License vs. Signatures

[Abdalla Alothman <abdalla at pheye dot net>]

This message was posted twice yesterday and didn't show up.

Salam,

On Friday 13 October 2006 05:49 am, Meor Ridzuan Meor Yahaya wrote:
>
> Why can't we have both? Is it too difficult? See below.

There is absolutely no reason why we can't have them both. It would be nice if
we can have both. However, the problem is with character coding. Not all the
symbols used in the 'uthmani standard are used in the Unicode character set.
So we can wait, or we can use what's available. Of course another option would
be to convince the Unicode people to add the missing symbols.

The other point we have to consider is when to use each standard. Many people
do not pay attention that in a single text unit, every character has a value
within that unit. So if you have an aaya in a sura in a file, and that aaya
contains a word that is followed by an optional stop, ج, superscripted above
it then it is followed by word2; we have to know that between word1 and word2
there is a special symbol that needs to be handled either by the program or
the user has to pay attention to it. If for example the two words are abc^ج
xyz, and the user enters "abc xyz" he wont get a result when he or she should
have gotten it. Why? Because there was a superscripted ج that wasn't handled.

Remember when we created the digest for the sample suras? We weren't able
to add or delete anything. It's because the entire unit is set to a specific
value. Everything in a text file is a character with a value, it's not just a 
matter of font....

> As I mentioned before, I'm not an Arabic speaker. Most people here
> does even know the differences between alef maksura and yeh. They
> don't even know why the yeh is not dotted (final and isolated). And
> they don't know how to spell. What they do know is they read from the
> Mushaf, which mostly in Rasm Uthmani. So, if they were to search for
> the word Ar-Rahman, al they know is it is spelled as alef lam reh hah
> meem noon. I believe in rasm imlaie it is spelled with alef, or maybe
> not, but you get what I mean. So, what do I get?

People would still need to bypass many symbols, including sajda. If the
'uthmani rasm is included in full in a standardized character set like
Unicode, the begin-of-sajda indicator has to be an included character. As you
know, this indicator is a line under a word or a set of words. This symbol by
itself would be a challenge for those working with character sets, because at
some instances such a character would be under one word, and sometimes more
than one word (i.e., you have something like an underlined whitespace, if that
was said correctly).

We are not talking about fonts here, we're talking about characters with a
given value. The text has to be displayed correctly whether it's in a high-end
wordprocessor or a terminal. So if you are in a wordprocessor, and you decide
to underline the words that indicate the beginning of a sajda, you will get 2
lines under the phrase. If you double underline the phrase, you must see 3
lines under the phrase. The first underlined space is a character with a
value which is part of the content not how it's decorated.

> Just a thought, not sure if this will happen or not. Let say a person
> type in the text using the African style. Then when they run it thru
> the routine, they found out that the text is tempered. It might give
> the impression that all the while, the text that they read is wrong.
> Of course, we do not want this to happen.

This can happen with or without the routine. In the reading of warsh, that
which is popular in some parts of Africa, the reading does not differ to the
extent where the text changes most of the time. I'll give you an example
before discussing your point:

In the reading from Hafs through 'aaSim, aaya 18 in surat alkhaf
(sura 18) is read:

law iTal't 'alayhim lawalayta minhum firaaran walamuli-ta minhum
ru'baa.

From warsh through nafi', it is read:

law iTala't 'alayhim lawalayta minhum firaaran walamulli-ta minhum
ru'baa.

The laam in walamul-ita is doubled in warsh. No problems so far. But if we go
to surat aal 'imraan aaya 57, 'aSim read it:

wa-amma alatheena amanoo wa'amiloo aSSaliHaati *fayuwafeehim* ujoorahum....

Nafi' read it:

wa-amma alatheena amanoo wa'amiloo aSSaliHaati *fanuwafeehimoo* ujoorahum....

This is a difference in a reading (all of them are considered THE QURAN) 
appears in the text in one letter only. It's yaa (y) from 'aasim and noon (n)
through nafi'.

Now consider the following cases:

1. A Muslim who is accustomed to warsh's reading buys one of the many
Quran software packages during his trip to Saudi Arabia. He starts the
software and searches for fanuwafeehumoo ujarhum in hope he will
get 3:57 as a result, but he doesn't get it because the reading used
in the software is that of Hafs.

2. Same case as above, but the same person goes to a web site.

3. Same case as above, but he looks in a PDF file downloaded from
the net.

So we ask: In what way is the license more beneficial in this case than a
digital signature? In a license, you have to have all of warsh's reading
before licensing it. In digital signature, you have to sign (create the sums)
for the reading, and you can't do that when the text is not available. In both
cases, the whole reading has to be present as text.

If you have both texts available, you can either license them (which makes
absolutely no sense and it is silly) or you can sign each text. You can't
license if you don't have the text, and once you have the text you can
sign it and use the sum to verify the text.

If the user who wishes to verify the text is a human being, you put big
signs that say: "Verify the Quran: Hafs through 'aasim." or "Verify the
Quran: warsh through nafi'." If it is a program, then it's a checkbox
and a button, or it could be a commandline option, and so on.

There are numerous schemes on which validation systems for the
Quran can be built using digital signatures. As we said, people with
the appropriate expertise can exercise their creativity.

Finally, if you are looking for the imlaa-ee Quran, you can find it here:

http://qurankareem.info/a/NormalQuran.zip

Salam,
Abdalla Alothman