Click \
These are the docs for XYZZY \
XYZZY is a magic word
\ This is a magic attribute
\
xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ ^^^ The whitespace causes an error Instead you have to do the following: \\xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ This problem only seems to crop up in the ADDRESS element; does anyone know why? Here's the relevant portion of the HTML 2.0 DTD: \ \ (The P element end-tag is minimizable.) --%!PS-Adobe /M{rmoveto}def/R{rlineto}def 10 10 scale 2/Times-Roman findfont 2 scalefont setfont 12 45 moveto 0 5 R 4 -1 M 5.5 0 R/C{currentpoint}def C 3 sub 3 90 0 arcn 0 -6 R 7.54 10.28 M 2.7067 -9.28 R -5.6333 0 R 9.8867 8 M 7 0 R%Shalom setlinewidth 0 -9 R -6 4 M 0 -4 R -31 -5 M C stroke moveto% Austin, Texas (Henry Churchyard http://uts.cc.utexas.edu/~churchh) show showpagexxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ > ^^^ > The whitespace causes an error I just tested this with the currrent DTD, and the space between paragraphs did not cause an error. (The second \ start-tag did, but I assume that was a typo.) >Instead you have to do the following: > >\\xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ > >This problem only seems to crop up in the ADDRESS element; does anyone >know why? > >Here's the relevant portion of the HTML 2.0 DTD: > \ > \ Looks fine to me. It may be that the HAL validator is using an out-of-date (*very* out-of-date) copy of the HTML 2 DTD; earlier versions may have had a problem with pernicious mixed content [*]. The HTML *3* DTD is known to have plenty of problems of this nature, especially in "Strict" mode. In fact, I checked with the HAL validator and did not encounter this error at the "level 2", but it did show up under the "level 3" setting. Maybe you were using the HTML 3 DTD? (Which, BTW, I do not recommend, for this and other reasons.) --Joe English jenglish@crl.com [*] Basically this means that elements are declared with mixed content but without allowing #PCDATA everywhere; in such elements, whitespace is (unintuitively) interpreted as data even where data is not legal. This is a well-known SGML pitfall.xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ : ^^^ : The whitespace causes an error : Instead you have to do the following: : \\xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ : This problem only seems to crop up in the ADDRESS element; does anyone : know why? : Here's the relevant portion of the HTML 2.0 DTD: : \ : \ : (The P element end-tag is minimizable.) My first thought is that the validator used the March HTML 3.0 DTD which I changed in the DTD I usually use: //3. Replaced ADDRESS mixed content model with the one from HTML v1.29 I just ran nsgml.exe on two files This one worked O.K. since it used the standard version 1.29 HTML.DTD \ \xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ However this one didn't because the DOCTYPE tells it to use the March HTML 3.0 DTD. \ \xxxxxxxxxxxx\
\xxxxxxxxxxxxxx\
\ as this snapshot shows. D:\\SP>sp address3.htm nsgmls -s -m d:\\sp\\catalog address3.htm address3.htm:1:16:E: delimiter `"' invalid: only delimiter `>', delimiter `[', PUBLIC', `SYSTEM' and parameter separators are allowed address3.htm:1:16:E: cannot continue because of previous errors I'm glad this topic will get more discussion and the above is helpful to someone. However I really just want to say THANKS for your software. I rely on htmlchek.pl to style check all my files and dehtml.pl so I can spell-check and catch _missing_ whitespace. I also use nsgmls.exe for more control, but htmlchek.pl cleans them up. htmlchek.pl is reliable in my experience. Thank you, Thank you, Thank you. frankText ... is preferred to \