LHAPDF6 PDF sets ---------------- PDF set data files are available from download from this directory. Each is a .tar.gz file, which contains a directory with the name of the PDF, in which the .dat and .info PDF data files are located. To use a downloaded set, expand the tarball into a directory listed in your LHAPDF_DATA_PATH environment variable, or into the share/LHAPDF directory of your LHAPDF installation directory. The set tarballs in the top level directory of this download area are those "officially" supported for production use, i.e. they have undergone detailed validation against version 5 of LHAPDF, or have been created and validated specifically for LHAPDF6 by the set authors. The tarballs in the @UNVALIDATED directory are those which have been migrated from LHAPDF5 but have not undergone this detailed validation process. Our experience is that the agreement between LHAPDF versions is typically very good -- better than a 1/1000 fractional difference in PDF value in almost all regions of the (x,Q) space -- so we actually encourage you to try them out and let us know if you encounter any problems... or even if you find that they work brilliantly, which will help us to mark them as validated! In the spirit of correctness, however, please do not use "unvalidated" PDFs for "production" use, e.g. large scale LHC experiment Monte Carlo production: if you desperately want to be able to use a particular PDF with LHAPDF6, then please let us know and we will increase its validation priority. Some PDFs may remain in @UNVALIDATED in perpetuity, since they are non-standard or otherwise not recommended for general use. In the @DEPRECATED directory, you will find duplicates of PDF tarballs with data-version number just before the ".tar.gz" part of the filename, e.g. MSTW2008nnlo68cl.1.tar.gz as opposed to the normal MSTW2008nnlo68cl.tar.gz file. These are old versions of PDF data files which have been updated in the main downloads area. You should use the "normal" form, but the deprecated versions are for checking effects of these bugs, and their presence serves as a notification that a previous version of the data file has been found to have problems. To avoid huge space wasteage old PDFs will only be archived here if there are significant physics changes: minor updates of metadata will also produce new data versions but will not be archived unless there are actual bugfixes. We will be reviewing this versioning scheme as time progresses: please let us know your thoughts on how we can best let you know of bugs in PDF data. Best wishes, The LHAPDF6 Team