Igor Tandetnik,
>>> Explain the problem you are trying to solve, *not* your proposed solution.
>>> <<<
What we are trying to achieve is to to find the minimum row id for each
unique Field Name in BLobLastNameTest where many rows can have the same
FIELDNAME but distinct BLOBS(Vertices Column).
Once we know the first/minimum row id of each unique fieldname group,
we would like to write a sqlite UPDATE Statement to accumulate all the BLOB
vertices of all the rows with that unique fieldname into the first(i.e
MIN(ROWID)) row's BLOB(Vertices)column for each unique fieldname group.
The reason we can't use select FieldName, rowid from BlobLastNameTest is
that it would slow our C++ Windows/LINUx/Solaris UNIX worker threads so much
that it wouldn;t be worth multithreading this whole process. We were hoping
that the SQLITE query processor and the appropriate indexes could accomplish
these previous 2 steps in a few minutes for a reasonable number of
BLOBLASTNAMETEST rows. Thank you for all of your help.
_______________________________________________
sqlite-users mailing list
[email protected]
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
