Roadmap #194
Replies: 8 comments 6 replies
-
|
Thank you, @justvanrossum for the road-map you provided in the meanwhile. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you, @davelab6 for your commit. I guess google is therefore supporting the project. That's wonderful to hear! |
Beta Was this translation helpful? Give feedback.
-
|
@signalwerk sorry my delay. I am only peripherally involved in Fontra, but I am funding development and setting high level strategic goals, so I am taking the liberty to reply :)
No, the financial model is the same as metapolator. In some ways this is metapolator 2.
You'll need to be more specific, but basically I would say, almost all libre font editors are/were GPL, and many web applications are GPL, while I agree that many web libraries do not have copyleft protections for users.
I'll answer those together: Generally I am quite against front end frameworks these days as I believe they rot too fast; but again I need you to be more specific and explain exactly what problem fontra has that you have experienced (as a user, or as a developer) tha ta framework or modern state manager would actually solve. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you @davelab6 for your Answer. I think my post was mainly a reaction/request to the tweet and for me to understand if it would be worth considering it. It really sounds like a dream job, and the project itself is just wonderful! But since we are half a year later, I hope you found someone for hire and therefore I'm going to close the comment. You basically answered my questions, and I guess I just have another perspective on it. If the project dies as metapolator died, then it doesn't matter what the roadmap/licence/technology is. Then it's driven by some people that see it as a paid job, and afterwards there is no community left. I can't speak for the community, but I can speak for myself why fontra is not attracting me a lot at the moment:
I will give you an example (doesn't matter whether it's realistic or not). Imagine you need Boolean-Operations in the Frontend-Code. This is really a hard problem and I saw only one robust and open implement in JavaScript so far. So, what do you do now? You start to implement it yourself. Sure. You can do a good Implementation in 2–3 days. But this will not lead to a robust implementation. You spend more time, more tests, more work. And in the end, you have spent a lot of time and energy on boolean operations. A very common problem, but now nobody will take over your implementations. Why? You have the wrong licence (non GPL-3 can't use it), you have it in your project and not separate, you didn't consider other cases in your first implementation (colours, outlines, ...). You included complexity in your project, and now you need to maintain it. You're doing stuff like that right now, but maybe it's not that obvious. If you don't use external packages you basically say; I have a unique case, and I'm willing to maintain all the complexity myself. But again; I completely respect your decisions and I love the project. And at one point the project might be so strong and advanced that all I said here sounds stupid, and I will happily use it again. Looking forward to that moment! Go Team Fontra! |
Beta Was this translation helpful? Give feedback.
-
I believe the position is filled but its a question for @JeremieHornus :)
Well, I don't really see a difference between paid-jobs and "community." Writing a complete editor is large-scale and intensive work, and intensive work requires basic social reproduction (rent, food, healthcare, etc) that has to be paid somehow. The problem with metapolator was that it in the funded period it did not develop sufficient functionality to become actually useful to anyone for any commercial purpose, such that their own usage activity neither created a self reinforcing feedback loop of self-interested development ("Today while using the product I encountered this bug and I fixed it, here's a PR") nor demand in the marketplace for development ("Today while using the product for a paid project I encountered this bug and I am seeking someone to fix it, here's a coin"). In both cases, there is public development, larger commercial interests involved, and people paid to work on the editor. It might be possible that Fontra will reach such "lift-off velocity" before "running out of runway", since Black Foundry is using it for one project they have contracted with Google for, and once that project is delivered, they may have other similar projects from other clients, and continue to fund development that way. Google's CJK project is not their first CJK branding project, after all, and if the Google one is seen as successful, it may be a strong case study for them to market themselves. And also, its possible that Just and Jeremie will make different decisions to me, Simon, Lasse, Peter and the others involved in directing Metapolator, such that even with all else being equal, they actually do product something useful in day to day workflows before their time runs out. Having a specific typeface deliverable will help focus, I think.
