Richnomore
Newbie
- Jul 31, 2010
- 10
- 3
I have a Pinterest account for 5 months now. Please let me beta test you bot, for I have just 87 followers..
OK
OK
One thing about browser bots is that they give a certain legitimacy to the actions being performed in the eyes of Pinterest. They took down their API, which means they don't want people accessing the site programmatically (not yet, at least). Also, speed isn't necessarily a factor on Pinterest. Again, a slower, more human-esque bot is actually more advantageous given their somewhat strict rate limits. As far as running multiple instances, it shouldn't be any problem.
Browser bots use more memory and resources on the users machine. Speed is the key, a proper bot will let you set realistic delays in between requests to avoid rate limiting and such. With raw requests you don't have the overheard of having to download the images each time a request to a page is made. A browser bot is going to download images and video every time, clogging up bandwidth for whatever else you might be doing on your computer. Who wants to buy a bot that's a resource hog?
You don't see car manufacturers saying, "our cars are slower than other manufacturers, we done this deliberately so you won't get caught speeding!".
Nice 5 posts you've made, mate. Seems like they're ALL hawking some bot that just follows and unfollows. Some nifty options, but $99 for that piece of junk? I'm tempted to code a free one just to see your bot crash and burn, if this is how you plan on promoting it. Move on, and if you threadjack again I'll report your post... again.We also just released our Pinterest bot for mass following and unfollowing... (VoodooPinner - Google it)
The browser visibility has barely any effect on performance, it is still constructing the DOM in memory, it is still downloading image data which is totally unnecessary .That's why the browser visibility should be toggleable. The tough part isn't downloading the images, any decent PC with a modern internet connection won't see much issue here. The real problem is painting to the screen, which is negated by making the browser invisible. Furthermore, how is speed the key when you're setting delays in the first place? The only difference in speed with two custom-delayed bots is the amount of time it takes to perform the actions, which is at most a second's difference (maybe more if your machine or connection is particularly sluggish). And one can simply account for that by setting a lower delay, no? Ultimately browser bots are more secure than request bots. If you're really going for top speed, then request bots take the advantage- but when you're automating the operation of websites that practice rate limiting (I've yet to bot one that doesn't), browser-based automation wins out in stealth and on a good PC with a good connection, loses out by less than a second in speed (which, as we've discussed, is irrelevant).
To use your own example, you don't see car manufacturers saying "Our car can go faster than their car" either, because there are speed limits. It doesn't matter how fast your car CAN go, it has no NEED to go over 80mph (but obviously they can, the same way a browser based bot is still fast enough to overstep rate limits).
Nice 5 posts you've made, mate. Seems like they're ALL hawking some bot that just follows and unfollows. Some nifty options, but $99 for that piece of junk? I'm tempted to code a free one just to see your bot crash and burn, if this is how you plan on promoting it. Move on, and if you threadjack again I'll report your post... again.
What do you mean by stealthed? A web server has no idea what you're using to view it on the client side, a web server speaks pure http, what needs to be stealthed lol? It's good software design philosophy to build apps efficiently and that have a small footprint.Secure in the sense that it's stealthed to the websites we're automating. Obviously I wouldn't make a bot to do Paypal transactions, but it's the most realistic appearance to the site, and therefore the most secure in regards to the accounts used in the automation process of said sites.
And yes, painting to the screen has a considerable effect on the amount processing power used by a program. Memory? No, because of the exact reasons you stated. But honestly if you have a PC where the memory used in WEB BROWSING (a single tab's worth of web browsing, I might add) is significantly altered to the point where it noticeable affects your other processes, you need an upgrade. I don't make my bots with this guy in mind:
http://www.thismachine.info/ apparently does. Don't question whether or not I know what I'm talking about when you ignore something as simple as user agent. I understand that this information can easily be faked, but nonetheless there are other factors that a script can take into account. For instance, the delay between the time you populate a field and the time you click a "Submit" button. Maybe the script records the click event for the image of the submit button as well as the anchor? It would be incredibly time consuming to find all of the little quirks in a website and account for them, and it's a lot more time-efficient (from a development standpoint) to just replicate real browsing as accurately as possible.What do you mean by stealthed? A web server has no idea what you're using to view it on the client side, a web server speaks pure http, what needs to be stealthed lol? It's good software design philosophy to build apps efficiently and that have a small footprint.
http://www.thismachine.info/ apparently does. Don't question whether or not I know what I'm talking about when you ignore something as simple as user agent. I understand that this information can easily be faked, but nonetheless there are other factors that a script can take into account. For instance, the delay between the time you populate a field and the time you click a "Submit" button. Maybe the script records the click event for the image of the submit button as well as the anchor? It would be incredibly time consuming to find all of the little quirks in a website and account for them, and it's a lot more time-efficient (from a development standpoint) to just replicate real browsing as accurately as possible.
Also, as far as scrapers go, webrequests are also less realistic in the server's eyes because I'm pretty sure it can be easily seen whether or not the images have been downloaded. And sure, some people disable images in their browser, so this isn't always a factor, but honestly; who uses Pinterest with images disabled?
I said that I know it can be easily faked. The fact of the matter is that we don't know what a script is and isn't recording, and for the sake of safety it's more secure for the accounts involved to operate via a web browser.Useragent is sent as a request header, did you not know?
If a script records anything and sends it to the server, it's sent via http, you just have to look at the requests and see what's going on.
How is a web request less realistic from the servers point of view? Pretty much EVERY piece of data sent to the web server is a http request!
I said that I know it can be easily faked. The fact of the matter is that we don't know what a script is and isn't recording, and for the sake of safety it's more secure for the accounts involved to operate via a web browser.
Look, obviously we have irreconcilable differences on this point of view. I think that the only advantage webrequests have is speed, which is irrelevant. You disagree. The reasoning behind that belief is beyond me since I've outlined clearly why speed has little to do with it, and we've digressed to a matter that's ultimately development philosophy (whether or not it's "safer" for the accounts used if one uses a browser, or not). I'm happy to leave it as an agreement to disagree, and would like to end the discussion on a positive note by once again posting the hilarious picture of the guy who thought computers in 2004 would have steering wheels:
P.S. I don't want to create any sort of non-amicable relationship, since you obviously know your shit and I have no desire to attempt to discredit you as a capable programmer. If you want to jot this down as a win in your book, you're free to do so.
Users can disable local shared objects using the Global Storage Settings panel of the online Settings Manager at Adobe's website.[SUP]http://en.wikipedia.org/wiki/Local_shared_object#cite_note-4[/SUP]
I'm going to quote Wikipedia here: