iraf.net

I came across another issue in doslit, having to do with the reidentification of arc lines. In my case I have two stars on the slit, and there
is a significant zero-point shift between the arcs of the two stars. doslit does the reidentification of the extraction of the second position arc based upon the first. But if the reidentification of comparison lines fails to find any
lines, or finds only one line, then you do not get the option to then
manually fit the spectrum. In other words, the "Fit dispersion
function interactively?" is NOT asked if the reidentification failed. The
log file contains an entry like this
iff0528c3iff0529c3.ms - Ap 2 0/13 0/0 0. 0. 0. INDEF
in that situation.

If I understand, your expectation is that if the reidentification fails you still want a chance to do things interactively. This should be tied to whether the interactive identification options is set since the task needs to do something reasonable when not enough lines are found when things are done non-interactively. All this is happening in a call to reidentify which is a pretty complicated task. If you would like to see this behavior changed I request that you set up an example with step by step instruction and all the par setting so that I can reproduce the behavior. I can then find where in the code the decision is being made and likely be able to have it go on to the interactive identify if the interactive option is active.Regards,
Frank Valdes

OK, will do. The problem is worse than you might realize, in that there is not even any indication (other than in the log file) that the reidentificiation failed.
All that happens in the terminal window (with verbose on) is that that particular reidentification isn't even mentioned. So, there is a bogus solution applied (one with an INDEF rms) but there's no indication of this in verbose mode. So, I don't think this is just an "interactive" vs "noninteractive" issue.I will construct a complete example and provide an image, as time permits.thanks!---phil