[Xesam] oscaf/xesam ontology merge: who is going to maintain/edit what?

[Urho Konttori]

ext Leo Sauermann wrote:
> It was Urho Konttori who said at the right time 09.06.2009 10:17 the 
> following words:
>   
>> I can be maintainer of NMM once we agree to include that in the official 
>> ontologies (and drop the NID3 and possibly even NEXIF while doing so). 
>> Also, by replacing the nid3, and nexif, we have 1 less ontology to get a 
>> maintainer for ;).
>>
>> Pros:
>> I designed that together with Evgeny Egorochkin and Mikael Ottela.
>> NMM is still not completely sanity tested and may still need quite a few 
>> changes, so I would be a good person to do that.
>>
>> Cons:
>> Quite busy until next autumn.
>>   
>>     
>
> I oppose to drop NID3 and NEXIF - they are quite sanity tested and good 
> because - they just copy ID3 and exif, two accepted industry standards. 
>   
And how does NID3 handle WMA files? How about m4a files? how about ogg?

It's really bad idea to start having format specific ontologies. It's 
like saying that we have a 'outlook' ontology and it should be fine for 
all 'outlook' type of apps, instead of having a messaging ontology.

> If NMM is not finished and you have no time to improve it, we would lose 
> compabiltiy with aperture and the existing code when using NMM.
>   
Well, NMM will be finished pretty soon. And cry me a river if someone 
looses support of a ontology that should never have existed.


> Once NMM can cover all that NID3 and NEXIF cover and is "sanity 
> proofed", we can discuss replacing NID3 and NEXIF, until then, I would 
> keep promoting that we maintin NID3 and NEXIF and fix bugs there.
>   
There is also no need to cover all. EXIF is just wonderful format, but 
it is intended mostly to be used during opening of the particular image 
file so that barrel distortion, lens effects, lighting and so forth can 
be nicely corrected by software. Most of it is not intended for 
categorization/searching type of cases. Thus, NMM has much smaller 
subset of it. It also has the support for a subset of XMP, which is 
essentially suffering from the same problem. I agree that the image side 
will probably be need to be extended, but it's much better to have a 
subset that is definitely needed than a format that offers tons of 
unnecessary fields. The same applies to NID3, just in a smaller scale.

Also, whereas nid3 is just a copy of the id3 to an ontology format, NMM 
has been designed for semantic desktop use.

> Once both are fine, we can make a decision to drop one and communicate 
> it. Note that dropping something (=making someone elses code break which 
> was fine before) is usually something hated by everyone in the process 
> (=because you force people to invest into something they could have for 
> free if the change did not happen), so I would not do this lightheartedly.
>   
Sure, we can keep those alive and just say that they have been 
superceeded by NMM, please move over there. It's totally fine way of 
keeping the support and moving on.
> but good, we need more ontologies, I wrote you down as maintainer of NMM
>   
Thanks!

And sorry about the flamy email. This really has not been one of my 
better days and I do apologise for letting it show.

Kind regards,
Urho


> best
> Leo
>
>   
>> Kind regards,
>> Urho
>>
>> _______________________________________________
>> Xesam mailing list
>> Xesam at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/xesam
>>
>>   
>>     
>
>
>