[daip] Comments from UK AIPS Users

ajc at star.rl.ac.uk ajc at star.rl.ac.uk
Thu Mar 22 12:01:45 EST 2001


Dear AIPS Supporters,

As part of its role of supporting the UK astronomy community in its computing
needs, the Starlink project recently conducted a survey of 'Software for
Astronomical Data Analysis' amongst the community (closing date 16 Feb
2001). A questionnaire was sent to all Starlink users (who are not 
necessarily users of Starlink software) and over 360 replies were 
received.

We thought you might be interested in the references to AIPS in the returns.
(A few references to AIPS++ also crept in.)

The attached document lists the relevant comments. The numbers in the 'Quest'
columns indicate different users - names have been removed to preserve
confidentiality. If you should want to reply to any specific points we can
probably arrange for contact to be made with the users. 

There are four sections:

1. Best and worst features of AIPS

2. General comments on AIPS

3. Reasons users changed to AIPS (nil return for this)

4. Reasons users changed from AIPS

We hope this may be of use to you.
 
Regards

--
Alan Chipperfield                     Email:  ajc at star.rl.ac.uk
CCLRC,                                Phone:  +44-1235-821900 ext 6165
Starlink Project,
Rutherford Appleton Lab.,
Chilton, Didcot, OXON, OX11 OQX
______________________________________________________________________
      This email represents the opinion of the author only and
      should not be construed as reflecting CCLRC's corporate view.   
----------------------------------------------------------------------
----------------------------------------------------------------------
   The contents of this email are sent in confidence for the use
   of the intended recipients only.  If you are not one of the intended
   recipients do not take action on it or show it to anyone else, but
   return this email to the sender and delete your copy of it.
----------------------------------------------------------------------
-------------- next part --------------
05.03.2001                                                              13:46:15

                       Report on Table: usage_features -- 1d
     1d -- 'Best and worst features?' (For the one or two applications which
            are most important to you, what are their best features and what
            are the worst?)

Soft  Best Features                          Worst Features

AIPS
00036 
      Command line interface, full record   Bugs in updates especially wrt LINUX
      of all processes performed, enormous
      flexibility and often very helpful
      (albeit idiosyncratic) documentation

00081 
      Familiarity                           User interface...

00115 
      Fast and generally reliable.          Not very simple documentation, no
                                            UNDO command!

00139 
      Comprehensive package for             User Interface is elderly and
      interferometric data reduction Vital  idiosyncratic
      for radio astronomy

00173 
      Powerful, can handle HUGE chunks of   lack of versatility; some bits are
      data                                  too primary

00216 
      fast, high quality algorithms         old 1960's still task input
                                            interface

00219
      Very powerful, 99.999% reliable       No good documentation. Cookbook is
                                            okay for learning, but reference
                                            manuals are needed.

00220 
      Widespread availability, so can be    Cumbersome internal image catalogue
      used to prepare and test image        system with insufficient length of
      archives for distribution throughout  names to permit distinct names for
      the community to ensure they will be  similar but  non-identical images.
      properly interpreted with coordinate  Generally rather cumbersome style of
      systems intelligible to a wide        operation throughout package.
      audience - especially useful for
      data originally synthesised and
      analysed with local software and
      subsequently required to be more
      widely distributed.

00223 
      Functionality                         Overwhelming

00245 
      It is very powerful (when you know    It is not always clear from the
      how to use it!).                      helps how to do something and it
                                            really requires you to be familar
                                            with it to get it to work well.

00374 
     generally reliable - few bugs.        No longer fully maintained.

00396 
      A near total package for radio        Outdated data input/output handling.
      astronomy.                            Very unreliable with DAT drives
                                            (here anyway)

00420 
      Suitability for purpose (radio        Old fashioned interface, especially
      interferometry).  Flexibility and     the online help.
      robustness of tasks, usable for
      non-radio-astronomy uses. Portable.

00486
      Easily accessible online help, nice   Awkward system management, getting
      command interface, excellent  image   tape support working can be
      display and hardcopy output.          problematic, cryptic error messages

00509 
      It works and it is a surprise if it   Very poor for data editing
      crashes

00517 
                                            Interaction with hardware is a real
                                            problem. eg Serious problems with
                                            tape drives at our site - cannot
                                            load data from the VLBA correlator.
                                            We've tried soliciting help - but
                                            no-one wants to know.

00532 
      Very comprehensive and flexible.      Some inconsistent terminology. Still
      Command-line interface is             a bit too VL(B)A specific. Some
      straightforward. Excellent automatic  documentation inadequate.
      records of most operations
      performed. Several layers of
      documentation. Well maintained in
      general.

00544
      Ubiquitous for data reduction         Self-calibration techniques not
                                            satisfactory

00547 
      Reliable, good applications, good     Difficult to write new applications
      documentation

05.03.2001                                                              13:46:16

               Report on Table: usage_comment -- 1b,2b,3b

     1b,2b,3b -- 'Any comments?' (about any of the software items in this
                                  Section)

Software
Quest  Comment

AIPS
00066
       It's so annoying that such an essential piece of software is so
       counterintuitive.

00163
       Dreadful user interface, but I guess that has been known for some time...

00396  
       Starlink could do more to install/support up to date versions of AIPS

AIPS++
00036  
       Although I do not use this at present it will become very  important in
       the near future when it is able to handle MERLIN and EVN data (and in my
       case spectral line data).

AIPS, AIPS++
00509  
       AIPS is currently the most important software package for MERLIN  users
       apart from Jodrell Bank 'internal' software.   AIPS++, not  included in
       the response since I have not used it, will presumeably  take over at
       some future date.

05.03.2001                                                              13:46:16

                       Report on Table: sources_change -- 1e
                                        (ordered by software changed from)

     1e -- 'Have you changed astronomical packages?' (Have you changed the
            choice of astronomical package you are using for your work 
            during the last two years? If so, what was the change and 
            why did you make it?')

            There are two reports for this question: one in order by software 
            changed FROM, the other in order by software changed TO.

From
Quest  To           Change Reason

AIPS
00081               
       MIRIAD       ATNF and BIMA analysis need Miriad; it has some nice
                    features which AIPS doesn't have for general image
                    processing.



More information about the Daip mailing list