Untitled Gurmukhi handwriting font

[Dan Reynolds]

Hello everyone!

I have missed you all very much. I have not been working on any libre font projects since March, but I have been keeping abreast with all of the discussions in this Group since then. Today I am pleased to announce that I’m starting a brand new project. Exciting times! Over the next few months, I’m going to be developing a handwriting font that will cover both Gurmukhi and Latin.

Since today was my first day working on the project, I spent it pulling a few resources together. Perhaps these will be of interest to some of you.

• First, I downloaded and re-read Emma Williams’s excellent essay, A Comparative Study of the Development of the Gurmukhi Script. This was part of her MA coursework on the MATD course at the University of Reading. Indeed, Emma was a classmate of both Dave’s and mine there. You can read her work, too. It is up at http://www.typeculture.com/academic_resource/articles_essays/
• I  purchased a book some time ago, Learn Punjabi through English in 30 days. This seems like it will be some help.
• Unfortunately, it does not have a lot of information about how the Gurmukhi characters are actually written, but Erin has an image of from a similar book, Learn Punjabi in 30 days, up on her website – http://1.bp.blogspot.com/_u7MaPZVBoag/SCpQBGQYtHI/AAAAAAAACPo/426ACTBEHG0/s1600-h/punjabi-howto.jpg; this shows a stroke-order for each character; it is helpful – I think – to compare and contrast this with the handwriting samples Emma takes from manuscripts in her dissertation.
• I also had a look at Linotype Gurmukhi – which Emma writes highly of in her dissertation – and the handsetting Gurmukhi typeface shown in the 1924 Reichsdruckerei catalog, which she does not write highly of (although, as she points out, that small specimen is more problematic from a typesetting point of view than anything else).

So, what is going to happen next?

• I’ve started practicing some Gujarati handwriting, although this is slow-going, of course. I will upload some photos later, I guess.
  
• There is not GitHup repository for the project yet, but I will post a link to it here once I create it.
  
• My goal is really to post on this thread every day that I work on this project. However, there will not be daily posts, as I won’t be working on this font every single day. Would that I could!
  
• This project is going to need a name.

Want to help?

• If you have any suggestions for other libre Indian-script fonts I should use to try to match my Gurmukhi to, I’d love to hear about them.
  
• Also, in case anyone is super-interested in discussion, I am thinking about making a “slower-style” handwriting font than a speedy/messy one. I guess I can try to make some handwriting sketches at the beginning of next week that will illustrate what I mean!
• I am not aware of any resources for Gurmukhi that are of the calibre of Naik’s Typography of Devanagari or Lambert’s Introduction to the Devanagari Script for Students of Sanskrit, Hindi, Marathi, Gujarati and Bengali, both of which I am fortunate to have copies of. If you know of any good texts, please drop me a line!

[Dave Crossland]

On 2 October 2015 at 08:02, Dan Reynolds <type...@gmail.com> wrote:

• Also, in case anyone is super-interested in discussion, I am thinking about making a “slower-style” handwriting font than a speedy/messy one. I guess I can try to make some handwriting sketches at the beginning of next week that will illustrate what I mean

I think slower is better as it allows you to work with more typographic forms; actual 'fast' handwriting in Indic scripts often looks really different to typographic forms :) 

[Dan Reynolds]

Yes, I think you are right!

Today:

I practiced my Gurmukhi handwriting a bit more. I also went out and bought several sorts of felt-tipped pens; I want to get the stroke to be the right thickness (for a 400/Regular sort of font), and then I will create artwork by hand and scan it in. 

Tomorrow:

More handwriting, with pens of different thicknesses.

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAEozd0yoaWMYsBJvC%3DJhDJ-_6_2qkm725PLv2R7Cuv%3DU4m4MvA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Dan Reynolds]

Today:

• Ordered a scanner from Epson.

• Ordered ScanFont from FontLab. Thomas Phinney is the president of FontLab now! Nice.

[Pablo Impallari]

Hi Dan,

Last year I've worked on 4 custom handwritten typefaces, each glyph having 3 or 4 pseudo-random alternates, and tested pretty much all soft available for tracing artwork: Illustrator, ScanFont, Cr8's Type3.2, Glyphs's tracing plugin, etc....
Ended up choosing the good old Potrace from command line http://potrace.sourceforge.net/

The results where quite good, but the main reason for choosing it was because it allowed for more 1) Flexible and 2) Faster workflow.

Just save your whole scanned page as BMP and Potrace will turnip into a quite good EPS in a split of a second.

Then, just open your EPS in Illustrator, ungroup, and copy/paste to FL or Glyphs and scale as needed.

If you have multiple instances of each glyph in your sheet of paper, for example "a a a a a a a a a a b b b b b b b c c c c", you can quickly pick the one you like the most and discard the others.

Also used this macro to remove excessive contour points without deforming your shapes
https://github.com/impallari/Impallari-Fontlab-Macros/blob/master/IMP%20Production/Cleanup%20tracings%20Current.py

You can set your "distance" variable to 4, 6, 10, 12 or whatever you want.

--

You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/cc57ae54-140d-4b91-bff9-864bbe3f458f%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Un Abrazo
Pablo Impallari

[Dave Crossland]

3 of the fonts Pablo designed, Caveat (R/B) and Caveat Brush (R), are now online:

www.google.com/fonts/specimen/Caveat

www.google.com/fonts/specimen/Caveat+Brush

[Dan Reynolds]

Hello everyone,

Unfortunately, I have not managed to post here after each day’s work session. I hope that this post is “better late than never.” Anyway, the day before yesterday, I wrote out some more sample text on paper in the Latin and Gurmukhi scripts with a few felt-tip pens of different thicknesses. Then I scanned these letters in, and auto-traced them with ScanFonts.

You are right, Pablo. I have used ScanFont before, several years ago, to scan in rather detailed Latin-script display letters. IIRC, it worked quite OK with that. It does not work well with Indian writing systems, however, as it cannot recognize what their shapes are supposed to be. Still, it does scan and trace them … you just have to reposition everything in resulting VFB file later.

Now, I need to preface this post by mentioning that the letters you see below are just the results of quick tests. The auto-trace is weird, and not what I want in a final font. The proportions of letters are not accurate, either, and I am well aware of that. Some letters you will see are too light, others are too heavy. Please forgive me. I figure it is better to share some of my process than to hide it.

This first image shows three Gurmukhi characters. uni0A1F is written with a 0.7mm Staedtler liner, while uni0A38 and uni0A39 are written with a Medium sized Staedtler Lumocolor pen. I also tried writing with a Sharpie Fine pen (not pictured), which gives strokes almost as thick as the Staedtler Lumocolor M, but it bleeds into my notebook paper, which the Staedtler Lumocolor M does not do. This auto-trace is with ScanFont’s normal setting, which gives way too many points and created an almost jaggy-bitmap texture, which for an actual handwriting font would be totally unacceptable IMO. Nevertheless, from an aesthetic perspective, I do like the way the stroke on the uni0A38 tapers off. Don’t think I’ll keep it, though. That’s glyph’s stoke is too light anyway. Would be OK for a Light font, maybe.

My second image shows nine Latin glyphs, which I scanned in at the same time as the glyphs in image 1. These letters are also traced with ScanFont’s normal setting, and have the same “problems” that the first image’s characters do. I had adjusted the sidebearings for each glyph; ScanFont’s algorithm set them too tight.

My third image is also from the same initial scan from which the glyphs in my first two images come from. These letters were written with the thinner pen, and I have adjusted the glyphs’ sidebearings post-import, just like in image number two.

I was not happy with the results of my first scan. So I went back to “the drawing board,” or my notebook. I wrote more text with a Staedtler Lumocolor M pen, and scanned that in. This time, in ScanFont, I used the “loose” setting. This results in far less nodes on the contours, but the letter shapes are still kind of dumb. I think that I will not rely on auto-trace in the end, but drawn new vector shapes overtop of scanned artwork.

Anyway, in this second text, I wrote all of my Gurmukhi letters without a headline stroke. Then, in FontLab, as I was repositioning each glyph, I drew in a rectangular “stroke” to use as the headline. This may not be the best approach. But every Gurmukhi character can’t have its own uniquely written headline-stroke, either. That will just look awful in any sort of handwriting font that I’m going to be able to make.

