For Approval: IPA Font License v1.0
bruce at perens.com
Thu Feb 12 19:17:44 UTC 2009
Given your explanation, It seems to me that the premise of the IPA font
license is that some utility for automatically dealing with difference
files in fonts will become a standard part of all computer systems.
Currently, this is not an automatic facility that is included with BSD
or Linux desktops that I am aware of, or Macintosh or Microsoft systems.
All of those systems have a list of font formats that they can presently
accommodate. On loading a combination of a font and a difference file,
the systems I have mentioned would fail to use the font. Currently, the
users of these systems would be required to operate command-line tools
of considerable sophistication in order to transform the font into a
The easy solution for this is to provide the font in the format expected
by these various systems, rather than expect the user to install special
software. A license that prohibits such a common and expected
operation would, I believe, be rejected. I do not know of any
OSI-approved licenses at present that make a similar restriction.
I urge you to take a wider view regarding the possible derivative works.
Consider, for example, the creation of a new Unicode font. Certainly it
would be to everyone's advantage for Unicode fonts worldwide to render
kanji according to IPA. Such a Unicode font would be derived from the
IPA font, and a number of other fonts as well. The overall size of the
combined work might hold many times the number of glyphs in the IPA
font. But the license you propose would restrict such a font to being
encoded as the IPA font and a much larger difference file, and it could
not be converted between formats for distribution. It seems that this
would make the use of IPA's font untenable for the creators of such
works. Surely it is to the advantage of IPA and the Japanese people for
this sort of work to be possible.
I understand your reluctance regarding the potential large impact of
small changes. Please understand that you are not the first to make such
an argument, and that Open Source licenses permit modification in other
potentially dangerous situations. For example, we have operating system
device drivers which can physically damage the computer equipment if
they are incorrect. If OpenStreetMap is incorrect, an application using
it can guide the driver of an automobile and its occupants into a
life-threatening situation. There are many such works under Open Source
licenses that permit the distribution of directly-executable versions of
The major restrictions on modification of the executable file, rather
than the source file, that OSI has been willing to accept in the past
are restrictions on trademark use and a requirement to rename derivative
I have not seen similar licenses accepted in the past, so I would
suggest that the prospect of OSI acceptance is unlikely if this license
is not modified. But I think this is an important project, so I urge you
to make the necessary modifications and obtain acceptance of the
license, or to use one of the existing approved licenses.
More information about the License-review