Font Quality Improvement Project — Alexei's Worklog

[Alexei Vanyashin]

Hi everyone!

I am happy to start working on the 'font quality improvement' project, 

dubbed 'Making Raisins'.

This thread will be used to post daily, or weekly updates on progress

for self-organisation and record-tracking purposes, and by no means spamming. 

If this amount of traffic will be overwhelming for you, feel free to mute the thread, 

or enable weekly summaries. 

My involvement will consist of 3 sprints of 4 week each, 

during which I will be touching different font projects.

My list of scheduled projects, and progress timeline can be tracked anytime

via this link:

https://github.com/google/fonts/issues/285

Let's start!

# 1. Warm-up, week 1, 2016-27-06

Today I summarised our teams work on Google Fonts new 2016 Latin and Cyrillic Encodings. 

Three new levels of character support were developed for each script. 

Core: the existing basic (Latin / Cyrillic) naming scheme, which has limited language support. 

1. Latin Plus: extends coverage to Vietnamese and all Latin-based languages. 

2. Latin Pro: adds more currencies, math symbols, and General Use Latin extensions.

3. Latin Expert: Adds Small Caps, Arrows, Ornaments, Fractions, Super- and Subscript. 

Cyrillic follows a similar naming scheme, but for different languages than Vietnamese.

By the end of the day, all files were pushed to git. Check them out here for more info:

https://github.com/alexeiva/fonts/tree/master/tools/encodings/GF%202016%20Glyph%20Sets

[Pablo Impallari]

Hey Alexei, would you like to add Cyrillic to Raleway Italics (Romans already have it)?
You are the man for it :)

--
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/42cacffb-65a6-4e24-b527-20affa63a407%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Un Abrazo
Pablo Impallari

[Alexei Vanyashin]

Hola Pablo!

Nice suggestion. I need to discuss this. 

I see that Raleway Cyrillics haven't been pushed yet. 

Is that only because the CYR italics aren't finished yet?

--
You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/CACpgx%3DO15D0c8ktiC7SCtK1hZong4UNFVYzTeWK8%3DFx8cf8m6Q%40mail.gmail.com.

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

--

-- 

[ alexei vanyashin typography ]

Avenue de Vaudagne 31

CH-1251 Meyrin 

[Pablo Impallari]

> Is that only because the CYR italics aren't finished yet?

Yes, I guess so...

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CADUrt2FEcNjK95Aomo8epg8fq0mh5ENVATugc_rfgKqCz9Znww%40mail.gmail.com.

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

[Dave Crossland]

That's right.

[Denis Jacquerye]

What do you mean by “all Latin-based languages”?

On Tue, 28 Jun 2016 at 05:12 Alexei Vanyashin <a...@cyreal.org> wrote:

Core: the existing basic (Latin / Cyrillic) naming scheme, which has limited language support. 

1. Latin Plus: extends coverage to Vietnamese and all Latin-based languages. 

2. Latin Pro: adds more currencies, math symbols, and General Use Latin extensions.

3. Latin Expert: Adds Small Caps, Arrows, Ornaments, Fractions, Super- and Subscript. 

Cyrillic follows a similar naming scheme, but for different languages than Vietnamese.

By the end of the day, all files were pushed to git. Check them out here for more info:

https://github.com/alexeiva/fonts/tree/master/tools/encodings/GF%202016%20Glyph%20Sets

 

--

[Andrew Cunningham]

Hi Alexi

On Tuesday, 28 June 2016, Alexei Vanyashin <a...@cyreal.org> wrote:

> Three new levels of character support were developed for each script. 
> Core: the existing basic (Latin / Cyrillic) naming scheme, which has limited language support. 
> 1. Latin Plus: extends coverage to Vietnamese and all Latin-based languages. 
>

Do you actually mean all Latin-based languages?

As far as I know no google sponsored font fits into this category. All fonts I have seen to date I would place in tbe "core" category.

And would require more extensive documentation on what glyphs would be required and which glyphs need to be made accessible via locl and which via cvNN features.

Also you would need to decide if it covers just modern orthographies or historical orthographies.

Such a mapping of usage in all parts of the world would be an extensive exercise in and of itself.

I have been working on some notes on african, se asian, pacifika and australian usage. Starting with African languages. A slow process in my spare time.




--

Andrew Cunningham
lang.s...@gmail.com

[Pablo Impallari]

I guess that by "all Latin-based languages" we mean "all Latin-based languages, as reported by DTL OT Master language checker" tool

--
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/CAGJ7U-UyR7z4_DrC6QAXTCP3QEQ8bv4znZxKSr6H0N3rXvbOnA%40mail.gmail.com.

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

[Andrew Cunningham]

In that case wording should be "selective Latin-based languages" rather than "all Latin-based languages" or some similar phrasing.

On 28 June 2016 at 15:26, Pablo Impallari <impa...@gmail.com> wrote:

I guess that by "all Latin-based languages" we mean "all Latin-based languages, as reported by DTL OT Master language checker" tool

--

Un Abrazo
Pablo Impallari

--

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/CACpgx%3DPyiBGFVK0_J9_96JRY%2BxhAVCac-8WjXGaqMx1ruCt7%2Bg%40mail.gmail.com.

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

--

Andrew Cunningham
lang.s...@gmail.com

[Andrew Cunningham]

or playing devil's advocate, and being more accurate to boot:

"a handful of Latin-based languages" instead of "all Latin-based languages"

--

Andrew Cunningham
lang.s...@gmail.com

[Pablo Impallari]

Andrew, when GF started, the 1st year, only basic Latin-1 was required.

This resulted in tons of users emailing me, on a daily base, asking "please add this glyph", "please I need this other one", etc.

So, for the 2nd year, I started adding all the glyphs previously requested by the users, even when they where not required by GF. For free.
The idea was to avoid having my inbox filled with those "please add X glyph" requests.

Gradually, I've constructed the "requested by users" encoding for my own use.
That worked pretty well for some time, but occasionally someone requested a obscure glyph.

I wanted that kind of emails to stop, so I went to the DTL Language Checker, and made a list of all the missing glyphs to include everything they have listed. This resulted in https://github.com/impallari/Impallari-Fontlab-Encodings/blob/master/Impallari%20Latin/Impallari_Latin_v008.enc

That list may not be perfect, but is much more inclusive that most Pro fonts from the big foundries. It's something in-between AL3 and AL4.
And since I started using that, never ever again got more users emails request for new glyphs.

Also, the served fonts are subsetted for smaller file size and speed, so many of that glyphs that we include in the fonts are ignored anyway.... as many of the OT features are also ignored. That optimized nature of web-fonts, give us little incentive for adding more glyphs, or OT stuff.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAGJ7U-XaSR2VZP_Wu_7AgstDa0uMJepJzB9uFw4DFfKBqhtMZg%40mail.gmail.com.

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

--

Un Abrazo
Pablo Impallari

[Andrew Cunningham]

Hi Pablo,

I understand what you are saying but there are issues around where you expect your font to be used.

DTL Language checker is a primative, lowest common denominator tool.

If you are developing Latin fonts for Europe thst is one consideration. If African another. If Asia another. If Australia anothrr. If Pasifika anothe. Then there are usage requirements in Nth America and Central/Sth America.

The issue is firstly around variant glyphs and variant usages for Latin script around the world.

Re majot foundries, as a rule I find most of their work relatively uaeless for the languages I routinely work with from various corners of the world. They creatr great fonts. But very limited fonts and the major foundaries are the last place I would look for typefaces for the projects I work on.

Ultimately fonts being done via Google will be used on the web and could be used anywhere in tbe world. But as a rule the fonts themselves are not global in nature even if user base may potentially be.

The reality is making a comprehensive latin font is complex, time consuming, and requires resources and information not readily available to most type designers.

And I can only think of one professional type designer I would turn to if I had complex Latin script requirements.

> To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CACpgx%3DMP%3DS%2BT7nF4jHEPYUCtFTQZ_yVXxH5x1BCQ9wdCcNps2A%40mail.gmail.com.

[Frank E. Blokland]

> DTL Language checker is a primative, lowest common denominator tool.

How impolite! We’ll organize a referendum this afternoon whether or
not we stay on this list!

