Title: Merit Double Spent - experiment and failed - may be we have a clue Post by: BitcoinGirl.Club on July 07, 2019, 08:58:12 AM I wanted to send some merit to this post: https://bitcointalk.org/index.php?topic=5161490.0
I had 8 sMerit left. I thought to send 3 sMerit for the topic. I knew there was an issue with sending merit in the send merit page (action=merit). Once you refresh, it spend the smerit again. I never experienced this so I was just curious to check it and was hoping may be theymos found a solution. So I refreshed the page and it showed this pop up of resending the data to the server. I clicked resend and it sent the same sMerit again. Broke my hurt knowing that it was not fixed, felt a bit angry too to lose 3 more smerits. Now I am left with 2 sMerits so I thought lets refresh it again knowing that this time there should be some message or any error that will tell me that there were not enough smerit (I was interested to see the message this time). But instead I saw this: This means, theymos tried to handle the event in somewhere in the code in the same page (action=merit) to stop double spending but it is not working. And while I was experimenting this, I had this idea in mind: Instead of bringing the user in the same page (action=merit) once they hit the send merit button, why not redirect the user in the topic page (or any other page but topic page is more logical). This way the send merit page (action=merit) is not going to load again with the same set of code when we hit the send merit button. Make sense? Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: hilariousetc on July 07, 2019, 09:16:54 AM I knew there was an issue with sending merit in the send merit page (action=merit). Once you refresh, it spend the smerit again. I never experienced this so I was just curious to check it and was hoping may be theymos found a solution. So I refreshed the page and it showed this pop up of resending the data to the server. I clicked resend and it sent the same sMerit again. Broke my hurt knowing that it was not fixed, felt a bit angry too to lose 3 more smerits. I think you've basically just done the digital equivalent of touching wet paint to see if it was wet even though there was a sign there saying do not touch - wet paint. https://i.kym-cdn.com/photos/images/original/001/272/716/3d9.jpg I thought about giving some merits to replace the ones you lost but touching wet paint when you know it's wet is a pretty silly thing to do. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: BitcoinGirl.Club on July 07, 2019, 09:41:28 AM I thought about giving some merits to replace the ones you lost but touching wet paint when you know it's wet is a pretty silly thing to do. You can't replace it once it's gone, can you? - unless theymos reverse the merit manually. Meanwhile I do not need it too. Yes, it was silly but it's the curiosity lol. Glad I am not stupid enough to pull a trigger to see how it feels to die 🤪By the way, I think we have a possible solution here to stop it. Not sure if this will get theymos's attention. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: hilariousetc on July 07, 2019, 09:51:00 AM I thought about giving some merits to replace the ones you lost but touching wet paint when you know it's wet is a pretty silly thing to do. You can't replace it once it's gone, can you? - unless theymos reverse the merit manually. Meanwhile I do not need it too. Yes, it was silly but it's the curiosity lol. Glad I am not stupid enough to pull a trigger to see how it feels to die 🤪By the way, I think we have a possible solution here to stop it. Not sure if this will get theymos's attention. Only theymos can return the merit but I'm not sure if he will or not and I don't think it's worth the time pursing and you should just accept the loss of merits and move on in this case. However, it would be helpful if the bug was fixed so it doesn't happen again in the future so you can keep pushing for a change/fix if you wish but it certainly wouldn't be on my list of priorities (for either you or theymos). Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: DdmrDdmr on July 07, 2019, 02:47:11 PM <...> You were 38 seconds too slow performing your experiment
:Alright, you now can't merit the same post with the same amount within 60 seconds. The timeframe is meant to cover the most frequent case scenarios, but does not cover all.Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: BitcoinGirl.Club on July 07, 2019, 03:22:25 PM <...> You were 38 seconds too slow performing your experiment
:Alright, you now can't merit the same post with the same amount within 60 seconds. The timeframe is meant to cover the most frequent case scenarios, but does not cover all.Anyway, this was a tricky move to stop double spending and to have the same page at the same time so that if a user wants to spend more sMerits for the same user without moving to any other page. If this is the cause then it's logical to come back to the same page however if this is not the case then the user can be redirected to the topic page once the event has been handled. In this case there will be no need for the extra code to prevent spending same amount of merit in 60 seconds. Only theymos can return the merit but I'm not sure if he will or not and I don't think it's worth the time pursing and you should just accept the loss of merits and move on in this case. However, it would be helpful if the bug was fixed so it doesn't happen again in the future so you can keep pushing for a change/fix if you wish but it certainly wouldn't be on my list of priorities (for either you or theymos). And I already said that I don't need those sMerit to have returned. With you, I also realize (agree) that it's worthless. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: IMadeYouReadThis on July 08, 2019, 05:03:39 AM Redirecting the user to the merited post after the merits are sent could be the best solution for this unwanted multi meriting problem.
60 seconds cool down time is also a better alternative but if a person merits a post and keeps the screen ON for 60 seconds doing some other work, and comes back after 60 seconds and refresh the page to update the content, the merit would be sent again without even getting noticed and without any warning on the screen. I would support Mr.Pamoldor's suggestion above, atleast he tried touching the wet pants, which requires high motivation anyways. ;D Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: DdmrDdmr on July 08, 2019, 11:32:11 AM <...> Regardless, the number of cases is small really in the big picture. I make it: 713 cases had sent the same amount of sMerit to the same post and person in <= 60 seconds, prior to the 29/12/2018 fix (60 second cooldown time). 431 cases had sent the same amount of sMerit to the same post and person in > 60 seconds and <= 600 seconds. 189 cases had sent the same amount of sMerit to the same post and person in > 600 seconds and <= 3600 seconds. Out of 209.332 overall sMerit TXs. You cant tell though if the intent was to remerit (i.e I gave 1, but perhaps I should have given 2), or the refresh page effect after periods of time over 1 minute. <
> Yes, according to your data this looks like only around 5% of the total merit transactions face such merit double sending problems approximately <
> Way less really. The <= 60 group is 0,34% of TXs, but should not take place any more. The 60..600 seconds group takes place in 0,21% of Tx, and the 600 .. 3600 seconds group only takes place in 0,09% of TXs.Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: IMadeYouReadThis on July 08, 2019, 12:47:11 PM <...> Regardless, the number of cases is small really in the big picture. I make it: 713 cases had sent the same amount of sMerit to the same post and person in <= 60 seconds, prior to the 29/12/2018 fix (60 second cooldown time). 431 cases had sent the same amount of sMerit to the same post and person in > 60 seconds and <= 600 seconds. 189 cases had sent the same amount of sMerit to the same post and person in > 600 seconds and <= 3600 seconds. Out of 209.332 overall sMerit TXs. Yo cant tell though if the intent was to remerit (i.e I gave 1, but perhaps I should have given 2), or the refresh page effect after periods of time over 1 minute. Yes, according to your data this looks like only around 5% of the total merit transactions face such merit double sending problems approximately I agree by looking at the numbers it is not that big of an issue and most of the generous merit sender's would not do such double spending mistakes often. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: Welsh on July 08, 2019, 04:56:22 PM Yes, according to your data this looks like only around 5% of the total merit transactions face such merit double sending problems approximately And some of these will be intentional. I've merited a post, and thought "actually this probably deserves a little more", and have then proceeded to reward more merit. In the grand scheme of things this really isn't a problem, and only affects such a small amount of users. I agree by looking at the numbers it is not that big of an issue and most of the generous merit sender's would not do such double spending mistakes often. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: IMadeYouReadThis on July 08, 2019, 05:39:45 PM Yes, according to your data this looks like only around 5% of the total merit transactions face such merit double sending problems approximately And some of these will be intentional. I've merited a post, and thought "actually this probably deserves a little more", and have then proceeded to reward more merit. In the grand scheme of things this really isn't a problem, and only affects such a small amount of users. I agree by looking at the numbers it is not that big of an issue and most of the generous merit sender's would not do such double spending mistakes often. Yaa, its overall not that big of an issue, and theymos's current solution of 60 second's warning seems to be working good. :D Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: TheBeardedBaby on July 08, 2019, 07:22:23 PM There it is: Alright, you now can't merit the same post with the same amount within 60 seconds. Edited, just saw Ddmr^2 comment above.... And your timing between the merited posts... Unlucky this time. Title: Re: Merit Double Spent - experiment and failed - may be we have a clue Post by: philipma1957 on July 08, 2019, 08:36:05 PM <...> You were 38 seconds too slow performing your experiment
:Alright, you now can't merit the same post with the same amount within 60 seconds. The timeframe is meant to cover the most frequent case scenarios, but does not cover all.Anyway, this was a tricky move to stop double spending and to have the same page at the same time so that if a user wants to spend more sMerits for the same user without moving to any other page. If this is the cause then it's logical to come back to the same page however if this is not the case then the user can be redirected to the topic page once the event has been handled. In this case there will be no need for the extra code to prevent spending same amount of merit in 60 seconds. Only theymos can return the merit but I'm not sure if he will or not and I don't think it's worth the time pursing and you should just accept the loss of merits and move on in this case. However, it would be helpful if the bug was fixed so it doesn't happen again in the future so you can keep pushing for a change/fix if you wish but it certainly wouldn't be on my list of priorities (for either you or theymos). And I already said that I don't need those sMerit to have returned. With you, I also realize (agree) that it's worthless. I am glad you did this thread. I gave you some merits. Please do not do it again. Btw I have spent around 20 merits testing it. |