Google Fonts: Your Questions, Answered

1235715

Comments

  • Dave CrosslandDave Crossland Posts: 1,389
    edited December 2013
    John, please take a look at the links to comparisons of signika and fjord, and explain to me how news/source is different, or the same, in its degree of derivation.
  • Paul van der LaanPaul van der Laan Posts: 242
    edited December 2013
    it is a fundamentally different object
    That is nonsense. The “materialisation” (for lack of a better word) of a libre font versus a non-libre font is 100% identical. It is the same kind of font data that can be transferred and rasterised on *exactly* the same devices and rasterizers. The *only* difference between the two is the accompanying license.
  • John HudsonJohn Hudson Posts: 2,955
    Paul's right. The idea that the terms of a legal usage agreement make two otherwise identical fonts 'fundamentally different objects' -- the equivalent of a technology change -- is a non-starter.

    I understand the point Vernon is trying to make, I think, but he's using the wrong language. The valid point is that for different users, the terms of the license agreement may be more important than other qualities associated with the typeface, i.e. openness may be the overriding criterion of value. Unfortunately, this might be true even to the degree that an open license for a shit font becomes more valuable than a restricted license for a great one (that's the point at which insistence on free software becomes pathological).
  • Ben BlomBen Blom Posts: 250
    edited December 2013
    Vernon: if design professionals are taking Libre webfonts off the web and sending them to their prepress bureau for print jobs, instead of buying your commercial fonts, then it's most likely it's because your commercial fonts are just not good enough.
    This suggests, that if the quality of Libre webfonts is similar to the quality of commercial fonts — then design professionals might use those Libre webfonts for print jobs. If both a Libre webfont and a similar commercial font are top quality, design professionals might use that Libre webfont for print jobs. Also, if a Libre webfont is top quality, and there is no similar commercial font, design professionals might use that Libre webfont for print jobs.

    Instead of saying that the quality of commercial fonts is too low, one might say that the quality of Libre webfonts is too high. Then, a good way to limit the damage, to limit the use of Libre webfonts for print jobs, is to keep the comparative quality of Libre webfonts low. Is this the intention of those who produce Libre webfonts?
    Vernon: Should we also worry about whether the few excellent foundries and designers in the world are effecting the sales of all other commercial fonts?
    Yes, we should, if they would give away their fonts for free.
    Pablo: the Libre license solves the demands of today's cloud based services
    This makes sense. How many different Libre font families are needed to satisfy the demand of those cloud based services? 20? 30? 40? So, when such a number of Libre font families has been created, will people stop making them? Do Libre font makers have an intention to stop when there are enough Libre fonts around?
  • Jack JenningsJack Jennings Posts: 151
    edited December 2013
    1. Google doesn't seem to have any qualms about killing entire industries by building a free alternative. Email providers, web analytics, et al. have be radically changed as an industry because Google has the ability to invest a vast amount of money into loss leader products. Google as a company doesn't seem to care about the small shops that close because their business is no longer sustainable. Concern for other people's financial well-being is not something Google has proven it cares about at all. I doubt that OSS licensing by itself is not destructive—anyone can give away what they like—but what is destructive is building a library of work from profits in another sector, storing on a CDN, and then giving it away access to that for free.

    2. While OSS licensing for fonts does solve the issue of bundling fonts with other OSS, this doesn't matter for a large majority of GFA/OSS licensed fonts because they can't be used for interface work. The majority of "Libre" faces are designed for the design context, not software, so the OSS license seems moot. (Besides that problem I am unsure really what issues are being faced by web services that OSS licensing solves; please elaborate.)

