Blog Commenting - My Project Question

[Jeff 0]

Hey Ubotters -

 

My Goal:

 

Okay, so I'm working on a bot that will troll around blogs all day long dropping comments for me. Here is the process I'm working with so far:

 

My Routine:

 

1) Subroutine 1: Visit google blogsearch & parse out a large list of keyword related blogs for commenting on. (Thanks to JohnB with help on this)

 

2) Subroutine 2: Visit each individual blog URL and determine whether or not it knows how to comment here. If so go to step 3, otherwise loop to next URL.

 

3) Subroutine 3: If there are comments, copy the most recent users comment and name and make the same comment with url and appended message.

 

4) Subroutine 4: If "your comment is awaiting moderation" is on page after comment, store URL as possible success.

 

Loop through all the blogs doing this routine.

 

Where I am:

 

Step 1, - Done.

 

Step 2, I'm having trouble wrapping my mind around. As you all know, there are SO many different variables that come into play when landing on a blog page of where to click, what to click on, and then what type of form is displayed, and everything in-between. It would be easiest (not BEST), but easiest, if perhaps I could find a way to only comment on wordpress blogs, that way steps #2 and #3 are more manageable.

 

Steps 3 and 4 I can handle.

 

Do any gurus out there have any advice or tips for me that might help me accomplish my goal? I would really appreciate your input. Thanks for taking the time.

 

- Jeff

[Frank 177]

Hey Jeff,

 

Ambitious project to be sure!!!

 

But what you will shoot for, especially if you are going for 100% automation is a percentage that will just work. Simple try to fill in fields for various platforms. If the field doesn't exist on the page, it will not affect anything.

 

Next. just shoot for as many as possible. If you get a 60% success rate and you create 1000 backlinks per day, you are still doing pretty darn good!

 

Also other hints - really, really think out the procedure before you try to generate one bubble of code. Take it manually and build out processess using subroutines and then push them all together in the end. Makes building much easier and troubleshooting in a month or two after, actually manageable.

 

Frank

[Jeff 0]

Hey Jeff,

 

Ambitious project to be sure!!!

 

But what you will shoot for, especially if you are going for 100% automation is a percentage that will just work. Simple try to fill in fields for various platforms. If the field doesn't exist on the page, it will not affect anything.

 

Next. just shoot for as many as possible. If you get a 60% success rate and you create 1000 backlinks per day, you are still doing pretty darn good!

 

Also other hints - really, really think out the procedure before you try to generate one bubble of code. Take it manually and build out processess using subroutines and then push them all together in the end. Makes building much easier and troubleshooting in a month or two after, actually manageable.

 

Frank

 

I've drawn out the whole process on paper and planned each individual smaller subroutine. So I've got a good plan. But I've never thought about it like you just said, just firing at several and hoping that I have a good percentage that hit. I think that will be much easier than trying to have a portion of the bot be a quite sophisticated "decision maker". Thanks for the tips, I think I'm going to implement this.

 

Has anyone else out there made something like this?

[Jeff 0]

Hey Jeff,

 

Ambitious project to be sure!!!

 

But what you will shoot for, especially if you are going for 100% automation is a percentage that will just work. Simple try to fill in fields for various platforms. If the field doesn't exist on the page, it will not affect anything.

 

Next. just shoot for as many as possible. If you get a 60% success rate and you create 1000 backlinks per day, you are still doing pretty darn good!

 

Also other hints - really, really think out the procedure before you try to generate one bubble of code. Take it manually and build out processess using subroutines and then push them all together in the end. Makes building much easier and troubleshooting in a month or two after, actually manageable.

 

Frank

 

Frank, quick question for you.

 

You said that if the field I'm searching for doesn't exist on the page, it will not affect anything. I'm having unpleasant results in regards to this when having a bunch of "choose by attributes" one after another, looking for different links on the page that either say "leave a comment", "# comments", or "comments #", to click on. So each of my "choose by attributes" I put in a row go like this:

 

Choose by attribute (outer html) "<A class=comment-link onclick=""href="*">* comments</A>" wildcards

 

Choose by attribute (inner text) "Leave a comment" exact match

 

Choose by attribute (inner text) "Comments *" wildcards

 

Click Chosen

 

The problem is that if it finds the first one, then it obviously doesn't find the following two, but an error pops up that says "No elements have been chosen! This command requires that elements are properly chosen first, using some sort of choose command. Command: click chosen". So the one it found, was then replaced by something not being found, and failing commences after that when it tries to click on nothing.

