Projects Google Fonts ought to fund in 2023?

John Hudson

@Mikhail Vasilev

Like this?

It works slightly better in Cyrillic, due to fewer ascenders in the lowecase, but still looks really weird. I can imagine doing this for some titling lettering on occasion, but not as something common in fonts.

Dave Crossland

Seems like something else the the typesetting engine should do for all fonts in all scripts that make sense for it

jeremy tribby

Newsreader is not quite Times, but it is a very good one on GF

Hin-Tak Leung

Mikhail Vasilev said:

...And on Windows particularly Times New Roman reads especially good, that is due to some "secret sauce" I beleive that makes it behave well in Windows apps. ...

I think it is not a secret - the hinting instructions on those "Microsoft core fonts" have been hand-tuned with Visual TrueType (or its ancestor in the 199x with a different name) against the Microsoft font scaler, by VTT expert(s). And they have embedded bitmaps at low resolution too.

I realize I am writing the exact same thing as yours, if you don't know what "hinting instructions", "VTT", "font scaler" are...

Johannes Neumeier

One thing I would really, really love to see some traction on is a production viable, multi-platform, open source font editor. As far as I am aware there are some projects in various degrees of staleness, so I doubt there is something concrete that can "just get funded". Maybe the creation of a publicly announced incentive could get this (gigantulan) ball moving? Or a series of hackatons on whatever seems the most likely contender? GF, if anybody, has the profile to make a dent here.

You can always dream...

Craig Eliason

Johannes Neumeier said:

One thing I would really, really love to see some traction on is a production viable, multi-platform, open source font editor. As far as I am aware there are some projects in various degrees of staleness, so I doubt there is something concrete that can "just get funded". Maybe the creation of a publicly announced incentive could get this (gigantulan) ball moving? Or a series of hackatons on whatever seems the most likely contender? GF, if anybody, has the profile to make a dent here.

You can always dream...

+1. For teaching, a viable, reliable, no-cost, cross-platform editor would be fantastic. 

Hin-Tak Leung

Something like unifying VTT and glyphapp?

Mikhail Vasilev

Florian Pircher said:

@Mikhail Vasilev you might also be interested in the Libertinus fonts. They don’t have a center-line cut (as far as I can tell) but there is a monospaced version which might fulfill the same role.

Hmmm, well, not quite in the same league as Times for me. Besides, on Windows it renders weirdly in e.g. Notepad++, maybe due to rendering issues, I am not sure. But same story with many other non windows-native fonts and WinAPI based apps. 

Mikhail Vasilev

John Hudson said:

@Mikhail Vasilev

Like this?

It works slightly better in Cyrillic, due to fewer ascenders in the lowecase, but still looks really weird. I can imagine doing this for some titling lettering on occasion, but not as something common in fonts.

Yes like this. Looks unusual of course due to strong habit. But after some time it is ok.

After all it was initially designed and optimized  for baseline placement. Ideally, caps and digits should be redesigned to be more natural part of mixed case flow. 

Also latin has sort of optical emphasis on the baseline due to lots of serifs on the baseline area (e.g. "mnhlk" glyphs). I think this effect can be reduced somehow but I am not sure how though. 

And e.g. the glyph "2" has heavy bottom part and lighter upper part, which makes it sit optically even lower than it is.

BTW Old Style seem to have less of mentioned effects, namely its caps and digits are more vertically balanced on their own. Although I percept Times (lowercase part at least) as a slightly better version of Old Style.

So it is not as simple as aligning along the centerline of course. 

Craig Eliason

Mikhail Vasilev said:

Yes like this. Looks unusual of course due to strong habit. But after some time it is ok.

After all it was initially designed and optimized  for baseline placement. Ideally, caps and digits should be redesigned to be more natural part of mixed case flow. 

Also latin has sort of optical emphasis on the baseline due to lots of serifs on the baseline area (e.g. "mnhlk" glyphs). I think this effect can be reduced somehow but I am not sure how though. 

In case it's of interest, my TypeWknd talk "Baselines are Overrated" discusses my experimentation along these lines. It involved a serifless design with all sorts of different glyph heights (and caps that center on a higher centerline). But we're running off-topic...

Dave Crossland

Johannes Neumeier said:

You can always dream...

We'll see where this leads

 https://www.youtube.com/watch?v=Xhw5iCTK1XI

Thomas Phinney

Fontra is still very early and primitive in most respects. What is lovely, however, is to see something intended from the beginning to work (1) in a web browser, (2) with variable fonts and (3) via collaboration.

Hin-Tak Leung

Fontra, as it  currently is, is a large scale collaboration tool. Interesting it is, it fits a rather different niche as "teaching, a viable, reliable, no-cost, cross-platform editor".

