On October 17, 2015, International Master Emory Tate was playing at the tournament in Milpitas CA. It was about an hour into the game when Tate had a heart attack and was pronounced dead later that night. The game was rated as loss for Tate.
IM Emory Tate was a vibrant presence in the chess world and is fondly remembered by many MSJE alumni. Below are a couple of photos taken by Chris Torres showing Tate in the MSJE multi-use room:
As we take time today to remember his impact on the game and community, it is deeply disappointing to see that FIDE deleted his profile ( https://ratings.fide.com/profile/2001365 ). A player’s FIDE profile serves as a record of their life’s pursuit and achievements. Deleting that profile page is erasing a piece of chess history. When will FIDE undelete profiles of deceased players?
I don’t know if FIDE will ever undelete any profiles. Some years back I wanted to look at IA-A Carol Jarecki’s tournaments and her profile was also deleted.
I understand that a lot of extra records in a database can make queries slower, particularly queries that do not go off of main keys and thus look through a larger portion of the database. Maybe there could be a separate deceased database that could be used for those types of searches.
Besides the cost of writing programs for a deceased ID database, the cost of additional storage might also be an issue. I’m not sure how much or how little room there is in the FIDE budget.
Some modern databases can be set up to do this type of segmentation so the query predicate excludes the scanning through the deceased/inactive based on context.
I am reminded of Wintel server requests using the AS/400 as a data server and taking about 20 times as long to do a query over a small set of records as a native AS/400 query would take over 25 times as many records (i.e. native queries were 500 times as fast). Granted that was in 2000 and we were able to speed things up a bit by creating logical views on the AS/400 that allowed the incoming queries to use views that excluded a LOT of records from the initial selection, but the native queries were still significantly faster when I retired in 2022. The native AS/400 queries had both the data and the query logic on the same machine without needing to first pull data from the data server (AS/400) to the wintel server for selection analysis.
2000 also saw the consultants wondering why we were complaining about the period need to reboot the wintel servers on relatively short notice (about 3-6 times per month) since that was just the expected need for such systems. After 6 months they realized the AS/400 had never had any reboot other than the scheduled one on Sunday morning, and that had been true for years. That is when they learned how different expectations were. For that matter there was still a need for periodic mid-day reboots even in this decade and the only way they avoided extreme outrage was by running multiple wintel servers at once so that when one was being rebooted the others could still handle the load, albeit at a reduced speed. When they finally move off the AS/400 to the cloud they’ll find out how quickly things run when a cloud connection is required for everything.
