On Wed, 21 Jan 2009 13:22:42 -0500, Gerhard Postpischil wrote: >Tom Marchant wrote: >> I spent a *lot* of time in the microfiche, reading the CVOL code. Whatever >> the reason was for concatenating the generation data sets in reverse order, >> I don't think it was for performance. > >The names were stored in the catalog in inverse order (the >"nnnn" portion was complemented; i.e. C'0123' became >X'0F0E0D0C'). This placed the entries with latest generation >first, and explains why it was easiest to retrieve them in that >sequence.
Yes, I'd forgotten that detail. Still, it begs the question, *why* was it stored that way? I suspect that it was a design decision to simplify the retrieval of the data sets in reverse order. I don't think that either forward or reverse order would offer a significant performance advantage. -- Tom Marchant ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