    3. TCP, SSH, and C are infrastructure for web services. Source Family is not.
  • John HudsonJohn Hudson Posts: 2,955
    John, please take a look at the links to comparisons of signika and fjord, and explain to me how news/source different, or the same.
    Do I have to? Without looking, I'm pretty sure that a) I'll see more difference than similarity, because that's how my mind works, and b) the similarity is only consequential because of imbalances in the price (and yes, it is more to do with the price than with 'openness'). The relative value of typefaces subsists primarily in their design, such that, all else being equal, a customer chooses one font over another because he prefers that design for his purposes. If this is the only measure of value, then only direct clones are an issue, because anything that is sufficiently different to be independently valued is legitimate competition (note that I am talking here only about design, not about how the design is made; derivative code is a different matter). Secondary value may subsist in the technical implementation, such that two implementations of the same design may be differentiated, and one considered by the customer superior to the other for his purposes. But the software licensing model used for the past thirty years introduces tertiary, non-design and non-technical value criteria, such that faced with two identical designs with technically equivalent implementations, the customer might sensibly make a choice based on differentiation of license terms. The problem with free software licenses in this respect is that they tend to elevate the value criteria of license terms above those of design differentiation or technical implementation. When that happens, similarity of design becomes a different kind of issue, because it encourages customers to go for what is similar but free -- and free as in gratis is a huge factor here; more so, I think, than 'openness' -- , as against what is better but costs and comes with some restrictions. Open software licensing distorts the value criteria of the font market.
  • vernon adamsvernon adams Posts: 82
    edited December 2013
    John, i agree with much of what you say there. I can't agree with the 'Kool-aid' though. Very funny :D but actually way, way off, and here's why; given the brief to create type that is able to fully engage freely with any infrastructure, interface and user, of the Internet, i don't see that a non-free font object could possibly function adequately. Sure, non-free fonts can function here and there on the internet, in their own way for some services, for some users, but that is not the brief. Do you understand why it's not the brief? You seem not to fully grasp it yet. It's because there is technology and services forming the infrastructure of the net, that would be slowed up and hindered by restrictive font licensing. That infrastructure, interface and user, functions easiest and quickest with a fully "Libre" font object. This is exactly why i ended up giving permission to Adobe and Monotype to 'ignore' the reserve font names (RFN's) in the licensing of my OFL'd fonts. My standard OFL fonts, that includes an RFN, were simply 'not free enough' to function in the way that those 2 webfont services wanted to use the fonts. If those fonts had not been fully free, they could not have functioned in that service. It has absolutely nothing to do with ideology at all.
  • Same goes for Typenerd ^^^^
  • vernon adamsvernon adams Posts: 82
    edited December 2013
    Paul's right. The idea that the terms of a legal usage agreement make two otherwise identical fonts 'fundamentally different objects' -- the equivalent of a technology change -- is a non-starter.

    I understand the point Vernon is trying to make, I think, but he's using the wrong language.
    True. I am using the word 'object' differently to you. I mean object as in something that is moved around, say as a data blob across the internet, that goes from one device to another and on and on. In ye olde Apple lingo we could call it a 'web object', it's the objects we all now use everyday. You seem to be using the word 'object' as in something you can pick up, or 'own', or that can only be carried around by humans, or exists in some physical 'original' form at least in one place. We could use 'web object' if it helps, or any other term that may help communication.
  • Vernon: Should we also worry about whether the few excellent foundries and designers in the world are effecting the sales of all other commercial fonts?
    Yes, we should, if they would give away their fonts for free.
    Oh. so it is ok to effect others by releasing more popular non-free fonts? but it's not ok to effect others by releasing free fonts ? Not sure i see the logic of that. Is your opinion based on some ideology?
  • John HudsonJohn Hudson Posts: 2,955
    Vernon:
    ...given the brief to create type that is able to fully engage freely with any infrastructure, interface and user, of the Internet, i don't see that a non-free font object could possibly function adequately.
    That wasn't defined as a 'brief' though. Your comment about 'joining the Internet age' implies that only open license fonts qualify for the brave new world, and suggests that being free is a prerequisite for fonts in the Internet. It isn't, as all sorts of commercial webfont licensing and subscription services are daily demonstrating. I do understand that in the Internet ecosystem there are companies doing business and providing services for whom openness is a priority for fonts and other software. And I understand that these business models and services are particular to the Internet, i.e. that they are different from previous models and wouldn't exist without the Internet. But it is a rhetorical leap to go from this to statements like 'joining the Internet age', as if all the other business models and services, and the non-free fonts and software that they utilise, that exist within the ecosystem are just anachronisms or somehow fundamentally at odds with the Internet. I know that many free software enthusiasts probably do think that, but that's why I used the term ideologues.

