Tracker log

Fixed in the code

Issue state: Unconfirmed → Confirmed.
Added 21 hour(s) ago.

Some time, after I have worked in OLEX2 for some time, when I drag and drop a new RES file into the GUI the HKL file against which the refinement should proceed does not get updated. It IS updated most of the time, but not always. The problem seems to occur when the new file name is similar to the old file name (such as test vs test_0m). Or is it possible it has to do with the unit cell parameters in the RES file?

Added 1 day(s) ago.

The Info->Reflection Statistics Summary lists many numbers up to the optimistic 12th decimal place.

Added 1 day(s) ago.

I have a suggestion for the very useful OLEX2 macro HTAB: in the instances when several H-bonding interactions are equivalent by symmetry only one needs to be reported. Example: a Donor-H group with site symmetry mm2 (occupancy 0.25) and an Acceptor with site symmetry m (occupancy 0.5). The unique N-H...O bond thus has three mates related by symmetry. HTAB correctly lists four D-H...A interactions:
HTAB N2 O1
HTAB N2 O1_$1
HTAB N2 O1_$2
HTAB N2 O1_$3

but of course they are all identical. If HTAB can be smartened up a little bit to not list redundant bonds it would be great.

Added 1 day(s) ago.

Thank you for pointing this out - fixed in code and will be fixed with the next update.

Issue state: Unconfirmed → Resolved.
Responsible manager: (UNASSIGNED) → Horst.
Added 2 day(s) ago.

In stable Olex2 1.2 (Compilation Info: 2012.05.09 svn.r2324) in a structure containing C, H, Mo, N O and P atoms the Mo is denoted as '...' in atom list. Issue reproduces both on Windows XP and on Windows 7.

Added 2 day(s) ago.

The issue is mostly solved in Olex 1.2, which provides a way to edit absorption correction info via the GUI and also notifies if confliction information is present in source files.

Added 2 day(s) ago.

Hi,

To which version are you referring? The 1.2 is supposed to reload the INS file if the refinement fails (and returns empty RES file). If it does not do it automatically (and 1.1 does not), you can always go File->Open menu (or press Ctrl+O) and pick the last INS file... Is this any useful?

Cheers,

o

Added 2 day(s) ago.

I like Olex but it is very unfortunate that it is not possible to re-enter an *.ins file, like you can in shelxtl.
When your refinement crashes you may lose your model or you have to enter the *.res-file of the former state, which is bad when you try to model a more complex disorder.
 

Added 2 day(s) ago.

Oleg,

Thanks. I could locate my old installation (sorry, I had chosen a different drive for installation last time) and could retrieve all the macros. I will make a new XLD file and "include" it in the macro.xld so I will not have this confusion next time!

Have fun,

Bhuv

Added 7 day(s) ago.

The custom.xld file currently is stored in the location of the olex2, so unless you uninstalled the previous version - it should be there. This will change in the future and the file will be stored in user folders. So you can simply copy that file from your older installation. Also you can indeed just have an include line in your custom.xld files and refer that to a file on the file system by using full path - this way you would need to modify macros in a single file vs having copies of the macros around.

Cheers,

o

Added 8 day(s) ago.

Hmmmm..... I updated to the latest version (1.2 from 1.2-beta) today. After installation I found that all my custom macros are gone. Is this normal? Does that mean, I have to copy my custom.xld to somewhere else, everytime I update the olex (lesson learnt hard way!)?

Or is it possible to have another custom macros file, say, bhuv.xld (in a directory elsewhere), and "include" that in addition to the custom.xld in the macro.xld after every update?

Thanks,

Bhuv

Added 8 day(s) ago.

Thanks for confirming that it is fixed in alpha, now it should be fixed in the release too,

Best wishes,

Oleg.

Issue state: Confirmed → Resolved.
Added 9 day(s) ago.

Yes, it runs !

Good night ! Klaus

Added 10 day(s) ago.

Hi Klaus,

I have updated the alpha version; despite not all of our intended fixes work in it yet (I will look into tomorrow), can you try and see if it at least runs on your system (just to be sure that it is exactly this bug is fixed)?

Cheers,

o

Added 10 day(s) ago.

OK, the issue boils down to the ',' as a floating point separator (vs '.') which causes an exception and later - "unclean" cleaning up code... So yes, changing the default C locale should resolve this issue, I will try to fix it in the code,

Cheers,

o

Issue state: Unconfirmed → Confirmed.
Severity: Medium → Important.
Responsible manager: (UNASSIGNED) → Oleg.
Added 10 day(s) ago.

OK, thanks, with yours and Rays comments it is clear what causes the problem, and completely unclear - how this happens. Can you try removing the ~/.olex2 and ~/.olex2data folders and running Olex2 again and see if the crash persists?

o

Added 11 day(s) ago.

I'm not very familiar with debugging.
First I've substituted the last line of the script with:
gdb ./olex2

Output: gdb: symbol lookup error: gdb: undefined symbol: PyUnicodeUCS4_FromEncodedObject

Then I've commented out the "export PYTHONHOME ..." and moved the libpython* files from olex2/lib to another location. So my local python 2.7 should be used.

Output of the script:

Starting program: /home/harms/olex2/olex2
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7fffeddfb700 (LWP 6383)]
[New Thread 0x7fffe7fff700 (LWP 6384)]
[New Thread 0x7fffe5508700 (LWP 6387)]
[Thread 0x7fffe5508700 (LWP 6387) exited]
[New Thread 0x7fffe54e7700 (LWP 6388)]
[New Thread 0x7fffde883700 (LWP 6389)]
[New Thread 0x7fffde082700 (LWP 6390)]
[New Thread 0x7fffdd881700 (LWP 6391)]
[New Thread 0x7fffdd080700 (LWP 6392)]
[New Thread 0x7fffcffff700 (LWP 6393)]

Program received signal SIGSEGV, Segmentation fault.
0x00000000009f0ccf in gxlib::TXGrid::Create(esdl::TTSString<esdl::TWString, wchar_t> const&) ()
(gdb)
 
Hope this helps.

Klaus

Added 11 day(s) ago.

This is odd - you are the second person (there was one anonymous too) to contact me regarding this crash, but I cannot reproduce it on any of my computers - just tried Suse 11.1x64 installed natively and older Suse installed in VirtualBox; after these reports I had tried it on Ubuntu 10.04 x64 and Mandriva 2011, but have not observed any problems... Are you using proprietary NVidia drivers in you Linux (not in the VirtialBox)? Can you change the script which starts olex2 so that you run by a debugger - any stack trace would be helpful...

Cheers,

o

Added 11 day(s) ago.

Dear Oleg,

yes, I'm using the start script. The export is present.

I'm using an ACER Travelmate with a NVIDIA graphics card.
The crash happens at the end of the script when it tries to start olex2 (./olex2): A short "flash", then "segmentation fault". It also happens with Ubuntu or SuSE in a "VirtualBox" of a Windows 7 host.

Klaus

Added 11 day(s) ago.