Apparently the author of the first truetype font editor, Apple's RoyalT in the 198x, works/worked for Google now, or at one time, in the android area. I think it was cross-platform, as some of it eventually became Microsoft's VTT.

Dave Crossland

Apple and MS collaborated on TrueType and I think there's various examples of code developed at Apple for TrueType entering Microsoft codebases.

Mike Reed's 2-person Skia company was famously acquired by Google in 2005 (https://www.crunchbase.com/acquisition/google-acquires-skia--ad6cccc9) and was the backbone of Android's rendering (https://en.wikipedia.org/wiki/Skia_Graphics_Engine). Skia is used in fontmake compilation for the removeOverlap() step, which has room to improve.

Mike recently left Google.

John Hudson

Apparently the author of the first truetype font editor, Apple's RoyalT in the 198x, works/worked for Google now, or at one time, in the android area. I think it was cross-platform, as some of it eventually became Microsoft's VTT.

I don’t think RoyalT itself was cross-platform, and as I recall the first generation of Windows TrueType fonts were made on Macs (by four Monotype staff sent to work at Redmond). VTT was based on a port of RoyalT’s hint compiler.

Hin-Tak Leung

Not having Mike Reed playing a more prominent role with Google Fonts is a missed opportunity there. But then, I guess he is/was probably way above Dave's pay grade :-). Maybe nor he want to be dragged back to old area of work from 30+ years ago.

Perhaps cross-platform is not the right description - code written 30 years ago made fewer assumptions about what is the host, and what libraries are or are not available, and more portable in that sense. On the other hand, things were not modular. Dynamic libraries / dlls weren't a thing; linux kernel modules (the ability to load/unload drivers on plug and play) wasn't there(? used much) until this side of the millennium. It was portable but not cross-platform.

Daniel Benjamin Miller

John Savard said:

Dave Crossland said:

DAMN THEY MOVE QUICK BRO

This tactic may prove counterproductive. If Monotype becomes for type what Hasbro is for boardgames, politicians may decide that giving strong legal protection to type designs is against the public interest

I am fairly sure that politicians don't really care (as people don't think about fonts so directly --- most people just use whatever is preloaded, after all). As for strong legal protection... in most places there is no law stopping you from redrawing many (most?) popular fonts. (Yes, there are design patents, but these expire, even for fonts that were patented. Minion's has expired, even.) A number of fonts on which Monotype makes a lot of money have FOSS clones or extemely close alternatives.

Monotype definitely has designs on being ... the monolith of type. But it's not so much due to strong legal protections on designs...

Dave Crossland

http://typeright.org/ tried to lobby usa politicians and had no luck.

Nick Shinn

For the “ought to fund” of philanthropy, or just the collector’s appetite (Edmund Fry and Giambattista Bodoni were great acquisitors of scripts), there are, despite the efforts of various type designers, still many rare and endangered languages with little or no font support. 

Perhaps other styles can emulate Noto’s scope?

On a personal note, I know it’s not very popular, but I might be persuaded to add bold and italics to Bellefair. Dm me with a proposal, if you like.

Dave Crossland

I agree that Noto's possible future could include working on pre-unicode-standard scripts, and supporting efforts to contribute to the Unicode process. 

After a business sans and serif (or low/high stroke contrast modulation) style, what is the 3rd most important style for any/every script? the 5th? 10th? I don't have a good answer today. Probably I think the 3rd style for Noto should be the "how to write" sort of handwriting style, since at the highest level Noto is about literacy, and a font to read a language in a UI (sans) or document (serif) is important for literacy, but then a font to 'write" a language is next. Maybe a sans mono is 4th, and there's already a Noto Sans Mono for Latin Greek Cyrillic. But then, are wedding fonts more important than scary fonts? etc

I'll add Bellefair to my list  

Nick Shinn

The problem with scripts (in Latin/Greek/Cyrillic at least) is that the generations who are growing up with keyboards rather than pens have difficulty reading cursive and connected scripts. And yet script fonts have a very useful presence in typography, everybody likes the hand-made, casual, organic effect.

Perhaps there is some way that a script typeface might educate such readers, say, by being available in both a disconnected and a connected version. Possibly with some kind of programmed interaction between the two?

John Hudson

Perhaps there is some way that a script typeface might educate such readers, say, by being available in both a disconnected and a connected version.

Microsoft have this for European writing systems in the Segoe Print and Segoe Script types, which benefited greatly from native expert input on Cyrillic and Greek handwriting.

Nick Shinn

Thanks John, I wasn’t aware of that.

I don’t know about the Cyrillic, but the Latin connected doesn’t match the plain version very well, it’s too bouncy, not neat enough. 

There is certainly design space for a Latin script face in which the two versions more closely match. 

