Pages:
Author

Topic: Bitstamp TX Fee Exploit - Fee 1.16% instead of 0.2% [Bitstamp Fixes on May 15th] - page 3. (Read 21154 times)

hero member
Activity: 798
Merit: 1000
legendary
Activity: 1148
Merit: 1001
things you own end up owning you
Not sure this will satisfy davout but;  Wink


Bitstamp have updated their minimum order amount to $5:

Quote
Dear Bitstamp clients,

On May 15th, 2014 the minimum trade amount will be changed to $5.00.

Bitstamp’s accounting system rounds trading fees to pennies. To greatly reduce the impact of this rounding for our clients the minimum trade amount is being raised to $5.00.

The $5.00 trade minimum will apply to both our web interface and our API.

Best regards,
Bitstamp team

https://www.bitstamp.net/article/bitstamp-minimum-trade-changing-to-5/




it wont, because no matter what Bitstamp would do he wouldn't be satisfied..... the reason is simple when you know who he is  Smiley
hero member
Activity: 798
Merit: 1000
Not sure this will satisfy davout but;  Wink


Bitstamp have updated their minimum order amount to $5:

Quote
Dear Bitstamp clients,

On May 15th, 2014 the minimum trade amount will be changed to $5.00.

Bitstamp’s accounting system rounds trading fees to pennies. To greatly reduce the impact of this rounding for our clients the minimum trade amount is being raised to $5.00.

The $5.00 trade minimum will apply to both our web interface and our API.

Best regards,
Bitstamp team

https://www.bitstamp.net/article/bitstamp-minimum-trade-changing-to-5/


legendary
Activity: 1372
Merit: 1007
1davout
Fixing it by applying the fee to the order instead of by transaction is also valid. The way this would be done is to reserve the full fee (based on order quantity) when an order is opened, but then credit back only the unfilled portion of the fee when the order is closed (canceled or filled). Round the first up, round the second down (except in the special case of zero filled). They get the generated rounding breakage they seem to want, but once per order.

It is indeed a much less scammy behaviour than what's currently happening.
That doesn't really work for small orders though, and is actually harder from a technical point of view, than simply increasing the precision of the stored amounts.

I think that'd be a good blog post subject, technical and accounting issues related to rounding and the best ways to solve them.
legendary
Activity: 2968
Merit: 1198
Has Bitstamp commented this thing at all? There are multiple ways to fix this... and being quiet makes it worse.

Actually there happens to be a single correct way to fix it.
For pretty much the same reason there is one single correct way to count.

Their answer is simply magical.

Fixing it by applying the fee to the order instead of by transaction is also valid. The way this would be done is to reserve the full fee (based on order quantity) when an order is opened, but then credit back only the unfilled portion of the fee when the order is closed (canceled or filled). Round the first up, round the second down (except in the special case of zero filled). They get the generated rounding breakage they seem to want, but once per order.



hero member
Activity: 552
Merit: 501
as if 0.5% for < $500 isn't ridiculous enough, the most honest and reliable exchange in my experience so far has to be btc-e with their 0.2% flat rate, why would you even use bitstamp in the first place  Undecided

That hasn't been my experience. Deposits at BTC-e were credited days or even weeks late. Mysterious three figure fees approaching 1% per deposit were deducted and customer service queries went unanswered. In fact I moved from BTC-e to Stamp for exactly these reasons and so far I have been very happy with them.  
legendary
Activity: 1064
Merit: 1000
Personally, I ditched Bitstamp for Bitfinex because stamp suck.
sr. member
Activity: 433
Merit: 260
Transaction Fee Malleability?


Allegation that DGEX.com also does this: https://nextcoin.org/index.php/topic,4705.0.html
legendary
Activity: 1372
Merit: 1007
1davout
Try to discuss the point

Your display of braindamage does not constitute a point in any way.


Bitstamp do not want to make changes because they will make less money from fees.

Quote
"It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"

+1
newbie
Activity: 11
Merit: 0
My typical 30 day trading volume is 300k at minimum.

And you didn't notice you were being ripped off?

If I had I would have been the OP. Shit happens. I guess that somehow as "obvious" as it must have been, none of the other handful of people who use stamp noticed it either. Our bad.

Still no reply to my comment on their facebook or the tweet I sent them.

Link to tweet? I will check quality and retweet.

https://twitter.com/NinjaBenjamin/status/463721726493589504
donator
Activity: 1218
Merit: 1079
Gerald Davis
Bitstamp do not want to make changes because they will make less money from fees.

Quote
"It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"
legendary
Activity: 2097
Merit: 1070
This exchange is straight up scamming it's customers.
legendary
Activity: 1190
Merit: 1001
Because changing your entire decimal precision could cause unintended bugs....

As opposed to what? Intended bugs?
Just because something is inconvenient or hard does not mean it should not be done, applying half-assed pseudo band-aids is exactly what you should not do.

