Pages:
Author

Topic: Are we stress testing again? - page 14. (Read 33192 times)

hero member
Activity: 784
Merit: 500
July 16, 2015, 12:18:38 PM
http://btc.blockr.io/tx/info/0c6429646bfe986c4d34d72f6592632cbbf0b6b890c42cb23ef06ba6456c1123

7 hours and 0 confirmations...

Total time 8.5 hours before 1 confirmation.

Lucky Guy..LOLZ  Cheesy that it took only 8.5 hours.

My transactions took more than 48 hours dude and it was so embarrassing to keep on waiting for 48 hours as I needed payment so badly but I was left with only one option and that was to keep on waiting till it gets confirmed.
staff
Activity: 3458
Merit: 6793
Just writing some code
July 16, 2015, 12:09:21 PM
Interesting.  So it is potentially possible that BitPay never gets that transaction, and I receive what I purchased for free.
Yes, but the merchant should never ship their products without having at least 1 confirmation.

When you delete a transaction from your wallet, doesn't mean the other nodes will delete it. Right?
The other nodes will not relay it and upon restart, they won't have the transaction.

Does Replace-By-Fee mean that you can replace the output-address? If so, then this "feature" just seriously breaks Bitcoin.
Full RBF means that you can replace the output addresses, but what is currently recommended for use is partial RBF. This requires that the inputs and outputs remain the same as the transaction it replaces but the only difference is the fee.

At some point, the network will "forget" you transaction.

Where is this officially documented?  Provide a link please.  I find it concerning that a transaction sent could be "forgotten" by the network for any reason, that seems like a huge problem.
I don't have a link, I will get back to you with that if I can find it. However, I know this after looking in the source code. Nodes will only relay transactions once when they receive it or when another node asks for it. The sending node will continue to broadcast the transaction while it is unconfirmed. If it stops broadcasting it, over time nodes will begin to not have the transaction in their mempools, either from restarts or other rules that people might set. Once enough nodes have dropped a transaction, a double spend can be attempted and if that double spend is confirmed, then the original will be dropped as a double spend.
newbie
Activity: 43
Merit: 0
July 16, 2015, 12:02:52 PM
When you delete a transaction from your wallet, doesn't mean the other nodes will delete it. Right?
Does Replace-By-Fee mean that you can replace the output-address? If so, then this "feature" just seriously breaks Bitcoin.
That is why a transaction is not a real transaction until it has at least one confirmation (is included into a block).  Before then, it is just a transaction "candidate"...

Merchants should never ship products before getting at least one confirmation.

In my case of having a merchant use BitPay I was just provided an address to send BTC, and once it reached the requested amount I received a receipt from the merchant saying I was all set.  It was a paid service, so my access has already been updated for what I paid for, and I am guessing if the transaction just doesn't go through BitPay will be out of luck.
legendary
Activity: 1386
Merit: 1009
July 16, 2015, 11:55:05 AM
At some point, the network will "forget" you transaction.

Where is this officially documented?  Provide a link please.  I find it concerning that a transaction sent could be "forgotten" by the network for any reason, that seems like a huge problem.
It will sit in some nodes' mempools, but won't be relayed. To my knowledge, transactions are relayed only once, when a node get it for the first time. It means that when some nodes are restarted, they won't have this transaction in their mempools anymore.

To counter that, the nodes that are sender/recipient of this transaction will rebroadcast it from time to time until it gets confirmed.
legendary
Activity: 2128
Merit: 1073
July 16, 2015, 11:53:08 AM
Where is this officially documented?  Provide a link please.  I find it concerning that a transaction sent could be "forgotten" by the network for any reason, that seems like a huge problem.
Use the Source, Luke!

It is documented in the source code. The mempool is a really lame implementation of https://en.wikipedia.org/wiki/In-memory_database . There's no mempool synchronization in the protocol, so eventually the non-confirmed transactions will get forgotten as users restart their nodes.
legendary
Activity: 1274
Merit: 1000
July 16, 2015, 11:19:28 AM
At some point, the network will "forget" you transaction.

