[Kst] healpix status

Ted Kisner tskisner.public at gmail.com
Sat Oct 29 22:33:52 CEST 2005

On Saturday 29 October 2005 08:04, George Staikos wrote:
> > 4.  There are no limits on the projection matrix size, so if you choose
> > one too big and use up all your memory, then shame on you ;-)
>   I can help put together this limit if you like.

ok- I've not looked at the code that computes the amount of available memory, 
but it seems like I could just check that value and not let the user create a 
matrix that exceeds it.  Of course there are separate issues when loading 
large healpix data files that require some temporary memory to store the map 
until it gets projected.

>   The legend code is going to change as well.  Probably not until 1.3
> though.

That's fine- my plan is to get the healpix datasource as functional as 
possible given the state of the code for 1.2.  For 1.3, I'd like to work on 
the new renderer so that we can have complete support for projections, etc.

>   Do you have any testcases?  It would be great to have some in tests/.  I
> can generate coverage reports if I have some samples.

I'll gladly make some tests.  Looking in the tests/ directory, it seems the 
idea is to make a small c++ program that links against libkst* and then 
create an instance of the datasource from an example input file.  Then I can 
create various KConfig's and pass them to the datasource to test all the 
features.  I guess the way to verify everything works is to call readMatrix 
and readField to make sure that the returned data is what I expect.  

So the only thing this doesn't test is that the actual config widget returns 
the proper KConfig- but that should be obvious...


More information about the Kst mailing list