> If Australia anothrr. If Pasifika anothe.

All those foreigners; we don’t understand a word we’re reading!

;-)

F.

> https://groups.google.com/d/msgid/googlefonts-discuss/CAGJ7U-XaSR2VZP_Wu_7AgstDa0uMJepJzB9uFw4DFfKBqhtMZg%40mail.gmail.com
> .

>>> For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
>> --
>> Un Abrazo
>> Pablo Impallari
>>
>> --
>> 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/CACpgx%3DMP%3DS%2BT7nF4jHEPYUCtFTQZ_yVXxH5x1BCQ9wdCcNps2A%40mail.gmail.com
> .

>> For more options, visit https://groups.google.com/d/optout.
>>
>
> --
> Andrew Cunningham
> lang.s...@gmail.com
>
> --

> 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/CAGJ7U-XLf_3wKE6xnnt9ba%3DVdwf-o1nAQMF_d2kq-BdvAvWKpQ%40mail.gmail.com.

[Andrew Cunningham]

Maybe I shouldn't have phrase it that way. It wasn't intended to be derogatory in any sense.

But technically it does cater for a certain set of languages. It isn't extensive and the nature of opentype makes it difficult to develop comprehensive tools. So in a technical sense it deals with a small subset of Latin script languages. Mainly languages of interest to the majority of its clients I assume since comprehensive latin aupport isnt the norm.

So in the sense of coverage ... the mathematical concept of lowest common denominator is an apt analogy.

The algs for determing language coverage is primitive in a techical sense. It is applying a simple technique to a complex question.

I wasnot attempting to malign the product just comment on its approach and comprehensiveness for the Latin script.

A.

> To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/20160628102328.Horde.65svdPLBY_yp0c9fFZh2PZz%40webmail.dutchtypelibrary.com.

[Dave Crossland]

Additionally I'd like to remind everyone to be polite and respectful
here, and that Alexei is not a native English speaker.

A kind word to suggest he could be more precise would be more welcome :)

Alexei's README file does list the languages he thinks each set supports:

https://github.com/alexeiva/fonts/tree/master/tools/encodings/GF%202016%20Glyph%20Sets#google-latin-plus-584-glyphs-total

This sets were developed in collaboration with Kalapi last week, and
started by reviewing Underware Latin Plus, Adobe Latin 4, and the
Impallari glyph set (used in Libre Franklin)

[Dave Crossland]

On 28 June 2016 at 06:23, Frank E. Blokland

<blok...@dutchtypelibrary.com> wrote:
> We’ll organize a referendum this afternoon whether or not we stay on this
> list!

It wouldn't be the first time I expect everyone to vote Remain and get
left holding the bag....

[Richard Fink]

On 28 June 2016 at 06:23, Frank E. Blokland
<blok...@dutchtypelibrary.com> wrote:
> We’ll organize a referendum this afternoon whether or not we stay on this
> list!

This is an American list. Formed by the people in perpetuity.  There is no secession allowed here. 

- rich

--
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/CAEozd0xA0vsVSByqWSRq8YSavZ51MGBrAR7JNzBWgwWRvycwSw%40mail.gmail.com.

[Richard Fink]

On Tue, Jun 28, 2016 at 5:50 AM, Andrew Cunningham <lang.s...@gmail.com> wrote:

>Hi Pablo,

>DTL Language checker is a primitive, lowest common denominator tool.

>Re major foundries, as a rule I find most of their work relatively useless for the languages I routinely work with from various corners of the world.

>They create great fonts. But very limited fonts and the major foundries are the last place I would look for typefaces for the projects I work on.

>The reality is making a comprehensive latin font is complex, time consuming, and requires resources and

>information not readily available to most type designers.

>And I can only think of one professional type designer I would turn to if I had complex Latin script requirements.

Andrew,

I condensed your reply to Pablo so as to focus only on certain statements. I also fixed your typos and/or spelling errors.

To be blunt, when I read your post, my bullshit alarm went off.   Loudly.

Your claims are completely devoid of specific examples. 

What languages do you routinely work with from various corners of the world that you can't find a font for?

What resources and information not readily available to most type designers are you referring to?

And, by implication, why is it that only you and that one other professional type designer can handle complex script requirements?

Are you joking around with us?

Your Linked In profile says you deal in:

"Digital publishing and Web globalisation and internationalisation solutions for minority and lesser used languages."

Well, I would love for you to talk specifics and let us know what edge cases you deal with in your work. And how it is that you're the only person in the world who's figured it all out.😀

Best regards,

Richard Fink

 

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAGJ7U-XLf_3wKE6xnnt9ba%3DVdwf-o1nAQMF_d2kq-BdvAvWKpQ%40mail.gmail.com.

[Dave Crossland]

Rich, I'd like to remind you to be polite and respectful here. :)

[Denis Jacquerye]

Hi Richard,

Andrew did mention African languages and also mentioned in wide terms languages from other continents.

The list is just too long. Where do you want that list to start?

Let’s take one example “edge case”: Hausa, spoken by about 50 million people in West African countries. It is available as a translation language on Google Translate. Several multilingual news sites like BBC World News, Voice Of America and others offer editions in Hausa.

The standard orthography requires Ɓ ɓ, Ɗ ɗ, Ƙ ƙ, R̃ r̃, Ƴ ƴ or ʼY ʼy, ʼ on top of basic Latin.

Very few fonts have those characters. In those that do the hooks are sometimes on the wrong side, too small or too big.

If you look at it the other way. The characters Ŋ ŋ are used in several languages and the uppercase has different preferred shapes according to experts. Most fonts only have one shape for the uppercase. The characters Ɛ ɛ, Ɣ ɣ, Ɩ ɩ, Ɔ ɔ, Ʋ ʋ, Ʊ ʊ, among many others, are very common in African orthographies and other orthographies, they are also often used with diacritics. Only a few fonts support those characters and even less support them with proper diacritic positioning.

Now if Plus, Pro or Expert sets don’t aim to support those, that’s fair. But let’s not confuse all Latin languages with only a fraction of that (I know, this has been clarified), nor edge cases with continental regions.

Cheers

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAGsb3GayXk%2BMj8ixvtD%2BKzCfc8qo5Wa_6549Ea29CV3m3E3FCw%40mail.gmail.com.

[Andrew Cunningham]

Hi Richard,

as Denis indicates there are a very wide range of issues,

My comments were kept general, because this thread, and a mailing list isn't the appropriate venue for a detailed dissection of typographic variation across the Latin script. The key issue is around wording of the levels identified .. and whether "All languages" was appropriate at that point.

But let's make it much more practical. I will restrict myself to the needs of Australian government departments and agencies at Federal, State, Territorial or local levels.

It is common to publish certain types of information in community languages. I may be working with indigenous languages of the Northern Territory which require the N-form of capital Eng (much like the usage in Northern Europe) but I will need the n-form of capital Eng for a range of West and East African languages, the ones of most import currently, as community languages, are Dinka and Nuer, both languages of South Sudan. Occasionally Moro is needed, a language form the Nuba Mountains, and they have a unique D-bar, that is more common with usage in the America's. But D-bar introduces another level of complexity, since Vietnamese support is also required.

Vietnamese languages which are written in the Latin script have a unique typographic approach to diacritic stacking. At the moment I am not just restricting myself to the Vietnamese language per se, but other Latin script Vietnamese ethnic languages as well. Although, it is worth noting that a font suitable for Vietnamese may not be suitable as a regional font across the languages of Vietnam.

The diacritic stacking conventions of the Vietnamese typographic tradition vary with the usage and requirements in Africa where some of the same combinations may be needed in various languages.

So as a minimum two different mechanisms for diacritic placement and sacking are required, one following VN typographic conventions, and one not following them. D-bar also comes into this conversation since the form needed by Moro is different than that used by the Vietnamese languages.

To add Denis' comments:

variant glyphs exist for Ŋ , Ɣ , Ɽ, Đ (as distinct from Ð or Ɖ), ƥ, đ, Ɲ, Ɓ  ... but that is just a series of examples not an exhaustive list.

