Hi Pascal,
FontBox is a small library within the PDFBox project. PDFBox is an
active Apache project, so getting patches committed there shouldn't be a
problem. Jeremias is a committer on PDFBox too. FontBox is no longer
hosted on Source Forge, as it was moved to Apache a few years back;
which is probably why there hasn't been a recent commit to the
SourceForge version. Last commit to the Apache version was still 2 years
ago, see: http://svn.apache.org/viewvc/pdfbox/
I think it makes sense to reuse code wherever possible, especially when
we are talking about fellow Apache projects. So I'm +1 to this proposal.
Thanks,
Chris
On 10/01/2013 09:12, Pascal Sancho wrote:
Hi,
unless I'm missing something, fontbox dev activity is quite slow
(latest commit was on 2007-10-01, 6 years ago, see [1]).
IMHO, introducing a dependency on such project witch will need some
improvement is not a good thing, unless we can ensure that submitting
patches to it will be applied on demand.
[1]
http://sourceforge.net/projects/fontbox/stats/scm?repo=CVSRepository&dates=2007-01-10+to+2013-01-10
2013/1/9 Robert Meyer <rme...@hotmail.co.uk>:
Hi All,
Unless someone else has been developing this in secret, I am planning to
make a start on adding support for OTF CFF (Open Type - Compact Font
Format). There are two choices I can see which are available and would like
to ask for your opinion. These are:
1) Using the implementation from fontbox and write adapter classes to allow
it to work with FOP.
2) Write a dedicated FOP implementation.
There are pro's and con's to each. Firstly, using fontbox will create
another dependency to FOP which is generally never a good thing. It will
also means if there is a problem with their implementation, we have to rely
upon them to commit the patch (either written by us or by themselves). I
don't know what their uptake is on committing patches, but unlike FOP the
control would be taken out of our hands.
Saying this however, using their implementation will cut the development
time as the majority of work will already have been done. There is also the
advantage that their implementation will have been around for a while and
will have benefited from subsequent use and have ironed out any bugs.
If you have any other comments for or against each option please post them.
Regards,
Robert Meyer
