Page 1 of 2

Changes to McIDAS-V nightly builds

Posted: Tue Jun 24, 2014 9:19 pm
by mug_admin
There will be several low-level changes made to McIDAS-V in the coming weeks to support newer Java versions. These changes will be made in McIDAS-V 1.5beta2, which will be available on the nightly installers page along with the beta1 builds. Note that we expect the beta2 builds to be relatively unstable, but please let us know if you run into any problems.

Thank you,
McIDAS User Services
mug@ssec.wisc.edu

Re: Changes to McIDAS-V nightly builds

Posted: Wed Jun 25, 2014 9:41 am
by hproe
Hi -

1.5beta2 of 25 June does not start up under Windows 8.1 Pro. No problem with start-up under Windows 7. However, in both cases no mcidasv.log is created.

HP

Re: Changes to McIDAS-V nightly builds

Posted: Wed Jun 25, 2014 12:05 pm
by Jon
Thanks for letting us know, H. P.. I'll find some way to get a Windows 8.1 installation to test against.

Re: Changes to McIDAS-V nightly builds

Posted: Thu Jun 26, 2014 12:32 pm
by hproe
I have been too fast (again :( ), 1.5beta2 starts up regulary and works well under Windows 8.1. I very much aplogise for this false alarm.

Most probably, I overlooked a javaw.exe that was still sitting in he system as a background process. The same happended this morning with 1.5eta1, i.e. McV did not go beyond the screen listing the outpur from runMcV.bat. I have never experienced such a silent misbehaviour before.

HP

Re: Changes to McIDAS-V nightly builds

Posted: Fri Aug 08, 2014 4:52 pm
by joleenf
Hi,

I am noticing navigation is off for my file in the preview window, but the display is fine in the Main Display. Version 1.5beta2 2014-08-05.

Screen Shot 2014-08-08 at 11.47.22 AM.png


ftp://ftp.ssec.wisc.edu/pub/ssec/joleenf/AUA88track.nc

Joleen

Re: Changes to McIDAS-V nightly builds

Posted: Fri Aug 08, 2014 5:25 pm
by bobc
Hi Joleen -

The Region tab I'm seeing is similar to your image:
Region.png
Region.png (8.78 KiB) Viewed 5643 times


Were you expecting to see a blue bounding box around the domain of your data (with Use Default Region), like you do with GFS CONUS data from the server:
Region2.png


Thanks -
Bob

Re: Changes to McIDAS-V nightly builds

Posted: Mon Aug 11, 2014 3:40 pm
by joleenf
Hi Bob,

Yes. I was expecting a bounding box or at least a view with the domain of my data highlighted in white. Is this showing my current view in the main display, or is it a bug?

Joleen

Re: Changes to McIDAS-V nightly builds

Posted: Mon Aug 11, 2014 4:11 pm
by bobc
Hi Joleen -

I wrote up Inquiry 1787 for this issue. I wrote this up as a bug, but it could be a development issue as well. I went back to McIDAS-V version 1.4 and saw the same thing, so it doesn't seem like this is directly related to the updated IDV/java versions in 1.5beta2. I ran a couple of tests with different projections in the Main Display, and it doesn't seem like the projection in the Region tab matches the Main Display, but it's just a generic global projection.

Thanks for noticing this -
Bob

Re: Changes to McIDAS-V nightly builds

Posted: Wed Aug 20, 2014 4:13 pm
by joleenf
Hi,

A few times over the past few days, the nightly I am using seems to run out of memory and become non-responsive. However, there is no indication that is an impending problem based on the memory reporting bar on the bottom left of the main display frame. In addition, in spite of the fact that I can't do anything within the software, the activity monitor does not indicate that the software is not responding.
Screen Shot 2014-08-20 at 11.12.16 AM.png
Screen Shot 2014-08-20 at 11.12.16 AM.png (6.28 KiB) Viewed 5577 times


Listed are two examples which caused the problem
Scenario A.
1.) Loading Suomi NPP data, displaying 2 different band using swath to grid
2.) Using that Suomni NPP data in a script and using a function to create a sandwich image.
vis=selectData() # select Formulas>Swath to Grid, then select I1 band
ir = selectData() # elect Formulas>Swath to Grid, then select I5 band
# create Sandwich (but sometimes, the double swath to grid formula selection is enough to make this happen before adding this user defined funtion.

Scenario B.
1.) Access GOES-14 other SRSO data.
2.) Click on absolute times.
3.) Select times 2014-08-18 from 18:48-18:50
# locked up. please see image below for memory reporting. It should also be noted that a script was run previously which created the image which is viewed in the frame below.

Screen Shot 2014-08-20 at 11.00.18 AM.png


I have the Suomi NPP data and the sandwich function:
ftp://ftp.ssec.wisc.edu/pub/ssec/joleenf/

Note, this memory usage/locking behavior is not happening every time, I was able to make an image with the NPP data. The region selection is similar to the region in the image above.

Joleen

Re: Changes to McIDAS-V nightly builds

Posted: Wed Aug 20, 2014 4:15 pm
by joleenf
Correction, it just took a long time for the system to respond after loading the GOES-14 data into the field selector ~15 minutes. 1.5Beta2 Build-date: 2014-08-19 09:30

Joleen