[Babase] Upload Programs Issues

Lacey Maryott lacey.maryott at duke.edu
Wed Jun 13 21:06:19 EDT 2007


Karl, (and everyone else)

    In working through some of the programs tonight, we came up with a 
"wish list" of things that we would like to ultimately become a part of 
the upload programs, as well as some questions. We'll try to present 
them in order of urgency or priority level.

1. In the WWWdiff program, there is no notification that the 2 files 
which are being compared have different numbers of rows.  The first set 
of files we tried had 302, and 304 rows, and the diff program came up 
with 305 rows, but never noted a discrepancy.  It also didn't show 
errors in every row subsequent to the start of the missing line 
discrepancies.  Even having notification that there are a different 
number of rows would give us enough information that we would know to go 
looking for a problem, it wouldn't need to be specific.

2. As we discussed on the phone, it would be helpful, and reassuring if, 
when using Psionload, the confirmation of upload would list the file 
which had been uploaded in babase.  Since there is no txt file being 
dumped into the folder, there is no way to confirm that the intended 
file was uploaded, and not a different one. Something like "060704P1 
successfully uploaded" would be helpful.

3. In addition to the last issue, we have found that as we move to only 
changing things within babase 2.0, it is increasingly important that 
some form of change tracking program be implemented.  In foxpro we 
always had a txt file for every file upload, which gave us details about 
when the given file was uploaded and by whom.

4.  The error messages which are produced by Psionload are a bit raw, 
and we have some suggestions on a couple of them. I'm sure we'll come up 
with more as we have only seen a few of the error messages thus far.

       "Error: Sname not on Biograph or UNKsnames."
        Our suggestion: "Error: Sname not found in biograph or UNKsnames"

       "Error: Pntid *****, Nghid *****, Neighbor after statdate blah 
blah (dead neighbor)"
        ---Pntid and Nghid are irrelevant as the file doesn't actually 
go into the system, and can't be searched, it can thus be removed.

5. Also in the Psionload program, it would be helpful if at the 
beginning of each error message, the name of the focal animal could be 
displayed.  This makes searching for the error to fix it much easier. 
Although, on a better note, the line numbers provided at the end of each 
error message do actually reference the line within the PTS file where 
the error is found, and that is a very helpful feature as well!! :)

On a somewhat different note, Leah and I are actively thinking about a 
way to document the changes we are making to the files as we go, 
equivalent to a conversion errors record... since these changes aren't 
being made to babase, we don't feel that the babase changelog is the 
proper place, but we are working on coming up with a systematic way of 
keeping track of those as well, so if you have any secret places 
something like this could hide, do let us know! :)

Have a good night

Lacey and Leah

-- 
Lacey Maryott
Alberts Lab
Department of Biology
Duke University
ph: 919-660-7306
fax: 919-660-7293
Lacey.Maryott at duke.edu 



More information about the Babase mailing list