|
Persona
|
|
Persona Testers- These websites give you real-time feedback about your
persona. The first link in each row takes you directly to the persona
testing web page. The second link (via google) takes you to a google
search, which should then lead you to the persona testing page.
Consider the Following Diagram:
A researcher is online and surfing the Internet. This computer has
a persona of:
- The employee's LAN (branch.agency.gov or division.company.com)
- dsl.town.state.isp.com (note that most ISP's name your persona to include a geographic indication)
The researcher is looking at a webpage (URL1) and clicks on a link which
leads to another webpage (URL2). With that simple click, this is some
of the information that is now available to the webmaster of
URL2:
- Remote Host: This is the Persona of your machine or the gateway your
requests pass through. The web server MUST have this information in
order to send the requested web page to you.
- http Referrer: This is the address of the web page you
were previously viewing (URL1) My
Check your person now
page talks about why this can cause a problem for you.
Another concern...
A researcher is online and surfing the Internet. The researcher enters some
"search terms" into a search_tool. The researcher then visits the sites
listed in the search tool's "hits". Look at the following diagram
to see what has just happened:
Thick Red Lines: The Webmaster at searchtool.com knows your “search
terms” and persona. For any search tool that you use, what do you know
about the organization (and webmaster) who runs that specific search tool?
Double Blue Lines: There is now a very good chance that the webmaster
of target.com also knows what search terms you have used to reach
them. How is this possible? Searchtool.com displays your "top ten hits"
on a search results page which may have a URL such as:
- http://searchtool.com/keywords=searchterms
By looking at the referring URL, the webmaster of target.com can now know
exactly which search terms you used to discover his site. In fact,
the URL of the search results page often contains all the additional parameters
you used to construct your search query. A webmaster can view the
exact same search results page as you, to see what other pages were suggested
to you, based on your search query.
Here are some example referral URL's taken from my own website statistics.
Click on them to see the search results that visitors used to find my site:
http://www.google.com/search?hl=en&q=isp+backbone+maps
,
http://search.yahoo.com/bin/search?p=russ+haynal
,
Some referral URL's are simply links from public web pages:
http://dir.yahoo.com/Computers_and_Internet/Internet/Maps/Network_Topology/
http://en.wikipedia.org/wiki/Internet_backbone
You must also be careful on how you name the web pages in your Intranet.
Here are examples of intranet pages that link towards my site:
- http://doc.uunet.ca:2001/tutorial/ispinfo.html
(you can tell why they link to me)
- http://www.oen.siemens.de/projects2/p203/pub/links.html
(notice the "non-revealing" URL)
-
http://insidefs.mcln.federal.unisys.com/misc/links/main.htm
.
A Manual way to suppress http_referrer...
Is there a way to avoid passing the referrer information along when selecting
a web page link? YES - Here are some relatively convenient
methods:
- Right-click on the link and select "copy link location"
This will place the Link's URL into the clipboard. Now you can paste the URL
into the browser's location area, and hit the Enter key on the keyboard.
- Right-click on the link and select "add bookmark" Now you
can select the link from the bookmark listing.
- I've also noticed that a referring URL does not seem to be passed along
when the referring web page is based on your computer. You may want to take
the web page, and "save as" to your hard disk. In the process, you
can also rename the page to something generic like: "page.html" This
should succeed in hiding the Referring URL, or at least give it a less obvious
address like: file:///C|/temp/page.htm
- Important Note: These three previous tips work ONLY if the URL is "direct" to the
web site. Be on the lookout for URLs that are "forwarding URLs".
In other words, the hyperlink takes you back to the search tool, and then
forwards you to the real destination.
To Automatically suppress http_referrer...
- Your local firewall may offer this option. Norton security Suite is
supposed to offer this feature, as does Zone Alarm Pro (from
Zonelabs) Directions for Zone
Alarm Pro: Open up Zone Alarm --> Click on "privacy" in the left column --> Click
on the "Main" tab along the top --> In the "cookies" section click on
"custom" --> in the "3rd Party cookies" section check box "Remove Private
Header Information"
- In Firefox, you can alter your browser using the following
steps:
- in the browser's address bar, type: about:config
- Scroll down to the line called: network.http.sendRefererHeader
- Right-click on the line and select "modify"
- Change the "2" to a "0" (zero) and then click OK.
