|
|
|
|
|
I Mattia, Very nice project! There is a very long time I not do a line catalog myself so I must try to remember some points. What is sure is the text catalog was added after I make the default line catalog, so text line catalog is probably untested. I look at the code and find the mapping between the text operation (start,vertex,end) and the numeric operation number is not implemented. This work when making the binary file because catgen do the replacement. But doing the replacement a run time from the text file is probably not good for the performance. So it is probably best if you replace this text by the number directly in the file: sed 's/start/0/; s/vertex/2/; s/end/1/' simplified.dat Then change the "line operation" field length to 1 in catgen and it work. I tested to modify a point in the file and it is immediately show on the next redraw of the chart. So this is the best mode to test a work in progress. Yes, single file binary catalog are only read from file the first time, then they are keep in cache for all the session. If this is really required I can add some checking of the cache validity using the file date for example. But this way to work is good for performance when the catalog is complete and no more change are done. Maybe it is best you use the text file catalog for the development and make the binary format only for release. The "maximum magnitude" is not used, probably any positive value as the same effect. If it not work with 0, this is probably a glitch in the catalog selection procedure. Also the Epoch field is of no use for any data without proper motion, but I let it for all the object type and this not hurt. You can probably advance with the change to use the numeric line operation for the text catalog. Let me know if there is something I can do to simplify your task. |
|
|
Thanks Patrick, using {0,1,2} instead of {start,end,vertex} made the trick. Another thing not mentioned in the manual about text catalogs is that the data source MUST be in the same directory of the catalog header: I used to have the data source in a subdirectory and it didn't work, when I copied the data source in the same dir it started working. About binary catalogs, it's fine they're kept in cache, but I would have expected that if I completely remove a catalog (by the "delete" button which removes the line) from the catalog list, it should refresh its contents if I reload it without quitting the program. I think this means that even if I remove a catalog from the list, Skychart always keeps the data in cache forever (?) |
|
|
Yes, the text data and the files .hdr and .info2 must be in the same directory to allow to install the catalog anywhere on any computer. If I record the full path, the catalog work only on the specific directory on the computer you build it. You are right I must delete the cache when a catalog is removed. This is done by: https://github.com/pchev/skychart/commit/39bb31c9c110283e199830e4af7ed9b498ae9dc5 With this change you can test a new version of a binary catalog by delete and add again. |
|
|
While testing the differences between line and spline types I've noticed that spline outlines are sometimes rendered quite odd. I don't know much about spline interpolation, but I would expect a single, smooth curve between two consecutive points, not two curves like the example image attached. Is the actual behavior correct? outlines.png (63,977 bytes) |
|
|
Yes, there is really something wrong here. I cannot find such an extreme case in my current set of lines. Can you please send me the data for this object. |
|
|
Here it is the source data file. The object in the above image is IC405, but also other objects outlines show some glitches. |
|
|
Thank you for the example data. This is fixed by : https://github.com/pchev/skychart/commit/f0d01834d6b5287fc12000498436804dbe46c957 |
|
|
Thank you Patrick, I confirm this is fixed, splines are drawn correctly now. |
|
|
Nice, let me know of the advancement of your project. |
- Anonymous
