Talk:Iran System encoding

• http://web.archive.org/web/20070223125421/http://sina.sharif.ac.ir/~roozbeh/farsiweb/iransystem.txt

gives a unicode mapping - i used this for the table —Random832 19:23, 18 October 2007 (UTC)

Table border colors

I took the liberty to remove colored borders from the layout table because they didn't add anything to the article and only made the table confusing. The information conveyed by them is already present in the table in the characters themselves and their shapes. Please notice that the encoding described here encodes visual forms rather than logical characters. However, the mapping table linked above here in talk page used logical characters instead and guarded them with ZWJ or ZWNJ characters on either side when needed to select an appropriate form. It could be better described by using Unicode presentation forms as it is actually done here in the article. It lacked any legend in article but according to the edit summary the border colors were added to denote combinations of these ZWJ/ZWNJ sequences. These aren't relevant to the subject of the article because it's just a peculiar way of denoting form of character, way that also depends on other forms that character might have. For example, orange border was used to mark an isolated form of a character (given with non-joiners on both sides in mapping table) and red was used to mark an isolated form of a character which otherwise doesn't have distinct initial form (so the non-joiner following it could be dropped because it would make no difference). — mwgamera (talk) 01:53, 16 May 2016 (UTC)

Hi, I wasn't particularly happy with the visual appearance of these colored borders, but I'm definitely not happy with removing the information they presented from the table. From your explanations above I take it that you are familiar with the language and this character set and code page and that you can derive some of the properties just by looking at the glyphs. However, I am not and only look at it from the viewpoint of a systems developer who would try to implement support for it in an operating system or application in some internationalization/localization project. For this, any "attributes", describing special properties of the particular glyphs are important to know to understand the encoding from a technical point of view.

If you don't like the colored borders, please find another way to incorporate the info into the article so that it can be understood by people looking for technical info to implement this.

--Matthiaspaul (talk) 16:51, 6 June 2016 (UTC)

I tired to argue above that they presented no information beyond what is already covered, so what info exactly? If you literally mean information about where to put joiners in Unicode to fake proper rendering, it sounds to me as relevant as marking every letter with acute in Latin-1 table with different color just because in Unicode it could be written with a combining acute character. The fact that «initial form of the ع» can be written in Unicode either as ع‍ or as ﻋ is not related to this encoding; these representations should display the same and all you need to convert between them are Unicode tables. What matters here is that this is what the code 0xE4 stands for; and currently the table uses ﻋ here, that is U+FECB ARABIC LETTER AIN INITIAL FORM, which accurately and completely describes it. It certainly should, however, be linked to the article about that character just like it's done for box-drawing characters. It should be easy and I might do that later.

However, I have to agree that currently the asterisks after the code, that I think are supposed to disambiguate cases where it's not that obvious as above (e.g. 0xF4 ﻡ is used both for isolated and final form as they usually look the same but Unicode has a separate codepoint ﻢ for its final form), aren't up to the job and don't really help much. And also they are incorrect (0xE2 is a final form, for example). What do you think about instead using something more akin to the system Random832 had used in [1] before converting it to the way it is now? That is, about marking which forms of the initial/​middle/​final/​isolated are covered by given glyph with some marks inside the cell rather than with color or borders? Using letters IMFX for it was pretty ugly, but relatively clear, I think. — mwgamera (talk) 00:32, 16 June 2016 (UTC)

Code Page 769

This is most likely Code Page 769. If not, what number is it?Alexlatham96 (talk) 01:07, 13 April 2018 (UTC)