 

It appears that if certain nodes in the path of operation don't fulfill, it stops the bot in its tracks. What am I doing wrong here?

[JohnB 255]

Jeff, another option is to use an if/then statement with the search page qualifier. Inside the search page you can place a contains qualifier and inside that you can place an either qualifier. This will then handle a search for multiple items. The contains qualifier will be based on the $document text. (The page inside the browser). This equates to a "comprehensive" search for the comments field(s).

 

John

[meter 145]

Jeff prepare for a headache. A lot of pages will crash Ubot with javascript/flash/java/popups/excessive size.

[Abs* 12]

Jeff prepare for a headache. A lot of pages will crash Ubot with javascript/flash/java/popups/excessive size.

 

Yes thats true - however you can get round this by disabling java in Internet explorer - I would also suggest disabling images or you will need to wait for the oh so many avatars to load first -

 

I already have a bot that does what you want to do however it allows users much more options and lets them use their own comments -

 

My suggestion to you would be to work on platforms at a time then merge them together using if else statements -

 

So the first thing would be to find blog platforms - then find the comment pages and code them into the bot - Then work through other platforms and do the same - It honestly isnt hard as you only have like 4 fields per a blog - One will be to fill the name - the other the email - then website then comment - and of course then you will need to hit the submit button -

 

Once you have all of your platforms coded - Place each step into a subroutine so that your organized -

 

So you could have something like

 

Sub - fill name

sub fill email

sub fill url

sub fill comment

sub hit submit

 

In each sub you will have if then else statements so that the bot knows what command to fire - If they dont exist then the bot will move onto the next list item

 

not sure if this makes sense -

 

thanks

 

abs

[Jeff 0]

Yes thats true - however you can get round this by disabling java in Internet explorer - I would also suggest disabling images or you will need to wait for the oh so many avatars to load first -

 

I already have a bot that does what you want to do however it allows users much more options and lets them use their own comments -

 

My suggestion to you would be to work on platforms at a time then merge them together using if else statements -

 

So the first thing would be to find blog platforms - then find the comment pages and code them into the bot - Then work through other platforms and do the same - It honestly isnt hard as you only have like 4 fields per a blog - One will be to fill the name - the other the email - then website then comment - and of course then you will need to hit the submit button -

 

Once you have all of your platforms coded - Place each step into a subroutine so that your organized -

 

So you could have something like

 

Sub - fill name

sub fill email

sub fill url

sub fill comment

sub hit submit

 

In each sub you will have if then else statements so that the bot knows what command to fire - If they dont exist then the bot will move onto the next list item

 

not sure if this makes sense -

 

thanks

 

abs

 

Abs - Thanks for taking the time to write out this comment, I really appreciate it. I'll be back in here to provide updates on my progress for anyone who is interested.

[Jeff 0]

The only way I can seem to distinguish between the different "Leave a Comment" "* Comments" "Comments *" etc. possibilities is to use Choose By Attribute.

 

The problem is that I can't use Choose By Attribute in my qualifier field of the IF statement. So I guess my question is, how am I to ever try one thing after another by nesting IF statements if I can't use the choose by attribute as a qualifier? Of course I could delete the qualifier place holder and just stick a Choose By Attribute in its place, but I don't think that will work.

 

I know this is probably a disconnect in my knowledge level I apologize in advance for my ignorance

 

Anything come to mind here?

 

Thanks

[meter 145]

Try to differentiate by searching html for names of fields you need to fill.

[Abs* 12]

HI Jeff - You will need to place a search page command in the If field -

 

So it would look like

 

IF> Search page (name=author)

Then > fill field -Keyword

Else> xxxxx

 

On many occasions you will be able to leave the Else statement empty - there are also many times when the else statement will be very useful to use

 

thanks

 

abs

[James J 3]

Actually you can use the "choose by attribute" as a qualifier.

Just delete all the place holders then click on the "then" node.

Then go into the browser window and use the choose by attribute

feature as normal. Ubot will then populate the area just above

your then statement with your "choose by attribute" node.

 

I've used this method several times in a bot I am currently

creating.

 

James

[meter 145]

That's a pretty neat tip James, thanks

[Biks 9]

What you really should be doing first is scraping for certain blog footprints. This way the attribute tags/names will be more consistent and your bot will work more likely. I've built a couple of auto commenter bots and it's surprising how many variations you encounter within just one platform.