    I'd like to go back to a comment I made earlier -- I can't remember if it is in this thread of the other one about Google Fonts ethics (which I'm happy not to revive, as this discussion seems much more productive). The breakthrough for me in understanding why Google Fonts favours open licenses came when I sat down with David Kuettel and he explained some of the reasons why it is useful for Google services for fonts to be licensed in this way. These were particular technical benefits, relating to Google's services (e.g. the more widespread the use of a webfont, the more end user systems it will be cached on, and hence the quicker Google services using that font can be deployed to that user), and while I doubt if I grasped all of the ways in which Google indirectly derives benefits from the GF fonts, I do at least understand that there are sensible business decisions involved. That is much, much more reassuring to me than listening to you or Dave talk about Libre software and the Internet age. I suspect it would be much more reassuring to some of my colleagues too.

    _____

    For me, the big negative issue with Google Fonts has always been the paltry amount of money they paid to have some of the fonts open sourced, which I think is exploitative when compared to the value Google derives from those fonts. Because that value is mostly indirect -- even more indirect than that by which companies like Microsoft and Adobe derived value from bundled fonts --, it is mostly invisible to the font makers. But it is real and substantial, and the contribution of font makers to that value should be appropriately rewarded.

    I would have no trouble with every font in the world being available under an open license so long as the companies and users that benefit from such licenses paid properly for that to be so. [Apart from anything else, that would put competition back where it belongs, in the area of design and technical quality.] Exploiting hobbyists and students to get stuff of dubious quality cheap is not a propitious start, but as I've said before, I give Google Fonts the benefit of the doubt at this stage, and am more interested in where they go than where they've been.
  • John HudsonJohn Hudson Posts: 2,955
    We could use 'web object' if it helps, or any other term that may help communication.
    Web asset is a useful term, I think. Object is too general a term, because a picture of a kitten is a web object, but generally speaking it isn't used as a tool and its ability to generate added value is constrained to particular situations. Fonts are tools, and generating added value is sort of in their nature, which is why we have more than one of them. Asset, in the military sense of something deployable, seems to me appropriate, and in this respect a web asset is indeed different from a desktop asset. But it still doesn't follow that all web assets need to be open licensed or that 'going everywhere' is some kind of sine qua non of the Internet.
  • John HudsonJohn Hudson Posts: 2,955
    BTW, I've also heard fonts referred to as 'web commodities'. That's really an unhappy thought, because commodification usually implies undifferentiation in the market. It seems to me that the only people who could view fonts as commodities are people who don't understand typography, or who have such minimal text display requirements that typography isn't even on their radar. Frankly, those people really do only need one font. As soon as you make a choice to use one font over another based on the typographic qualities of that font, you're not dealing with a commodity.
  • Vern wrote: “How and why does the existence of large server of Libre webfonts run by google for the purposes of styling and moving text on the web, have anything to do with helping or not helping the market for commercial fonts?? What's the connection between a web search giant getting a mass of free webfonts into use on the internet, and, you (or anyone else), selling fonts?”

    On the off chance that anybody else finds this confusing....

    The fact that Google is serving them up as web fonts is not the point as far as commercial f0nt makers feeling threatened. Google and others also make them available for download.
    Libre fonts can be used for anything, including all of the uses of retail proprietary fonts. About the only commercial font uses they are not good for are (1) replacing a font specifically commissioned to be uniquely proprietary for a single customer or (2) replacing fonts whose exceptionally high retail price gives them a degree of exclusivity (such as Dutch Type Library's fonts).

