On Fri, May 30, 2008 at 2:04 PM, Alan <[EMAIL PROTECTED]> wrote:
> It looks like they probably hacked a style sheet or put in a chunk of php
> code to alter common words and add links to the porn sites.

Knowing the subject matter of the article and looking at the placement
of the links within the grammatical constructs used, I think it's
deliberate substitution of URLs in hyperlinks created by the article's
author rather than substitution for common words. The link titles and
placement in juxtaposition to important points suggests to me that
it's not an automated linking of common words. I'd expect to see a
more random placement of links were it an automagic substitution of
common words with hyperlinks.

E.g., there are many phrases in the article that are clearly intended
by the author to be hyperlinked, such as the link from the word "here"
in "[t]here is a compilation being tracked in here[.]"

I can conceive of a malwareware tool that automagically substitutes
URLs from a common list of porn URLs once a web apps' editing
permissions are penetrated. Of course the hyperlinks could be manually
edited by a Third World worker earning a pittance. But a fairly
powerful plain text editor's script for automating such substitutions
after cutting and pasting the page content from an HTML editing form
to an editing application would seem to be a fairly trivial hack to
keep expenses even lower. After processing, just cut and paste the
altered content back to the HTML form. Then on to the next page to be
defaced, without even bothering to save the altered text in the text
editor.

Non-obligatory and only marginally relevant thread drift warning:

Lord knows I've written enough similar scripts in my time to batch
process a LAN's archive of WordPerfect document templates, document
assembly document "parts," and documents for common misspellings,
markup errors, incorrect or updated citations, etc. I will never
forget an incident in the WordPerfect 5.1 days when I learned that
about half of 20 paralegals working on summarizing documents produced
in discovery had not followed instructions and had been cutting,
pasting, and revising the form from completed summaries rather than
working from the template for the form. And of course errors got
dragged into the process, making the resulting database far less
useful for search purposes because of replicated errors like the
misnaming of data fields.

My "BigClean" script stepped through over 60,000 files to fix 54
common replicated errors we identified in the document summaries
resulting from the cute and paste defects . On a state of the art DEC
DX 2-50, the script ran for nearly 40 hours but did the job without a
hitch once debugged. Working around the WordPerfect 5.1 reluctance to
process the next file without rereading all remaining filenames in the
directory into memory was the major headache in the script
development, although I can't remember what my hack was to work around
that issue.

I later adapted the script to correct common misspellings of people's
names that appeared in the documents and a few other similar variables
that got in the way of the Isys 3.1x search engine's usefulness. Our
experience was one of the major reasons the Isys development team
added synonym ring capabilities to Isys so users could easily
compensate for common misspellings and the like without having to
correct the indexed documents and to update the search index files.

Isys is a tremendous tool that I really wish would run on Linux/Wine.
E.g., I've got virtually all data on my system that my colleagues and
I wrote or acquired since 1984. Isys is way out ahead of similar FOSS
tools that run on Linux such as Apache Lucene and its derivatives. In
the current version, it indexes over 200 file formats and data types
and that's just for starters on its advantages.  I regard it as the
single largest advance in word processing since the storage of
editable documents on magnetic media.

If anyone is interested, I wrote a fairly thorough review of Isys back
in 2000. <http://www.llrx.com/features/isys2.htm>. The limitations on
the number of documents that can be indexed has since been removed,
with the number now limited only by the data storage capacity of the
local system.

Best regards,

Paul
_______________________________________________
EUGLUG mailing list
[email protected]
http://www.euglug.org/mailman/listinfo/euglug

Reply via email to