Miriad Future

From CARMA
(Difference between revisions)
Jump to: navigation, search
(Infrastructure)
(Infrastructure)
Line 22: Line 22:
 
== Infrastructure ==
 
== Infrastructure ==
  
* usage of '''ptrdiff_t''' in fortran. Tools like ratty and intf2c have been updated for this. This would allow memory allocation above 8GB.
+
* usage of '''ptrdiff_t''' in fortran. Tools like ratty and intf2c have been updated for this. This would allow memory allocation above 8GB. Note as of about Jul/Aug 2009 this is in a confusing state as the newly submitted versions of ratty, flint and intf2c by Sault are slightly different. This will hopefully be reconciled in an upcoming merge. Note the new version will use the '''ptrdiff''' type, not '''ptrdiff_t'''.
  
 
* the old style shell scripts that [[Installation from source|install]] and maintain Miriad will be replaced by an [[Autoconf based installation|autoconf]] based system. Currently this is mostly working, except for WIP and RAD and the hkmiriad/ZENO tools
 
* the old style shell scripts that [[Installation from source|install]] and maintain Miriad will be replaced by an [[Autoconf based installation|autoconf]] based system. Currently this is mostly working, except for WIP and RAD and the hkmiriad/ZENO tools

Revision as of 09:08, 9 September 2009

The CARMA bugzilla (you need an account for this) will contain a list of bugs and enhancements requests that are closer to being implemented than this list.

Contents

Tasks

  • uvimage currently just writes a visbrick image with no WCS. Once that is written, tools can use this to intelligently write region files for flagging using uvflag
  • uvclean: a new proposed task that would use the clean algorithm and for major cleaning cycles would subtract from the original visibilities (or the gridded one, not decision taken on this)
  • maxen: measure=cornwell does not seem to work (moved to bug 632)

Library

  • cleanup or clarify the MAXDIM, MAXDIM2, MAXCHAN, MAXCHAN2, MAXANT, MAXANT2 confusion? Or use better memory management?
  • python/ruby interfaces (SWIG): there are 3 python interfaces and 1 ruby interface:
    • AIPY, Aaron Parsons' python interface [1]
    • the WSRT python interface (in $MIR/src/scripts/python)
    • Peter Williams' python interface (via git), see also [2]
    • Dave MdMahon's ruby interface [3]
  • uvio enhancements

Infrastructure

  • usage of ptrdiff_t in fortran. Tools like ratty and intf2c have been updated for this. This would allow memory allocation above 8GB. Note as of about Jul/Aug 2009 this is in a confusing state as the newly submitted versions of ratty, flint and intf2c by Sault are slightly different. This will hopefully be reconciled in an upcoming merge. Note the new version will use the ptrdiff type, not ptrdiff_t.
  • the old style shell scripts that install and maintain Miriad will be replaced by an autoconf based system. Currently this is mostly working, except for WIP and RAD and the hkmiriad/ZENO tools
  • a set of procedures to use generated (uvgen, imgen) and teal (CARMA, SMA) telescope data to ensure that Miriad keeps on working as it used to work.
  • attempt to keep some level of correspondence with the ATNF version of Miriad
  • switch the current CVS to SVN, possibly in a neat environment for bugtracking etc.etc. An example could be the trac

Other

  • BIMA compatibility: some tasks (e.g. listobs) have turned from very BIMA specific, to more general or perhaps somewhat CARMA specific. A task like gapply has a bima= keyword that will need to be used in case of BIMA calibration. There may be some tasks (e.g. uvwide) that need to be patched to make it work again for BIMA data
Personal tools