My fifth and final image just shows more text generated with the Latin characters that resulted from my second scan. Until I start working in Glyphs, it is going to be easier set longer passages of test text in the Latin characters than the Gurmukhi one. Also – for posterity’s sake – I realize that the ! and ? here are just whack ;-)

Dave: Thanks for linking to Pablo’s Caveat fonts. I downloaded them, had a look, and found them helpful. I never would have found them without your link!

For those of you in São Paulo for this year’s ATypI conference, have a great time!

Dan

[Erin McLaughlin]

Hey! Sorry that I haven't been following this -- it might be too late to be useful, but actually Gurmukhi handwriting looks a lot different than the forms used in type --> http://www.billie.grosse.is-a-geek.com/handw.html  Just wanted to make sure you were aware!

-E

[Dave Crossland]

On 18 October 2015 at 01:54, Erin McLaughlin <soyle...@gmail.com> wrote:

Gurmukhi handwriting looks a lot different than the forms used in type

I'm okay with this, as I believe the texture treatment gives the type an informal feeling, similar to https://www.google.com/fonts/specimen/Cabin+Sketch , while being appropriate for a wider range of document types than a true handwriting font

[Dan Reynolds]

Dave, Erin,

I have been practicing my handwriting for a bit, and am going to finally move back into vector drawing on Thursday. But while I have been writing on paper, I have been thinking about your points regarding the difference between handwritten and typographic Gurmukhi characters.

Here are five scans of some of my handwriting sheets, only mildly edited in PhotoShop to removed the grimmest of my errors ;-)

Each scan shows three “hands.” The top three lines on each scan show a sentence of Gurmukhi text written in an approximated handwriting style. I have shorted the extenders used in handwriting quote a bit, because I don’t think that these are all that practical in digital fonts – here I am not writing about the Gurmukhi script specifically … I mean that for many scripts “in general” (of course there are exceptions, though).

The mid-three lines are the same text in a handwriting style that is mimicking typographic forms. The bottom lines are Latin-script. Except for the first sample here, the English text and the Gurmukhi text have the same meaning.

These samples don’t represent what I think the final font will look like! I must write that I know that the shaping of several characters both in the Gurmukhi and the Latin scripts are errors, and not what I would consider ideal. However, I think it is worth showing you these images anyway, because it helps move this discussion along.

I think that the more informal handwriting style (top three lines of each image) “looks” better, and also would match Latin better (no headline!). However, I think a handwriting font that basically looks like “handwritten typographic forms” (middle three lines of each image) would be more useful, and is the direction I will go in.

What do you think?

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAOgqPYd6ceFbtRiz1EJReCg4RYzS3xJ8QDrUYiWieyhbP4KggA%40mail.gmail.com.

[Dave Crossland]

Hi

I agree that a headline would be better; this is then a hipster-type not a 'true' handwriting type, which I expect to be popular. 

In the latin the double story 'a' and 'e' crossbar that joins on the left helps indicate this also. 

Great stuff! 

Cheers,
Dave

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/6F8B4455-8D0F-4DF5-B40F-0A50318B7811%40gmail.com.

[Dan Reynolds]

I have not been posting enough. At the moment, I am still scanning and auto tracing things, because it is so quick and satisfying (even though the resulting letters are too uneven). Just two of the A4 sheets I posted before result in 349 separate glyphs. That is too many, so I am going to have to stop this kidding around. I only need one or two forms for the /a, really, not 26!

Anyway, here is a PDF and a screenshot showing what I mean.

[Catherine Leigh Schmidt]

Super interesting project and discussion about "handwriting". Don't know anything about Gurmukhi and cool to see the big difference between handwriting and typographic forms — especially the form without a headline! Brings up some interesting questions about what users might expect from a handwriting font; with Latin we often prefer fonts with typographic shapes with a handwriting texture to true handwriting.