    Of course there is utility and value in an open license, whether in a purely web font context or otherwise. But commercial font makers who are concerned about the impact of libre fonts are not crazy to be concerned.*

    Personally, if there were any clear metrics available to measure the effect,** I would be willing to bet real money that the total amount of revenue g0ing into the pockets of type designers will decrease as a result of all this. But I can't guarantee it—Ben Blom is right to say “we don’t know”—although the balance of probabilities seems clear to me.

    * At least mostly not crazy to be concerned—DTL shouldn’t worry!

    ** I know of no solid metrics or intelligence on the amount of money spent on fonts each year. Any real sources of info would be welcome.
  • The breakthrough for me in understanding why Google Fonts favours open licenses came when I sat down with David Kuettel and he explained some of the reasons why it is useful for Google services for fonts to be licensed in this way. These were particular technical benefits, relating to Google's services (e.g. the more widespread the use of a webfont, the more end user systems it will be cached on, and hence the quicker Google services using that font can be deployed to that user), and while I doubt if I grasped all of the ways in which Google indirectly derives benefits from the GF fonts, I do at least understand that there are sensible business decisions involved. That is much, much more reassuring to me than listening to you or Dave talk about Libre software and the Internet age. I suspect it would be much more reassuring to some of my colleagues too.
    John, i can understand you find the 'sensible business decisions' framing easier to acknowledge. The angle from 'sensible business decisions' makes a lot more sense to you, and i assume it's the kind of area you mainly deal with in your profession. However it does not mean that those 'sensible business decisions' are the only prime motivators at work. I can also appreciate the business decisions, but David (Kuettel) has some a fascinating ideas and approaches to what these free font 'web assets' can do, how they function, and the what they mean. It's not only 'sensible business decisions' that are at work with libre webfonts, it's also some fairly 'crazy business decisions' too :-) It's the 'crazy' that most businesses probably steer clear of, and I suspect that it's the 'crazy' that helps you to disregard what you call 'idealogy' and 'pathology'.
  • I would have no trouble with every font in the world being available under an open license so long as the companies and users that benefit from such licenses paid properly for that to be so.
    How on earth would that work? How would payment be made? in which currency? and where? at end user level? at web developer level? At isp level? At font-server level? At publisher level? And even if you kludged that logistical nightmare into a working pay-for-it system, how would you combat piracy? who is going to police it? The font designer? the web developer? The publisher? The font server? The NYPD? And even if you put into place an effective policing and license enforcement system, what would happen if a bunch of people started to release Libre fonts? :-) Why do you even think such a system could work? Belief in that sort of system is pathological. Mind you if the mega corporations Adobe and Monotype were to send me $0.001 every time their servers served one of my fonts, i would certainly sign up for their £ibre font licen$e programme.
  • Ben BlomBen Blom Posts: 250
    edited December 2013
    --- Should we also worry about whether the few excellent foundries and designers in the world are effecting the sales of all other commercial fonts?
    --- Yes, we should, if they would give away their fonts for free.
    Vernon: Oh. so it is ok to effect others by releasing more popular non-free fonts? but it's not ok to effect others by releasing free fonts ? Not sure i see the logic of that. Is your opinion based on some ideology?
    When you are in business, you can only survive in the long run, if you can, at least, recoup the costs of the products you sell. There are many different definitions of “costs” — but if you give your products away for free, not temporarily, but permanently, then you surely cannot recoup your costs. So for a normal business, giving its products away for free permanently, is not sustainable.

    There are different kinds of competition. One of the normal characteristics of competition is, that it affects other players in the market. Affecting other players in the market, is, in itself, OK. It’s part and parcel of competition. I Think most kinds of competition are OK, even if it’s quite fierce, like temporary deep discounting. The competition which I feel is unfair, is cut-throat competition, also called ruinous or destructive or predatory competition. This is competition where a player in the market offers its products for prices permanently below the costs of making them. Now when a player in the market offers its products permanently for free, while its competitors need to, at least, recoup their costs to be able to survive — then this predatory player might, in the end, eliminate all its competitors. In such a situation, there is no level playing field. I think what this predatory player does, is not OK.