Few people want to write the whole thing, and are content enough with someone else's foundation and their scripting API, as long as it offers sufficient functionality to get their immediate needs me. Only either 'true believers' of the "libre software user freedom movement" find this ethically unacceptable - e.g. MFEK.org - or people with a LOT of subsidized time, retirees or students in FontForge and TruFont - are willing to write a whole thing. The rest (glyphr, runebender) are personal projects to use a font editor as a vehicle to learn and explore something else (web app development, rust GUIs), and as hobby horses are subject to whims and winds.
Perhaps it is different enough, as above.
Sure, I know for example Microsoft folks won't contribute. Maybe Adobe also may have some concerns. But GPL allows private forks; its proprietary forks that are prevented, and it seems unlikely to me that MS or Adobe would want to offer one. Perhaps you are right and an ally will silently approach and flee without any discussion, but my experience with Font Bakery (which had at some point a GPL dependency) is that in fact if they are bringing resources to the table and license is the only concern, they will raise it and we will discuss it.
I don't find anything to disagree with in this, however, at this stage, Fontra is a "sketch" phase for how it should be - Just is exploring what can work, and isn't actually seeking any contributions (AFAIK.) As the project matures into another phase, declaring (and setting up linters or other automation) for standards is certainly a good idea for me. Type safety checking without requiring a rewrite to typescript is good in a phase of performance optimization, but in sketching, seems to me that it is helpful and productive to allow duck typing.
We'll see, but I would suspect that for this specific example it would be good to consider kurbo's wasm, which has more and more b-ops functions appearing. It is also trivial to take some functionality out into its own repo with its own license, once out of the sketch phrase.
I'm curious if you can name the specific cases :)
Appreciated! We are all in Paris this week for ATypI, BTW |
Beta Was this translation helpful? Give feedback.
-
|
@davelab6 thank you for your extensive answer and caring so much about Open Source in the field and font production.
I think the main difference is; if you build on top of a community, you need to have enthusiasts and people who care. That's different to paying someone. You can have a kick-start with one-time-money, but later you need a sustainable model. Ideally, with people who care and earn money. That's probably the ideal solution, and I hope fontra can reach that level before running out of google-funding.
I guess that's a cultural and historical discussion. I respect all the GPL-Believes, and it brought us many good. I just have a different understanding nowadays – but that's most likely only a matter of preference and not worth an additional argument. We can read the pros and cons in every libre-project.
It's a good sketch! I might have read it just as more than a sketch. But definitely you can go wilder if you see it as a sketch.
You're right. Ideally, it would be cool not only to break out certain parts, but also to bring in other parts that already have a community.
It's basically the whole code-base that is stating it's an individual case nobody had before. Or as you named it, and that's probably also true here: it's a sketch.
I'm based in Switzerland and usually can't attend to ATypI. But let's hope a Libre Graphics Meeting will be our next chance to meet. |
Beta Was this translation helpful? Give feedback.
-
|
FWIW, |
Beta Was this translation helpful? Give feedback.
-
|
Thanks @justvanrossum - that's an excellent reason for transform.js to exist, I think - and your first commit (per https://github.com/googlefonts/fontra/blame/main/src/fontra/client/core/transform.js) did explain this on line 1, albeit in a terse way. I've made PR #506 to pull this fragment of lore into that comment line :) |
Beta Was this translation helpful? Give feedback.
-
|
Fontra is a WIP tool, that's publicly open for anyone interested to try out. I wouldn't state it's a sketch. |
Beta Was this translation helpful? Give feedback.
-
Then it has already matured from the last time I discussed the development strategy with @justvanrossum as I'm pretty sure that is how he explained it at that time :)
That would be excellent - although I guess it remains to be seen if you will continue public development – and I guess that depends on our ability to grow a user-to-developer/contributor community, such that private development by BlackFoundry wouldn't make sense because public development by the rest of the community would continue, and it would become expensive to keep up. |
Beta Was this translation helpful? Give feedback.
-
|
@justvanrossum @BlackFoundry thank you for your answers and further explanations. I don't want to rock the boat at all. My first post was in response to get a feeling if I should apply to the job-offer and for that, I only had the code to build my opinion and wanted to get more insides. Sorry if I seam to be ungrateful. I'm very excited about fontra and I appreciate all your time, money and sweat you're investing in it. I wish the project only the best. |
Beta Was this translation helpful? Give feedback.