[Dave Crossland]

On 14 November 2015 at 00:52, Catherine Leigh Schmidt <cathls...@gmail.com> wrote:

cool to see the big difference between handwriting and typographic forms — especially the form without a headline! Brings up some interesting questions about what users might expect from a handwriting font; with Latin we often prefer fonts with typographic shapes with a handwriting texture to true handwriting.

Catherine, glad you say this because in fact (afaik) this _is_ hand-drawn typographic forms - real Indian handwriting seems to me to usually look quite unlike typeforms :) 

Dan, obviously some scaling issues on some glyphs but it looks really great! :D

I think a default and 3 alts is the max reasonable, and for a web font, probably 1 or 2 alts is a good tradeoff for 'real feel' and filesize. 

[Dan Reynolds]

Thank you Catherine and Dave for your comments!

Attached is a PDF showing yesterday and today’s progress:

• I have placed auto-traced letters in the background layer of my Font Editor’s cells

• Based on those, I have drawn the following letters in a slightly more “typographic form” – G a i m n o u z 

• More than anything, this redraw process was to remove the awful “auto-traced” look of my scanned handwritten letters. But I also have been re-sizing and harmonising my letters to one another. This creates a result that is probably in between a “handwriting” font and an “informal” sans

• I am sorry about the /o. I am going to have to think about what to do with it.

Here is what I am going to start working on from tomorrow:

1. Fix several errors obviously visible in this PDF, including

• /a too dark

• /i tail too long

• /z outstroke needs to be more curvy

• /z too wide

• Spacing is still wonky

2. Add the following 14 Latin glyphs: A E H I O S T V d e h l s period

3. Move from FontLab to Glyphs and add in my first Gurmukhi characters

Best,

Dan

[Dave Crossland]

Hi Dan

I must admit, I feel a little nervous about the bowl of the a not connecting. I think having it slightly overlapping the stem, would be better, for a mass market audience.

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/F142EF70-C487-434F-9ACF-7B8385272E14%40gmail.com.

For more options, visit https://groups.google.com/d/optout.

––
I’m moving! New address (valid from November 21, 2015):

Siegfriedstraße 14
12051 Berlin / Germany

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAOgqPYcVNHQ_V3W9_KiuiQX0WBnardeTwtCd7oF1s49RB0bVaQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/F142EF70-C487-434F-9ACF-7B8385272E14%40gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Cheers
Dave

[Dan Reynolds]

OK, will change that today, too!

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAEozd0zVMXz3gg7k00nmM0VZARcfsyPFbss9wBUti8uM%3DaLm-w%40mail.gmail.com.

[Dan Reynolds]

Today:

– Updated /a so that it closes

– Edited /i and /z

– Added 15 Latin glyphs: A E H I O S T V d e h l s period comma

Next steps (by Tuesday):

– Fix the currently available glyphs (see notes about PDF below)

– Move to Glyphs and add in my first Gurmukhi characters

Notes to attached PDF:

– If Dave did not like the open /a I posted yesterday, I doubt he will like today’s /A

– /S and /s should be wider

– /e is also too wide

– Stroke contrast is not yet the same across all the glyphs

– Not all glyphs really have the same “weight” yet either

– /E isn’t really handwritingy enough

– Spacing still not jiving the way I’d like it to

[Dan Reynolds]

Happy Thanksgiving, everyone!

I usually cook a giant turkey for way too many people to celebrate the holiday. But this year I am moving, and it will still be weeks before I have a working kitchen again. Ask me about turkey-cooking in January1 Or on next year’s Thanksgiving.

[Meir Sadan]

Hey Dan,

So nice to see this font taking shape. In this last PDF the lowercase e gets repeated a lot in the text, with its irregular protrusion becoming very prominent – I like it as a feature but I'm not sure I would necessarily want such a common letter have such an irregular feature?

I don't know anything about Gurmukhi but it looks nice and even to me.

Is this project on GitHub?

Happy turkey-less turkey day!

[Dan Reynolds]

On 26 Nov 2015, at 14:48, Meir Sadan <meir...@gmail.com> wrote:

Is this project on GitHub?

Not yet! But soon.

