In article <[EMAIL PROTECTED]>, "Alexander H. Montgomery" <[EMAIL PROTECTED]> wrote:
> On 2008-03-24, at 11:29 AM, Christiaan Hofman wrote: > > > > On 24 Mar 2008, at 7:16 PM, James Howison wrote: > > > >> > >> On Mar 24, 2008, at 1:44 PM, Adam M. Goldstein wrote: > >>> > >>> On Mar 24, 2008, at 1:29 PM, Daniele Pontillo wrote: > >>>> Hmm, > >>>> and what about all the scientific publishers? > >>>> In my case, there are thousands of medical journals that could > >>>> benefit > >>>> from this issue. > >>>> Dan > >>>> > >>> > >>> I think this is what Christiaan means by "suppliers of PDFs," that > >>> is, > >>> the publishers on on-line articles and the like. > >>> > >>> If publishers in medicine aren't doing this, it's going to be a > >>> long, > >>> long time before anyone else does. Many of the important innovations > >>> in on-line databases of articles and other publications come from > >>> the > >>> medical world, because that's where the money is; they have better > >>> funding than other sciences, and certainly, than humanities. > >> > >> After I babbled about this for ages, without actually taking action > >> [1], I got a few calls from publishers interested in doing it, > >> including Nature. The stumbling blocks at the time were: > >> > >> a) Code to insert the XMP elements into the PDF (although I think > >> they > >> could have licensed Adobe's stuff easily enough) > > > > That's right, it requires you to have Acrobat Pro. IMHO, that > > requirement just kills XMP from ever to become useful. For me personally, this reduces the chance of embedding XMP on my own to zero. It should be a complete non-issue for publishers, though, since they're typically using Adobe Distiller/Acrobat to produce the PDF anyway. Even if they're not, considering the subscription rates of most scientific journals, I'd think they could afford it without breaking the bank! > There's a Java library to do it which JabRef uses; other > implementations haven't been coded yet. I don't know if the Java > implementation is hacked in or follows some official spec. There's also one for Perl, but ISTR it's a hack. The PDF spec is a monster, and inserting data in a PDF stream is non trivial. > >> b) A universal format for citation metadata to actually insert into > >> the file. AFAIK this still doesn't exist, they all have their > >> limitations. The feeling was that BibTeX was too limited field-wise, > >> Endnote was proprietary and hard for others to use. MODS was > >> considered to be coming close, but I haven't tracked things since > >> then. > > > > That was my point. Without an agreed upon standard there is no way to > > pass info in a reliable way, and nobody will do it. > > Endnote's import format (from Refer) isn't too awful, actually, but > like BibTeX, is showing its age. MODS XML looks promising, but > implementations are sparse. I agree with Christiaan that the lack of a standard is going to prevent this from happening any time soon. If/when it does, though, BibDesk can (mostly) import MODS XML, and I think the last XMP toolkit provides enough for us to read XMP from PDF files. ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Bibdesk-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bibdesk-users
