From: William P. <wil...@ya...> - 2010-05-24 18:26:44
|
On May 24, 2010, at 9:58 AM, Hilmar Lapp wrote: > The stack trace means that there is a situation when two taxon labels are being compared that either one of the two TaxonLabel objects, or the taxonLabel property of one, is null. Thanks all. Looking at the user's *original* data files, when I try to open them in Mesquite, not only is there a taxon label mismatch between matrix and trees, but Mesquite kicks out an error saying something about detecting a reticulation in the tree (???). At any rate, I cleaned up the original files so that Mesquite no longer has any issues. After getting Rutger's feedback, I tried deleting all matrices and trees from the problem study (which should delete all associated taxon labels), and then completely re-uploading with fresh, and fixed, datasets. But the error persisted. So in the end I created a new submission, uploaded the files, and that works fine. > Indeed, the TaxonLabel.taxonLabel property property is nullable (why?), so this seems to be a legitimate situation and the fact that the code assumes that it can never be null is a bug. I fixed this and committed it to trunk. I haven't had a chance to rebuild, so I'm hoping it compiles. Thanks for doing that -- unfortunately I went ahead with creating a new submission and deleting the old files before seeing that you made this fix. On May 24, 2010, at 11:01 AM, Jon Auman wrote: > Instead of a tracestack, I now get : > > Access denied > Sorry! your action is not authorized. > > > -jon Yeah, that has to do with me converting the submission back to "in progress" so that I can mess with it. But now it has been deleted, so you should get a "no results" return. bp |