[Erin McLaughlin]

Hey Dan! I LOVE the Latin, but the Gurmukhi looks too small to me! Compare a short Latin word "sang" or "Hose" with a short 3-letter Gurmukhi word. They should feel equally large.  I'd also suggest finding places to give the Gurmukhi more fluidity (like the outstrokes of your Latin letters) - maybe you can make a less-rigid ੀ ?

[Dave Crossland]

On 26 November 2015 at 20:55, Dan Reynolds <type...@gmail.com> wrote:

Not yet! But soon.

 

Also, apropos the subject line, what family name do you propose? :)

Sorry about the kitchen kerfluffle! :( 

[Dave Crossland]

On 26 November 2015 at 20:48, Meir Sadan <meir...@gmail.com> wrote:

the lowercase e gets repeated a lot in the text, with its irregular protrusion becoming very prominent – I like it as a feature but I'm not sure I would necessarily want such a common letter have such an irregular feature?

I agree, I think it would be great as an alternate but the default glyphs should be quite conservative to drive adoption :) 

[Dan Reynolds]

Dear Meir and Sadan,

You’ve convinced me! I do want to keep the “hipster” feeling to this design that Dave mentioned below, but maybe keeping a diagonal crossbar on the /e is enough for that? The top line of this image is a new “round” /e, while the bottom line has the /e that I previously posted.

Best regards,

Dan

[Dan Reynolds]

I’m going to Frank’s non-Latin day in Paris on Tuesday. I will ask for some suggestions! Otherwise, I will probably name it after some Punjabi food item. I must confess that I have not eaten enough Punjabi cuisine …

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAOgqPYeQfAUJ%2Bb6NwJxNjo0nqTv4tguBsE_4Lgu4Yz3x7fu3zg%40mail.gmail.com.

[Dan Reynolds]

Dear Erin,

OK, I am digging this comment. How about this 20% higher headline height? I will think about the  ੀ and fluidity next. Hopefully the next PDF I post (Wed. or Thursday) will go into that.

Thanks,

Dan

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at http://groups.google.com/group/googlefonts-discuss.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/ba6534a8-cc5c-47d2-ba32-a4fa2ccc34c2%40googlegroups.com.

[Dan Reynolds]

Hello everybody!

Long time, no posts. I hope you all had a happy new year. I am excited about 2016.

This project is now called Moga, after a city in the state of Punjab in India. I have finally created a repository for Moga on GitHub. Here is the link – https://github.com/typeoff/moga

The PDF posted to this thread is also available there, in the test folder. The font used in the PDF is also in the test folder, and the Glyphs file – should anyone be interested in looking at this so-still-work-in-progress source– is in the sources folder.

So. Here is the deal with my PDF:

• Page 1 shows all characters so far (not many). There are two lowercase /e variants at the moment.
• I have scaled-up the Gurmukhi glyphs somewhat since November.
• I hope that my  ੀ is less rigid (Erin?)
• Page 2 has placeholder Gurmukhi text and a Latin lowercase test. The Gurmukhi text was generated at adhesiontext.com with the eight Gurmukhi base characters I have so far. I randomly added in a bunch of  ੀ glyphs, turning real words into presumably not-real words. Latin-script words have been parachuted into the Gurmukhi column, too, so show relative size.
• Page 3 is all Latin text, with a column lowercase and a column in uppercase.

Oh, the bugs to be fixed! Here is what I have on my list already:

• Make the Gurmukhi letters, here and there, a little more fluid (sorry, Erin!).
• The top of the ਓ is too flat, I think, and the horizontal stroke inside the counter needs to grow a bulbous “terminal” I think, even though that element is not really a stroke terminal.
• The /D and the /O are too light.
• The overshoots on /B and /S do not seem right.
• That /E still does not look handwriting-y enough to me, and I do not like it.
• The right side of the /M is weird.
• The “teeth” at the bottom of the /W (particularly on the left) are/is too heavy.
• The new /e does not look like it has been written with two strokes.
• /k is too narrow.
• /r is too wide, and the top right – the arch, I mean – is too light.

What is to be done?

