[sword-devel] Devanagari text displays different in SWORD than in the source IMP file

[David Haslam]

Just read this,

"At www.valmikiramayan.net the Sanskrit-English edition of the Valmiki
Ramayana by Sri Desiraju Hanumanta Rao et al. is offered as html files.
These html files were created using the old fonts "Sanskrit 98" and
"Translit 98" of the former "Itranslator 98". While the Ramayana html files
were readable with old versions of Microsoft's Internet Explorer, they are
not readable with the latest versions of Internet Explorer due to
Microsoft's faulty new IE font handler expanding zero-width characters to
spaces: "

Found in this page,
http://www.sanskritweb.net/sansdocs/index.html#RAMA
http://www.sanskritweb.net/sansdocs/index.html#RAMA 

IE fails some of the web standards, unlike Firefox.

-- David


David Haslam wrote:
> 
> Or rather the front end rendering engine...
> 
> Tim wrote me as follows,
> 
> Take a real close look at the first word on Mark 1:1 and you’ll see that
> the text is rendered differently on BPBible than on both Bible Desktop and
> Alkitab Bible.  The correct rendering is from BPBible.   When you see the
> hylanta below the word, that is a good indicator that the word is being
> rendered incorrectly – though not in all cases.  I didn’t catch this when
> I first did the module creation.  I took the demo over to the Nepal Bible
> Society and they immediately noticed that words with half letters were not
> displaying correctly.  
> 
> According to Peter von Kaehne, developers of these Bible front ends have a
> choice over what rendering engine they are going to use.  Apparently some
> are better at rendering complex scripts than others.
> 
> -- David
> 
> David Haslam wrote:
>> 
>> Hi Tim,
>> 
>> By pasting your reply into SC Unipad, I have found the actual difference.
>> 
>> After the Virama, U+094D
>> 
>> The correct one has     U+200D Zero-width joiner
>> The Sword one has      U+200C Zero-width non-joiner
>> 
>> My guess is that this is a bug deep in the Sword engine. 
>> 
>> We need Troy's expert eyes to find the root cause.
>> 
>> -- David Haslam
>> 
>> 
>> Tim Chase-13 wrote:
>>> 
>>> Hi David
>>> 
>>> I have tried several different Unicode fonts and still the display of
>>> some words are incorrect.  For instance you might want to try to display
>>> the first word of Mark 1:1 of the text that I sent you.  The correct
>>> rendering of this word is परमेश्‍वरका  but the way SWORD renders the
>>> text, it looks like परमेश्‌वरका .  which is an incorrect rendering of
>>> the 4th character in the word which should be a half character.  You'll
>>> notice the hylanta below that character position. That is a half
>>> character marker but, it should be invisable on this word.  Any words
>>> that have half characters have the hylanta showing in SPW.  Try BPBible
>>> and the words will render correctly.  
>>> 
>>> I know the differences are slight but they do make a difference.
>>> 
>>> Tim
>>> 
>>> 
>> 
>> 
> 
> 

-- 
View this message in context: http://www.nabble.com/Devanagari-text-displays-different-in-SWORD-than-in-the-source-IMP-file-tp25137030p25327851.html
Sent from the SWORD Dev mailing list archive at Nabble.com.