[Egyptian] Using hieroglyphs in a Word processor

[Marwan Kilani]

Hello Bob,

thank you for that.
Your document is very interesting, but at the same time highlights a
problem with control characters that I pointed out a while ago, but which
has still not been addressed/answered, as far as I know.

The point is: as already said, whether you use control characters or
ligatures, groups will be displayed correctly *only* if they are present,
as precomposed glyphs, in the font.

Control characters do not make things up on their own.

Now, the problem I see is: what happens if the group you want to display is
not present in the font?

With ligatures embedded in the font, essentially nothing happens: the
single characters would be displayed just one after the other without being
grouped together (or they would be grouped together in simpler groups, if
some simpler group combining at least some of the signs would be present in
the font). It would not be visually too fancy, but it would still be
readable and will still be overall ok as text.

On the other hand, your file shows very well what would happen in such
cases with control characters: the result would be broken control
characters popping out here and there between your actual hieroglyphs.
Essentially, the editor would not display only the proper hieroglyphic
signs, but it would display the hieroglyphic signs *AND* the control
characters, that will appear as unrecognized/unprocessed glyphs.
Essentially the result will be those sequences of hieroglyphs + control
characters that you highlighted in yellow in your proposal.

I think this a huge problem: leaving aside secondary issues like the fact
that such broken sequences would be ugly (much uglier that just ungrouped
hieroglyphs) and that no journal editor will ever accept to publish a text
with broken control characters, leaving aside that, the huge problem I see
is that such sequences of unprocessed signs+control characters would
essentially make the text unreadable and thus useless.

The examples appearing in your paper are relatively simple, and therefore
is relatively easy to visually focus on the hieroglyphic signs, to ignore
the unprocessed control characters and just to read through them. It's ok.
But one can also assume that those simple groups will be quickly integrated
into the font.

But what about complex groups (which i think is the main reason people here
want to use control characters in the first place)?

What about ramesside groups?
If your ramesside group is not present as a precomposed glyph in your font,
your text processor will not display it correctly. Whether you use
ligatures, or control characters, it won't display correctly.

Now, if you use font-embedded ligatures, this won't be a big problem,
because the unrecognized ramesside group will just be displayed as a plain
sequence of hieroglyphs, which perhaps could be reorganized into simpler
groups (that are more readable than plain hieroglyphs). It won't be as nice
as the proper desired group, but it will work and the resulting text will
still be readable and usable.

But what about control characters?
What it will display will just be the raw sequence of hieroglyphs *AND*
unprocessed control characters.

That means that the users will see just a string of a lot of random
unprocessed/unrecognized control characters with some hieroglyphs scattered
among them (and if you think that Stéphane etc's proposal requires 10
control characters to code a group of just 3 signs like Htp, you can
imagine how many control characters you will need to code even just a
"simple" ramesside group).

The result would just be graphically unreadable, and probably completely
undecipherable for the large majority of egyptologist who won't know how to
read through,to mentally "parse" and to mentally "compose" the control
characters (which again will be very numerous, nested one into the other,
etc).

If you add that some of the control characters you (pl, not necessarily you
Bob) are suggesting require to modify the inputting order of the signs, you
would just end up with a completely broken text, not only with tens
(literally tens) of unprocessed control characters scattered everywhere,
but also with proper hieroglyphic signs displayed in totally random and
totally wrong order (like having the hand-d displayed before the cobra-D in
group Dd, with a broken control character between the two). it would just
be near to impossible to easily make any sense out of such a text.

Now, this in my opinion this is a huge problem, because it is a problem
that is very likely to occur.
Consider: it will occur if your font does not have a precomposed glyph for
the group you want to display.
It will also occur (and this is the important part) if your font had the
precomposed glyph, but the text editor (whatever will be) will not be able
to to process your control characters.
And this is something that is *very likely* to happen, especially if you
are considering to introduce "exotic" control characters that have no
parallel in other already encoded languages.
Because let be honest: very mainstream text processors such as office and
open office or browsers like safari still have problems in processing
widely used control characters and complex scripts.. and well, egyptian
definitely does not have the commercial weight for us to hope in a fast and
bug-free implementation of exotic, hieroglyphic specific control characters.

This means that until text processors and browsers won't have implemented
the recognition of control characters, you will have very high
probabilities of displaying broken text with unprocessed control characters
that no one will be able to use in any way because it will be just crazy to
read and to make sense of them (or even almost impossible, if you include
control characters that will mess up the sequence of the hieroglyphs
themselves).

So my question is very simple:
How are you (all) planning to deal with such a scenario?

In how many practical cases would be more useful to have a broken string
with visible control characters, rather than a readable plain sequence of
signs, possible recomposed into simple ligatures?
You are right, you will lose the information about the original precise
spatial organization of the signs, but again, in *practice*, in how many
cases the preserving exact information about the precise order will be more
important that just displaying the text in a readable way?

And note: I am seriously asking these questions, as I think they are
important questions that requires precise answers.

Best

Marwan

























On Fri, Jul 29, 2016 at 3:35 PM, Bob Richmond <bobqq at live.co.uk> wrote:

> Hi All
>
>
>
> Before we get to discuss what is needed to get consensus for UTC-related
> issues I thought it would be a good idea to get back on track about what
> this is all about.
>
>
>
> So I drafted this note about one fundamental point – hieroglyphic in Word
> processing. Maybe I’m wrong but I suspect this is one key part of what many
> Egyptologists want to be able to do. PDF is attached.
>
>
>
> I first did experiments along these lines back in 2015 and successful
> prototyping coloured my thoughts on how to approach the topic.  So its old
> news to me but possibly new to some of you – I’d very little time available
> to dig into this before lunch in Cambridge!
>
>
>
> Bob
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> Egyptian mailing list
> Egyptian at evertype.com
> http://evertype.com/mailman/listinfo/egyptian_evertype.com
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://evertype.com/pipermail/egyptian_evertype.com/attachments/20160730/cedd9dd9/attachment.htm>