• I need your feedback :-)
• I need to spend more time with the side-bearings on like, every glyph.
• Edit the current Gurmukhi glyphs & add more of them.
• Add in some Latin-script numerals.
• Post a new entry here, pointing you to the files, after I have updated them. I am thinking that this will happen on Saturday or Sunday.

Thanks,

Dan

On Sunday, November 29, 2015 at 9:39:18 PM UTC+1, Dan Reynolds wrote:

Dear Erin,

OK, I am digging this comment. How about this 20% higher headline height? I will think about the  ੀ and fluidity next. Hopefully the next PDF I post (Wed. or Thursday) will go into that.

Thanks,

Dan

On 27 Nov 2015, at 18:14, Erin McLaughlin <soyle...@gmail.com> wrote:

Hey Dan! I LOVE the Latin, but the Gurmukhi looks too small to me! Compare a short Latin word "sang" or "Hose" with a short 3-letter Gurmukhi word. They should feel equally large.  I'd also suggest finding places to give the Gurmukhi more fluidity (like the outstrokes of your Latin letters) - maybe you can make a less-rigid ੀ ?

On Thursday, November 26, 2015 at 7:55:07 AM UTC-6, Dan Reynolds wrote:

On 26 Nov 2015, at 14:48, Meir Sadan <meir...@gmail.com> wrote:

Is this project on GitHub?

Not yet! But soon.

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-discuss+unsub...@googlegroups.com.
To post to this group, send email to googlefonts-discuss@googlegroups.com.

[Dan Reynolds]

Here is another work update:

Today, I saved new in-progess files in the GitHub repository for the Moga project (https://github.com/typeoff/moga). Included is the PDF that I am also attaching here for all of you.

This PDF, showing my progress since Thursday, is all about more Latin key glyphs. I spent my energy there, rather than on the Gurmukhi glyphs. So of my issues mentioned below are solved, some aren’t yet. Of course, more glyphs also bring more issues with them, too :-D

About the PDF:

• The font has one Stylistic Set, a few f-ligatures, and a narrow /f that comes before certain other glyphs. On page 1, you can see the ss01 variant glyphs, they are second in the string (e.g., the second /A, /M, /N, /e, and /s). In most cases, these glyphs are not in the font because I think that they could add to the textual whimsicality or other informality. Rather, they are there because of my own still-indecisiveness.

• You can see an /f and and /ä in the last line on page 1, too. This /f is my narrower. alternate f. It gets subbed in by a Contextual Alternates (calt) feature.

• Page 4 is a copy of Page 3, except that all text is set with ss01 selected. This means that all /A, /M, /N, /e, and /s instances are the *.ss01 version of those characters, not the standard /A, /M, /N, /e, and /s.

I hope to post my next PDF here on Thursday. This will be devoted to new Gurmukhi glyphs.

Thanks,

Dan

[Dan Reynolds]

Hello!

Here is an update to the Gurmukhi. I have a better text sample in this PDF. It is not “proper” text yet, but at least it is proper words, with vowel marks, etc.

Just two pages. Page two is set with all ss01 applied to all the Latin & Devanagari text in Moga. In the Gurmukhi, only the ਓ is different in ss01, but new to the Latin are single-storey /a and /g glyphs. Which variants of glyphs do you all prefer?

All files are also over on GitHub. I will try to post another update here over the weekend.

Dan

[Dan Reynolds]

Hello,

My updates in the last seven days seems so minor, visually, that is seemed unlikely to be of much benefit to upload any of my test files. But in addition to lots of little changes to the form and spacing of glyphs, this PDF has some new glyphs in it. The biggest change you’ll will notice here is the addition of oldstyle figures to the font. You can see them peppered into the text of page 3 of this PDF. What do you all think of my super top-heavy 8?

Up next on my plate:

• Gurmukhi figures
• Latin lining figures
• More Gurmukhi base characters
• Last of the Latin letters (cap X and Z, etc.), plus more Latin diacritics

Thanks,

Dan

[Dan Reynolds]

Hello Dave, Erin, everyday out there …

Ugh! I have been wanting to post a new PDF all week. Sorry! Instead, I have been just been fidgeting with glyphs, instead of posting my progress :(

I am hoping and planning on publishing a proper “version 1.0” of this font at the end of next week. Wow! In the mean time, here is a long-delayed PDF. Updated files are available one GitHub, too.

