Mailinglist Archive: opensuse (3964 mails)

< Previous Next >
Re: [SLE] html symbols question ..
  • From: Randall R Schulz <rschulz@xxxxxxxxx>
  • Date: Fri, 12 Nov 2004 14:10:01 -0800
  • Message-id: <200411121410.01262.rschulz@xxxxxxxxx>
Jalal,

On Friday 12 November 2004 13:41, jalal wrote:
> Randall R Schulz wrote:
> > Jim,
> >
> > On Friday 12 November 2004 09:00, James Hatridge wrote:
> >> Hi all...
> >>
> >> A few weeks ago someone wrote that to stop the bots from getting
> >> your email address off your web site you should use symbols
> >> instead of letters ie "&#252;" means the German ΓΌ.
> >
> > I'm not sure how helpful that technique really is, since HTML
> > character entities (the proper term for these things) are designed
> > for machine processing.
>
> It seems to work reasonably well. I had a couple of email addresses
> for testing and the one in clear text got a hell of a lot more spam
> than the encoded one.
> However... the smarter bots will look for encoded Mailto:'s. So it
> helps to mix encoded and none encoded text together, just to make it
> difficult. e.g.:
> &#109;ailto&#058;someone&#064;somewhere&#046;co&#109;

But if you really want to shut out the bots, create an image file (GIF,
PNG, JPEG, etc.) that displays the email address. If you want to be
even more sure, include some graphical obfuscation that a person can
easily disregard but which will confuse OCR software. I just signed up
for on-line payment of my telephone bill, and one step of the process
requires the user to read a series of alphanumeric characters in an
image and enter them in text box. The original image from which the
human user must transcribe those letters and numbers has the appearance
of yellow graph paper. I take it that is something that OCR software
has great difficulty dealing with.


> jalal


Randall Schulz

< Previous Next >
Follow Ups