Questions For Boting/Automation

I'm a little confused with this.
Okay say you are coding a twitter follower bot and have a list of people you want to follow, such as.
@alice
@bob
...Etc
Which is the best way to go about this?

1.
Have the bot go to twitter.com/Alice then click follow and do this for the rest of the users.
Possible Problems - Could leave mad footprints because users typically click from a certain page to go navigate to a users page,Not just directly go there (at-least not in mass), i feel some websites record this.

2.
Tweek the html code from a single persons follow button.
Such as

Into the below to follow alice.

Possible Problems - Twitter could detect that bot is on single page mass following.

Or should i just scrap the idea of following from a list and just go to X user's followers page and/or the search bar then scroll and follow from that?

I guess at the end of the day it's better to build something that emulates human behavior, such as scrolling/viewing photos/using search bar, as opposed to something that just "logs in ,mass follows then logs out".


Another random question.
Can websites such as Facebook/Instagram/Twitter etc record mouse movement?

Perhaps i am just being Paranoid.
Any insight or advice would be appreciated

I have a single threaded bot that does this. You are more than welcome to try it out.

sevnthaxiom said:

I have a single threaded bot that does this. You are more than welcome to try it out.

Click to expand...

...

My take on this,

Right now twitter, facebook, instagram etc.... they don't record the mouse movements and carefully track everything being human or not, HOWEVER you will want to just pay attention to acting within normal thresholds.

In other words, I think having the bot go to the page on twitter and press follow is perfectly fine, but twitter may notice an account following 4000 people per day, you just have to test the limits.

You are being overly paranoid in thinking they are watching *that* closely. Although im sure its within their reach they aren't detecting mouse movements of their millions of users at this time. Maybe one day tho lol.

There's 2 ways you can automate websites. By simulating user actions or using http requests.

Simulating user actions aka automating the browser is super easy but can lead to leaving footprints. You can use selenium and any language of your choice to automate chrome, firefox, internet explorer and headless browsers (phantomjs, etc).

When I can I always use http requests. This means I make a piece of code in which I tell the program which url to send a request to, which type of request it is (GET, POST, DELETE, etc), which parameters will be sent (for instance when you signup on a website), etc. It is quicker, you can add any header you want, etc.

Which way are you thinking to choose?

mnunes532 said:

There's 2 ways you can automate websites. By simulating user actions or using http requests.
Which way are you thinking to choose?

Click to expand...

Selenium. Chrome and/or Phantomjs
Besides useragents/proxy/windowsize
What footprints can be left?

yellowcat said:

Selenium. Chrome and/or Phantomjs
Besides useragents/proxy/windowsize
What footprints can be left?

Click to expand...

Websites can track your mouse and you can't simulate that using selenium. For instance, you can't create google accounts using selenium, they know you are using some sort of automated actions and won't let you proceed or always ask for phone verification.

Whoa that was a super cool site you posted. Too bad you deleted it ;P

https://panopticlick.eff.org/ I deleted it because I didn't think it was the right website I wanted to share. If you scroll down you'll see plenty of info a website can get from your browser.

mnunes532 said:

Websites can track your mouse and you can't simulate that using selenium. For instance, you can't create google accounts using selenium, they know you are using some sort of automated actions and won't let you proceed or always ask for phone verification.

Click to expand...

It's funny when I'm working on Google stuff ( yes folks clients do ask me to work nice with Google - but i don't ever send reports OR stor data in gmail or gdocs) and I click quickly on the same button and repeat the same thing very similarly, their potsmoking Rank Brain sends me recaptchas to prove I'm not BartoBrain.

What language should I be using to manage https requests?

TheTyper said:

What language should I be using to manage https requests?

Click to expand...

Any language. It depends on you.