2010-06-18T04:57:51  *** sbmiller has joined #opticks
2010-06-18T05:00:17  *** goffena has joined #opticks
2010-06-18T05:03:13  *** tclarke has joined #opticks
2010-06-18T05:03:13  *** ChanServ sets mode: +o tclarke
2010-06-18T05:21:18  *** dsulgrov has joined #opticks
2010-06-18T06:55:55  *** yuguess has joined #opticks
2010-06-18T07:10:36  *** yuguess has quit IRC
2010-06-18T07:56:18  *** yuguess has joined #opticks
2010-06-18T08:56:26  *** gmartin_cn has joined #opticks
2010-06-18T09:05:25  *** yuguess has quit IRC
2010-06-18T12:07:44  <goffena> .
2010-06-18T12:07:44  <goffena> .
2010-06-18T12:08:09  <goffena> HdfImporterShell::validateDefaultOnDiskReadOnly()
2010-06-18T12:09:29  <goffena> Is not marked as virtual in HdfImporterShell header file
2010-06-18T12:09:48  <goffena> Is there any reason why it is not marked as virtual?
2010-06-18T12:10:16  <goffena> HdfImporterShell inherits from RasterElementImporterShell, and RasterElementImporterShell marks validateDefaultOnDiskReadOnly as virtual.
2010-06-18T12:10:28  <tclarke> it doesn't need to be...it's virtual in the base class so it's automatically virtual in HDFImporterShell
2010-06-18T12:11:34  <goffena> I understand it doesn't need to be
2010-06-18T12:11:36  <tclarke> we try and mark the sub-classes as virtual to avoid this sort of confusion
2010-06-18T12:11:40  <goffena> i thought it was common convention to mark it as virtual anyways
2010-06-18T12:11:42  <tclarke> but we haven't changed some of the older classes
2010-06-18T12:11:44  <tclarke> this one just got missed
2010-06-18T12:11:46  <goffena> ok.
2010-06-18T12:11:51  <goffena> i only care, because i almost didn't realize it was a virtual method.
2010-06-18T12:11:57  <tclarke> we do have an ongoing JIRA issue to bring the whole codebase up to our coding standards and we do a bit at a time
2010-06-18T12:12:01  <tclarke> so it should get fixed at some point
2010-06-18T12:12:03  <goffena> as i was looking at the derived class and not seeing it marked as virtual.
2010-06-18T12:12:11  <tclarke> that's why we try and mark them that way...but it's a recent convention on our part (a year or so old)
2010-06-18T12:12:16  <goffena> my question is answered.  thanks.
2010-06-18T12:12:16  <goffena> .
2010-06-18T12:14:12  <tclarke> just talked to dave...he's got a patch with a lot of that stuff updated...it's been waiting for us to have time for the massive code review that will be required
2010-06-18T12:14:21  <tclarke> we may start pushing smaller subsets of it in each release...starting with the SDK classes
2010-06-18T12:15:55  <goffena> good to know
2010-06-18T12:27:15  <goffena> .
2010-06-18T12:28:39  <goffena> is there anything in the HdfImporter architecture that will prevent subsetting by band for BSQ data?
2010-06-18T12:28:45  <goffena> other than the aforementioned validate method?
2010-06-18T12:30:47  <kstreith> thinking...
2010-06-18T12:32:26  <goffena> keep in mind i understand we wouldn't have official support
2010-06-18T12:32:31  <kstreith> it won't work right now
2010-06-18T12:34:44  <goffena> in a limited test case (hdf5 file 6 datasets), it seemed to work
2010-06-18T12:34:44  <kstreith> the Hdf5Pager uses .activeNumber()
2010-06-18T12:34:49  <kstreith> it must use .onDiskNumber
2010-06-18T12:35:03  <goffena> how big of a change is it?
2010-06-18T12:35:11  <goffena> 5 minutes, 1 hour, 1 day, 1 week.
2010-06-18T12:35:14  <kstreith> depends
2010-06-18T12:35:18  <kstreith> 5 minutes to enable it for BSQ
2010-06-18T12:35:23  <kstreith> technically the Hdf5 library would let us enable for BIP and BIL
2010-06-18T12:35:29  <kstreith> but the Hdf5Pager would need a bit more work to enable the right hyperslab selection
2010-06-18T12:37:11  <kstreith> i think changes to the validate and the Hdf5Pager would get it to work
2010-06-18T12:37:28  <goffena> if i submit a patch Monday, what version of Opticks could we get the patch into?
2010-06-18T12:39:21  <kstreith> that's a dsulgrov question
2010-06-18T12:41:11  <dsulgrov> hard to say without seeing the patch
2010-06-18T12:41:34  <dsulgrov> we can certainly evaluate the patch in the context of the issue to come up with a level of effort
2010-06-18T12:42:01  <dsulgrov> that level of effort to integrate the patch will be the factor that decides in what release it can be implemented
2010-06-18T12:43:28  <kstreith> i don't think it would be binary breaking so it could go into 4.4.0 or 4.4.1
2010-06-18T12:43:35  <kstreith> don't think if it missed 4.4.0 it would have to hold till 4.5.0
2010-06-18T12:46:27  <goffena> spoke with tjohnson.  tjohnson guessed it was probably to late 4.4.0, but 4.4.1 would be ok.
2010-06-18T12:46:27  <okibot> New news from nightlybuilds: EXTRAS-30-Windows-Public #8 (SUCCESS)
2010-06-18T12:47:52  <goffena> So i need to create/submit a patch, and create a JIRA issue, correct?
2010-06-18T12:48:23  <kstreith> correct
2010-06-18T12:48:29  <kstreith> send the patch to me
2010-06-18T12:50:36  <goffena> will do
2010-06-18T12:50:37  <goffena> thanks
2010-06-18T13:52:47  *** dsulgrov has quit IRC
2010-06-18T14:15:00  *** kstreith has quit IRC
2010-06-18T17:19:16  *** gmartin_cn has quit IRC