So from previous posts it is clear that some fairly major "under the hood" changes relating to how the image headers etc. are read have been made: this will sometimes (for example with Rigaku instruments) use the dxtbx, which was in xia2 and has been moved to cctbx to allow use in other projects.
Thanks to some work from Nat Echols and James Parkhurst this is now working correctly in the nightly build from cctbx.sf.net: N.B. you need to get a recent (i.e. last night) "unreleased build" for example from
http://cci.lbl.gov/cctbx_build/show_results.cgi?build_tag=2013_03_26_0005
If you download and run the installer for this you should find that xia2 from subversion will work: this can be used for processing data from a Rigaku A200, for those of you who want this. The next release of xia2 will need an updated cctbx but if you are using the current release these changes should not affect you.
Tuesday, 26 March 2013
Tuesday, 19 March 2013
dxtbx starting to pay off: Rigaku A200
So, as of now if you have a Rigaku A200 and use xia2 (and cctbx) from subversion, the latter can be used for the former! However today this is a little technical to use - I will try to figure out an updated release mechanism tomorrow. Unless I can figure out a compiled code free way of bundling dxtbx into xia2...
Certainly there will be a dependency on downloading a cctbx nightly build. I will now migrate all of the image handling code over to dxtbx which will make maintaining the whole system a whole lot better.
Certainly there will be a dependency on downloading a cctbx nightly build. I will now migrate all of the image handling code over to dxtbx which will make maintaining the whole system a whole lot better.
Monday, 18 March 2013
dxtbx for image header reading
Over the next week I will pull out the old way of reading image file headers and add in a new one based on dxtbx - two comments:
- please send me an image from your detector so I can be sure xia2 will still work - to xia2.support@gmail.com ideally
- the subversion version of xia2 may appear broken
I will send an all clear when this is done. It should make xia2 much more useful and easier to maintain in the future.
- please send me an image from your detector so I can be sure xia2 will still work - to xia2.support@gmail.com ideally
- the subversion version of xia2 may appear broken
I will send an all clear when this is done. It should make xia2 much more useful and easier to maintain in the future.
Monday, 4 March 2013
dxtbx: the diffraction experiment toolbox
The diffraction experiment toolbox in xia2 has now been migrated to cctbx. This is a two edged sword: it is a good thing as it can be used for other projects, but it is a problem as well as from the next release you will need a cctbx with the dxtbx contained within it and correctly configured.
This is necessarily a breaking change as the API for the dxtbx has changed since import into the cctbx, so I will need to modify xia2 to use the new API, and will remove the old dxtbx.
From later on today the svn version of xia2 is likely to be "broken" therefore if you have an older version of cctbx.
This is necessarily a breaking change as the API for the dxtbx has changed since import into the cctbx, so I will need to modify xia2 to use the new API, and will remove the old dxtbx.
From later on today the svn version of xia2 is likely to be "broken" therefore if you have an older version of cctbx.
Friday, 1 March 2013
-migrate_data
Anyone use -migrate_data or even know what it does? Wondering about culling some ancient and probably useless command-line options to make the code base cleaner, and this is one on the hit list. If you want to keep it then please send a comment to xia2.support@gmail.com otherwise I will clear it soon...
Subscribe to:
Posts (Atom)