Issue 633: Usability idea: when we fail to open a Skyline document, check to see if it actually IS a Skyline document

issues
Status:closed
Assigned To:Guest
Type:Defect
Area:Skyline
Priority:3
Milestone:4.3
Opened:2019-02-28 10:23 by Brian Pratt
Changed:2019-06-21 09:30 by Matt Chambers
Resolved:2019-06-10 14:31 by Brian Pratt
Resolution:Fixed
Closed:2019-06-10 14:31 by Brian Pratt
2019-02-28 10:23 Brian Pratt
Title»Usability idea: when we fail to open a Skyline document, check to see if it actually IS a Skyline document
Assigned ToGuest»Brian Pratt
Type»Defect
Area»Skyline
Priority»3
Milestone»4.3
Lately we seem to get a lot of support requests where users are trying to just fire up Skyline then open a mass spec raw file. We issue an error like "the file contains an error at line 1 column 1", which only makes sense if the file is actually XML data.

Instead, we should:
 scan the first few hundred bytes to see if it contains "srm_settings" and "Skyline"
    if so, then throw that line/column error, fine.
 if not, see if we can identify it as mass spec data.
   if it is mass spec data, suggest they use File > Import Results, or if they don't have any targets, suggest they look into SeeMS.
Failing all that, issue an error like "this file does not appear to be a Skyline document."

2019-04-03 16:34 Brian Pratt
Part of this is now implemented - we no longer just show the bare xml parser error. The bit about trying to identify raw data isn't there, but we do now offer a more friendly message about the file not being a skyline doc. Additionally if it's a .skyd file we'll find the .sky file next to it and load that without complaint.

2019-06-10 14:31 Brian Pratt
resolve as Fixed
Statusopen»resolved
Closing this - the important part is done, I doubt we'd ever get to the second half of this idea (about people trying to directly open raw data files instead of .sky files).

2019-06-10 14:31 Brian Pratt
close
Statusresolved»closed
Assigned ToBrian Pratt»Guest

2019-06-21 09:30 Matt Chambers
Milestone4.3»
Just randomly looking at issues. The second part would be pretty easy for pwiz-recognized files. Just call ReaderList::identify on it and see if it returns non-empty. But that would add a ProteowizardWrapper dependency, where maybe you don't have one right now.