Dear sir,
Fist of all, I'd like to thank STIX for releasing its fonts for public
use. We've been distributing it as a default component in Fedora
Linux since Fedora 10: http://fedoraproject.org/
They will probably percolate from Fedora to Red Hat Enterprise Linux
this year.
We are also very happy about your decision to drop a custom licence for
the standard OFL font licence.
I regret that you didn't change your font naming to something more human
friendly during the beta. Applications have been tested against Times
New Roman for years, there is no reason to stick to Postscript-like
whitespace-free names any more.
In a related problem, STIX declares a lot of different font names,
consuming a lot of precious space in application font lists. Could you
please try to cut down a little here? Or at least provide guidance to
distributors on the minimal font set they can distribute as default that
is sufficient to display engineering text without taking so much place
in font lists?
Lastly, a plain text *.txt licensing file would be appreciated, as the
current PDF one is a bit bulky.
Now, as part of Fedora's ongoing QA work, we run automated sanity tests
on all the font files shipped in the distribution. It is very important
that all our fonts, especially those installed by default, pass those
tests with minimal warnings. Our ability to collect font problems in the
field is limited (most users do not report the font problems they
experience) and we have to rely on automated testing.
Moreover we've found out that most foundries do not test for the same
problems as us. Therefore our testing usually complements nicely the one
performed by font authors, and makes it a free service to thank them for
releasing Free/Libre/Open fonts.
This testing found several problems in STIX fonts. This is not surprising
for such a young font set, but we'd be grateful if you could look at them
and confirm we made the right choice in selecting STIX as a default
typeface.
I've attached the full test logs to this message. As a quick summary, we've
found two classes of problems that need fixing your side:
* Various fontlint errors.
Fontlint is the font testing suite created by the fontforge
project.
http://fontforge.sourceforge.net/fontlint.html
Fedora, OpenOffice.org, and others, have been adding new tests
to fontlint whenever a bad font broke an application on our
platforms.
(see the fontlint data files in attached logs)
* Partial script coverage.
We understand that creating a typeface that covers the whole
Unicode range is an enormous amount of work and not something
that can reasonably be asked of most font authors. Therefore our
test limits itself to detecting almost-complete scripts, and
only triggers when less than ten glyphs are necessary to
complete one or more of them (very often it triggers on the few
diacritics necessary to support minority languages).
Thus our test identifies areas where only a little more work can make
the font files useful for many more people. Please use this
opportunity to improve your fonts.
(see the fc-query data files in attached logs, and the explanation in
summary.txt)
We also run unicode block coverage tests, but they are a
lot less useful in practice
(see the unicover data files in attached logs)
I hope this feedback will be of some use to the STIX projet. Once again,
thank you very much for releasing this wonderful free/libre/open font
set!
Best regards,
--
Nicolas Mailhot
Fedora fonts special interest group
http://fonts.fedoraproject.org/
Dear all,
As requested by many during the STIX fonts beta review phase, the STIX
Fonts project changed its licence to the OFL (was using a custom
one-of-a-kind licence initially). The change is effective in the STIX
fonts 1.0.0 release, which has been pushed in Fedora today:
http://koji.fedoraproject.org/koji/buildinfo?buildID=186404
This should be fine as the OFL is Fedora's preferred font licence (with
the GPL, provided the standard font exception is not missing)
http://fedoraproject.org/wiki/Legal_considerations_for_fonts#Preferred_Lice…
Best regards,
--
Nicolas Mailhot
Dear sir,
Fist of all, I'd like to thank ParaType for releasing PT Sans for public
use. We've started distributing it as a default component in Fedora
Linux since Fedora 13: http://fedoraproject.org/
Depending on the feedback we receive it may find its way in derived
Linux distributions (OLPC, Red Hat Enterprise Linux, CentOS, Oracle
Enterprise Linux, etc).
I must say that your release of PT Sans under the OFL font license helps
a lot. The OFL is a licence we know and respect, and our legal people
are not happy about one-of-a-kind licences such as the PT Free Font
Licence you used at first.
Now, as part of Fedora's ongoing QA work, we run automated sanity tests
on all the font files shipped in the distribution. It is very important
that all our fonts, especially those installed by default, pass those
tests with minimal warnings. Our ability to collect font problems in the
field is limited (most users do not report the font problems they
experience) and we have to rely on automated testing.
Moreover we've found out that most foundries do not test for the same
problems as us. Therefore our testing usually complements nicely the one
performed by font authors, and makes it a free service to thank them for
releasing Free/Libre/Open fonts.
This testing found several problems in PT Sans. This is not surprising
for such a young font, but we'd be grateful if you could look at them
and confirm we made the right choice in selecting PT Sans as a default
font.
I've attached the full test logs to this mail. As a quick summary, we've
found three classes of problems:
* Bad font metadata.
PT Sans font naming does not respect Microsoft's WWS font naming
guidelines. It is a big problem for us as Linux does not perform
automated font name fixing like Windows Vista or Seven. We
suggest the following naming change in your OpenType/WWS
metadata fields:
PT Sans Narrow, Bold → Pt Sans, SemiCondensed Bold
PTN77F.ttf [paratype-pt-sans-fonts]
PT Sans Narrow, Regular → Pt Sans, SemiCondensed
PTN57F.ttf [paratype-pt-sans-fonts]
See also:
http://blogs.msdn.com/text/attachment/2249036.ashx
* Various fontlint errors.
Fontlint is the font testing suite created by the fontforge
project.
http://fontforge.sourceforge.net/fontlint.html
Fedora, OpenOffice.org, and others, have been adding new tests
to fontlint whenever a bad font broke an application on our
platforms.
(see attached logs)
* Partial script coverage.
We understand that creating a typeface that covers the whole
Unicode range is an enormous amount of work and not something
that can reasonably be asked of most font authors. Therefore our
test limits itself to detecting almost-complete scripts, and
only triggers when less than ten glyphs are necessary to
complete one or more of them (very often it triggers on the few
diacritics necessary to support minority languages).
Our test identifies areas where only a little more work can make
the font files useful for many more people. Please use this
opportunity to improve your fonts.
(see fc-query results in attached logs, and the explanation in
summary.txt)
I hope this feedback will be of some use to ParaType.
Best regards,
--
Nicolas Mailhot
Fedora fonts special interest group
http://fonts.fedoraproject.org/
Dear all,
At the request of various free/libre OSS organisations, ParaType has
agreed to release the PT Sans font under the OFL in addition to its
one-of-a-kind custom font licence. PT Sans was created with funding from
the Russian Federal Agency for Press and Mass Communications.
The OFL version is now the one packaged in Fedora:
http://koji.fedoraproject.org/koji/buildinfo?buildID=186380
This should be fine as the OFL is Fedora's preferred font licence (with
the GPL, provided the standard GPL font exception is not missing)
http://fedoraproject.org/wiki/Legal_considerations_for_fonts#Preferred_Lice…
Regards,
--
Nicolas Mailhot
Fedora fonts special interest group
http://fonts.fedoraproject.org/
Hi,
Fist of all, I'd like to thank you for releasing the wonderful Yanone
Kaffeesatz for public use. We've been distributing it as an optional
component of Fedora Linux (http://fedoraproject.org/) since 2007 and
it's been a pleasure to see it evolve.
Your latest re-licensing from CC-BY to the OFL is very much appreciated
(like other free and open source organisations, we consider the OFL one
of the best existing font licences). However I note that your
latest .zip release still does not include detached OFL.txt and
fontlog.txt files. That's a pity as many users won't dig in font
metadata for information and our own sanity rules forbid adding
licensing documents when they're not provided by the original author
(it's too easy to accidentally betray the author's intent when you write
those in his place).
Therefore, I'd suggest adding those files to make sure you're justly
credited when your work is redistributed by third parties.
Secondly, as part of our ongoing QA work, we've been running automated
checks on the font files included in Fedora Linux. Yanone Kaffeesatz
seems to fail some of those. Could you take a look at it please? I've
attached the detailed logs to this message, they include the output of
fontlint and of the coverage tests.
P# t17 t19
1 4 4
Total 4 4
P# SRPM RPM EVR Arch
1 yanone-kaffeesatz-fonts yanone-kaffeesatz-fonts 0:20100514-1.fc14 noarch
Test explanation:
t17. Warning: fonts that do not pass fontlint sanity checks
☛ Font upstream task
Fontforge's fontlint¹ test suite found problems in some files included in
the package. Those problems may not be obvious and only manifest as
strange behaviour in specific applications (making them hard to debug).
For that reason it is recommended to report those problems upstream and
get them fixed, even if the font file seems to work fine most of the time.
You can ask help about specific fontlint errors on:
https://lists.sourceforge.net/lists/listinfo/fontforge-users
¹ http://fontforge.sourceforge.net/fontlint.html
t19. Suggestion: fonts with partial script coverage
☛ Font upstream task
Some font files included in the package are missing a few glyphs to be
accepted by fontconfig as covering one or several scripts. Therefore they
could be made useful to more people with only a little effort.
Many scripts differ by only a few glyphs and it is unfortunately common
for font authors not to notice they stopped just short of full support for
some of them.
To check a font file script coverage, run:
$ FC_DEBUG=256 fc-query font-file
and look for lines like:
script-id¹(number) { list-of-unicode-codepoints }
For example
“mi(2) { 1e34 1e35 }”
means fontconfig will accept the tested file for Maori if codepoints 1e34
and 1e35 are added.
fontconfig is used by a lot of applications on many platforms so ignoring
its opinion on a font is a mistake.
Best regards,
--
Nicolas Mailhot
Fedora fonts special interest group
http://fonts.fedoraproject.org/
Hi,
2010/7/18 Hans de Goede <hdegoede(a)redhat.com>:
> Hi,
>
> On 07/18/2010 11:51 AM, Parag N(पराग़) wrote:
>>
>> Hi,
>>
>> On Sun, Jul 18, 2010 at 6:33 AM, Kevin Fenzi<kevin(a)scrye.com> wrote:
>>>
>>> On Sat, 17 Jul 2010 23:35:26 +0200
>>> Sven Lankes<sven(a)lank.es> wrote:
>>>
>>>> On Fri, Jul 16, 2010 at 10:09:37AM +0100, Paul Flo Williams wrote:
>>>>
>>>>> I imagine that Nicholas raised this problem while doing his periodic
>>>>> fonts-incorrectly-packaged-in-non-font-package checks. The request
>>>>> to separate out the fonts from poker3d-data has been long since
>>>>> closed WONTFIX:
>>>>> https://bugzilla.redhat.com/show_bug.cgi?id=477443
>>>>
>>>> I just went through all bugs blocking F11-new-font-rules that have
>>>> been closed recently through the bugzapper EOL procedure and reopened
>>>> those which are clearly still valid.
>>>>
>>>> A rather large number of packages never had a single reply from one of
>>>> the maintainers.
>>>
>>> Sad. ;(
>>>
>>>> Do we have any provenpackages who volunteer to be CCed on bugs that
>>>> have patches attached? (Not that there currently are any such bugs
>>>> with patches but reopening I have found some low hanging fruits which
>>>> I'm planning to look at in the future).
>>>
>>> We should be careful here... if the packagers are really gone and no
>>> longer maintaining we should orphan the packages so people can take
>>> them over, not keep drive by maintaining them without being very
>>> involved. :) IMHO.
>>>
>>
>> I think those packages whose spec's need to be re-written should block
>> F14Target and volunteer's are welcome to post patches/fix them.
>> Recently I come across few packages which need to follow current fonts
>> packaging guidelines. I start working on those but found
>> openfontlibrary.org is almost broken and for those fonts it is
>> upstream. I am now searching those fonts if have some other publisher.
>> If i will find any other foundry then will use and import it in Fedora
>> otherwise my suggestion will be to retire those packages whose
>> upstream are gone now.
>
> Unless I read the thread wrong openfontlibrary.org will return. Also I see
> no reason to drop packages just because there upstream is gone.
Current fonts guidelines asks for using foundry name in font
package. If based on initial import where upstream URL is no longer
exists but its allowed to use that and use oflb as foundry name then I
am more willing to work on it.
Please tell me if this is ok. If you think its ok I will submit
new, to be renamed packages for package review.
If this is not ok then let those bugs be open forever.....
> Esp. for
> something like a font. If its a nice font and reasonably complete what sort
> of maintenance would you expect from upstream ?
Here problem is not maintainer required from upstream but should
foundry oflb is ok to use for a dead URL?
I mean once Rembrand had
> finished the Night Watch, it was well finished. I don't see how a font is
> much different.
>
I don't understand this...
Parag.
> Regards,
>
> Hans
>
I have been investigating bug 536920, which is a crash in fontlint when
running against a font file included in poker3d-data.
This font file is neurpoli.ttf, which turns out to be one of the Larabie
fonts. Larabie fonts cannot be packaged because of the restrictions on
modification:
https://fedoraproject.org/wiki/Licensing:Main#Bad_Licenses_4
I imagine that Nicholas raised this problem while doing his periodic
fonts-incorrectly-packaged-in-non-font-package checks. The request to
separate out the fonts from poker3d-data has been long since closed
WONTFIX:
https://bugzilla.redhat.com/show_bug.cgi?id=477443
But actually, it seems that poker3d-data couldn't have this font separated
out anyway.
Should I be raising a bug against poker3d-data along the lines of "you
need to remove non-free font from package"?