Things I need to do soon:

1. Scale the font up. It is too “small on the body” at the moment – the Latin AND the Gurmukhi together, I mean (not that I think the scripts are the wrong size in relation to each other anymore). See the two columns of page 4 of this PDF to see what I mean.

2. Add these missing Gurmukhi glyphs: ਐ ਙ ਜ ਝ ਞ ਟ ਠ ਢ ਥ ਧ ਨ ਪ ਫ ਬ ਭ ਯ ੴ

3. Plus Gurmukhi figures, and a few other miscellaneous characters.

4. Fix eth, X, x, some of the f-ligatures, etc.

Thanks again,

Dan

[Dave Crossland]

Hi

Thanks Dan - looks good, I'll be posting notes on a bunch of PDFs including this one shortly :) 

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.

Visit this group at https://groups.google.com/group/googlefonts-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/5F69E443-2A06-4E49-9E5E-EB9EA58B6636%40gmail.com.

For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.
Visit this group at https://groups.google.com/group/googlefonts-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/00c646b9-7f41-43e7-9495-1184d1a777c4%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

<20160121 02 Moga.pdf>

--
You received this message because you are subscribed to the Google Groups "Google Fonts Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to googlefonts-dis...@googlegroups.com.
To post to this group, send email to googlefon...@googlegroups.com.

Visit this group at https://groups.google.com/group/googlefonts-discuss.
To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/5F69E443-2A06-4E49-9E5E-EB9EA58B6636%40gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Cheers
Dave

[Dan Reynolds]

Hello everyone,

I have now finished the font to an extent advanced enough that I believe it is time for me to publish Version 1.0.0 now!

I am not sure how much this actually “counts” as publication, because there have been functioning and updated font files on GitHub now for weeks, but I doubt that anyone downloaded them. Everything about the repo certainly looked un-done. As indeed it was – until now.

https://github.com/typeoff/moga/

I have attached a PDF here, which is also available on GitHub. It shows the font’s functionality much more superbly than any my previous test documents did. Thank you Gunnar for the excellent PDFs you published on your repo! Those was a big help over this past week.

I need some help though, folks, for a few tiny things (that I am aware of … maybe you see bigger problems that I don’t):

– Page 1 of the PDF, in the signs & modifiers display: I would have shown everything with the dotted circle glyph, but most of the zero-width Gurmukhi marks did not have the feature written upon font-generation so that they “snap” to the anchor-positions placed on the dotted circle (all except for the udaat and yakash glyphs, which DO attache properly). I suspect this is might be a Glyphs bug.

– So I showed most of the marks with ਕ instead, but udaat and yakash do not “snap” to the Gurmukhi base characters. Again, the anchors seem to be named & placed correctly.

– The tippi + bindi sequence does not lead to the proper substitution (i.e., the tippibindi glyph), at least not in InDesign CS 5.5. The second glyph just replaces the first, no matter in which sequence I type. Note that I cannot get this to work in InDesign CS 5.5 with Atithi or Langar, either!

– ਫ਼ plus the rakar-form of ਰ does not cause a substitution to a dedicated fa_ra-gurmukhi glyph for me in InDesign CS 5.5 (where this PDF was generated). Instead, the rakar-form of ਰ is just positioned below the glyph.

– The yakash is not snapping into its proper in InDesign CS 5.5 position, either, although the anchors seem to be set correctly to me. I do get this working in other applications, though.

That is it! Thank you very much for your helpful comments over the past few months.

Dan

[Dan Reynolds]

Hello again,

Speaking of OpenType feature encoding questions …

Some of you may know about my website, typeoff.de. The fonts I show on the site are all webfonts, and users can test the fonts live on the site, typing in their own text, changing size, line-height, color, etc.