For serif fonts, in Ɔ and  ɔ, the serif could be on the bottom or the top depending on typographic preferences of the language in question

If you look at handwriting and how it might inform an italic face there are interesting adoptions. I find the the open-o in handwritten sources form Sudan and South Sudan quite interesting, looking more like a rounded version of R rotunda

But back to more practical examples. For government publications it is useful to retain a consistency in typeface across communications. Completely changing the Latin typeface from one language to the next is less than ideal. So fonts capable of supporting both African, European and Vietnamese requirements are preferable.

Regarding spelling in my previous emails, and thus a very personal comment, I was responding on a mobile phone. I have a medical condition I am receiving treatment for which impacts on my typing on small touch devices. I should have waited till I was at a computer to send the email, but was away from computers for an extended period.

A.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAGsb3GayXk%2BMj8ixvtD%2BKzCfc8qo5Wa_6549Ea29CV3m3E3FCw%40mail.gmail.com.

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

--

Andrew Cunningham
lang.s...@gmail.com

[Eben Sorkin]

Thank for the further and better particulars!

If you find that you somehow have time to create any sort of guide to solving one or more of the problems you touch on here that would be great because many type designers would like to support more people and cultures than they do now. There kinds of documents can have a huge impact on what’s available.

-e.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/CAGJ7U-Va1s3RBasqgZYPYB%2BJ2Cqjfp5NUF2sG%3De-v0zHJZZXmQ%40mail.gmail.com.

[Dave Crossland]

Thanks for the detailed explanation Denis and Andrew :)

I note that initially many Google Fonts only supported 220 glyphs for Western European languages and in the last 3 years mostly Adobe Latin 3 with some Adobe Latin 4.

The development of these sets started with AL4 and added in others from Underware Plus and Impallari Latin, and they haven't been 📌 down :)

I asked for transliteration and Asian and African Latin to be intentionally excluded, because with 100s of families to process from 220 glyphs upwards it isn't possible to make a change across the collection if the sets include them.

I can see exceptions might be possible  where a locale could be supported for a very low marginal cost of effort.

And of course I expect to come back and further extend the fonts in future years :)

[Andrew Cunningham]

Hi Eben,

I am aware of the need for supporting documentation, and intend to pull something together. But it is a huge , time consuming task. And for African languages will be using the networks I have in place to pull together as much as we can find. No time scale available. This type of work was stuff i was funded for while my previous employer was still interested in community languages, but that I am continuing with the limited resources i have at hand.

Similar documentation is needed for other scripts , not just Latin. Myanmar is a huge area that needs serious documentation on glyph variants and also positioning of characters (which can change form language to language). There is no Myanmar font adequately covers all Myanmar script languages.

Newari and Nepali variations in Devanagri script.

The various minority languages in Ethiopia that sometimes have glyph variations. And more complex if you add historical usage by scripts that now use Latin script.

We are working on a Unicode proposal to disambiguate Eastern and Western Cham, and still trying to decide what to do about Eastern Cham before orthographic/spelling normalisation and reform occurred.

Also trying to document Cham Arabic script orthographies in Cambodia and Vietnam.

Along with trying to figure out what to do with the huge variation seen in some of the west African scripts like Mende Kikakui, Vai and others.

And the list goes on ...

There are some resources for Latin that are good, but not complete.

We will do what we can.

A.

To view this discussion on the web visit https://groups.google.com/d/msgid/googlefonts-discuss/2CD884F8-1ABC-4417-A05A-B994086DC7F1%40eyebytes.com.

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

--

Andrew Cunningham
lang.s...@gmail.com

[Andrew Cunningham]

Hi Dave

On 29 June 2016 at 12:25, Dave Crossland <da...@lab6.com> wrote:

Thanks for the detailed explanation Denis and Andrew :)

I note that initially many Google Fonts only supported 220 glyphs for Western European languages and in the last 3 years mostly Adobe Latin 3 with some Adobe Latin 4.

The development of these sets started with AL4 and added in others from Underware Plus and Impallari Latin, and they haven't been 📌 down :)

I asked for transliteration and Asian and African Latin to be intentionally excluded, because with 100s of families to process from 220 glyphs upwards it isn't possible to make a change across the collection if the sets include them.

This is perfectly reasonable, and should be documented up front. It makes life simpler.

I assume South, Central and North American (non-English, non-Portuguese and non-Spanish) usage is also included in that exclusion?

I can see exceptions might be possible  where a locale could be supported for a very low marginal cost of effort.

And of course I expect to come back and further extend the fonts in future years :)

actually I see little need for extending many of the fonts. It would be easier and sufficient to just have one or two that are comprehensive and flexible.

 

[Eben Sorkin]

On Jun 28, 2016, at 11:14 PM, Andrew Cunningham <lang.s...@gmail.com> wrote:

actually I see little need for extending many of the fonts. It would be easier and sufficient to just have one or two that are comprehensive and flexible.

I get that having just one or two super capable examples would solve a lot of problems in one fell swoop. And it makes sense that that would seem attractive. But I would argue that calling that ‘sufficient' is myopic or short sighted. 

I am not saying that it would not be worthwhile to have the fonts you describe just that giving people the tools to make improvements and help more people can have a simply huge impact. Adam’s Twardoch’s article about the needs of the Polish language has probably had an impact on thousands of typefaces by now.

This kind of impact is worthwhile because it isn’t always enough to communicate clearly. Sometimes to communicate effectively you need a particular feeling or set of cultural associations. This may not matter as much in engineering, public policy, law or medicine but it is very important for commerce, human expression, and even to help give a sense of validity or a feeling of belonging or being real worthwhile or authentic for people whose languages are not supported.

-e.

 

--
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/CAGJ7U-XgNkdJKo1-Rdbd0bz%3D3C0G-qwrU8i8cPwKQroyxkF6XQ%40mail.gmail.com.

[Alexei Vanyashin]

Andrew, Denis: 

my mistake, I did mean that the Google Latin Plus 

set supports all Latin-based languages *which exist in the Latin Expert set*.

In other words wide Latin language support starts from Latin Plus, 

while further sets add mostly math symbols, fractions, small caps, and the like.  

Thank you for your insightful details on the languages. 

I can say that Pinyin and (Arabic) transliteration sets is something we excluded intentionally 

to limit the scope.

Now, If I may, I would like to continue posting my daily update. 

¶ #2. Warm-up, week 1, 2016-28-06

Updated Encoding project: switched from uni to nice names.

Started working on improving [or cleaning-up] the Jura font project. Migrated sources from SFD to UFO.

Original source files for the 4 weights were done without interpolation. Starting from Light, 

Bolder weight were added by offsetting the Light, and so on... Source files are in TT format.

As a result, no two masters are compatible, and the proportions and widths of characters are

inconsistent across styles. The charset is different in the 4 styles too.

Instead of fixing all this, I decided to work from scratch, and put the original file in the background for reference. 

By the end of the day I completed a single master in basic Latin a-z, and used Glyphs App filters (offset, round corners)

to generate 8 new styles programmatically. This was done as an exploration to see 

how far a single hairline master can be expanded. The Bolder weights of course require manual reworking. 

In overall the new approach seams viable, and much faster than fixing the master compatibility for each glyph.

[Richard Fink]

Andrew, Denis,

Thank you both for your explanations.

Andrew - I, too, have a disability called wet macular degeneration in my right eye which makes it useless for reading or keyboarding, all the work is done by my left eye, so I understand. And if your lack of details led me to an incorrect assessment, I happily apologize for any rudeness. There are trolls lurking out there everywhere, maybe I've gotten too paranoid. 😵  

If I might remind everybody - nobody's ever been in the position we find ourselves in today. On the web, technically, we can support any and all writing systems instantly. What we don't have yet is the knowledge needed - knowledge codified and understandable by ordinary people - to do it as a practice.

I have an article about the state of webfonts publishing on A List Apart in a few weeks. I'm working on a follow up article now which focuses specifically on multi-script/muli-lingual webfonts that will publish probably in August sometime.

I can use some readers to check my facts and comment on early drafts so if any of you hear from me shortly, I hope you'll oblige me. 

Regards,

rich

--

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/ce44f9b4-afee-4917-b201-9206e321f267%40googlegroups.com.

[Dave Crossland]

On 29 June 2016 at 00:08, Alexei Vanyashin <a...@cyreal.org> wrote:

In overall the new approach seams viable, and much faster than fixing the master compatibility for each glyph.