Where is this officially documented?  Provide a link please.  I find it concerning that a transaction sent could be "forgotten" by the network for any reason, that seems like a huge problem.
hero member
Activity: 714
Merit: 500
July 16, 2015, 11:17:10 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
You can create a double spend with Bitcoin Core using console commands. Also, since some nodes won't relay or accept transactions with low fees, changing the fee get the double spend accepted into some mempools when the original did not. Otherwise, nodes that enabled Replace-By-Fee will allow double spends as long as the fee is greater than the original and nodes that are Bitcoin XT will relay double spends.

Say I don't go the double spend route.  Will the low fee transaction be cleared out the of the mempool at some point, or will it just eventually get processed?
At some point, the network will "forget" you transaction. You just need to either delete it from your wallet or not run your client for a few days. It will be as if the transaction never happened.
When you delete a transaction from your wallet, doesn't mean the other nodes will delete it. Right?
Does Replace-By-Fee mean that you can replace the output-address? If so, then this "feature" just seriously breaks Bitcoin.
newbie
Activity: 43
Merit: 0
July 16, 2015, 10:57:41 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
You can create a double spend with Bitcoin Core using console commands. Also, since some nodes won't relay or accept transactions with low fees, changing the fee get the double spend accepted into some mempools when the original did not. Otherwise, nodes that enabled Replace-By-Fee will allow double spends as long as the fee is greater than the original and nodes that are Bitcoin XT will relay double spends.

Say I don't go the double spend route.  Will the low fee transaction be cleared out the of the mempool at some point, or will it just eventually get processed?
At some point, the network will "forget" you transaction. You just need to either delete it from your wallet or not run your client for a few days. It will be as if the transaction never happened.

Interesting.  So it is potentially possible that BitPay never gets that transaction, and I receive what I purchased for free.
staff
Activity: 3458
Merit: 6793
Just writing some code
July 16, 2015, 10:29:13 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
You can create a double spend with Bitcoin Core using console commands. Also, since some nodes won't relay or accept transactions with low fees, changing the fee get the double spend accepted into some mempools when the original did not. Otherwise, nodes that enabled Replace-By-Fee will allow double spends as long as the fee is greater than the original and nodes that are Bitcoin XT will relay double spends.

Say I don't go the double spend route.  Will the low fee transaction be cleared out the of the mempool at some point, or will it just eventually get processed?
At some point, the network will "forget" you transaction. You just need to either delete it from your wallet or not run your client for a few days. It will be as if the transaction never happened.
newbie
Activity: 43
Merit: 0
July 16, 2015, 10:25:24 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
You can create a double spend with Bitcoin Core using console commands. Also, since some nodes won't relay or accept transactions with low fees, changing the fee get the double spend accepted into some mempools when the original did not. Otherwise, nodes that enabled Replace-By-Fee will allow double spends as long as the fee is greater than the original and nodes that are Bitcoin XT will relay double spends.