John Hudson

To be clear: the two Segoe handwriting types were not intended to be disconnected and connected versions of the same design. Indeed, what they indicate about the differences between writing connected script and disconnected letters is part of what makes them interesting.

There is certainly design space for a Latin script face in which the two versions more closely match. 

Indeed. But I think the connected version could easily end up looking artificially static, like many ‘script’ types, lacking the characteristics of actual writing.

Nick Shinn

It would be better to work the other way around (basing the disconnected version on the connected)—similar to how Matthew Carter began with bitmap shapes for Verdana.

At any rate, both versions should be developed in tandem—as with the ideal scenario for developing Latin-Cyrillic-Greek types—rather than finalizing one and accommodating the others to that.

John Butler

Add full JALT support, glyph palettes and HZ-algorithm H&J multi-line composers to Inkscape and Scribus so we can stop paying Symacrodobe subscription fees for these capabilities.

Hin-Tak Leung

Hmm, I think that suggestions along the line of: "commercial software X does this useful thing, but I don't want to pay the price. Can Google please make open-source software Y good enough to replace X so I can save some money..." is not a good argument though.

Firstly, I think that clever and hard-working people at commercial software X that actually makes some features their customers are happily to pay for, is a reasonable agreement and transaction on both sides.

If you want a feature in Inkscape and Scribus, are you willing to pay the Inkscape and Scribus people for their work yourself, however small your donation/ sponsorship might be? I believe both projects accept donations.

As my comments above regarding font editors and Glyphapp , how about proposing that Google pays Symacrodobe "well enough" to open-source their work and make it free? That way software enthusiasts can improve on it in directions not driven by pure commercial needs, and Inkscape /Scribus people can learn from it , and the people at Symacrodobe get compensated for their hard work properly.

John Butler

This thread began with a representative of Google, from time to time the richest, most powerful and resourceful company on the planet, soliciting requests as to how they might allocate upcomng time and money in the realm of type and typography.

The OpenType specification is now over twenty years old. I personally attended the first OpenType Jamboree hosted by Microsoft sometime back during the Clinton Administration. The page layout components of the TeX typesetting system are older still, as is the HZ typesetting algorithm adopted by Adobe in InDesign. Good-looking H&J is the heart of all good typesetting.

Adobe today is not the company it was in 1998. In some respects (Source Sans and Source Serif) they, like Microsoft with its WSL and embrace of non-Windows environments, have progressed. I have been in computers long enough to observe software innovations begin with private funding and then later, after their creators have been rightly recognized and bought out, been dispersed into open source tools and libraries around which newer, even more sophisticated proprietary tools can grow. Sampo Kaasila’s Visual TypeMan was quoted to me at $15K at one time before Microsoft eventually bought it and adapted it into Visual TrueType, which they now distribute free of charge. Inkscape does probably 80% of what Illustrator can do, sometimes more tediously and obscurely, likewise the GIMP versus Photoshop, and Scribus and XeTeX versus InDesign. Adobe’s AFDKO is given away free as well.

In my opinion, many glyphs and features in modern commercial, free and open source fonts are inaccessible to ordinary folks without a Creative Cloud subscription or the time to gain workable competence in XeTeX or Scribus. To this day, Microsoft Word does not render the true small caps in an OpenType SMCP feature, not even through the feature selector in the Advanced tab. I can get froofy ct and st ligatures in Word, but the small caps are still spindly and artificially rendered. I seem to recall LibreOffice handles this one aspect much better than Word.

The point of my comment was that Google can likely more efficiently finance the further development of existing open source illustration and layout projects, to more benefit for Google Fonts type users, than other alternatives. In fact I seem to recall Inkscape had some major recent advances thanks to a Google Summer of Code event. Being the giant company they are, just like Microsoft and Adobe, I’ve had my grumbles with Google’s left hand while praising their right hand.

Paying Adobe or Microsoft to open-source some of their software is simply not practical from their lawyers’ standpoints, having to chase down original parties (or their heirs!) from ancient contracts and code licenses that probably no current Adobe or Microsoft employee even has working knowledge of. Not even unlimited money could untangle such a knot.

Hin-Tak Leung

Argh, I see I possibly made a mistake - "Symacrodobe" is some kind of derogatory pseudonym for Adobe & friends, that I am not aware of.

That said, my proposal (sent private to Dave) indeed involves going towards partially untangling the historical Adobe + Microsoft + Apple cross-licensing agreement mess, and allow them to open-source some more of their age-old tools further, including Typeman / Visual TrueType, further.

John Butler

IMO Adobe changed for the worse after the Macromedia merger, almost Symantec-like in their juggernautery, hence my coining. I’m sure there’s plenty of good stuff still happening inside Adobe somewhere, but their subscription model does not constitute progress.