I agree - I expect this could be needed for Quicksand and Nunito

[Dave Crossland]

On 28 June 2016 at 23:14, Andrew Cunningham <lang.s...@gmail.com> wrote:

This is perfectly reasonable, and should be documented up front. It makes life simpler.

I assume South, Central and North American (non-English, non-Portuguese and non-Spanish) usage is also included in that exclusion?

Yes, where they need glyphs to be drawn; however, if a locale can be made using only anchors/components, it has a very low marginal cost of development - Glyphs App can construct them automatically and the designer can review and quickly reposition them as needed - so I'm open to including them. 

I expect Kalapi and Alexei to provide some insight into which glyphs require drawing and which can be constructed semi-automatically in the coming days

[Dave Crossland]

On 28 June 2016 at 23:57, Eben Sorkin <eb...@eyebytes.com> wrote:
> I would argue that calling that ‘sufficient' is myopic or short sighted.
>
> I am not saying that it would not be worthwhile to have the fonts you
> describe just that giving people the tools to make improvements and help
> more people can have a simply huge impact. Adam’s Twardoch’s article about
> the needs of the Polish language has probably had an impact on thousands of
> typefaces by now.
>
> This kind of impact is worthwhile because it isn’t always enough to
> communicate clearly. Sometimes to communicate effectively you need a
> particular feeling or set of cultural associations. This may not matter as
> much in engineering, public policy, law or medicine but it is very important
> for commerce, human expression, and even to help give a sense of validity or
> a feeling of belonging or being real worthwhile or authentic for people
> whose languages are not supported.

I agree; its not going to be much fun making slide decks and school
reports if you speak Hausa and use Boko -
https://en.wikipedia.org/wiki/Boko_alphabet - if you only have Noto
Sans and Noto Serif to choose from.

Rather what I would expect Hausa users to need is support for say the
top X fonts in the Google Fonts collection.

[Jacques Le Bailly]

Alexei,

I noticed that caronalt is not working to create lcaron or dcaron.

I tried caroncomb.alt and it works. How do you want to proceed ?

[Kalapi Gajjar-Bordawekar]

Hello Jacques,

Nice catch! When you say that 'caronalt' does not work, are you referring to Glyphs not auto-generating feature code?

In any case, the glyph name should be updated to 'caroncomb.alt' in the .nam files (as you've pointed out). Can you open an issue here so that it can be recorded and tracked? 

Best, 

Kalapi

[Alexei Vanyashin]

Jacques, 

Thank you for spotting that, we will track this issue.

¶ #3. Warm-up, week 1, 2016-29-06

AM Thanks to feedback from Jacques, and Thomas 

I adjusted spacing and revised the design concept. 

Later I set up a proper Unified-Font-Repository for this project, 

which is live at https://github.com/alexeiva/Jura

It includes documentation and project roadmaps. 

PM Some fixes on the 2016 Encoding.

When starting a new project one must keep

the end-goal in mind. My end-goal is seeing the new font 

in web, so I created a web test page where I can check how the font performs in all styles. 

As a starting point for the webpage I cloned the magnificent Wei Huang's Work Sans Specimen, 

which in turn is derived from Claus Sorensen's Inknut Antiqua Specimen. 

So everything is a copy of a copy.

I only have the basic a-z glyphs to show for all weights, 

here they are live at https://alexeiva.github.io/jura/

Technically a folder called webfontspecimen inside the main github font repo

is deployed to the gh-pages branch via 'git subtree push --prefix webfontspecimen origin gh-pages'.

To further automate the process I plan to add a git hook that would fire up on every commit.

I think this practice might be useful for other open-source projects. 

So today no new glyphs were designed, the day was entirely devoted to various technicalities, 

and setups. As the saying goes, Russians are slow to saddle up, but ride fast (c) Bismarck.

[Dave Crossland]

On 29 June 2016 at 23:58, Alexei Vanyashin <a...@cyreal.org> wrote:
> Technically a folder called webfontspecimen inside the main github font repo
> is deployed to the gh-pages branch via 'git subtree push --prefix
> webfontspecimen origin gh-pages'.
> To further automate the process I plan to add a git hook that would fire up
> on every commit.

Nice! Please document in the project readme :)

[Frank E. Blokland]

Okay, we stay then. ;-)

We have been discussing the option to make the language-check files
also user-definable in OTM. I’m not sure whether that will satisfy you
completely, because of your comment concerning the
primitive-in-a-technical-sense algorithm. In any case, the wishlist is
pretty long though, so it could take some time before support for
user-definable files will be available.

> https://groups.google.com/d/msgid/googlefonts-discuss/CAGJ7U-XLf_3wKE6xnnt9ba%3DVdwf-o1nAQMF_d2kq-BdvAvWKpQ%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 https://groups.google.com/group/googlefonts-discuss.
>> To view this discussion on the web visit

> https://groups.google.com/d/msgid/googlefonts-discuss/20160628102328.Horde.65svdPLBY_yp0c9fFZh2PZz%40webmail.dutchtypelibrary.com

> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>
> --
> Andrew Cunningham
> lang.s...@gmail.com
>
> --
> 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/CAGJ7U-Vmaz0Z_wD2X5cBmtYu0qpOOumN%2BdftcixtoWjSZqpb-w%40mail.gmail.com.

[Alexei Vanyashin]

¶ #4. Warm-up, week 1, 2016-30-06

AM: Started with a 2 hour team conference call, 

where I presented the Jura redesign roadmap: https://github.com/alexeiva/jura/blob/master/documentation/Jura%20Redesign.pdf

Many hotfixes to GF 2016 Encoding:

—fixed Glyphs compatibility issues (caronalt > caroncomb.alt)

—fixed typo in glyphs filter lists

—wrote troubleshooting guide on unicode clash errors in Glyphs.

Web Specimen page(https://alexeiva.github.io/jura/): 

—added text resize and spacing sliders. 

Project planning: 

—revised list of Cyrillic fonts. 

—Added list of glyphs script that may be helpful for production.

I did. You will see it in my next commit. 

[Dave Crossland]

On 30 June 2016 at 23:37, Alexei Vanyashin <a...@cyreal.org> wrote:
> —wrote troubleshooting guide on unicode clash errors in Glyphs.

Looking forward to this - where is it? :D

[Alexei Vanyashin]

Here:

https://github.com/google/fonts/blob/GF2016-fix/tools/encodings/GF%202016%20Glyph%20Sets/TROUBLESHOOTING.md 

[Alexei Vanyashin]

A Quick update on the 2016 Encodings.

The latest files are on this 'GF2016 branch':

https://github.com/google/fonts/tree/GF2016-fix/tools/encodings/GF%202016%20Glyph%20Sets

Recent changes:

Latin Plus:

• moved germanbls.alt to recommended list
• removed CR, and NULL

Latin Pro set:

• added emptyspace ∅ (math)

Other:

• Added Glyphs List Filter Troubleshooting Guide
• Added List Filter for Vietnamese unique glyphs

[Alexei Vanyashin]

¶ #5. Warm-up, week 1, 2016-1-07

AM: Again, started with a 2 hour team conference call,

where issues with encodings and other technicalities were discussed. 

 

Right after that, I resolved them and posted in the new branch, see in my previous e-mail for a link. 

PM: I uploaded a new personal project to git:

https://github.com/alexeiva/cyr-o-pedia

It is a fork of https://github.com/huertatipografica/devanaguide

With additional functions and support for Cyrillic. 

Next up: Working on a Generic ‘Donor’ font with Math Symbols, Ornaments and such, 

which my colleagues can reuse for extending other font’s charsets.

The glyphs will be made as easily-customizable smart components.

That’s all. Have a nice weekend, 

[Dave Crossland]

Thank you Alexei! :)

[Alexei Vanyashin]

About the Cyrillic unicode visualisation I mentioned yesterday:

It's now running on heroku, thank very much for this suggestion, Dave!

