[osis-users] Work osisIDs

[Chris Little]

>
> I'm looking into how osisIDs can be used to unambiguously and precisely
> identify an exact edition/revision of a work. The manual states:
>
> The parts of an osisID may contain any mixture of numbers, letters, and
>> underscores.
>> […]
>>
> The prefix to an osisID must contain at least one letter, number or
>> underscore, that may be followed by any
>> number of letters, numbers or underscores, separated by periods, and
>> concluding in a colon ":"
>>
> The example prefixes found in the manual are:
>
> Bible.en.ABS.CEV.1995
> Bible.en.NIV.1984
> Bible.en.KJV
> Bible.KJV
>
> The following illustrates the components:
>
> Bible(.$lang)?(.$publisher)?.$abbr(.$year)?
>
> It would seem that all components are needed to unambiguously identify a
> work. Furthermore, in order to precisely identify a particular revision,
> some more granularity should be possible (these numbers I am making up):
>
> Bible.en.ABS.CEV.19950123123000
>
> And not only ISO timestamp, but also a revision number (or edition number)
> seems like it would be necessary:
>
> Bible.en.ABS.CEV.19950123123000.r273
> or
> Bible.en.ABS.CEV.ed2.19950123123000
>
> Is there some standard for constructing these more granular osisIDs to
> precisely identify works? Having a standard format would be ideal so that
> the information can be parsed out of the osisID.
>
> Thanks!
> Weston
>
>    

These would certainly be legal. Only the format used by 
Bible.en.ABS.CEV.1995 has been defined, though, so I'd recommend keeping 
all additions to the workID format to the right of that. For example:

Bible.en.ABS.CEV.1995.0123123000
Bible.en.ABS.CEV.1995.r273
Bible.en.ABS.CEV.1995.0123123000.r273


--Chris