Re: DOCBOOK-APPS: representing and rendering symbol characters
Bob Stayton <[EMAIL PROTECTED]> writes: > The entity renders properly in IE 5.5, but appears as a "?" in > Netscape 4.75 on Windows and Netscape 4.76 on Linux. > I didn't try a Mac. > Jirka, are you sure you saw these entities rendered properly > in "mozilla for Windows" (was that Netscape 4.7)? No, by Mozilla he probably meant Netscape 6 or Mozilla proper. In Netscape 4.x all bets are off, and web designers should not target that browser due to poor standards compliance, IMHO. -- .Adam Di [EMAIL PROTECTED]http://www.onShore.com/>
Re: DOCBOOK-APPS: representing and rendering symbol characters
/ Bob Stayton <[EMAIL PROTECTED]> was heard to say: | The entity renders properly in IE 5.5, but appears as a "?" in | Netscape 4.75 on Windows and Netscape 4.76 on Linux. | I didn't try a Mac. It doesn't surprise me that these don't render properly on all platforms with all browsers. For one thing, the browser needs access to a font that includes the symbol. | Jirka, are you sure you saw these entities rendered properly | in "mozilla for Windows" (was that Netscape 4.7)? Even Mozilla on Linux gets it right. They must have added a Unicode font. | Were you perhaps thinking | that Netscape supports the utf-8 character set? If | you render your output as utf-8 and add this: | | This whole issue is so murky... UTF-8 is a scheme for encoding a sequence of characters into a sequence of octets. UTF-8 has nothing to do with whether or not ∞ is rendered. In UTF-8, the character at Unicode position 8734 could be represented as a sequence of three (or possibly four, I don't recall the precise algorithm) octets. But after parsing the document, that sequence and the sequence ∞ have the same meaning: they represent Unicode character 8734 which is the infinity symbol. | The HTML 4.0 spec has been around since 1997, so why haven't | these standard character entities been implemented in Netscape? HTML 4.0 explicitly calls out[1] the fact that not all user agents may be able to display all characters. | So I think Erik's original question still stands: how do you render | the symbol font characters universally on all HTML browsers? You can't. | It looks like image files are the only universal form. They aren't. Lynx and wn can't display graphics. Be seeing you, norm [1] http://www.w3.org/TR/1998/REC-html40-19980424/charset.html#h-5.4 -- Norman Walsh <[EMAIL PROTECTED]> | Why shouldn't things be largely http://www.oasis-open.org/docbook/ | absurd, futile, and transitory? Chair, DocBook Technical Committee | They are so, and we are so, and | they and we go very well | together.--Santayana
Re: DOCBOOK-APPS: representing and rendering symbol characters
Erik Hennum wrote: > However, Netscape 4.7 on Windows doesn't render Α > > In fact, in my test document, Netscape 4.7 seems to render > less than half of the generated Unicode characters. Internet > Explorer 5.5 seems only a little better. For instance, neither > seems to render any of the Unicode characters generated > from iso-amsn.ent. I this case, the easiest way would be use entities in XML, and at the start of XML file redefine it replace with tags to insert corresponding pictures. Once there will be better support for Unicode in browsers, you will just switch off your declarations. Do something like this: %myentities; ]> ... Use common entities like &Agr; ... myent.ent File: "> - Jirka Kosek e-mail: [EMAIL PROTECTED] http://www.kosek.cz
Re: DOCBOOK-APPS: representing and rendering symbol characters
> From: Bob Stayton <[EMAIL PROTECTED]> > > > Erik Hennum wrote: > > > > > Most of the character set from the symbol font could be mapped > > > to ISO character entities. However, rendering the entity seems > > > problematic. > > > > > > The HTML XSLT scripts pass through the ISO entities in the > > > generated HTML. The mainstream browsers seem to have > > > only limited support for the ISO entities. > > > > [EMAIL PROTECTED] wrote: > > > > Which XSLT processor did you use? In generated HTML there should not be > > entity references (except for chars from ISO Latin 1), there should be > > only character references in form &#code; where code is Unicode/ISO10646 > > character code. If your browser supports Unicode (this for example true > > for IE and Mozilla on Windows, probably also other browsers/platforms) > > there should not be problem with displaying these character glyphs. > > And connecting this to Docbook, the conversion of characters > from Framemaker to Docbook XML needs to map them to the ISO entities > listed in the Docbook iso-*.ent files included with > the Docbook DTD. For example, in iso-tech.ent you see: > > > > The FrameMaker infinity character > should be converted to "∞" in your Docbook XML > files so you have a nice mnemonic entity. > The ent declaration identifies it properly by > its Unicode character in XML, and Norm's stylesheets > should render that Unicode character in whatever output format > you choose, including HTML. I decided to double-check this myself, and found that your browser mileage may vary. The DTD makes ∞ into &x#221E; and Norm's stylesheet for HTML renders that as ∞ which is a valid HTML 4.0 character entity. This is all completely correct processing. But when the load the resulting HTML into various browsers, you get various results. The entity renders properly in IE 5.5, but appears as a "?" in Netscape 4.75 on Windows and Netscape 4.76 on Linux. I didn't try a Mac. Jirka, are you sure you saw these entities rendered properly in "mozilla for Windows" (was that Netscape 4.7)? Were you perhaps thinking that Netscape supports the utf-8 character set? If you render your output as utf-8 and add this: then it probably will render. But the HTML 4.0 entity ∞ does not. And HTML using utf-8 encoding will not work on other older browsers. The HTML 4.0 spec has been around since 1997, so why haven't these standard character entities been implemented in Netscape? So I think Erik's original question still stands: how do you render the symbol font characters universally on all HTML browsers? It looks like image files are the only universal form. bobs Bob Stayton 400 Encinal Street Publications Architect Santa Cruz, CA 95060 Technical Publications voice: (831) 427-7796 The Santa Cruz Operation, Inc. fax: (831) 429-1887 email: [EMAIL PROTECTED]
Re: DOCBOOK-APPS: representing and rendering symbol characters
> Erik Hennum wrote: > > > Most of the character set from the symbol font could be mapped > > to ISO character entities. However, rendering the entity seems > > problematic. > > > > The HTML XSLT scripts pass through the ISO entities in the > > generated HTML. The mainstream browsers seem to have > > only limited support for the ISO entities. > > Which XSLT processor did you use? In generated HTML there should not be > entity references (except for chars from ISO Latin 1), there should be > only character references in form &#code; where code is Unicode/ISO10646 > character code. If your browser supports Unicode (this for example true > for IE and Mozilla on Windows, probably also other browsers/platforms) > there should not be problem with displaying these character glyphs. And connecting this to Docbook, the conversion of characters from Framemaker to Docbook XML needs to map them to the ISO entities listed in the Docbook iso-*.ent files included with the Docbook DTD. For example, in iso-tech.ent you see: The FrameMaker infinity character should be converted to "∞" in your Docbook XML files so you have a nice mnemonic entity. The ent declaration identifies it properly by its Unicode character in XML, and Norm's stylesheets should render that Unicode character in whatever output format you choose, including HTML. bobs Bob Stayton 400 Encinal Street Publications Architect Santa Cruz, CA 95060 Technical Publications voice: (831) 427-7796 The Santa Cruz Operation, Inc. fax: (831) 429-1887 email: [EMAIL PROTECTED]
Re: DOCBOOK-APPS: representing and rendering symbol characters
Erik Hennum wrote: > Most of the character set from the symbol font could be mapped > to ISO character entities. However, rendering the entity seems > problematic. > > The HTML XSLT scripts pass through the ISO entities in the > generated HTML. The mainstream browsers seem to have > only limited support for the ISO entities. Which XSLT processor did you use? In generated HTML there should not be entity references (except for chars from ISO Latin 1), there should be only character references in form &#code; where code is Unicode/ISO10646 character code. If your browser supports Unicode (this for example true for IE and Mozilla on Windows, probably also other browsers/platforms) there should not be problem with displaying these character glyphs. - Jirka Kosek e-mail: [EMAIL PROTECTED] http://www.kosek.cz