Now, it's live at: https://cyrill-o-pedia.herokuapp.com/index.php

On Friday, July 1, 2016 at 11:14:56 PM UTC-4, Dave Crossland wrote:

Thank you Alexei! :)

[Alexei Vanyashin]

¶ #6. Sprint 1, week 2, 2016-04-07

1. Today I pushed the first version of MathFont on Git.

2. Tomorrow I will look for ways to improve MathFont,

make Plus, Pro, Expert lists for a Latin version of Cyrill-o-pedia

3. Questions. How can the MathFont be improved?

[Daniel Johnson]

Hi, original creator of Jura font here.

It has occurred to me to mention one philosophical tidbit that guided me when I was first creating Jura. It all started with the Kayah Li range which I created for FreeMono (whose Latin range is a clone of Courier). Insofar as was possible, I tried to use curves copy/pasted from the glyphs in that range (though, because Kayah Li does not have full diagonal strokes, I could not use this idea throughout). This is particularly apparent in the upper right corner of "g" or the upper left corner of "p", and also explains the curved terminals on the "z". It also explains the rather sharpish corners of the tail of the "g" and the "y". I think the initial redesign as shown on slide 19 of the "Jura Redesign" deck loses some of that in the heavier weights.

As an aside, I believe if you compare the stroke width of my original Light weight, it should match FreeMono exactly.

[Alexei Vanyashin]

Hi Daniel, 

Thank you very much for your feedback. I have looked into FreeMono. I understand where you are coming from.

This definitely clarifies many design-decisions for me. I will keep the new design closer to the sources. 

This week I have to put Jura aside in favour of quick-fix projects.

I will get back to drawing Jura from next week.

Until soon, 

[Alexei Vanyashin]

¶ #6. Sprint 1, week 2, 2016-05-07

1. Today.

a. Admin: Call to sync up with group. Collect feedback on MathFont, etc.

b. Tech. Further Encoding refinements:

— Moved all .case glyphs to an Optional list, since it heavily depends on the design. 

c. Planning. Picked Cyrillic families to work on, as 'Raisins', and 'quick-cleanup'.

d. Design. Started working on BadScript, 1 day project. Goal is to extend to Cyrillic Plus.

— took time off to develop a Custom GlyphsData.xml for Cyrillic Plus, which will 

speed up generating new glyphs from components in all new fonts. 

2. Tomorrow.

— Fix missing glyphs for Vietnamese support.

— Finish Badscript improvements. 

3. Any Questions?

[Alexei Vanyashin]

¶ #6. Sprint 1, week 2, July 6, 2016

1. Today:

Finished 1-day improvement project on Badscript font

— Added Cyrillic Characters for Google Cyrillic Plus support

— Reshaped odd-looking glyphs

— Improved Spacing

— Added core kerning pairs

[ Github link ]

2. Tomorrow:

— Fix Encoding Issues

3. No questions so far

[Alexei Vanyashin]

I just tried the new fontdiff for Badscript. 

It's helpful.

[Dave Crossland]

On 7 July 2016 at 12:09, Alexei Vanyashin <a...@cyreal.org> wrote:

I just tried the new fontdiff for Badscript. 

It's helpful.

Great!

This is available from https://github.com/googlei18n/fontdiff

There's also http://graphicore.github.io/charset-inspector :) 

[Alexei Vanyashin]

Hey Pablo, 

At the moment we are exploring options for Cyrillic Raisin projects, 

so I have examined Raleway's Cyrillic charset. 

Green is for existing characters, Orange — Cyrillic Plus, Red — Cyrillic Pro, 

and Purple — Cyrillic Expert (which includes Small Caps).

I think it is viable to have Plus support level, and additionally Small Caps from the expert set, 

since they are present in Latin. 

The existing Cyrillic will need some fine-tuning, and spacing adjustments (It is too tight at the moment).

Altogether, I think this would be a great project to work on. 

[Pablo Impallari]

Great!

--
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/5f01a64a-b8af-45df-affa-4ae3e37d1bd5%40googlegroups.com.

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

[Alexei Vanyashin]

¶ #6. Sprint 1, week 2, July 7, 2016

1. Today:

I started working on Poiret One with the blessing of it's author Denis Masharov. 

Denis is keen on keeping track of my progress, and asked for a final review before pushing the update live. 

So here I am posting today's results.

My main goal is to extend Poiret into the Google Cyrillic Plus set. I started with anchors placement for Cyrillic glyphs:

'

Then I made 'spare parts' for making smart component glyphs. 

The Ҵ glyph is semantically a conjunction of T and Ц.

I would want to use Ц as a base character and add just the top horizontal

bar of T, but not the whole glyph. 

For this purpose I created a smart component glyph named _part.topbar-cy, 

and placed a _center anchor in its middle.

In Ц I also added a center anchor at the top left corner. 

Now I can dynamically control the width of the horizontal bar

using the smart component slider:

This component will be handy in many other Cyrillic glyphs, such as Ҡ,ҡ.

For further automation I edited my local copy of GlyphsData.xml 

to include the center anchor automatically in Ц.

If you want to use this file for your workflow, you can download it from 

the google/fonts repo.

Along the way, I spotted a few stroke inconsistencies, which I slightly tweaked, 

preserving the original character.

Here is an example:

By the end of day one, I was able to cover most of the planned encoding range:

The latest work-in-progress files for this project are on my github.

2. Tomorrow I will continue on Poiret One.

3. No questions so far. 

[Alexei Vanyashin]

¶ #6. Sprint 1, week 2, July 8, 2016

1. Today:

[Admin] Started with an admin call to sync up on the clean-ups checklist, 

and we discussed problems with vertical metrics exceeding the bbox, 

in SigmarOne.

[Design] Later I continued my work on Poiret One, and as per second day, 

completed the full Cyrillic Plus encoding with alternate localised forms for 

Bulgarian, Serbian/Macedonian, Chuvash, and Bashkir languages.

The green glyphs are newly designed:

 

Here is my favourite Cyrillic glyph: the Abkhasian Ha (U+04A8).

Recently it's design guidelines have changed. According to the new rules, 

the bowls need to be very similar in height. Such is the case in Poiret One:

N.B. And here is bit of my design process. I start with a line, 

and offset the stroke for best results. 

Arial, on the other hand, still follows the old (outdated) convention

While Charter is up-to-date.

Now, here is the lineup of Cyrillic localised forms. 

The ones with red circles are not exported, 

because they are unnecessary in this design context. 

For example the triangular version of Л is customary 

for Bulgarian-style, but in our case the standard design is identical.

Similar for Д. Alternates for Н, and И(generally) are required for seriffed designs.

All in all, this is a question of stylistics, that should be resolved on a case to case basis.

The Bulgarian style of Cormorant goes another step further, with alternates for /М

introduces. This isn't normal practise, but it does work better in the given context.

If you are curious on Bulgarian-style, you will find this thread interesting.

Having filled the encoding I started revising the basic shapes.

There was a lot of work in almost every glyph: 

— fixing the contrast

— balancing the shape

— readjusting related component glyphs, (new anchors, component glyphs, etc.)

— respacing glyphs

I did this process for both Cyrillic, and Latin, in total touching all 664 glyphs.

My previously made characters were updated automatically

with this revision, because their bases-glyphs are components. 

Denis (Original Author) was happy with the improvements I showed on З(ze-cyrillic), 

as an example, and granted me full rights to fix whatever I can. Great!

Next-up kerning time!

The original file has 2027 pairs, but no classes. Ouch!

I tried to compress kerning in Glyphs, but the results were hideous:

some weird pairs like H A -200 popped up. 

My next step was to create kerning classes for 664 glyphs. 

But wait. TAAFT. There's an (app) script for that! 

Results of which need to be tailored to your font.

I do this manually, but along the way, 

I am working on an update of this script with more 

accurate values for Cyrillic classes.

By the time of writing, I pushed a commit with a backup of my current work.

Before the final push I need to revisit the kerning and check how it behaves in 

new glyphs...

...Or, maybe get some sleep?

[Alexei Vanyashin]

Update:

The kerning is now fully reworked.

• Total pairs uncompressed: 2046
• Total pairs after compression auto+manual: 403
  
• Total pairs in Release v.1.1: 532
  

This concludes the 2.5 day font cleanup project.

Latest files: Poiret One v.1.1

Changelog:

• Converted to Glyphs format
  
• Expanded Charset to GF Cyrillic Plus
  
• Added Tabular figures
  
• Major design and spacing improvements
  
• Rechecked every glyph for stroke weight consistency
  
• Fully reworked kerning, added manual pairs. 
  

