https://bugzilla.redhat.com/show_bug.cgi?id=812207
--- Comment #2 from Shriramana Sharma samjnaa@gmail.com --- Created attachment 601282 --> https://bugzilla.redhat.com/attachment.cgi?id=601282&action=edit Glyph for Marathi-style L_LA (and for headstroke)
Hello Pravin. Sorry for the delay in responding to your offlist mail. Please find attached a suggested glyph for L_LA ल्ल which is based on the Marathi LA shape. The glyph is the one with postscript name "marathi_la_vir_la".
I have also included some rough work glyphs which were intermediary stages in arriving at the final glyph in case someone wants to edit them further. You can ignore them when copy-pasting to trunk.
I have also included the sample positioning of the below-base VS-U VS-UU (the reduced versions which stand before "glyph244" in the font and are meant to be attached below conjuncts) for the new glyph.
As you can see, I have just managed to position the vowel signs below the new L_LA glyph but above the windescent level. My first draft was just by doing a copy - paste - remove headstroke and stem - correct points sequence of the existing LA glyph below itself, but to give space for the below-base VS-U VS-UU to not go below the windescent level, I had to vertically shorten the glyphs. I have attempted to do so without distorting the curvature.
Note however that the below-base RA + VS-U/UU glyphs (which come after "glyph239" and before the above-mentioned reduced VS-U/UU glyphs) do not have enough space below the new L_LA glyph. Since there are no real words which have the sequence L_L_RU L_L_RUU, I did not feel it justifies squashing the LA-shape even further (it may get distorted). Any sequences involving L_L_RU L_L_RUU in Marathi style may be rendered as LA_halant L_RU/RUU. I hope I'm clear on this. Relevant lookups [or whatever you call it in OpenType] may need to be added.
As another issue, along with the above, I have also included a substitute glyph for A8FB DEVANAGARI HEADSTROKE. I found that its glyph in the Lohit Devanagari font does not correspond to the description in the original proposal (http://std.dkuug.dk/jtc1/sc2/wg2/docs/n3488.pdf pp 2 and 3), which says that it should be at the height of the headstroke of other characters (but not connect to it). I have just pushed down the glyph at the height of the headstroke of क. Please patch that also. (I didn't want to open a separate bug for a trivial matter.)