Re: Use case driven work

On 2015-02-07 10:47, Phil Archer wrote:
> Hi Frans,
>
> My 2 cents in line below.
Thank you. Actually, I think this type of guidance is worth more than 
two cents.

I have also placed some in line comments...

Greetings,
Frans
>
> [snip]
>>
>> 2) At the moment, use cases come from WG members only. We are a diverse
>> group, but I don't know if we are a good cross section of the people
>> that will have to work with what we come up with. It would be a pity for
>> real and important requirements to go unnoticed because of that. Would
>> it be a good idea to ask for comments on the use cases from other,
>> related communities, at some point before the requirements are defined?
>> We could ask people to send comments to the public comment mailing list
>> (public-sdw-comments@w3.org) for instance.
>
> The WG *MUST* solicit comments from external stakeholders and *MUST* 
> take note of what they say. That doesn't mean the WG has to agree with 
> them but does have to respond to comments and the public comment list 
> is there for that purpose. Those comments may themselves be a new use 
> case or a modification of an existing one. The WG makes the decisions 
> but the process ensures that those decisions are informed by the wider 
> community.
It is good to see this stress on getting influence from external 
stakeholders. Still, soliciting comments could be done in various ways. 
We could post a message on the wiki stating that the WG welcomes any 
comments. That would be a rather passive method. A more active method 
would be for each of us to approach the  communities of which we are a 
part to draw attention to the use cases and the requirements that we 
think follow from them. I think an active approach is in order.  
Standardizing space and time on the web is a fundamental thing and I 
guess it directly affects the work of millions of people (and many more 
indirectly).

Of course timing is important. We can not expect the world to check the 
wiki for new developments each day. So what is a good time to attract 
attention to the plans? I think that would be around the time when we 
think the use cases are complete and we start drafting requirements...

>
>>
>> 3) It is my experience that in purely use case driven system design
>> important requirements can be left out. That especially goes for
>> requirements that do not directly come from user stories, but from
>> common sense design principles. Principles like modularity, keeping
>> things as simple as possible, separation of concerns. In my mind those
>> things are very important. In case this kind of requirement does not
>> follow from the use cases, do we still have room to add them or to keep
>> them otherwise in mind?
>
> Here are some examples of UCR docs:
>
> LDP http://www.w3.org/TR/ldp-ucr/
> Web Audio http://www.w3.org/TR/webaudio-usecases/
> CSV http://www.w3.org/TR/csvw-ucr/
> SKOS http://www.w3.org/TR/skos-ucr/
>
> They all begin with an intro that sets out the general problem to be 
> solved (one might think of that as a preçis of the charter). The 
> intros also set the direction of the work to be undertaken and might 
> identify the technologies to be used. That context can be used to 
> guide the kind of requirements needed without necessarily ending up 
> having to write a bunch of requirements that justify fundamentals such 
> as the use of HTTP.
>
> IMO the discipline of deriving requirements from use cases, preferably 
> real world use cases with real data, is a good one. Ideally, the data 
> associated with those use cases can then form the basis of a text 
> suite that is published alongside each standard. Then the WG is 
> confident that the new spec meets the requirements and developers can 
> test their implementations against it.
>
> Incidentally, Deirdre Lee, editor of the Data on the Web Best 
> Practices WG's UCR, has gone as far as to create a little bit of RDF 
> to describe which use cases link to which requirements - a process 
> that certainly helped clarify my thinking when I've contributed to 
> that doc. See
> https://www.w3.org/2013/dwbp/wiki/images/2/2f/UCR.ttl
> http://w3c.github.io/dwbp/usecasesv1.html
That's interesting! Whether done in HTML or RDF, linking things is a 
good idea. Looking at the use cases for the Data on the Web Best 
Practices WG, I see the case of the Dutch base registries is present in 
both that WG and ours (here 
<http://w3c.github.io/dwbp/usecasesv1.html#UC-DutchBasicReg> and here 
<https://www.w3.org/2015/spatial/wiki/Working_Use_Cases#Dutch_Base_Registry_use_case_.28Best_Practice.29>). 
So not only could we link use cases to requirements (and tests and 
implementation examples) within our WG, we could also link to use cases 
and requirements in other working groups.

>
> HTH
I had to look that one up, but I think it does :-)
>
> Phil.
>
>>
>> Regards,
>> Frans
>>
>> ------------------------------------------------------------------------
>> Frans Knibbe
>> Geodan
>> President Kennedylaan 1
>> 1079 MB Amsterdam (NL)
>>
>> T +31 (0)20 - 5711 347
>> E frans.knibbe@geodan.nl
>> www.geodan.nl <http://www.geodan.nl> | disclaimer
>> <http://www.geodan.nl/disclaimer>
>> ------------------------------------------------------------------------
>>
>


------------------------------------------------------------------------
Frans Knibbe
Geodan
President Kennedylaan 1
1079 MB Amsterdam (NL)

T +31 (0)20 - 5711 347
E frans.knibbe@geodan.nl
www.geodan.nl <http://www.geodan.nl> | disclaimer 
<http://www.geodan.nl/disclaimer>
------------------------------------------------------------------------