> Today I downloaded a book in HTML format and converted it using the > parser from CVS. No visible problems during the parsing, but when I > tried it in the viewer most of links would end up in the external > link dialog. Taking a closer look at the Plucker document it turned > out that the records were in a "random" order. They have to be in an > increasing UID order or the viewer can't find the records. I don't > know how long this problem has been in the code, but it has to be > fixed before we can release a new version of Plucker...
Sorry, Mike, I don't quite understand this. > On Sat, May 04, 2002, Michael Nordstr�m wrote: > > Taking a closer look at the Plucker document it turned > > out that the records were in a "random" order. > > Some more investigation reveals that the reason it looks like they > are in "random" order is because some of the records has not been > assigned the correct UID, but instead seems to have got "random" > values. The UID that's part of the PDB format has the correct value > for each record, but the viewer only cares about the UID inside the > record... You're saying that the Plucker UID value stored inside the data record header (the first two bytes of the record) have values other than those expected by the links? Or are you saying that the UID values of the records in the documents are simply not monotonically increasing? There's no requirement in the DB format spec that the UIDs of records physically laid out later than other records must also have higher UID values. Do we need to add this requirement? Better point me at this book, too, so that I can reproduce the problem. Bill
