It's getting to that stage where io_lib is well overdue for a major overhaul. Not least is the point that the ZTR format can cope with much more than the core Read structure allows, so I propose that the new format has a totally different Read structure (with backwards compatibility mode perhaps).
It is also very inefficient when reading lots of formats due to the insistance on reading from disk. In-memory decoders would streamline the process immensly.
Finally there's the issue of how to store data from different types of instruments or different styles of traces. Eg pyrosequencing data or SSCP analysis. Even the standard AB1 format has additional data (raw vs processed) which we'd like to support.
I'm at the planning stage now, but please contact me (jkbonfield at users.sourceforge.net) if you have pet peeves you'd like addressed or if you have general feedback on things to include.
Log in to post a comment.