Date: Tue, 16 Dec 2008 14:25:27 +0000 From: "Mark-Jan Nederhof" To: "Ancient Egyptian Language List" Subject: Re: AEL annotated sign list On Tue, Dec 2, 2008 at 7:28 PM, Serge Rosmorduc < serge.rosmorduc@qenherkhopeshef.org> wrote: > > For reasons which escapes me, a number of the signs proposed in the > supplement to the catalogue of the Egyptian hieroglyphic printing type have > been given slightly different codes in the Manuel de Codage. Irreparable damage was caused by the lack of any documentation by the makers of the MdC sign list that would explain why certain signs were included and why they received certain names, a poor tradition, which I'm sorry to observe, is continued with the EGPZ. This is in addition to the irreversible confusion caused by introducing conflicting Gardiner names, and avoidable proliferation of signs due to 'ligatures'. Most issues in the previous messages of this thread are still open, but let me pose another question, which should interest everyone working on software that renders hieroglyphic. This question is simple: How many dimensions does a glyph have? In my Revised Encoding Standard, there are special operations that work on the actual shape of signs, but the basic dimensions are defined to consist of a width and a height, and these are the main parameters that determine formatting in context. There is a natural unit for these dimensions, let us call that 1 EM, which is the height of A1, as well as the height of most signs in the A, B and C categories. Many broad signs such as F46 (intestine) are also 1 EM wide, at least in most fonts. If the height or width of a glyph does not equal 1 EM, it is usually less than 1 EM. However, there are a few signs that appear with height or width slightly bigger than 1 EM, and in many cases the excess seems to match parts that 'stick out', such as the tails of animals. This is for example the case with E24 (panther), where the distance between nose and heels of the hind legs seems to be exactly 1 EM. The tail makes the glyph somewhat wider than 1 EM. Similarly for I11 (cobra), the glyph appears slightly bigger than 1 EM, the excess being the end of the tail going downward. So, if we encounter E24 or I11 in context, would we expect that the tails would stick out from the quadrat in which they occur, and would this always be the case? In other words, is formatting to ignore the excess parts? Also for a small collection of signs that are less than 1 EM wide and less than 1 EM high, there seems to be some part that typically sticks out. For example, E15 (sitting canine) often has the tail sticking out below the base line (in horizontal text). Does this mean that each sign should have 2 + 4 parameters that determine its dimensions, with width and height excluding the parts that stick out, plus the extra parts above, below, to the left and to the right? Related question: should one standardise these dimensions? One should allow for font designers to draw signs in different ways, and A2 may in some fonts be 0.8 EM wide and in some other fonts 0.7 EM wide, but it may not be unreasonable to demand that it is always 1 EM high and have width somewhere between say 0.5 EM and 1 EM. Automatically formatting an encoded text using an aberrant font that does not conform to such reasonable assumptions may go very wrong, both in MdC and in RES. Comments anyone? Mark-Jan ==============================================================================