 

[Dave Crossland]

On 9 July 2016 at 17:22, Alexei Vanyashin <a...@cyreal.org> wrote:

This concludes the 2.5 day font cleanup project.

Did you work on the Latin also? I guess that perhaps it needs another 1-2.5 days

[Alexei Vanyashin]

I did. With regard to the limited time constraints.

In Latin I was able to rebuild the accents, and fix the stroke inconsistency.

For this extra time, I could make a 'raisin' from this. 

Denis promised to send latest improved Bold version 

for reference on Monday. 

--
You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/CAEozd0ygvbVnLXGicfkfKQmbkSVEmKppwszm_Pf66kZrcopNSg%40mail.gmail.com.

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

--

-- 

[ alexei vanyashin typography ]

Avenue de Vaudagne 31

CH-1251 Meyrin 

[Dave Crossland]

On 9 July 2016 at 18:37, Alexei Vanyashin <a...@cyreal.org> wrote:

I did. With regard to the limited time constraints.

In Latin I was able to rebuild the accents, and fix the stroke inconsistency.

For this extra time, I could make a 'raisin' from this. 

Ah, okay great! When you say 'rebuild' do you mean that your .glyphs file now supports Latin to the Plus set?

[Alexei Vanyashin]

I rebuilt the components of existing glyphs using anchors. The source file was a decomposed TTF.

Latin Plus support was far beyond my 2-day time-box.

Manual kerning was also something beyond this time limit. 

However for the extra time (+2 days), I would be happy to extend Poiret One to Latin Plus.  

>Did you work on the Latin also? I guess that perhaps it needs another 1-2.5 days

Do you mean for expansion, or are there any other quality concerns with Latin?

I tried too fix them in this batch.

 

[Dave Crossland]

On 9 July 2016 at 19:49, Alexei Vanyashin <a...@cyreal.org> wrote:

However for the extra time (+2 days), I would be happy to extend Poiret One to Latin Plus.  

>Did you work on the Latin also? I guess that perhaps it needs another 1-2.5 days

Do you mean for expansion, or are there any other quality concerns with Latin?

I tried too fix them in this batch.

I think spending another 2 days would be great, its a very popular single-style font and expanding the Latin to Latin Plus to go with the Cyrillic expansion would be great. 

[Alexei Vanyashin]

Alright! I'm on it. 

Here is the current situation: Gray = todo

--

You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/CAEozd0xHn-M09Hpf4AndCpmgrw6ZLiQ_yidNrTXWKbLhcx8Gig%40mail.gmail.com.

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

[Alexei Vanyashin]

¶ #6. Sprint 1, week 3, July 11, 2016

1. Today:

a. [tech] I made a PR with the following changes on the GF Encoding:

• Added missing Vietnamese characters (Thanks Jacques)
• Added .sinf glyphs to Pro set (Thanks Pablo, #316)

I have researched this problem, there are two possible solutions for Scientific Inferiors, and Subscript Figures:.

1. They can be identical — Easy method

2. They can deviate. This is the recommended approach, which I have

described in details in the new tutorials section. 

Here are the differences.

• Added foursuperior (Thanks Kalapi)
• Added tutorials section with tips to speed up the charset expansion process

b. [design]

Poiret One — Extension to Latin Plus:

I completed all 'drawn' type glyphs, which require the most time,

including currencies. 

German Capital Sharp S, and its neighbour shapes:

Drawn glyphs, and currencies:

The Bulgarian-style ve is much wider than the international-style one. 

This is a result of applying the general idea of wide round shapes, 

versus heavily compressed squares. 

I think this will be one of the peculiarities of this art-deco typeface.

Here is the lineup of the symbols category. 

The existing shapes @, &, and some math signs, 

might need facelifting. Is there anything that pops out?

2. Tomorrow: Continue work on Poiret One

What needs to be done:

— Lining Figures, and Superiors

— Vietnamese accents

— Generating composite glyphs

3. Questions:

 Any thoughts about the changes in Encoding?

[Dave Crossland]

On 12 July 2016 at 00:02, Alexei Vanyashin <a...@cyreal.org> wrote:

 Any thoughts about the changes in Encoding?

Sounds good to me!

[Alexei Vanyashin]

PS Image of overall progress on Poiret One, June 11, 2016

 

[Alexei Vanyashin]

I have a suggestion to add these glyphs in the Plus Set.

Glyphs App will do the magic then.

- IJacute.loclNLD

- ijacute.loclNLD

 

Any thought on this? 

[Alexei Vanyashin]

¶ #6. Sprint 1, week 3, July 12, 2016

1. Today:

Elaborated on issue #316, 1

Made a PR on Lobster to add some glyphs that 

where are new in the GF encoding update. 

Poiret One, day 4: Completed GF Latin Plus:

Now I am waiting for the Vietnamese part to be reviewed by Nhung, 

While working on Vietnamese I found the alternate `narrow` 

accents feature handy.

In Glyphs, create a new accent glyph with the .i suffix, 

and it will automatically be places for narrow glyphs like i.

In my case I created circumflexcomb.i, acutecomb.i, among others, 

and used them where accents where tight. 

On this image. Left: original circumflex accent. 

Right: Its narrow version.

2. Tomorrow:

Help Thomas with OT features rebuilding problem due to the naming scheme switch(from afiiXXX to nice-names) in latest Lobster version. 

Fix any issues with Vietnamese.

3. Questions:

I am looking for some input on issue #316.

[Alexei Vanyashin]

¶ #6. Sprint 1, week 3, July 13, 2016

1. Today:

Mainly I was nitpicking and finalising Poiret One.

— fixed issues with Vietnamese (Thanks Nhung!)

— Fine-tuned symbols, punctuation marks. 

— respahed some characters like German Sharp S, etc.

— added extra narrow accents for i-based glyphs

— semi-manually added kerning classes to all new glyphs from GF Latin Plus

— rechecked kerning using Pablo Impallari's Wonderful Font Testing Page

— ran preflight scripts for checking

And, finally pushed version 1.1. on github.

Also examined Lobster OT features, and made a small fix for Thomas. 

That wraps up the 5 day process on Poiret One. 

(which ironically started as a 2 day project).

[Dave Crossland]

On 13 July 2016 at 21:08, Alexei Vanyashin <a...@cyreal.org> wrote:

— fixed issues with Vietnamese (Thanks Nhung!)

Where were these issues discussed? :) 

 

That wraps up the 5 day process on Poiret One. 

(which ironically started as a 2 day project).

(All good! It was a 2 day Cyrillic project, and a 2 day Latin project, plus a day for other activities :)  

[Alexei Vanyashin]

Issue was discussed here:
https://github.com/alexeiva/poiretone/issues/1

--

You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/CAEozd0xqPi4qjF33G8uOF%2Bs29nmHS5PBCpGe%2BhukJYm1v4bp4Q%40mail.gmail.com.

[Dave Crossland]

On 13 July 2016 at 23:12, Alexei Vanyashin <a...@cyreal.org> wrote:

Issue was discussed here:
https://github.com/alexeiva/poiretone/issues/1

Perfect!

[Alexei Vanyashin]

Today I am resuming work on Jura, and I will be posting my daily updates in it's own thread.

[Omer Ziv]

Thanks for all these updates Alexei - they're a good read.

Please add more screenshots for those of us who don't visit github that often ;)

On Tuesday, June 28, 2016 at 12:12:13 AM UTC-4, Alexei Vanyashin wrote:

Hi everyone!

I am happy to start working on the 'font quality improvement' project, 

dubbed 'Making Raisins'.

This thread will be used to post daily, or weekly updates on progress

for self-organisation and record-tracking purposes, and by no means spamming. 

If this amount of traffic will be overwhelming for you, feel free to mute the thread, 

or enable weekly summaries. 

My involvement will consist of 3 sprints of 4 week each, 

during which I will be touching different font projects.

My list of scheduled projects, and progress timeline can be tracked anytime

via this link:

https://github.com/google/fonts/issues/285

Let's start!

# 1. Warm-up, week 1, 2016-27-06

Today I summarised our teams work on Google Fonts new 2016 Latin and Cyrillic Encodings. 

