Report to ATUC, 2005-11-27 Software & Computing -------------------- * Neil Killeen departed ATNF on 5 Sep 2005 (two years leave of absence). * ASAP-1.2 was released on 30 November 2005. Some notable changes are - support for frequency-switched observations - building quotients from matched _S/_R pairs or 1*_R/n*_S automatically - numerous plotting improvements, including noninteractive plots - easier customisation, through $HOME/.asap/ configuration directory and support for user-written plugins - GUI interface for Mopra users to monitor data quality We plan to create an asap-users majordomo list, for release announcements and general discussion. Instructions for joining will be available on the ASAP home page. * Miriad support ATNF elected to reduce support for MIRIAD in the late 1990's, so that AIPS++ would get new features first and build up its user base. Hoewever demand for MIRIAD has remained strong, so we are now looking at how to improve support above the "bugfixes" level. To us, this means finding resources to write new tasks and add support for new instruments such as CABB. Over the last few years this has mainly been done by Bob Sault. Bob Sault will depart ATNF in the first half of 2006. Mark Calabretta will become the principal maintainer at ATNF. Vince McIntyre will continue as assistant maintainer. With Bob's departure, we are losing significant capacity for adding new tasks or features to miriad (eg support for CABB). In particular the finding and fixing of subtle problems relating to data acquisition problems will require us to find effort from somewhere. We are canvassing options for this at present. We are also looking into merging ATNF's code with the CARMA version, so the maintenance effort can be shared. The amount of work involved in merging appears reasonable. The benefits to be gained are not clear-cut however, ie at present we are unsure how much support CARMA is providing. More formal time-based releases may be needed because of the extra coordination required when more than one institution is involved. Presently, it is done by updating individual tasks and rebuilding the tar archive. End users can do updates almost automatically. It may be better to move to a time-based release system with email notification and a list of changes. The advantages and disadvantages of this are being considered now. A 'bleeding edge' version would be available too, possibly via the existing 'mirimport' update mechanism. Work is going on now to get miriad working on 64-bit Intel-like architectures (IA64, AMD64/EMT64) in the ATNF version. This should help users needing to create images larger than 2Gbyte. No work is going on to add 24-bit support to XMTV. We don't have available the expert effort this task will need. A public bug tracker will be available in January or February. This will enable users to review the status of bug reports and add their own. This will complement, not replace, the 'mirbug' mechanism. The at_bugs page http://www.atnf.csiro.au/computing/at_bugs.html will also remain. Question for atuc: is a miriad-users majordomo list useful? * Public Workstations we may need to delay upgrading a few months, owing to budgetary restrictions. * Computer support CSIRO IT is moving most of its staff off-site, to the Delhi Rd location. Fortunately, the Business Support staff (Tony Ambaum, Helen Stibbs) will continue to be based at Marsfield. Hristo Nikolov left in October 2005. We are working with CSIRO IT on a service level agreement, but progress has been slow. A significant problem with their service delivery model is that it can take hours to get responses to simple requests, because of the number of hands the requests pass through before getting to the person who will actually do something. We hope to get agreement that ATNF visitors can approach CSIRO IT staff directly for assistance with minor matters (e.g. password resets). Question for atuc: would a "fault reporting" system for the Marsfield site, similar to that at the observatories, be useful to have?