-
You could have just said so :) If you're still vaguely interested, please send your credentials to [email protected] and perhaps we can talk. |
Beta Was this translation helpful? Give feedback.
-
I thought that was clear from my very first sentence in that whole thread (6 months ago), since I reacted to the tweet when you were looking for someone. But I see… I made myself not clear enough. 🤓 |
Beta Was this translation helpful? Give feedback.
-
You are very welcome :)
Perhaps, although I suspect that since @justvanrossum has designed Fontra with several plugin APIs, then what we will see is that the majority of the "people who care" user-developers' code is written to those APIs, just the same as with RoboFont, GlyphsApp, and Fontlab, rather than working on internals; and if any of those 3 editors was to turn all their private code public under a libre license, it would suffer the same problem: Most people enjoying writing GlyphsApp extensions are not sufficiently fluent with Objective C that they could go out and successfully apply for a salary job as a software engineer at some mac app shop.
I agree, in that, when things are made for sale, rather than for direct use, there is always shift in priorities. Sometimes this is a large and kind of grotesque degradation of the nature of the thing, where the power of marketing/sales/position compels us to use the things even though as we spend time with it we realize it is junk, that it sells well but most customers are grumpy, disillusioned, and (sooner or later) it has high churn and no long-term loyal customers or future. Sometimes this is very subtle, and the company enjoys a strong and happy community around it that is just as good as the best libre software projects. Surely we all know of proprietary applications in the font editor space and elsewhere we have gone through phases of these situations.... and since I work with 100s of type designers, I am always amused when I hear the diametric opposite opinions about the same editor product within a short period :) But, conversely, when things are made purely for immediate direct use, and no attention is paid for 3rd party users, then it also becomes a different kind of grotesque degradation of the nature of the thing... users who show up expecting the developers to care about them are just as grumpy and churn away disillusioned, because the developer does not care at all about them. No documentation? Read the source code! No basic feature I expect? Make it and send a PR! Can I pay you to add the thing I want? No, money and libre software should not mix! (Or worse :) Ultimately my point here is that "cultivating a community of enthusiastic people who care" is 100% orthogonal to the use of money to advance the project.
I agree that the ideal solution will be adoption by enough established foundries that they are self-interested to continue development publicly, and BlackFoundry is adopter #1. I hope we'll find a #2 before 2024 :)
:)
It seems it already matured beyond that stage, and so I recommend to the team to start thinking about carving things out - but this is probably also not a priority anytime soon, unless there is a direct request for it to respond to.
I would expect harfbuzz to be an early such part to bring in :)
Any other examples? :)
I hope there will be another successful LGM soon, but it seems the community has dispersed back into the individual projects; the 2021 event (now 25 months ago) had very few proposals (so few no committee was needed to triage them and they were all accepted) and no one has stepped up in 2021 or 2022 to run an event since then.... The GIMP community seems smaller than ~10 years ago, and is mainly a project of 2 people now, per https://www.gimp.org/news/2023/01/29/2022-annual-report - who appear to receive $1-1.5k USD/month on Patreon each, which keeps them going. If $50k year is "sustainable" for GIMP, I hope that it will be sustainable for Fontra too :) (Then again, the GIMP bitcoin wallet is currently worth over $500k, and was worth over $1M a couple years ago.... so maybe money is not the limiting factor there :) |
Beta Was this translation helpful? Give feedback.
-
I saw this tweet today that you’re looking for a web developer. I applaud this decision and also your effort! This is a great project!
tl;dr
Backstory
When I saw the video of the fontra announcement a few weeks back I was super excited, but I just had a brief look at the code and my first reaction was: love the idea, love the effort, but the current state was not only excitement (who am I to judge anyway...).
After the tweet today, I thought: the shape of the next generation font-editor should be nothing but awesome. So, I installed today the code and had a look at it.
What I really like
What I was concerning
Thank you!
Please do not take my comment here in any form as an attack or rant. I respect this project and the people behind it a lot!
Beta Was this translation helpful? Give feedback.
All reactions