Goal of the project

[raphaelbastide]

Now I can take time to dive back into the project. I feel the need to point my first intentions for Unified Typeface Design:

Unified Typeface Design is a proposal for the standardization of typeface design, first in an open source context. It also aims for the promotion of open source typography by introducing a transversal and flexible classification. Technically, UTD is a folder architecture to organize font sources, inspirations and references. It is also a YAML file containing useful meta informations about the typeface and its repository.

METADATA.yml contains the data useful to build a large independant database of typefaces. With this kind of identity informations, it can be easy to search and browse fonts by license, similar fonts, authors, keywords… METADATA.yml also reflects the folder structure and can link to the font files, github repository etc…

Since the recent @davelab6’s useful participation, I am wondering if METADATA.yml should or should not duplicate the font informations already present in the font file. Indeed, I think that this project should’nt be open source exclusive ; users must be free to fill a METADATA.yml for the font they own or not. Here is an example:

Most of the font I use are Open Source but I really love Radim Peško one’s. It is not impossible that in the future I’ll buy one of the RP’s font for a project. I should be able to build an identity card on my computer of this commercial font to let me chose it when I am searching for the right font. I don't own it, but I know it, so I should be able to find it back. I don’t need any compiled files or sources for a typographic casting.
If at any moment I find that Helvetica is similar to Free Sans, I should be able to find or create the Helvetica METADATA.yml, with at list one specimen image.
And what if the connection is lost between the METADATA.yml and the actual font file containing the license, authors, version…

I am afraid the deletion of those informations will limit the project. Il also believe that UTD should be cross-licenses and independent from the source and/or compiled files. I see UTD like a tool for type designers, font users and developpers because duplicating the font informations into the METADATA.yml also means an easyer way to access to the data (writing and reading).

I also need feedback about this idea so feel free to critic and share your questions and opinions.

[davelab6]

I think its okay to have 2 aspects to the project. The larger one is how to organize the source files of a font family. A subset of that is how to organize the binary files of a font family. Indeed, there are many libre fonts without source files (like DejaVu, Droid, Open Sans, Noto, Liberation....)

I should be able to build an identity card on my computer of this commercial font to let me chose it when I am searching for the right font.

I'm curious: What software will you use to search for the right font? :) What software will index, display and search these 'ID cards'? :)

what if the connection is lost between the METADATA.yml and the actual font file containing the license, authors, version…

The files are primary. If they are lost, the METADATA.yml isn't worth much, I think.

duplicating the font informations into the METADATA.yml also means an easyer way to access to the data (writing and reading).

Duplicating metadata inside the fonts means that the connection can be lost more easily. I would prefer to make accessing the data inside fonts easy for everyone; there are many libraries that make it easy to do this, in many languages, and this project can provide value to the users you see it for by giving them a nice list of the libraries and example code to do that.

[raphaelbastide]

I should be able to build an identity card on my computer of this commercial font to let me chose it when I am searching for the right font.

I'm curious: What software will you use to search for the right font? :) What software will index, display and search these 'ID cards'? :)

Such a software dosn’t exist yet. There is I think, two kind of software that must be designed:

1. At first, on the side of the type designer, a software that can help to build those ID cards. For now, it can be nothing more than a form to fill. Maybe Bakery will do that? (not sure I got what bakery is yet).
2. Then, one or several services or software will be needed to search and display the fonts data. For a first prototype, it can be a piece of code that browse a GitHub repository containing around several UTD font pakages.

In the future, I thik UTD can inspire web services such as http://openfontlibrary.org/, http://theleagueofmoveabletype.com/, http://typedia.com/, http://fontsquirrel.com/, independant foundries, and all the font collections that suffer from a lack of folksonomy (licenses, tags, font relations). GWF also needs this kind of tool now it is a huge collection.

That is why UTD must be flexible enouth to ensure those différents goals.