Say I don't go the double spend route.  Will the low fee transaction be cleared out the of the mempool at some point, or will it just eventually get processed?
staff
Activity: 3458
Merit: 6793
Just writing some code
July 16, 2015, 10:20:20 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
You can create a double spend with Bitcoin Core using console commands. Also, since some nodes won't relay or accept transactions with low fees, changing the fee get the double spend accepted into some mempools when the original did not. Otherwise, nodes that enabled Replace-By-Fee will allow double spends as long as the fee is greater than the original and nodes that are Bitcoin XT will relay double spends.
hero member
Activity: 714
Merit: 500
July 16, 2015, 10:17:30 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
Which client can "override" a transaction aka double spend it?
How do they trick the nodes, that just won't accept a double spend in their mempool?
legendary
Activity: 1386
Merit: 1009
July 16, 2015, 09:15:39 AM
SebastianJu, Xialla, I understand what you mean. You wanna say it's a Bitcoin problem, as if its adoption by everybody (even totally clueless) is the main goal. That's not what I want, because I like when people make educated decisions, especially about their money, that's why I disagree with you. Ok, that's understandable, it's an ideological disagreement, so no point in arguing.
legendary
Activity: 1386
Merit: 1009
July 16, 2015, 09:12:28 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
The BTC has not really left your wallet until the transaction is included in a block (confirmed). It means you can 'override' your transaction, but it depends on your client.
newbie
Activity: 21
Merit: 0
July 16, 2015, 08:41:58 AM
these spams are really frustrating  Angry
newbie
Activity: 43
Merit: 0
July 16, 2015, 08:33:27 AM
I am not sure if this has been said in this thread, and it seems to be a lot of back and forth from the pages I read, so I figured I would just ask.

If I sent a transaction and didn't set the fee high enough, will the transaction eventually roll back?  The BTC have left my wallet at this point, and it was for a paid service, so I am wondering if I have to worry about the transaction not going through.

It was through BitPay, so the vendor got their money I am sure, but I didn't know if the transaction itself would revert back because of the backlog. 
legendary
Activity: 1036
Merit: 1001
/dev/null
July 16, 2015, 07:41:44 AM
It's not a Bitcoin problem.

..is hard to imagine impact for new users. but for majority of them them, it looks like this:

* downloaded and installed some wallet
* somehow more or less secured their wallet, read lot of stuff about it, how to backup it, etc
* created and verified exchange account, wait 1-2 days
* transfered fiat to exchange, wait 2 days at least
* bought BTC at market price
* initiated transfer from exchange to their wallet

and now, they have to wait another bloody 1 or 2 days, even they somewhere read, that it takes ~10 minutes in average. believe or not, but they are not aware about stress tests, blockchain site or amount of confirmations. their VERY FIRST experience is, that they have to wait in really uncomfortable situation (because of no experience, they are scared that money/BTC are lost), for endless hours.

sorry, but I can't agree with you. it is bitcoin problem and it is quite huge one...
legendary
Activity: 2674
Merit: 1083
Legendary Escrow Service - Tip Jar in Profile
July 16, 2015, 07:11:55 AM
Quote
Right, but its a Bitcoin Problem. At the end its all Bitcoin and when bitcoin has a problem then it doesnt matter anymore that the mainclient has a way to prevent this. You see all these threads on here and on reddit? Imagine how many persons found bitcoin and... it does not work. Great experience. And surely not helpful for adoption. While adoption is the only thing that could help miners to earn considerable amounts of fees in the future.
It's not a Bitcoin problem. Go approach alternative clients' devs and complain to them about fees. Because the fee market exists, and they have to deal with it.

Quote
So? What does it matter when you name the bank only differently? He wanted a currency that isnt controlled by somebody. It doesnt matter who controls it. How its named.
Bank? I don't care how you name it. If it's a trustless bank, then I'm in. Trustless means control of ownership and double-spend prevention.

I wonder what your solutions are to the underlying problem.

I wonder if you really dont understand what i mean. When new adopters come to bitcoin and their experience is that the transactions doesnt work then they wont divide into normal client and alternative client. Its BITCOIN then that makes a bad impression. Sure the other clients devs can change that but at the moment the spam attack means bad things for adoption.

Sure, miners manipulating fees or maybe blocking transactions of some companies totally are getting a more serious risk the more mining gets centralized and miner work together in order to raise their profit. So trustless ok, but still, bitcoin is in risk, or at least some features of it.
legendary
Activity: 1386
Merit: 1009
July 16, 2015, 06:41:57 AM
Quote
Right, but its a Bitcoin Problem. At the end its all Bitcoin and when bitcoin has a problem then it doesnt matter anymore that the mainclient has a way to prevent this. You see all these threads on here and on reddit? Imagine how many persons found bitcoin and... it does not work. Great experience. And surely not helpful for adoption. While adoption is the only thing that could help miners to earn considerable amounts of fees in the future.
It's not a Bitcoin problem. Go approach alternative clients' devs and complain to them about fees. Because the fee market exists, and they have to deal with it.

