P.S: I pushed the code the other night if anyone wants to take a look. There are still some issues that I'm working on and I'm trying to get the layer's imagespecs filled in.
On Fri, Jun 24, 2011 at 11:21 PM, Daniel Wyatt <[email protected]>wrote: > I don't know yet if there's anything to really be gained by using their > library. > A pugixml impl may be more feasible. > Either way, as it stands now, I'll use what's already there. > > On Fri, Jun 24, 2011 at 10:39 PM, Larry Gritz <[email protected]> wrote: > >> Hmmm... remind me what problem are we supposed to be solving by >> incorporating Adobe's library? Is there some XMP data we're not reading >> correctly now that an OIIO-using application would need? If it doesn't fix >> a bug or significantly extend functionality, and do so with less effort to >> swap out the library than to fix our code, I think we should leave well >> enough alone. >> >> -- lg >> >> >> On Jun 24, 2011, at 7:21 PM, Daniel Wyatt wrote: >> >> Ok. I'll use the existing stuff for now. >> Quick note: their library requires expat (which I guess would be redundant >> since we have pugixml) and zlib. You have to place the source in >> third-party/expat/lib and third-party/zlib/zlib which is a little odd. Not >> sure if linking to a shared object is possible or not. >> >> On Fri, Jun 24, 2011 at 10:07 PM, Larry Gritz <[email protected]> wrote: >> >>> Let me look at this library and get back to you. The regex is admittedly >>> awkward, and we've talked about removing it and using actual XML parsing now >>> that we have PugiXML as a dependency elsewhere. It's not entirely clear to >>> me what the tradeoffs are of using their library, but I'll poke around. Use >>> what's in OIIO for now; we can always swap out the implementation of the XMP >>> stuff later. >>> >>> -- lg >>> >>> >>> On Jun 24, 2011, at 5:34 PM, Daniel Wyatt wrote: >>> >>> As for Exif/XMP, I wanted to ask a question about that: >>> Is there reason not to use Adobe's XMP library ( >>> http://www.adobe.com/devnet/xmp.html)? >>> It is BSD licensed. >>> I saw the current regex hackery and just thought it might be easier. >>> It /is/ kind of a pain to build and required adding in some missing >>> includes (cstring at least) but it might be worth looking into. >>> I can use the existing code for now though. >>> >>> >>> >>> -- >> Larry Gritz >> [email protected] >> >> >> >> _______________________________________________ >> Oiio-dev mailing list >> [email protected] >> http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org >> >> >
_______________________________________________ Oiio-dev mailing list [email protected] http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