Three new levels of character support were developed for each script. 

Core: the existing basic (Latin / Cyrillic) naming scheme, which has limited language support. 

1. Latin Plus: extends coverage to Vietnamese and all Latin-based languages. 

2. Latin Pro: adds more currencies, math symbols, and General Use Latin extensions.

3. Latin Expert: Adds Small Caps, Arrows, Ornaments, Fractions, Super- and Subscript. 

Cyrillic follows a similar naming scheme, but for different languages than Vietnamese.

By the end of the day, all files were pushed to git. Check them out here for more info:

https://github.com/alexeiva/fonts/tree/master/tools/encodings/GF%202016%20Glyph%20Sets

[Alexei Vanyashin]

On Friday, July 15, 2016 at 10:56:30 AM UTC-4, Omer Ziv wrote:

Thanks for all these updates Alexei - they're a good read.

Thank you, Omer,  

Please add more screenshots for those of us who don't visit github that often ;)

Absolutely, I will do that!

 

[ken p]

् ा ि ी ु ू ॅ े ै ॉ ो ौ ़ ं ः

a̩  ä  i    ï  u  ü ă à  ȁ  ŏ  o  ȁ: å aː 

I need extended copy able,paste able,convertible Unicode Latin letters with diacritics to write Indian languages in Roman script.

[Alexei Vanyashin]

Hi team!

I created a Glyph's script that automatically sets kerning classes for the GF Latin Plus encoding. 

Download it here.

It is a modification of Georg Seifert's script. The classes were created specifically for this encoding from scratch, fully reworking the original version. I am currently adding classes for punctuation, so I will update this script later today. 

It works for all glyphs with no kerning class. If you have a class assigned already, it will produce no action. 

-a

[Alexei Vanyashin]

Okay, the script is ready, and officially released in a separate repo here:

https://github.com/alexeiva/alexei-glyphs-scripts

There were some drawbacks however. I wanted i and idotless to have different classes for situations like this:

But due to the nature of Glyphs App or the rest of the code of this script, i is inheriting all the classes from idotless.

Ideally I wanted idotless to share the left class with /n. For now I don't have the time and knowledge to resolve this. This can be easily fixed by updating the classes in the font. 

Here is how I split the kerning classes for punctuation. Each color represents a separate class. White means the class is unique. 

To achieve this I setup the classes manually in the font, and exported this data to CVS. 

From that point I made a regex to feed this data into the script. 

I hope this script will be useful for the team to build classes fast. Note that it needs to be customised. For example if the /g is a single-story design, its classes can be changed to /o, /u. The default value is /g,  /g for a double story scenario. 

-a

[Alexei Vanyashin]

But due to the nature of Glyphs App or the rest of the code of this script, i is inheriting all the classes from idotless.

I found the answer to this. [Git discussion]

1. If Hbar is decomposed is gets the value from the script, i.e. Hbar

2. If Hbar is composed, it's inherits the value from H

 I wonder how I can have the script do option 1 without decomposing glyphs.

[Alexei Vanyashin]

I wonder how I can have the script do option 1 without decomposing glyphs

Update: the script no longer auto-assigns keys from existing components. Instead all keys are generated from the list. Thanks to Georg for the tip.

[Alexei Vanyashin]

¶ #6. Sprint 2, week 6, August 1, 2016

Yesterday[during my 5 hour flight] I resumed work on Badscript One with the goal of expanding into the GF Latin Plus set.

There were many problems with outlines: they require optimisation. For this I used the 'Delete Nodes and Try to Keep shape' script. This process localising a group of nodes, and running it selectively. Unlike the simplify path function from FontForge, this involved manual actions for each glyph individually instead of running as a batch for the whole font. 

Before, after

I completed a large chunk of GF Latin Plus, starting with accents. 

It's not too bad!

Today (Tuesday): Complete GF Latin Plus encoding, finalise font. 

[Dave Crossland]

On 2 August 2016 at 15:43, Alexei Vanyashin <a...@cyreal.org> wrote:

Yesterday[during my 5 hour flight] I resumed work on Badscript One with the goal of expanding into the GF Latin Plus set.

There were many problems with outlines: they require optimisation. For this I used the 'Delete Nodes and Try to Keep shape' script. This process localising a group of nodes, and running it selectively. Unlike the simplify path function from FontForge, this involved manual actions for each glyph individually instead of running as a batch for the whole font. 

Sounds good. Script is simple - https://github.com/mekkablue/Glyphs-Scripts/blob/master/Paths/Delete%20Nodes%20and%20Try%20to%20Keep%20Shape.py - seems the 'removeNodeCheckKeepShape_' method is a proprietary one of Glyphs. 

[Alexei Vanyashin]

I think the quality is better than FF's simplify nodes, but than the required manual time equates. 

A ported version of FFs 'simplify' might be useful for Glyphs, but I couldn't find one. 

--
You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/CAEozd0wqnApwROCmueNL9wtH0dNGiD3KiB7U_%3DNSPn5d4bNKiQ%40mail.gmail.com.

[Alexei Vanyashin]

¶ Sprint 2, week 6, August 2, 2016

1. Today I completed the whole GF Latin Plus set. For this project kerning classes were set manually. This is a case were my script wouldn't be helpful. 

2. Tomorrow will start work on Marmelad expansion. 

3. Question: Nhung, could you have a look at Badscript Vietnamese?

-a

[Nhung Nguyen]

I think they're good already. Excellent job, Alexei!

 Although there're some small issues:

- The breve and circumflex in Abreve/accents and Acircumflex/accents, Ocircumflex/accents should be moved a little bit to the right so they'd be in the exact center of the character. Vietnamese eyes are sensitive to this xD

[Alexei Vanyashin]

Nhung, 

Thank you very much for the feedback. I have fixed this. 

I got your Libre-Franklin file, will make you an XML early next week. 

--
You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/341ee8d2-de30-4fc7-9028-2fa1cb09c6b9%40googlegroups.com.

[Alexei Vanyashin]

¶ Sprint 2, week 6, August 3, 2016

Today I started working from the outset: the first thing I did was to convert source files to Glyphs format, regenerate new ttfs, and run font-bakery checks. Later I updated vertical metrics, as suggested by font-bakery. Along the way I submitted some information on the version hotfix bug, and this issue was resolved by the end of the day. It's great to see how Dave and Felipe react so fast, and on-spot. 

For the design part, my main focus today was combined marks, and GF Cyrillic Plus. This charset is now complete:

 

 

Tomorrow: 

— localised Cyrillic Glyphs

— start expanding GF Latin Plus

[Nhung Nguyen]

Thank you so much Alexei, they look great now! :3

[Alexei Vanyashin]

Nhung, thank you for your sensitive eyes, and professional look on Vietnamese!

On Thu, Aug 4, 2016, 12:50 Nhung Nguyen <vns.won...@gmail.com> wrote:

Thank you so much Alexei, they look great now! :3

--

You received this message because you are subscribed to a topic in the Google Groups "Google Fonts Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/googlefonts-discuss/OIBrUNekQlo/unsubscribe.
To unsubscribe from this group and all its topics, 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/d334712b-4e48-4d8f-b351-fe306c6b2e8e%40googlegroups.com.

[Alexei Vanyashin]

¶ Sprint 2, week 6, August 4, 2016

1. Today, as planned I completed localised Cyrillic glyphs. They add localised support for 

Bulgarian, Serbian, Macedonian, Chuvash, and Bashkir languages.

After that I worked on Vietnamese:

And the better part of GF Latin Plus:

2. Plan for tomorrow:

— Complete the few remaining glyphs marked red. 

— do kerning, and prepare font for shipping

— meet with Raph at campus to discuss future work on Inconsolata

3. Questions:

Again to Nhung: could you please have a look at the Vietnamese part? Thanks in advance for your helpful and detailed feedback. 

-a

 

[Nhung Nguyen]

Hi Alexei!

No problems, I'm honored to do it :"3 Excellent job as always! There's only a tiny little small problems:

[Alexei Vanyashin]

Nhung, thank you very much for spotting that!