Quote
So? What does it matter when you name the bank only differently? He wanted a currency that isnt controlled by somebody. It doesnt matter who controls it. How its named.
Bank? I don't care how you name it. If it's a trustless bank, then I'm in. Trustless means control of ownership and double-spend prevention.

I wonder what your solutions are to the underlying problem.
legendary
Activity: 2674
Merit: 1083
Legendary Escrow Service - Tip Jar in Profile
July 16, 2015, 05:41:49 AM
I cant believe that some claim that bitcoin is working as intended. Its simply not normal that the forum, reddit and all other places, are full of persons that are annoyed about bitcoin. What should you do with a currency you cant use?
That's the problem with all these people. It's not that they can't use Bitcoin, it's that they don't know how to do it properly.
Hint: pay high enough fees and it works. In this particular situation, Bitcoin works as intended.
PS. It might not work when Bitcoin actually saturates, and that need to be addressed. But this doesn't have much to do with this stupid attack.

And? You await that every adopter has to know these rules? Thats stupid to await and you most probably will agree. People simply want to use a currency. And even with knowing, i would have paid a higher fee, i only didnt know that there was another attack going on.

I see you dont like that attack too but what you say is that some scammers can enforce that all bitcoiners need to pay higher fees. You think thats worthwhile? Miners are greedy they want always more and they will have MUCH less when the next halving comes. And the loss in profits can never gained back with fees. That would simply be impossible. So the higher they force the fees the less likely will adoption become. I mean would you want to use a currency that is ruled by miners at the end? Wouldnt be such a big difference than using banks anymore. It should be a free currency and it shouldnt happen that someone forces everyone to his rules. Guess you think so anyway. Tongue
They might not know all the rules, that's understandable. That's why Bitcoin Core has the fee estimation mechanism that aids in selecting correct fee. It's bad that other clients do not have one, and that mechanism might not be perfect, but it's not a protocol problem.

Right, but its a Bitcoin Problem. At the end its all Bitcoin and when bitcoin has a problem then it doesnt matter anymore that the mainclient has a way to prevent this. You see all these threads on here and on reddit? Imagine how many persons found bitcoin and... it does not work. Great experience. And surely not helpful for adoption. While adoption is the only thing that could help miners to earn considerable amounts of fees in the future.

Quote
Nah... i cant get this being called a test. They dont use testnet or something, they attack a life working economy. Its like ddosing paypal or something. Payments cant happen or are delayed. And everyone would see that its an attack. It doesnt matter what intentions the attacker have. The result is an attack.
JorgeStolfi has already explained why it's not an attack, at least not a malicious one. You might not agree, of course.

Yes, i dont agree. At this point many start to doubt that its only to force a higher block count. It might be to force higher fees.

Quote
And >3 Cent for a tx is still nothing? What currency do you want? Satoshi wanted a currency free of banks. When you cant pay someone a cent for something, only because you would pay 4 in fact then this is no replacement for cash. And as far as i see we want adoption and look pitying to fiat. But fiat has a minimum of 1 Cent and it has zero fee to pay it to someone.
Satoshi wanted a currency free of banks, but it doesn't mean it should be free. I can understand if you somehow want Bitcoin to be everything: trustless, anonymous, robust and cheap. But reality is that there are trade-offs between these properties. And we might never see all them at once. I personally value its trustlessness the most.

So? What does it matter when you name the bank only differently? He wanted a currency that isnt controlled by somebody. It doesnt matter who controls it. How its named.

Seeing that its possible to force all bitcoiners using higher fees is somewhat disturbing. And it seems it doesnt even cost much to do so.
Pages:
Jump to: