Troglodyte text-processing techies
If, several years ago, I had to divide techies into two camps, I would do it along these lines: binary and text. The binary camp would be exemplified by Microsoft and Apple; the text camp would be Unix.
For instance, aside from AppleScript, VBScript, and batch files, most programming on Apple/MS was done using compiled binaries; those binaries stored things like preferences in binary files.
On the other hand, Unix programmers generally prefer to use human-readable shell scripts when possible.
That "text culture" spills over into things like the design of the Internet's components: email, USENET, the Web, etc.
This "culture" leads to all manner of problems: big human readable text files must be parsed, and mistakes are usually discovered at run-time, rather than at compile-time. To a computer, a text file is just a subtype of a binary file. The difference is that the specification of a text file is almost always less precise than that of a binary file. And, text files can use different character sets, languages, and the like; these problems are not found in binary files.
This "culture" also lead to the use of HTML to do really weird things. New keywords had to be continually added to HTML, and despite all the Rube Goldbergian features added to it and XUL and XML and XWHATEVERTHEFUCK, nothing written in HTML today would look as good or have as much functionality as a well-written 80s Mac program.
The defense of text is sometimes stated as a battle between closed Apple/MS and "open systems." Supposedly, binary files are more amenable to "embrace and extend." That argument is completely bogus; text specifications have been "embraced and extended" as much as binary specifictions. Any argument that points out the advantages of something being human-readable is countered by the fact that viewers and compilers can be written that will convert any binary file into a human-understandable format. The binary side of things can be used for computers, they're good with that. The human-readable side is readily available for humans.
This article exemplifies the problems with this "script kiddie" approach to how computers should process information.
As you're reading that article, imagine how much simpler things would be if the people who had invented the Web had been familiar with how GUI-oriented computers are programmed, instead of being script kiddies with the belief that their way is best.