 

¶ Sprint 2, week 6, August 5, 2016

1. Field work day

Today was the greatest day ever! 5AM took off from SoCal to campus. The start of my six hour commute coinciding with today's hangout call... I imagine time difference isn't something easily manageable for a team that works cross-continents. My side, during last month calls for me started 7AM, which was a bit early for me, but I quickly got accustomed to this. This week the calls were moved one hour later, which was great news. And really makes a difference...if you like to witness the Pacific dawn like me. I admit the dawn was indeed something unregrettable.

I was lucky that the connection on my phone allowed to participate in the call while driving, and not miss out on latest news. However I had some difficulty observing what Marc was showing with the masters of glyph H. If only this was shown during my stopping on a red. 

Yesterday Daniel Johnson advised me to take the scenic 101 via Santa-Barbara. Unfortunately this admirable idea wasn't possible without turning up late for today's meeting. Well, there is always a next time. 11AM I arrived, and was warmly greeted by David Kuetell. Shortly after, Raph showed up, and we immediately dived into details of project Inconsolata. Conversation was highly insightful, touching the goals, and issues of the current state. Raph explicitly scoped out the project goals and priorities. 

Similarly to my first steps on Jura, I started outlining the features of Inconsolata. As I see it, if Jura originates from the Kayah Li part, the starting point of Inconsolata lies in Raph's spiro curve. We agreed that in today's production settings drawing spiro curves wouldn't be viable, and bezier curves would work fine as a replacement. However the nature of Inconsolata's design, and parameters of its curves are directly connected to this method of drawing. My thoughts: the visual appearance of spiro curves should be mimicked in beziers, and serve as the main graphical idea of Inconsolata. 

Raph's insights on true italic versus slanted fonts in coding environment was invaluable. It occurs that syntax highlighting italicises certain parts of code during typing, while the text style may change back again to roman if it is updated to match a certain pattern. During this multiple dynamic style change if the roman, and italic designs are different, the flashing of styles will be distracting. So a less disambiguate design would be preferred. Simpler put, italic and roman styles side by side should look consistent, and closely related. This does not directly advocate for a slanted approach. It may also be a mix. 

Raph mentioned that Light weight would be good to have, for styling comments in code. As an exploration of possible expanding directions, I proposed an inverted cut that would match the darkness of a normal cut, but on a dark background. This is something that hasn't been in done coding fonts, AFAIK. 

We also discussed Inconsolata's trademark features: double-story g, inclined vertical bar of the roman t, influence from Franklin Gothic, and the American tradition. 

To sum it up our meeting provided a lot of food for thoughts on the project scope. Later we would need to specify what would happen in the allocated time-box. 

After Raph left, David showed around the campus. Our discussion was more general. If my teammates recall, some time ago Dave Crossland asked about our preferred T-shirt sizes. I decided to follow up on this matter, just out of curiosity. My curiosity was awarded with many shiny new team T-shirts. So, please catch up with me to claim yours :)

Suddenly while promenading on campus my xiami band started vibrating to signal that the daily 5km(3.1 mi) goal was reached. And all those steps were made onsite. This gives some information on the size. 

In total, I spent an amazing 4 hours at campus, and this day will be unforgettable for me. To answer Dave's question, I haven't seen the Kegerator, but have bumped into Behdad while climbing the wall. Nor did I manage to do any actual drawing work on Marmelad today, so I am shifting it for tomorrow. Instead, I would like to share some photos.

-a

Discussing Inconsolata with Raph Levien

Discussing Inconsolata

With David Kuettel in the shiny new team T-shirt

Hanging out on Google hangouts

This one is self-explanatory

The unregrettable dawn after hangouts call on route to campus

[Alexei Vanyashin]

Over the week-end I finalised Marmelad: added missing characters, kerned, fixed Vietnamese, and vertical mertics (thanks font-bakery & HT's script 'Lowest & Highest Glyphs').

Git repo link

Prior to kerning Cyrillic, I added new kerning classes to my script. Then I exported kerning pairs from my previous projects using Marc's script — this became the 'Core kern pairs kerning text' for GF Latin & Cyrillic Plus. Using both scripts saves a lot of time. Thanks Marc!

[Alexei Vanyashin]

On the week-end I checked Cyrillic team's work:

Arvo is ready, has been expanded to GF Cyrillic Plus. 

— Git repo link for Arvo. 

I have uploaded Roman weights. Will upload Italics later today, after another check. 

Yanone Kaffesatz — has minor design issues with Serbian б, ₴. Sent back for reworking. 

Today had a call with Olga, original author of Lora to walk though Glyphs-related questions. 

[Dave Crossland]

Will your team use Github directly? :)

[Alexei Vanyashin]

On Monday, August 15, 2016 at 7:12:15 PM UTC+3, Dave Crossland wrote:

Will your team use Github directly? :)

Orders accepted! I will make a little voluntary-compulsory training for the comrades. First comment from Ivan: "Installed everything, but don't get it"

So, it may take some time to set up, but surely would be helpful for our workflow. 

Today I uploaded these completed projects on git:

• Arvo
  
• Yanone-Kaffesatz

With Arvo it was a little tricky, since Anton only sent me the Roman masters. Italic sources were converted from otf/ttf, which involved various clean-ups(of existing Latin) during finalisation. 

I have made preparations/briefings for this weeks new projects. Bitter, and Play have started. 

Some more progress made on Raleway (Cyrillic fork). This week is final for Cyrillic, as planned. We figured out the workflow with Pablo here.  

-a

[Alexei Vanyashin]

Status for today:

Raleway Cyrillic Italic — 1st master, Light is complete. 

I posted a suggestion to change the Italic angle for rounds here. 

[Alexei Vanyashin]

Today I pulled two of my team members on git, and they made their first commits. Hoozay!

'Hacking' into Olga's notebook using skype and hangouts to introduce her to the world of github.

Olga is expanding the glyphs set's of Lora, and also improving existing glyphs. Ivan is finalising Raleway Cyrillic Italics. 

[Alexei Vanyashin]

Today I created the Making Vietnamese better guide, as suggested by Nhung. Thanks Dave for spotting that typo! Now I am curious how this will affect our Vietnamese workflow. 

During the day I had some back-n-forth file exchanges with my team over github. I reviewed the work on Raleway, made some corrections, and sent it back. Updated Olga's glyph lists for Regular, and Bold. 

Bitter: this weekend Regular, and Regular Italic will be pushed to git. Play is in work.

Announcement: I have updated Cyril-o-pedia with GF Cyrillic glyphs sets. Also created lists for GF Latin glyph sets which will live on a separate site. 

[Alexei Vanyashin]

Today — planned next 2 weeks for Cyrillic projects, inspected project sources, updated schedules. 

Raleway: Helped Ivan with some tech issues. He is finalising the problematic Medium Italic. 

Lora: exchanged with Olga on localised Bulgarian glyphs. 

Bitter: Regular, Regular Italic are ready. Bold, and Bold Italic — next week. 

Next week I will do some QA work on finished Cyrillic projects, and prepare next projects for work. 

[Alexei Vanyashin]

¶ Sprint 2, week 6, August 22, 2016

There are a lot of completed Cyrillic projects piled up, and reviews awaiting my attention. This week I will be addressing all this, doing QA work, and then pushing completed projects on git. 

— Posted a report on Play Cyrillic Regular here.

— Reviewed Ivan's work on Raleway Italics. Made some corrections, and pushed final version to git. Next up on Raleway: kerning. 

Some of the issues I found and fixed:

— Reviewed Olga's work on Lora.

Top: Before. Bottom: Two versions of Ӻ I proposed, and adjustments on other glyphs:

After some discussions we agreed to go with this version of GHE WITH STROKE AND HOOK.

— Reviewed Manvel's work on Bitter. All good. 

— I reshaped the most problematic glyphs in Rubik, and passed the further expansion work to team.

-a

[Alexei Vanyashin]

¶ Sprint 2, week 6, August 22, 2016

— Reviewed Bitter Regular, and Bold. Will push tomorrow.

— Reviewed Nikita's work on Rubik Bold (UC). All good. 

Top: Before. Bottom: After

— Planned fixes for tomorrows update on Inconsolata: 

— Make separate font files with dotted zero and straight quotes as per development request from British Columbia Ministry of Education

— Make amends following Raph's and Nigel's requests.

-a