WD-entities-961125

Hello!

This is a comment on the Working Draft "Additional named entities for
HTML" from 25-Nov-96 (http://www.w3.org/pub/WWW/TR/WD-entities-961125).

I am working on an HTML to PostScript converter - html2ps. Hopefully,
I will be able release a new version of this program next month. This
version will have support for the new entities that are defined in this
Working Draft.

When I implemented these new entities in my code, I noted a few errors
in the draft:

* A typo on line 319:

<!ENTITY Iacute CDATA "&4amp;#205;" -- capital I, acute accent -->

should be

<!ENTITY Iacute CDATA "Í" -- capital I, acute accent -->

* Line 399:

<!ENTITY fnof     CDATA "À"  -- latin small f with hook, =function, =florin, U0192 ISOtech -->

must surely be:

<!ENTITY fnof     CDATA "ƒ"  -- latin small f with hook, =function, =florin, U0192 ISOtech -->

* There are two different definitions of 'theta' on lines 436 and 455:

<!ENTITY theta    CDATA "θ" -- greek small letter theta,  U03B8 ISOgrk3 -->
<!ENTITY theta    CDATA "ϑ" -- greek small letter theta symbol,  U03D1 NEW -->

perhaps the second should have been called 'thetav'.

* Line 508:

<!ENTITY sum      CDATA "−" -- n-ary sumation, U2211 ISOamsb -->

should be:

<!ENTITY sum      CDATA "∑" -- n-ary sumation, U2211 ISOamsb -->

* Lines 517-518:

<!ENTITY and      CDATA "⊥" -- logical and, =wedge, U2227 ISOtech -->
<!ENTITY or       CDATA "⊦" -- logical or, =vee, U2228 ISOtech -->

have numerical entities that do not match their Unicode numbers.
Since furthermore ⊥ is used for ⊥, I suppose the above
two lines should be:

<!ENTITY and      CDATA "∧" -- logical and, =wedge, U2227 ISOtech -->
<!ENTITY or       CDATA "∨" -- logical or, =vee, U2228 ISOtech -->

* Line 527:

<!ENTITY asymp    CDATA "≅" -- almost equal to, =asymptotic to, U2248 ISOamsr -->

has the same numerical entity as ≅. To match its Unicode number
this should be:

<!ENTITY asymp    CDATA "≈" -- almost equal to, =asymptotic to, U2248 ISOamsr -->

---

Since my program produces postscript, it is of course easy to generate
the Adobe Symbol characters. There is one major problem (that I have
not yet solved), though: how to display the symbol characters within
preformatted text.

It is simple in postscript to change the font metrics to get a fixed
width font out of the Symbol font. But since some characters are quite
wide (for example horizontal arrows), one has to choose a very small
font size to avoid getting those characters overlapping the surrounding
characters. Using this font size other characters, such as Greek
letters, look ridiculously small. Another alternative is to scale
characters individually, but this is a bit messy. Have this problem
been discussed?

I have generated an HTML table with the entity definitions from the
Working Draft, and created a postscript file with html2ps. This might
be useful for you as an illustration of the entity definitions. You
can get the PS file from

  http://www.tdb.uu.se/~jan/demo/entities.ps

(Incidentally, this table illustrates the use of the proposed TABLE
element THEAD, to get a running table header for multipage tables.)

Note that I have not corrected any of the errors mentioned above, and
I may have misunderstood some definitions (e.g. ≅ and ≈).

If you think this may be of some use for you, I will be glad to reformat
the PS document if you prefer. My program is not yet in the shape to be
used by others, though.

Best regards,
Jan Karrman
Dept. of Scientific Computing
Uppsala University
Sweden
jan@tdb.uu.se

Received on Monday, 21 April 1997 06:25:50 UTC