Forum Moderators: DixonJones

Message Too Old, No Replies

Load of Newbie questions

Please indulge me

         

Powdork

5:54 am on Sep 21, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Hi
I'm doing some work for a company that has some interesting tidbits on their site. First, they use LiveStats and either that company is way behind the times or they possibly don't exist and thats why their hosting company provides it for free. They list Google as a domain but not as a search engine (Yahoo, Looksmart are the only SE's listed). Anyway, when I first logged in the stat program told me that there were currently 45 visitors online which seemed way high. Then I looked further and most of these were AOL users all coming out of Virginia. Further review showed that 60% of the sites users were coming from AOL out of Virginia. I've read that this is not unusual but is there a program which avoids this? Also, the following js code is on the site.

<Script> if(navigator.appName.indexOf('Netscape')==-1) {document.write('<DIV ID=Pere_logo STYLE=position:absolute;top:1px;right:1px;visibility:show;z-index:9>'); document.write('<A HREF=http://www.xxxxx.com/linkpage.asp?beno=26847 TARGET=_new ><IMG SRC=http://www.xxxxx.com/images/directorylink_corner.gif ALT=www.xxx.com BORDER=0></A>'); document.write('</DIV>'); } else{document.write('<LAYER NAME=xxx_logo top=1 visibility=hidden ALIGN=right>'); document.write('<A HREF=http://www.xxxxxx.com/linkpage.asp?beno=26847 TARGET=_new><IMG SRC=http://www.xxxxxx.com/images/directorylink_corner.gif ALT=www.xxxxxx.com BORDER=0></A>'); document.write('</LAYER>'); } function setVariables() {imgwidth=62; imgheight=62; if (navigator.appName == 'Netscape'){ horz='.left'; vert='.bottom'; docStyle='document.';styleDoc=''; innerW='window.innerWidth'; innerH='window.innerHeight'; offsetX='window.pageXOffset';offsetY='window.pageYOffset'; } else { horz='.pixelLeft'; vert='.pixelBottom';docStyle=''; styleDoc='.style'; innerW='document.body.clientWidth';innerH='document.body.clientHeight'; offsetX='document.body.scrollLeft'; offsetY='document.body.scrollBottom'; } } function checkLocation(){ objectXY='branding'; var availableX=eval(innerW); var availableY=eval(innerH); var currentX=eval(offsetX); var currentY=eval(offsetY); x=availableX-(imgwidth+8)+currentX; y=availableY-(imgheight+3)+currentY;evalMove();setTimeout('checkLocation()',10); } function evalMove() { eval(docStyle + objectXY + styleDoc + horz + '=' + x);eval(docStyle + objectXY + styleDoc + vert + '=' + y); } </Script>

Anyone see this before? Even know its java script the hidden links scare me re Google.
You can sticky me if you need the xxx's filled in.

tedster

8:29 pm on Sep 21, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



You're right to worry about Googlebot. It will not see links added with documnet.write()

The script is out of date, and it assumes that there is no version of Netscape except v.4. It will blow up your page on Netscape 6+ by writing in <layer> tags, from what I see.

The AOL numbers are crazy - my guess is that the stats packages is confounded by AOL's dynamically assigned IP addresses. Most log analysis program have this problem, by the way - you really need cookies to track AOL sessions well.

The deal here is that just one human user will show in your logs as 6 or more different IP addresses, even for a single page (each image GET will have a unique IP, for example).

Slade

9:40 pm on Sep 21, 2002 (gmt 0)

10+ Year Member



tedster is right about the AOL users, I just wanted to add one more tidbit.

What you're seeing as AOL users in VA is probably the whole east coast(maybe even more), as all AOL users touch the internet through a great big caching proxy server. The IPs you're seeing are probably from a datacenter in VA.

Powdork

9:58 pm on Sep 21, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



You're right to worry about Googlebot. It will not see links added with documnet.write()
\
Are you saying this can get them in trouble with Google? Neither humans nor Gbot can see the link.

The script is out of date, and it assumes that there is no version of Netscape except v.4. It will blow up your page on Netscape 6+ by writing in <layer> tags, from what I see.

My concern is that the url is not the same as the domain in which it resides.

The deal here is that just one human user will show in your logs as 6 or more different IP addresses, even for a single page (each image GET will have a unique IP, for example).

Thats good. I'd hate to think they had gotten that popular.

tedster

10:29 pm on Sep 21, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Are you saying this can get them in trouble with Google? Neither humans nor Gbot can see the link.

No, there's no trouble - it's just that the link might as well not be there as far as Google is concerned, and their algo untilizes linking to a significant degree.

Neither humans nor Gbot can see the link.

There's nothing in the code I can see that should hide the link from humans - at least humans with javascript turned on in their browser. The code is superimposing a linked image on the page, and doing it differently depending on the browser.

My concern is that the url is not the same as the domain in which it resides.

I can't really address that - the site can link to whatever domains they choose. Do you think this was a stealth edit added to the page by a third party? It's not out of the question.

Powdork

1:10 am on Sep 22, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



There's nothing in the code I can see that should hide the link from humans - at least humans with javascript turned on in their browser. The code is superimposing a linked image on the page, and doing it differently depending on the browser.

Oops, I guess I was looking at it in Dreamweaver, where only the tag shows up and then went and saw 1px and hidden and kind of panicked.

I can't really address that - the site can link to whatever domains they choose. Do you think this was a stealth edit added to the page by a third party? It's not out of the question.

Noone at the company claims to know about it and several people were able to edit the site. I followed the link and it looks as though it may be some kind of pay for referral service which leads me to believe someone is trying to make some money on the side. I checked on webdev.archive.org and the modifications must have been made this year (why don't they have 2002 results, or am I in the wrong spot?) Well, it looks like its nothing to worry about anyway. Thanks