Actually managed to get some work done today - What I've got done will be going up in CVS tomorrow soon's I test it to make sure I haven't broken anything obvious. esearch and esummary web interface and xml results are now properly split into separate files. The "generic" xml parser got extended such that it should now handle text results as well as filehandles (needed for the NCBI BLAST interface) I THINK I've worked out all the quirks I needed to get around that I noticed in my earlier NCBI BLAST tests (i.e. when getting XML format results, the "QBlastInfo" part of the response doesn't exist as it does in text or HTML format results, got the "send POST query to webserver" part split off and working through the postbox class, etc. etc.). Haven't got the actual XML results parser written yet, but as it's just an extension of the "generic" XML parser, I don't think that'll take long. The updates to esearch, esummary, and the generic xml parser should all be on CVS tomorrow morning (US West Coast time) along with HOPEFULLY a functioning NCBI BLAST set of code. Nico - can you give me some hints as to your planned structure for filetype modules in the seqIOWrite section? One of the things I can work on while on the road are output modules for Phylip and Clustal (for example). I'd also like to work up a simplistic "compare the quality of the match between two aligned sequences" algorithm, just for quick automated "throw out bad matches" sort of applications. More tomorrow. Bed calls.