[Nepomuk] Data stored in database - how to manage, backup and merge?

[Sebastian Trueg]

Karl Günter Wünsch wrote:
> Hello,
> On Thursday 03 December 2009, Sebastian Trueg wrote:
>> There is the nepomukshell in kde playground which allows to clumsily
>> browse the resources in Nepomuk.
>> Then there is sopranocmd for purists.
>>
>  
>> We have no backup service yet. It is on the ever so long TODO list though.
>> For now the ugly way is to export the whole DB using "sopranocmd export"
>> (for details on how to use sopranocmd with Nepomuk see [1])
>>
> 
>> This is not implemented yet but was the topic on the last Nepomuk
>> workshop. It will be tackled in KDE 4.5.
>> So for now there is no way to merge the data unless manually which would
>> be a pain.
> 
> So to sum it up: There is no management tool whatsoever which would allow an end user access to his own data, there is no backup/restore tool that would allow safeguarding the data against accidental loss and there is no merge tool for data that may have differing origins but must be merged to be useful? How about separating discardable information (like the strigi index which holds little value) from the persistent data which can't be recreated from other sources in an emergency?
> Do you care to explain to someone who is trying to use the semantic desktop for more than casual searches how this is supposed to be reliable? If persistent semantic information about items is stored in Nepomouk, there has to be an infrastructure in place to deal with these things. As long as a strigi index was the only reason for the data storage in Nepomuk that's no problem, such an index can be regenerated, but now that Nepomuk is being used to hold information beyond that scope - especially as it will hold information which can not be easily recreated - Nepomuk has to be considered a system critical application which must not lose information and for which there has to be a sophisticated management tool. 
> As a short term remedy all I can think of is a description on how to backup and restore the database with the tools you mentioned - or with the tools belonging to the database. With the recent switch from redland/sesame2 to the virtuoso backend (a database for which some distributions do not even have an install package) I am willing to bet that there are only a few people who know where the database is kept and how to handle a backup/restore situation!

That is all well known. And I can only repeat my offer: I am happy about
any help with the system and am more than willing to give you pointers
on where to start.

And btw: that tone you are using is not appropriate.

Cheers,
Sebastian