[evlatests] Naming of .evla files

[Bryan Butler]

it will really be up to rich to implement this, since it is his code that 
assigns the name of the evla script file.  it is transparent to the operator, i 
believe (unless it is bound up in the fetch process).

it might have to await rich's return from his vacation for a real fix.

	-bryan


On 6/29/07 23:36, Walter Brisken wrote:
> Hello Operators and testers,
> 
> Due to some overlooking of two modes of operation, tipper scans and fast 
> switching, some information about these scans is not passed on to IDCAF. 
> This problem does not affect mosaics or normal point mode. The 
> implications are:
> 
> 1. No calibrator code gets assigned to calibrators
> 2. Because of this antsol will not generate output when expected
> 3. The project code assigned to these scans _could_ differ from other
>     scans in the project causing problems retrieving the data
> 
> #1 has no permanent side effects
> 
> #2 will affect the operators ability to judge quality of data for fast 
> switching scans using antsol.  D10 will still function though, but you 
> won't be able to tell easily if you are expecting fringes or not.  Thus 
> you should not be alarmed unless there are no fringes for several minutes 
> and don't be surprised if they come and go.
> 
> #3 is the most problematic.  In this case the fast switching scans and 
> tipper scans will be assigned project codes using different algorithms. 
> The former will be based entirely on the filename of the evla script and 
> the later will be based on the observe code the observer gives the file. 
> It seems that the filenames made in the dynamic queue are OK and the two 
> algorithms give the same name.  In the cases where the operator gives a 
> name to the file it is imperitive that entire project code be included in 
> the filename.  In Linux there are not restrictions on the file lengths 
> like there were on the modcomps, so I recommend making filenames that 
> might look like:
> 
> 003ab1248-000.evla		for AB1248
> 998ag627-000.evla		for AG627
> 
> and so on.
> 
> The first three numbers are the last three digits of the LST day as usual. 
> These are followed by the entire project code (case does not matter) and 
> finally -000 , -001, -002, ... for the version number (however that is 
> created these days).
> 
> This message unfortunately preempts another message that someone else will 
> likely send out that will attempt to standardize the naming convention. 
> For now the above will do.
> 
> Sorry for this confusion so late on a Friday night, especially right 
> before I skip town...
> 
> -Walter
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests