Today, I realized that the Golden Database ratings data file (Feb 06) doesn’t seem to contain information about established versus provisional players. Could I be mistaken about this? If a tournament has limitations on provisional players’ eligibility for certain prizes, you might not be able to rely wholly on the electronic database, but instead have to double check the web or the paper copy. I thought at one point the Player Type column used to have E’s and P’s for established and provisional.
Yes, it appears that the type fields are blank in the February Gold Master file.
I found a second problem with the February Gold Master file, involving some records that have blanks in the expiration date field. I’m not sure what caused that, I’ll have to check the program.
I should be able to fix both problems.
However, we’re in the middle of the update cycle for the June Supplement today, so I can’t generate a new Gold Master file at this time. Does it make more sense to wait until we’re ready to release the June supplement files and release a new Gold Master then, one that corrects both problems?
The Gold Master format has another weakness, it does not have a place for the provisional game count. That’s not something I can correct.
I’m hoping to finalize a new rating supplement update file format around the same time that I finish the rating reporting file format, so that the pairing program authors can update their programs to deal with both of the new files at the same time.
This new format will have a longer name field and will also contain status information (duplicate ID, deceased, etc.)
Mike,
Since it seems you will be phasing in a new format for the reporting tdexport, thexport, and tsexport files, does this mean that you will be phasing out the old format? Or will you continue to support both?
The local major tournaments use DOS-based ChessTD because of, let’s just say “particular ways” the TD likes to do pairings. I’ve been asked to write a Windows version, but I’m dragging my feet as usual. Perhaps a sunsetting deadline of format A would energize me.
The existing formats will continue to be available for several years, though I suppose that eventually we will want a sundown date on them.
Similarly, we will need to accept the old dBase rating report files for several years once programs can prepare the new format.
The problems with the Gold Master format have been corrected and a new Gold Master has been generated along with the August supplement file, both of which should be released in about a week.
Now I notice that in the August Golden Database, some members’ information is duplicated. I first noticed when I looked for Robby Adamson, but it seems to extend to countless others. One of the duplicate entries seems to lack the E/P information while the other has that information.
Sorry to nitpick.
The copy I have of the August 2006 gold master file has 638,770 records on it.
I downloaded the copy from the website and checked it. It is larger than the original file and has 639,770 records. 1000 IDs have gotten duplicated somehow. I don’t know when or how this happened, but as I recall I did correct a formatting problem with the gold master format in July, and I’m not sure at this point if my copy is intact in those other fields.
I’m going to have to do some testing of the program that generates the supplement files before I can build a new gold master.
I’ve removed the August gold master file from the website for now.
A corrected version of the August Gold Master file is now available.