When you have zero experience in a field, and no valid point to make, your best course of action is usually to refrain from talking.

My point is that the order limit can be place immediately; while the more permanent solution of higher precision decimal rounding can be tested and implemented over a longer period of time.

So your solution would be what?

a) Leaves the system running as is (stealing fees) for months while safely developing and test the change in rounding.

b) Rush out a change in decimal precision without adequate testing and hope it has no bugs?

Try to discuss the point rather than semantics and ad hominem
legendary
Activity: 1372
Merit: 1007
1davout
Has Bitstamp commented this thing at all? There are multiple ways to fix this... and being quiet makes it worse.

Actually there happens to be a single correct way to fix it.
For pretty much the same reason there is one single correct way to count.

Their answer is simply magical.
hero member
Activity: 518
Merit: 500
Has Bitstamp commented this thing at all? There are multiple ways to fix this... and being quiet makes it worse.

See https://bitcointalksearch.org/topic/m.6590228
legendary
Activity: 1511
Merit: 1072
quack
Has Bitstamp commented this thing at all? There are multiple ways to fix this... and being quiet makes it worse.
hero member
Activity: 798
Merit: 1000
I have received a response from Bitstamp. I request that you all re-read my starting post, as I changed it to better reflect the situation as it is.

Bitstamp did reply to my inquiry about their transaction fee calculation.

Quote
The rounding up which occurred in your case was in accordance with our notice provided in the Fee schedule section of our website. Fiat currencies, unlike Bitcoin, have two decimal places, therefore the smallest fee Bitstamp can charge for its services can never be below $0.01.

Bitstamp does not sell or buy bitcoins nor does it manipulate how orders are executed. If a sell was executed for one client (you in this case) this means that another Bitstamp client had his buy order executed on the opposite side of the trade.

As you are probably aware that "If you try to commit an order with price greater or lower than 20% of the current market price, our system will ask you for additional confirmation." (available in our FAQ: https://www.bitstamp.net/faq/). Due to such actions, it is sometimes possible that a user can perform a trade which is lower than $1 in value when placing an order which is higher than the market price at the time when it was executed.

They further explain their rounding policy, which is fine, but did not explain (yet) why the order limit is then not $5 so that the exploit as mentioned in this post cannot happen. The fact remains that an order limit of $1 in combination with the rounding off to $0.01 causes any order of $1 or smaller to be charged five times the fee.

Whether or not the low $1 order limit is intentional in order to harvest these fees or not, I will leave upon you (the community) to decide. The willingness of Bitstamp to change the order limit will show us as a community if it is intentional or not.

You should point them in the direction of davouts solution as well which is probably fairer.

Yes we know fiat currencies has 2 decimal places but a system can store more than two. Just use more decimal places and round on withdrawal.
hero member
Activity: 518
Merit: 500
I have received a response from Bitstamp. I request that you all re-read my starting post, as I changed it to better reflect the situation as it is.

Bitstamp did reply to my inquiry about their transaction fee calculation.

Quote
The rounding up which occurred in your case was in accordance with our notice provided in the Fee schedule section of our website. Fiat currencies, unlike Bitcoin, have two decimal places, therefore the smallest fee Bitstamp can charge for its services can never be below $0.01.

Bitstamp does not sell or buy bitcoins nor does it manipulate how orders are executed. If a sell was executed for one client (you in this case) this means that another Bitstamp client had his buy order executed on the opposite side of the trade.

As you are probably aware that "If you try to commit an order with price greater or lower than 20% of the current market price, our system will ask you for additional confirmation." (available in our FAQ: https://www.bitstamp.net/faq/). Due to such actions, it is sometimes possible that a user can perform a trade which is lower than $1 in value when placing an order which is higher than the market price at the time when it was executed.

They further explain their rounding policy, which is fine, but did not explain (yet) why the order limit is then not $5 so that the exploit as mentioned in this post cannot happen. The fact remains that an order limit of $1 in combination with the rounding off to $0.01 causes any order of $1 or smaller to be charged five times the fee.

Whether or not the low $1 order limit is intentional in order to harvest these fees or not, I will leave upon you (the community) to decide. The willingness of Bitstamp to change the order limit will show us as a community if it is intentional or not.
legendary
Activity: 1372
Merit: 1007
1davout
Fees should be charged per buy or sell order, not by each split up transaction. It will not be difficult to change it. Having developed an exchange, changing fee structure should be trivial.

Think more, I already gave the correct solution.
What you're suggesting is strictly equivalent to charging the fee at the transaction level, that is because you need to be able to handle partial fills properly.
hero member
Activity: 658
Merit: 500
Fees should be charged per buy or sell order, not by each split up transaction. It will not be difficult to change it. Having developed an exchange, changing fee structure should be trivial. Bitstamp do not want to make changes because they will make less money from fees. The only way to pressure them to make modifications is to stop using Bitstamp.
Pages:
Jump to: