IDEA -- Automated sig camp join tool - Feature and functionality Suggestions

[NyeFe]

You didn't tell me what you thought of the idea, do you like it, would you use it, what don't you like about it?

To be honest with you.. maybe i'm wrong, but i think that it will take you a lot of work, but you will not get out of it profits.

Tell me please, for who really it will be most useful in your opinion: for users, campaign managers or for advertisers?

In my opinion all of them like to have everything under their own control, and your service will not get much attention.

For me, the advertiser, along with the user would be the most important.

The details does sound a lot, but in reality it's probably 1 days worth of algorithms and another day worth of testing.
How can one get attention without making any action?

I'm sure there's someone out their who would rather get paid weekly, for posting, instead of searching for the latest signature campaigns, each time one ends.

Advertisers always employ manager - your service can't change that, because posts quality is most important and here have to be manual control.

It can be useful tool for campaign managers, when they are looking for quality posters, but as i know for them, once again, most important is posts quality.

For users.. hmm they have this thread:
https://bitcointalk.org/index.php?topic=615953.0
and if they are active, they watch very carefully service section, what is enough to be well informed.

I like your idea, but like i wrote, from this project you will gain only few +1 and many thanks

Edit: Nobody knows in which direction will go signature campaigns. It is an option that will go to the level of unnatural spam without having to check the quality (some sig campaigns are doing it now). Then indeed your service can be used directly by the less demanding advertisers.

I like your help and responses, they're really helpful.

If more people feel the same way as you, then i might stop the project, which would save me the stress of its downfall.

[achow101]

This is an interesting idea. It could help to curb some spam depending on how the payments work. You could make it so that users are paid at a fixed rate and the user doesn't know the terms of the payment, just that if they are chosen to be in a campaign, they will get at least X Bitcoin per week (or month or whatever). That would help to curb spam as then there isn't an incentive for people to spam to meet their post requirements.

There is, of course, the huge security issue. The problem is that many users may not feel comfortable with entrusting your service with their login credentials, especially since having the login gives full access to the account. This could potentially lead to privacy leaks and loss of the account if your service were to be hacked. The risk goes up for higher ranking members and those with green trust.

Lastly, there is also a technical issue. Bitcointalk only allows you to perform a login action once every 45 seconds per IP. If you need to change the signatures and avatars on a lot of accounts, you're going to need a lot of IPs or servers. They will also IP ban you if there are more than one request per second per IP.

[NyeFe]

This is an interesting idea. It could help to curb some spam depending on how the payments work. You could make it so that users are paid at a fixed rate and the user doesn't know the terms of the payment, just that if they are chosen to be in a campaign, they will get at least X Bitcoin per week (or month or whatever). That would help to curb spam as then there isn't an incentive for people to spam to meet their post requirements.

There is, of course, the huge security issue. The problem is that many users may not feel comfortable with entrusting your service with their login credentials, especially since having the login gives full access to the account. This could potentially lead to privacy leaks and loss of the account if your service were to be hacked. The risk goes up for higher ranking members and those with green trust.

Lastly, there is also a technical issue. Bitcointalk only allows you to perform a login action once every 45 seconds per IP. If you need to change the signatures and avatars on a lot of accounts, you're going to need a lot of IPs or servers. They will also IP ban you if there are more than one request per second per IP.

Thanks for the comment knight, your idea gave me hope. You managed to solve the spamming issue. Because of some pf your comments, I've been able to find some of the other solutions.

For the payment method, your suggested fixed price & hidden terms would be the way to go.

Because of the login system which you mentioned above, we won't require users to entrust us with their login detail. Instead, we would bypass the whole system totally, so the user would only need to give us their cookie_id which is already logged_in. So with this method, even if we did get hacked *hypothetically* the hacker wouldn't be able to change account settings such as password, email...

Hopefully more people like you & feinter will show their interest so we can solve any issues and find a way to meet everyones requirements, to get this project Rock 'n' Rolling

[hermanhs09]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

[NyeFe]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

Thanks for show interest herman, i really appreciate it.

Knights login issue was a good one. There's different ways to actually login to bitcointalk, which can bypass all of this issues entirely, the best thing is that these other methods wouldn't require the storage of login credentials (username/password) but would still provide the same access to edit sigs.

[achow101]

Because of the login system which you mentioned above, we won't require users to entrust us with their login detail. Instead, we would bypass the whole system totally, so the user would only need to give us their cookie_id which is already logged_in. So with this method, even if we did get hacked *hypothetically* the hacker wouldn't be able to change account settings such as password, email...

AFAIK cookies can't really be revoked without the admins going into the server and deleting all of the cookies, so that is an issue. If you were to be hacked, the attacker could pretend to be that person even if the user changed passwords. At the very least it would be a nuisance and at worst, the attacker could use high ranking and potentially positive trust accounts to scam others.

It would be a lot easier if Bitcointalk had something like OAuth 2.0

[NyeFe]

Because of the login system which you mentioned above, we won't require users to entrust us with their login detail. Instead, we would bypass the whole system totally, so the user would only need to give us their cookie_id which is already logged_in. So with this method, even if we did get hacked *hypothetically* the hacker wouldn't be able to change account settings such as password, email...

AFAIK cookies can't really be revoked without the admins going into the server and deleting all of the cookies, so that is an issue. If you were to be hacked, the attacker could pretend to be that person even if the user changed passwords. At the very least it would be a nuisance and at worst, the attacker could use high ranking and potentially positive trust accounts to scam others.

It would be a lot easier if Bitcointalk had something like OAuth 2.0

That would be an issue if we were to be hacked. It would be an issue if any service (coinbase, btc-e, xapo, coindesk) were to be hacked, we understand the important of following fully strict security procedures.

The important thing (once the security procedures are taken care of), is that we've been able to work together, to find a suitable solution, which would actually allow automating signature campaigns.

Bitcointalk is merely a clean, default install of SMF forum, as you already know. It would be easier if it did have it, but I wouldn't hold my breath for an OAuth 2.0 API to come out anytime ever

[Newcoins2020]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

If you let the renter deposit some money, it could be successful, since he/she will be more careful. The only problem that 'is left' is for a renter to be a scammer. You could bypass this by having them rotate accounts though.

[NyeFe]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

If you let the renter deposit some money, it could be successful, since he/she will be more careful. The only problem that 'is left' is for a renter to be a scammer. You could bypass this by having them rotate accounts though.

In our case, the renter would actually be the server. We wont actually give the account to advertiser. Instead we'll actually just edit the sig camps for them, when they've paid for a new signature campaign (all without them having actually access to the account).

This in my opinion is probably the safest option.

[wuvdoll]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

If you let the renter deposit some money, it could be successful, since he/she will be more careful. The only problem that 'is left' is for a renter to be a scammer. You could bypass this by having them rotate accounts though.

In our case, the renter would actually be the server. We wont actually give the account to advertiser. Instead we'll actually just edit the sig camps for them, when they've paid for a new signature campaign (all without them having actually access to the account).

This in my opinion is probably the safest option.

Good idea for automated thing it will have a flexibility in work and such. But it also have a lot of time to work the project can have a lot of error and require maximum effort.
The site should maintain its uptime and other generation will know this forum honestly the idea is as crazy as it is it's great to have such an implementation.

[NyeFe]

I think this is a good idea, however as KnightDK said there are many restrictions imposed on by the bitcointalk forum that doesn't allow you to do exactly what you want. Also, the storage of so many login credentials at one single place could prove to be disastrous.

But having said that, account renting is a very good idea. Also make sure that the renter deposits some money to compensate for any losses incurred with the account(bans, negative trusts, etc.)

If you let the renter deposit some money, it could be successful, since he/she will be more careful. The only problem that 'is left' is for a renter to be a scammer. You could bypass this by having them rotate accounts though.

In our case, the renter would actually be the server. We wont actually give the account to advertiser. Instead we'll actually just edit the sig camps for them, when they've paid for a new signature campaign (all without them having actually access to the account).

This in my opinion is probably the safest option.

Good idea for automated thing it will have a flexibility in work and such. But it also have a lot of time to work the project can have a lot of error and require maximum effort.
The site should maintain its uptime and other generation will know this forum honestly the idea is as crazy as it is it's great to have such an implementation.

Hi wuvdoll,

I'm really happy you understand, and find this concept really helpful. Hopefully more people would join to have signature campaigns automated for them.

[NyeFe]

[Update]

A preview has been released.

To support everyone, and to make it as simple as possible, we decided to use browser extensions instead of Desktop apps.

[achow101]

[Update]

A preview has been released.

To support everyone, and to make it as simple as possible, we decided to use browser extensions instead of Desktop apps.

How exactly does a browser extension with with this? I thought it was supposed to just be a website.

[NyeFe]

[Update]

A preview has been released.

To support everyone, and to make it as simple as possible, we decided to use browser extensions instead of Desktop apps.

How exactly does a browser extension with with this? I thought it was supposed to just be a website.

Noooooooo We're not providing junk! This is some hardcore stuff. This is real. This is powerful.
We're trying to distinguish our service from the false PHP clutter out there


How it works, is that it keeps you logged-in and updated (on our server) as long as you're logged-in on bitcointalk.

From there, your sigs will be updated automatically, and payments will made on the bitcoin address on your profile (weekly).
On the server side, it checks every hour that each account (member in a sig) is following the rules or else, they will be banned for 24hrs ...

[achow101]

[Update]

A preview has been released.

To support everyone, and to make it as simple as possible, we decided to use browser extensions instead of Desktop apps.

How exactly does a browser extension with with this? I thought it was supposed to just be a website.

Noooooooo We're not providing junk! This is real. This is powerful.


How it works, is that it keeps you logged-in and updated (on our server) as long as you're logged-in on bitcointalk.

From there, your sigs will be updated automatically, and payments will bitcoin_address on the bitcoin address on your profile (weekly).
On the server side, it checks every hour that each account (member in a sig) is following the rules or else, they will be banned for 24hrs ...

Oh I see. This way it pretty much never risks users accounts even if you are hacked. It would require a malicious browser extension.

[NyeFe]

[Update]

A preview has been released.

To support everyone, and to make it as simple as possible, we decided to use browser extensions instead of Desktop apps.

How exactly does a browser extension with with this? I thought it was supposed to just be a website.

Noooooooo We're not providing junk! This is real. This is powerful.


How it works, is that it keeps you logged-in and updated (on our server) as long as you're logged-in on bitcointalk.

From there, your sigs will be updated automatically, and payments will bitcoin_address on the bitcoin address on your profile (weekly).
On the server side, it checks every hour that each account (member in a sig) is following the rules or else, they will be banned for 24hrs ...

Oh I see. This way it pretty much never risks users accounts even if you are hacked. It would require a malicious browser extension.

That's the key actually, nice spot. Even if we're hacked, the user accounts will never be risked.

At most, a user would just need to logout and back in, and any access to anyone will be gone instantly. Its just amazing.

[NyeFe]

Somewhere on this thread we agreed that it wouldn't have been a good idea to show the price that an advertiser/campaign manage is paying the users, because it would lead to spamy posts. Does anyone else agree with me?

[achow101]

Somewhere on this thread we agreed that it wouldn't have been a good idea to show the price that an advertiser/campaign manage is paying the users, because it would lead to spamy posts. Does anyone else agree with me?

I was thinking that the users should be kept in the dark about all aspects of their payment except for how much they are paid and how frequently. So all the extension would do is just calculate an amount that the user will probably be paid based upon the user's past posting behavior and display that as well as the next payday. In this way, I think that people have less of an incentive to spam because they do not know how much they need to post and they know that they will probably get the calculated amount if they don't change their behavior.

Also, maybe there should be something in the extension that tells the users which posts they made are considered spam and won't be paid for so that people can learn to make less spammy posts.

Lastly, I think we should also try using Google's Prediction API: https://cloud.google.com/prediction/ and use its machine learning capabilities to have it be better at detecting spam than a simple bot. I'm going to try to make that happen.

[NyeFe]

Somewhere on this thread we agreed that it wouldn't have been a good idea to show the price that an advertiser/campaign manage is paying the users, because it would lead to spamy posts. Does anyone else agree with me?

I was thinking that the users should be kept in the dark about all aspects of their payment except for how much they are paid and how frequently. So all the extension would do is just calculate an amount that the user will probably be paid based upon the user's past posting behavior and display that as well as the next payday. In this way, I think that people have less of an incentive to spam because they do not know how much they need to post and they know that they will probably get the calculated amount if they don't change their behavior.

Also, maybe there should be something in the extension that tells the users which posts they made are considered spam and won't be paid for so that people can learn to make less spammy posts.

Lastly, I think we should also try using Google's Prediction API: https://cloud.google.com/prediction/ and use its machine learning capabilities to have it be better at detecting spam than a simple bot. I'm going to try to make that happen.

Wow knightdk! I really appreciate the help

For the first part, do you mean like keeping a record of the average posts per week  (for at least 2 weeks) a user makes & using those figures to calculate how much they receive for each campaign? (if its a pay per/post campaign)

For the second part, I could add another link "History" which has weekly logs of all the posts that where rejected by the application or the campaign manage, with explanations. That's a great idea.

For the last part, I haven't actually considered Google's API as an extension to this. I but if you can make it happen, and if it's successful, then it would provide a really intelligent, unbeatable spam detection feature.

[achow101]

Wow knightdk! I really appreciate the help

For the first part, do you mean like keeping a record of the average posts per week  (for at least 2 weeks) a user makes & using those figures to calculate how much they receive for each campaign? (if its a pay per/post campaign)

Yes. It could crawl through the users post history and see how many good posts they made in each time period. Then average it and use that post number for calculating an estimated pay for pay per post campaigns. For fixed rate, they would just see the fixed rate.

Knowing the post history of a user would help with assigning campaigns since you would know how frequently they post and where and whether that matches a campaign's restrictions.

For the second part, I could add another link "History" which has weekly logs of all the posts that where rejected by the application or the campaign manage, with explanations. That's a great idea.

For the last part, I haven't actually considered Google's API as an extension to this. I but if you can make it happen, and if it's successful, then it would provide a really intelligent, unbeatable spam detection feature.

I've already started: https://bitcointalk.org/index.php?topic=1514300.0