Hi Rasmus,
I'd say that's exactly how it happened. In the inital email I sent I
mentioned that I had been switching the projects between my work machine
(linux) and my home machine (mac). Strangely, when I switched between
systems over the weekend, and them switched back, the second set of
files were not created.
Best wishes,
Pryank
Rasmus Fogh wrote:
> Dear All,
>
> The NameAmpping is indeed created afresh by Analysis, and so can be
> deleted with impunity. It is just a timesaver instead of creating the
> mappings in every session. What must have happened is that the
> NameMappings got created twice, maybe on different computers, and later
> moving the project back and forth put two diferent files with different
> unique IDs but the same key into the same directory. Does that sound
> plausible?
>
> Yours,
>
> Rasmus
>
> ---------------------------------------------------------------------------
> Dr. Rasmus H. Fogh Email: [log in to unmask]
> Dept. of Biochemistry, University of Cambridge,
> 80 Tennis Court Road, Cambridge CB2 1GA, UK. FAX (01223)766002
>
> On Fri, 22 Jan 2010, Wayne Boucher wrote:
>
>
>> Hello,
>>
>> That bit with the "loading attempt failed" and "re-trying" is indeed normally
>> output if the project has moved. So the loader first tries in the locations
>> it used last time and if that fails it tries based on the current directory
>> (if I remember correctly). But I guess it could be triggered otherwise.
>>
>> And if I remember correctly with the NameMapping stuff, it gets generated
>> again if it gets lost and no other package imports it so I think removing the
>> files ought to cause no problems (but Rasmus and/or Tim might correct me on
>> that one).
>>
>> Wayne
>>
>> On Fri, 22 Jan 2010, Pryank Patel wrote:
>>
>>
>>> Hi Wayne,
>>> Thanks for that. I noticed the timestamps between the two xml files in the
>>> NameMapping dir, so after backing them up I deleted the older files (the
>>> '11' files, it seemed more logical somehow) and kept the '14' files. The
>>> project opened fine. Then, my curiosity got the better of me and I changed
>>> the files over, so I had the '11' files in the NameMapping dir and moved
>>> the '14' files out. The project opened, but it took a little longer than
>>> usual and there was this message in the window:
>>>
>>> For Python API documentation, tutorials and CCPN web applications visit:
>>> http://www.ccpn.ac.uk/
>>> First loading attempt failed - has project moved?.
>>> Re-trying, skipping cached TopObjects: GL_VENDOR = NVIDIA
>>> Corporation GL_RENDERER = Quadro NVS 290/PCI/SSE2
>>> GL_VERSION = 3.2.0 NVIDIA 190.42 >>>
>>>
>>>
>>> However, it has not caused any other problems as far as I can see, and I
>>> don't think it's a bug (or very interesting).
>>>
>>> Pryank
>>>
>>>
>>>
>>> Wayne Boucher wrote:
>>>
>>>> Hello,
>>>>
>>>> Well the problem is that the project is invalid! So for some reason (and
>>>> I haven't a clue how this might have happened) there are two xml files in
>>>> AGR2/ccpnmr/NameMapping/ (and similarly in AGR2_backup), which is allowed
>>>> but unfortunately they have the same key (name="AGR2"), which is not
>>>> allowed.
>>>>
>>>> It looks like the NameMappings are not used for much (Tim might know about
>>>> this) since if they were then it should have triggered an error in
>>>> Analysis even ignoring the backup situation.
>>>>
>>>> So after you start up Analysis if you had done:
>>>>
>>>>
>>>>
>>>>>>> top.project.nameMappingStores
>>>>>>>
>>>>>>>
>>>> it would load those files and so would have given you the same kind of
>>>> error as for the backup. Or if you had done:
>>>>
>>>>
>>>>
>>>>>>> top.project.checkAllValid()
>>>>>>>
>>>>>>>
>>>> it would have triggered this again.
>>>>
>>>> I'd keep a copy of the files somewhere, just in case, but I think it's
>>>> safe just to remove either of the two files, at the operating system
>>>> level, and do so in both AGR2/ and AGR2_backup/. Obviously remove the
>>>> same one in both cases.
>>>>
>>>> I just tried removing the one with a 14 in it rather than 11, and analysis
>>>> then seemed to open ok, and should function ok, including the backup. (And
>>>> it looks like the backup does do a load after all, when the backup
>>>> directory exists.)
>>>>
>>>> Wayne
>>>>
>>>> On Fri, 22 Jan 2010, Pryank Patel wrote:
>>>>
>>>>
>>>>
>>>>> Hi,
>>>>> I am seeing the error message below in the console 15mins after opening
>>>>> the project (automatic backup is set every 15 minutes in my project).
>>>>> However, it does not occur when I do a manual backup of the project.
>>>>> Also, after 30 minutes, it backs up the project fine (I've tested the
>>>>> backup after 30mins, but not after 15mins yet).
>>>>>
>>>>> I have a linux system in the office and a new Macbook Pro at home. I'm
>>>>> wondering if it's related to the fact that I occasionally tar and copy
>>>>> the project directory between the two systems since I've only noticed it
>>>>> recently. I have not checked yet if it is occurring on the other system,
>>>>> but I can do so this weekend.
>>>>>
>>>>>
>>>>>
>>>>>>>> Error loading file for: <ccpnmr.NameMapping.NameMappingStore
>>>>>>>> ['AGR2']>
>>>>>>>>
>>>>>>>>
>>>>> Reading: <open file
>>>>> '/home/pryank/nmr-projects/AGR2-P40/AGR2/ccpnmr/NameMapping/AGR2+AGR2_pryank_2009-06-05-11-44-41-975_00001.xml',
>>>>> mode 'r' at 0x1b2e5580>
>>>>> Last xml tag read: NMAP.NameMappingStore.nameMappings
>>>>> Parser state was: postprocessing data
>>>>> Object stack was empty
>>>>> Backup error Fri Jan 22 09:30:27 2010: Cannot add child - key already in
>>>>> use:
>>>>> ccpnmr.NameMapping.NameMappingStore:<memops.Implementation.MemopsRoot
>>>>> id:96376400>:AGR2
>>>>>
>>>>>
>>>>> Best wishes,
>>>>> Pryank
>>>>>
>>>>>
>>>>>
>>>>>
|