I have created a new page on my site for Mota. It is not live yet (sorry!), because I will have to make two image before I make the page live. Otherwise, I will break something :-/

Anyway, in testing the site, I have noticed that neither Chrome nor Safari are performing the Gurmukhi below-the-base substitutions. The positioning of below-the-base glyphs like the halant and yakash looks just right (better than InDesign … ). But the below forms of ha, ra, and va (the rakar) are not subbing in. Nor do my rakar-conjuncts. I think that this is sub-optimal. Text is rendering “OK,” I guess. But not as I would like. See the lower right-hand column of my screenshot.

Thanks,

Dan

[Erin McLaughlin]

Wow, Dan, it's great! You've done so much! Cooool!

I don't have too many suggestions, but here they are:

-- The design might look more natural if you were able to make the subjoined ਰ (and other characters) join with the characters above -- or at least with the "normal" ones that have vertical stems and aren't round on the bottom. I know it's more work, but it might be worth it! Its likely that someone writing this by hand would make a little loop/flick when writing a character. Right now, I think that some of these gaps (with subjoined ਵ and ਹ) look a little tight, like they should either be farther away, or should be joined completely.  Just a thought :)  

-- I think your Uu vowel sign might look more natural if the lower one is drawn slightly differently than the top one (right now they look like identical shapes). perhaps more flat or not as wide or something. 

-- This might be pointless, but you might want to increase the left sidebearing of Gurmukhi numerals 8 and 9 so that, in the rare event they're used next to a letter, they won't clash. (wow, nit-picky erin! lame!)

-- The comma (and quotes) stick out to me. The fact that it has a taper and a point, it doesn't relate to your other small marks (like Gurmukhi vowel signs). You might want to reconsider a new shape that has a blunted end.

-- Might also want to make the () a bit looser, as they are rather snug against the Gurmukhi. Punctuation spacing is a difficult balance in these things :-/ 

-- You'll want to space the Danda under the assumption that someone will use it without inserting a space beforehand. So I think you might want to go a tiny bit looser on the left side.

I think that's pretty much it! Gurmukhi looks good to me! :)

Thanks!

Erin

[Alessia Mazzarella]

Hi Dan,

I tested your tippi bindi in Indesign CC and it works fine, if that can be of any help. Not sure why the substitution doesn't happen in the 5.5 version, have you tried to select Punjabi as a language? 

I also tested ਫ਼ and the rakar-form of ਰ and the substitution does not happen.

Alessia

[Lasse Fister]

>
> I have created a new page on my site for Mota. It is not live yet
> (sorry!), because I will have to make two image before I make the page
> live. Otherwise, I will break something :-/
>
> Anyway, in testing the site, I have noticed that neither Chrome nor
> Safari are performing the Gurmukhi below-the-base substitutions. The
> positioning of below-the-base glyphs like the halant and yakash looks
> just right (better than InDesign … ). But the below forms of ha, ra,
> and va (the rakar) are not subbing in. Nor do my rakar-conjuncts. I
> think that this is sub-optimal. Text is rendering “OK,” I guess. But
> not as I would like. See the lower right-hand column of my screenshot.
>

Hi, this is just a shot in the dark. Did you set the "lang" attribute of
the element that contains the Gurmukhi text to something appropriate?
E.g. lang="pa" for the Punjabi language. This is a guess supported by
wikipedia, I don't know what the right language is for your case. The
shaping engine may need that help to do the right thing, also depending
on how you features are written.
In my experience with Arabic script, Safari was not very good, but
Chrome should be just as good as Firefox, as they use both Harfbuzz.

[Dan Reynolds]

On 13 Feb 2016, at 16:17, Lasse Fister <la...@graphicore.de> wrote:

Hi, this is just a shot in the dark. Did you set the "lang" attribute of
the element that contains the Gurmukhi text to something appropriate?
E.g. lang="pa" for the Punjabi language. This is a guess supported by
wikipedia, I don't know what the right language is for your case. The
shaping engine may need that help to do the right thing, also depending
on how you features are written.
In my experience with Arabic script, Safari was not very good, but
Chrome should be just as good as Firefox, as they use both Harfbuzz.

Oh! I will try this. I have script tags for Gurmukhi (gur2 and guru), but not language tags for Punjabi (PAN). The Gur2 and Guru language tags at the moment are just dflt.

Thanks,

Dan