A note about Google's Cached search links....
Some researchers incorrectly assume that clicking on Google's "cached"
in a search result will only make hits to Google, and make no hits to the search
target's website. Truth is, clicking on "cached" will almost always result in
unusual hits to the the target web server. Consider the following sequence of
events.
- You click on Google's cached link for a search result
- The Text of that web page downloads from Google's cache to your
computer.
- Your computer displays the text of that cached web page in your
browser.
- The text web page (now in your browser) will now begin to download any
embedded multimedia on that page (embedded graphics, animations, flash
plug-ins, sound, video etc)
- Such embedded multimedia, is downloaded directly from the target
website to your browser. Google only caches the text of the page, any
graphics, sounds, etc. are downloaded by you from the target
website.
- The Target webserver now has hits from your persona asking for just the
multi-media of a web page, but has no hits from you for the text of
the page itself. The webmaster now knows that you are viewing the text
of the web page from some other resource (such as.Google's cache or a
language translation site)
- If your browser leaks http_referrer, then the downloaded graphics will
leak a referrer along the lines of::
google_cache_/_URL_of web_page_from_cache_/_Your_search_terms_used_to_find_page
Google does offer a "Cached text only" version of the web page, but normally
you can't get to the text-only cached page until you first view the regular
version of google's cached page (which by then will have downloaded the embedded
multimedia from the target site). Here is the work around:
- In the Google search result, Right-click on the cached link.
- Select "copy shortcut" or "copy link location" depending on your
type of browser. (This copied hyperlink is for the regular cached copy of
the web page which will include multimedia)
- click into the web browser's address box and paste-in the cached URL you
just copied
- Now You need to EDIT the URL by adding the following text onto
the end of the URL: &strip=1
- Now hit enter on the Keyboard, and that should take you directly to the
"text only" version of Google's cached copy of the web page. This
means that the target website will not see any hits from your
research. However Google will still know that you are viewing the
cached copy of the web page.
This cutting/pasting/&strip=1 technique will only only work IF google is not
"hijacking" its own search results. Google has this annoying tendency of
running a javascript on their search page as you click on a search result or
cached link in the search results. This javascript re-designs the
hyperlinks to lead back to Google so google can log what search links you are
clicking on. If the &strip=1 technique is not working, you need to disable
javascript in your browser. In a work environment ask tech support for
help.
A note about Google's "Instant
Preview" (and where is "cached")
You may have noticed that google search results no longer display a "cached"
option directly in Google's search results. This change happened as
Google implemented a new feature called "instant preview". In your google
search results, hover your mouse over a search hit that you are interested in.
While hovering over the hit... A pair of >> arrows should appear just to
the right of the hit. Now hover your mouse over the >>, and a small screen
shot of the web page should appear. I can confirm that this "instant
preview" is causing live hits on the target website from Google, as Google goes
out in real time to grab the web page text and all the graphics from the
target website. Repeat.. Displaying the "instant preview" of a search
hit causes live hits to the target web server from Google. The Target web
master will know that his web page is bring viewed in a google search.
Google also decided that the "cached" link should now be displayed in the
"instant preview" window pane. There are many researchers who are angry
with google about this as shown on
this thread.
So right now you can no longer get to google's cached link to
copy/paste/&strip=1 without invoking the "instant preview" pane (which will
cause live hits to the target web server) Options include:
- Hack your browser to display the cached link in Google's search results
(do NOT do this at work) (
http://userscripts.org/scripts/show/114455 )
- Enhance your browser a plug-in designed for text only caching (do NOT
do this at work)
https://addons.mozilla.org/en-US/firefox/addon/cachemachine/
- Wish and hope Google listens to hundreds of angry users
- Try to find a suitable alternative to Google for text-only cached .
Gigablast does offer this using the
identical /copy/paste/&strip=1 technique described above for Google, but
Gigablast is a fraction of the size of Google, so this is not much of an
alternative
s