    There are very old stories of well-intentioned people who help a small baby creature grow, and who, in the end, are devoured by this creature when it gets bigger. Perhaps there are some similarities between such well-intentioned people, and the people who sell their fonts to Google — perhaps not.
  • PabloImpallariPabloImpallari Posts: 777
    edited December 2013
    This is exactly why i ended up giving permission to Adobe and Monotype to 'ignore' the reserve font names (RFN's) in the licensing of my OFL'd fonts. My standard OFL fonts, that includes an RFN, were simply 'not free enough' to function in the way that those 2 webfont services wanted to use the fonts. If those fonts had not been fully free, they could not have functioned in that service. It has absolutely nothing to do with ideology at all.
    There is a real little gem hidden inside that paragraph.
  • Dave CrosslandDave Crossland Posts: 1,389
    edited December 2013
    Instances of software freedom are numerous. You miss the forest for the trees, sir :)

    > the other one about Google Fonts ethics

    What thread is that?
  • John HudsonJohn Hudson Posts: 2,955
    edited December 2013
    Vernon, I never suggested that I thought there was a practical mechanism by which every font in the world could be open sourced and the font makers properly compensated in each instance. I said that I wouldn't have any concerns or objection to all fonts being open sourced under such conditions. My point was that I am more concerned with font makers being able to make a living and share in the financial value that their work generates for other people than I am in license terms per se.

    Please take time to read what I write carefully, and ask questions if it doesn't make sense. Note also that I used the term pathological to refer to a very specific circumstance, i.e. the choice to use something bad but open rather than something good but restricted. This is pathological in the same way that eating cheap junk food rather than buying good ingredients and cooking a meal is pathological: it sacrifices the good to the convenient. We want a typographically healthy Web don't we? So I wasn't suggesting that all free software is pathological, only that a particular condition that can arise in the choice of free software is.
  • John HudsonJohn Hudson Posts: 2,955
    Vernon:
    If those fonts had not been fully free, they could not have functioned in that service. It has absolutely nothing to do with ideology at all.
    But now you are talking about specific fonts in specific services, which is what I've been talking about all along. Do you see the difference between talking about making fonts work in particular services and talking about them 'entering the Internet age'?
  • PabloImpallariPabloImpallari Posts: 777
    edited December 2013
    How many different Libre font families are needed to satisfy the demand of those cloud based services? 20? 30? 40? So, when such a number of Libre font families has been created, will people stop making them?
    Personal opinion only:
    I guess the results will be dictated via user research methodology. But yes, I think there is a good number now.
    At an ATypI presentation about the then-new Google Fonts user interface, I heard that when When GF first launched with only 15 fonts, the Nº1 user request was: "Add more fonts". There are about 700 now (some not yet published) and my focus has shifted to "improve the ones we already have" and I'm working on that. Many of my fonts are being reviewed and improved (contours, spacing, kerning, hinting), and many also expanded into families, or adding more language coverage. I may add a few more fonts, but that's not the main focus right now.
    Do Libre font makers have an intention to stop when there are enough Libre fonts around?
    It's a possibility... Nothing last forever. However, as John Hudson said, many more organizations (other than Google) will require Libre fonts in the future as the world tends to get unified, so... we don't know.
  • there are sensible business decisions involved. That is much, much more reassuring to me than listening to you or Dave talk about Libre software
    John, you wrote in the previous 'Why Google Web Fonts aren't really open source' thread on January 18th 2012, some 2 years ago now:
    I'm sympathetic to the utopian social intent of the libre software movement, but when multibillion dollar corporations encourage individual creators to give their work away free I think this is cynical exploitation.
    Are you now less sympathetic to the utopian social intent of the libre software movement?
    the contribution of font makers to that value should be appropriately rewarded.
    How do you suggest determining the price, other than through the normal market mechanism? Do you think there is a moral aspect?