 

A sneaky thing I sorta got to work is scraping the authors name off the page, insert their NAME into your spam post. (I always come back to this blog to see what <AUTHORS NAME> has written.) How can this be spam when you mentioned the authors name? :-) Note: Scrapebox DOESN'T do this. Blogengine blogs have a tag for authors name. Still, my bot didn't always work because of the subtle differences between all the versions and themes.

 

The next step down was to scrape the domain name and insert THAT into your spam comment. (I always come back to <DOMAIN NAME> and see what has been posted! Great stuff!) As an example most Wordpress blogs don't include the authors name.

 

At this point I'm not sure if hammering comments is very effective. I can't see how Google is just ignoring, seeing that Scrapebox is cheap and ubiquitous. Can anyone still vouche for comment spamming?

[meter 145]

Ranked 11 sites last week on the first page with only Scrapebox spam on autoapprove lists. Yes, it still works great Just picked up a domain for an ultra-competitive niche in credit cards. Will be unleashing 10k/comments a day on it, along with other stuff ^^

[Abs* 12]

Actually you can use the "choose by attribute" as a qualifier.

Just delete all the place holders then click on the "then" node.

Then go into the browser window and use the choose by attribute

feature as normal. Ubot will then populate the area just above

your then statement with your "choose by attribute" node.

 

I've used this method several times in a bot I am currently

creating.

 

James

 

Yup - no follow links still work a treat guys - You should check out my bot that already does this - autocommentbomb.com

 

someone just posted a review last week how they got their site to page one with the autocommentbomb

 

Regarding the choose by attribute placed inside of the IF command then yes this deff works and this is what I was using previously - However one of the Ubot updates made this useless - It was important to scrape page in the IF node -

 

Its all back to usual now and works a treat with the choose by attribute in the If node however I would suggest you stick to scrape page as this would be the most precise way to do it - and as Ubot brushes up on the coding then dont be surprised if all your bots coded in this manner will need altering

 

Just my 2cents to work with scrape page as it will eliminate any issues in the future

 

thanks

[Biks 9]

OK, the next question is - does it make a difference if your comment link is on a related post by topic? I suspect that it's all a myth, just blast away. How would Google qualify if a blog entry is about WEIGHT LOSS or ACNE if the author is writing about multiple things in one post? I suspect the only thing that matters is the PR or trustworthiness of the site. Still - ANY link must be worth something, right? This is why blasting thousands of random links with comments still works?

[meter 145]

Biks, while I agree with you that comment links on related entries SHOULD be worth more than comment links from random websites, I get the exact same results spamming targetted blogs and random blogs. For example, a lot of the random blogs I hit are non-English (chinese, filipino), and yet google still appearts to be giving me decent linkjuice from them (I'm getting to Page 1 after all).

 

So it's really a mystery to me how it works, but the good news is any link will do

[Jeff 0]

Actually you can use the "choose by attribute" as a qualifier.

Just delete all the place holders then click on the "then" node.

Then go into the browser window and use the choose by attribute

feature as normal. Ubot will then populate the area just above

your then statement with your "choose by attribute" node.

 

I've used this method several times in a bot I am currently

creating.

 

James

 

James -

 

THIS is what I was looking for the whole time. Thanks for the tip, this is the way to get this thing to work without it getting halted by an error every time.

 

There are a lot of issues so far that I've been able to overcome and when I'm done, I'm going to share them right here.

 

When given a list of wordpress blogs, my bot can step through them one at a time and comment all day long, mixing it up in all sorts of ways. The tricky part is getting a list of blogs who's platforms are consistent with one another (e.g. all wordpress, or all blogger etc. etc.).

 

Also, another tip, I like to comment spam with large paragraphs of text, as I feel like it makes the author more prone to accept the nice large chunk of new content for their page

 

Everyone else, thanks for the comments, I'll be back in with updates.

 

Jeff

[meter 145]

Protip: Scrape already approved comments, and resubmit them. Surprising how well it works

[Jeff 0]

Protip: Scrape already approved comments, and resubmit them. Surprising how well it works

 

I actually already thought of trying to do that, but I couldn't figure out how to scrape one of the other comments and the name of the commenter!!

 

Do you know how to do this, say on a wordpress blog?! If so I would greatly appreciate a point in the right direction...

 

Thanks man