCURRENCY EXCHANGES in 2014
http://www.usacryptocoins.com/thecryptocurrencytimes/uncategorized/list-of-the-best-and-worst-cryptocurrency-exchanges/

Too bad i cannot find that screen with the top exchanges sorted by volume. Mintpal had so much volume in early 2014 that it was being compared with Bitstamp and Bitfinex.
________________________________________________________________________________________

Speaking about MetaTrader, i think it looks absolutely horrendous and its success had to do more with the fact that it was released in 2002-2005 and also that it had a great marketing campaign behind it. I remember that before adblock existed, i was receiving hundreds of pop-ups per day about the MetaTrader platform.
Another reason could be:

Why is MT4 monopolizing the Forex Biz?
http://www.forexpeacearmy.com/community/threads/why-is-mt4-monopolizing-the-forex-biz.23937/

________________________________________________________________________________________

There are some elements that you can take from Bitstamp and implement in graphene, but i'm afraid that "Tradeview" would be the last. It looks great, but it's clunky and it hangs often when you are dealing with great volume.
I agree with copying Poloniex too. Their last UI update made a night-and-day difference.

[mf-tzo]

Clean, simple and professional like bitstamp and the same time having a trader option for advanced graphs and indicators. Integrate margin trading like Polo which is very simple and easy to use. Integrate  some binary options as well and you will have created the best exchange!
Metatrader is very popular for professional traders. The interface should be simple as hell like bitstamp but should have an option to trade with metatrader if one chooses to.