[Gllug] Interesting article on the MS monopoly

Gordon Joly gordon.joly at pobox.com
Fri Apr 30 13:33:58 UTC 2004


At 9:46 +0100 2004/04/30, Tethys wrote:
>Nix writes:
>
>>Like in er, C?
>>
>>/* Try using a visible slash star in here. */
>>
>>/* Plus there are other hidden *tricks*??/
>>/ Is this in the comment? /* No, it wasn't. */
>
>No, not like in C. BTW, using trigraphs to make obfuscated code is
>considered cheating -- it's too easy :-)
>
>But no, in C, the only invalid characters are the end of comment markers.
>The same isn't true in XML. For example, the following is (to me) a fairly
>natural way (and immediately obvious to the eye when scanned) to do a block
>comment in XML:
>
>	<!-- This is a comment.
>	---- So should this be.
>	---- And this...
>	-->
>
>However, the "--" at the start of the second line is invalid within a
>comment, despite requiring a trailing ">" to make it an end of comment
>marker. This has bitten me elsewhere, for example when commenting out
>text that includes a double dash to simulate an em dash in ASCII -- like
>the one I've just used.
>
>XML could also *really* use a #if 0/#endif equivalent for block commenting
>large volumes of text, regardless of whether they contain other XML comments.
>
>The failings of XML in this regard have evolved into a tradition of not
>using any markers to indicate that a line is part of a block comment. Try
>looking at a tomcat config file[1] and working out which lines are comment
>and which are not (without having to parse the file by eye).
>
>Tet
>
>[1] conf/server.xml, for example... bleurgh!



This is a hang over from SGML where comment delimeters where simply

--

as I recall. It was a sop to SGML and XML doing similar things...

Hmmm. Am I stating something very obvious?

:-)

Gordo.

-- 
"Think Feynman"/////////
http://pobox.com/~gordo/
gordon.joly at pobox.com///
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug




More information about the GLLUG mailing list