    Thomas Phinney wrote on the old thread on January 21st, 2012:

    I think this whole area is pretty complicated.

    Google is indeed not paying the "going rate" for serious, professional
    quality fonts. But as a result, they mostly haven't been getting
    serious, professional quality fonts. Some are better, some are worse.
    Does that mean they are not paying enough, or instead that there is
    room for a variety of pay rates and quality levels?

    Are type designers like doctors where some minimum standard must be
    required? I think not.

    Don't get me wrong. I wish Google were paying a bigger bounty per
    font. I just don't think it is necessarily immoral to offer to pay
    less than what a mega-corporation pays a skilled, name type designer,
    as long as you don't expect the same level of craftsmanship.
  • Jack wrote,
    3. TCP, SSH, and C are infrastructure for web services. Source Family is not.
    John wrote,
    Fonts are tools
    Well, are fonts tools or not? :)
  • vernon adamsvernon adams Posts: 82
    edited December 2013
    Please take time to read what I write carefully, and ask questions if it doesn't make sense.
    durrr... ok. i'll try and keep up with your blazing intellectual powers.... :p
    Vernon, I never suggested that I thought there was a practical mechanism by which every font in the world could be open sourced and the font makers properly compensated in each instance. I said that I wouldn't have any concerns or objection to all fonts being open sourced under such conditions.
    You mean you would have no objection to a totally unfeasible pipe dream? (ok, no wonder i'm not keeping up). Actually I thought maybe you were resurrecting David Berlow's 'permissions table' idea from 2009-ish. Or something similar.

    The Permissions Table. What happened to that idea? Why was it never adopted? Which reminds me...

    The biggest danger to non-free fonts is not Libre fonts. The biggest danger to non-free fonts is the fact that so many non-free fonts are now being exposed, almost totally unprotected, ready to be downloaded straight from the browser to the desktop, by anyone who wants a copy. Imagine i-tunes without the pay wall (just steal whatever music you want, just a few clicks) and you pretty much have the situation that people are putting their fonts in, courtesy of the large non-free webfont servers. It's exactly the situation that many warned back in 2009 should never happen. But it has happened! I don't understand why designers tolerate that situation. It's some deep irony that people use forums like this one to scaremonger on libre fonts, but say nothing against the real direct danger to their livelihood. On these non-free webfont services, type designers are expected to earn from a % of sales of their fonts, but the hosting corporations are pretty much looking the other way to the fact that they are actively exposing these fonts to browse-by theft. I wonder what the ratio is of a font paid for / to font ripped via the browser? If it's anything like the % of loss that the music industry (what's left of it) complains about, then the losses from non-free webfont services is likely massive.
  • vernon adamsvernon adams Posts: 82
    edited December 2013
    me:
    If those fonts had not been fully free, they could not have functioned in that service. It has absolutely nothing to do with ideology at all.
    Hudson:
    But now you are talking about specific fonts in specific services, which is what I've been talking about all along. Do you see the difference between talking about making fonts work in particular services and talking about them 'entering the Internet age'?
    John, do you not see that "making fonts work in particular (web-based) services" is exactly "'entering the Internet age" ?? You know? In the same way that type was re-drawn, recut and re-licensed so it could enter the machine age.
  • I'm also just going to make up a bunch of shit. Everyone pretend this post is longer and I used the word #libre a lot.
  • Jackson, how have your views on web fonts changed over the last 3 years?
  • DaveCrossland:

    Over the last three years my outlook on web fonts has changed from cautious (curious to see how the technology and models evolved) to oh-shit-have-to-do-something (forced by customer demand to invest a ton of time and money in offering web fonts) to, only very recently, somewhat optimistic (seeing use, sales, and standards improve).

    Overall, my general view hasn't changed much. I think "web fonts" are a temporary issue. They are significant part of the painful transition the type industry is being forced through as type technology, licensing models, and the customer base changes from an outdated DTP licensing model to a yet-to-be-resolved 21st century digital model.

    How was that for making shit up?
Sign In or Register to comment.