On Thu, 17 Feb 2011, Vicente Guerra wrote: Hi,
> > I change this lines to solve the problem at hbdbsort.c,clipper may work in > > this way, but I think it's a bug that should be corrected > If Clipper works in "bug mode", fixing it would cause > index corruption on shared environments. The whole current sort code (SORT ... command) for DBF based RDDs has to be replaced by the new one practically written from scratch for three reasons: 1. current code has many small incompatibilities to Clipper (i.e. string comparison but not only) and they have to be fixed 2. it works correctly only with few field types so it's not fully functional 3. it's _very_ slow. In practice it's impossible to use it with very large DBF files due to time cost. It's necessary to redesign it from scratch and create new sorting engine, i.e. sth like the ones I created for indexing (CDX, NTX, NSX). It's very long on my TODO list but it has been never critical for me (I've simply used temporary index with COPY TO as workaround) and I haven't implemented new SORT() method for DBF RDD so far. Any one who wants to create new SORT() code for DBF RDD is greatly welcome. best regards, Przemek ------------------------------------------------------------------------------ The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb _______________________________________________ xHarbour-developers mailing list xHarbour-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/xharbour-developers
