[okfn-discuss] Wikipedia as SaaS by the OSSD?

[Mike Linksvayer]

Hello Bastien,

On Fri, Dec 21, 2012 at 9:20 AM, Bastien <bzg at altern.org> wrote:
> Mike Linksvayer <ml at gondwanaland.com> writes:
>>> I see two problems with this "orthogonal" definition of SaaS:
>>> (1) it doesn't really match the common meaning of the term and
>>> (2) it is so broad than I don't see any website that it does *not*
>>> cover.
>>
>> Indeed, any website could be considered either Open Software Service
>> Definition conformant, or not...
>
> That's a bit of a tautology, right?  :)

Unless you define some websites as not software services, in which
case you get nothing (or an exception) instead of conformant or
not-conformant. I guess this is your point, that some sites (maybe
including Wikipedia) aren't software services, thus OSSD not relevant
to them, and OSSD related documents shouldn't give them as examples,
let alone the only example.

> If every website is a "software as a service" then all the discussions
> about software as a service are of little value IMHO.

Hmm, I did not suspect the reason, but indeed, I have found little
value in most SaaS discussions in the world. ;-)

For me, an inclusive, even tautological, definition of software
service shouldn't be an obstacle to focusing on the most
interesting/strong examples, though perhaps providing Wikipedia as an
example is an illustration that it is an obstacle.

>>> I agree FSF's definition can also have some problems, but at least
>>> it's narrow enough to limit the topic.
>>
>> ...SaaS can't be rigorously defined.
>
> Why?  A definition can be formal and operational, or only one of them,
> or none.  The definition of a triangle is both formal and operational,
> because it is a non-ambiguous description, and this description refers
> to some things we recognize as such in the real world.
>
> The definition of SaaS can be formal (i.e. "any computation of your data
> that is done by a server accessible through the internet") but not 100%
> operational (i.e. "if this is a computation that you would normally do
> by yourself.")  I acknowledge that the second part is not operational,
> because it's relative to what users are used to, can evolve with time,
> etc.  But still, the definition is formal and refers to something less
> general than the current OSSD.

I guess I meant operational, but admit I was speaking way too loosely
to throw in the word "rigorously". Mea culpa, joke's on me.

"Your data" interpreted broadly could include request necessarily sent
to any website, and used in computation to retrieve appropriate
document, but admittedly this is extreme.

But even where a site is only a software service to a reader in a
facile way, might be a software service to a writer. Consider
wordpress.com.

>> https://www.gnu.org/philosophy/who-does-that-server-really-serve.html
>> provides some illumination on a particular dimension ("your
>> computing"), and certainly services implicating your computing are far
>> more interesting than most others for OSSD conformance. But many
>> things in that essay ruled out as doing your computing are also
>> extremely interesting for OSSD conformance, including Wikipedia.
>
> I'd be curious to see people really arguing that Wikipedia is "software
> as a service".  (And yes, the FSF has no problem with Wikipedia.)

I haven't considered the question seriously. But it could be seen as a
super fancy (and not so fancy in some respects) word processor, and
folk seem to think google docs are the canonical unambiguous software
service, because they used to word process on a local computer. I
don't remember details, but I recall wikis and collaborative editing
in general, and Wikipedia in particular, was discussed extensively (as
pertinent examples, not questioning their relevance) among people
meeting to discuss what eventually became the Franklin Street
Statement.

>> I don't see any need to narrow the topic. Perhaps in part because I
>> don't see the Open Definition group isn't going to be certifying
>> particular service instances as OSSD conforming any more than it
>> certifies particular works as Open Knowledge Definition conformant
>> (and a work can certainly be published under an OKD conformant license
>> and not be an OKD conformant work). And in part because I see some
>> social benefit, perhaps small, of comprehensive openness, even if a
>> site is just publishing information, analogous to a local program
>> which just displays a single artwork (such things used to be pretty
>> common). It isn't incredibly important that such things be open, but
>> they ought to be, still.
>
> I agree with the openness.  It's also good to keep things simple and
> straightforward when you want to _inform_ people at large.
>
> I just don't see the value of a definition that does not help anyone
> distinguishing SaaS from any website -- or free SaaS from any website
> using free softwares and delivering free content (e.g. all the wp
> websites using CC-by-sa.)
>
> I think this will confuse people.
>
> Even if OD is not delivering certificates, I assume its goal is to
> promote openness, and it can do so by providing a clear picture of what
> is SaaS, then of what is free SaaS.

I see the primary value in providing a clear picture of what is needed
to be Open, rather than what is needed to be a Software Service, which
I see as a lot more confusing. But this scoping discussion is good to
have as a prelude to considering what to do with the OSSD going
forward, and I'd be happy to be wrong about where the most potential
for promotion and confusion sits. Please continue to tell me I'm
wrong; I have learned a bit so far from this friendly discussion.

>>>> We could use more examples, but honestly I'm not sure anything
>>>> relating to Wave is intuitive for many people. And there's a
>>>> distinction between software which is amenable to powering an
>>>> OSSD-conformant service, and a service itself (eg
>>>> MediaWiki/Wikipedia). Maybe Kune.cc is OSSD-conformant; I haven't
>>>> looked closely.
>>>
>>> Kune is just a website to facilitate online collaboration.
>>>
>>> I think it is OSSD-conformant, but I'll forward the question to the
>>> developers and let them reply if they can.
>>
>> I thought Kune is software, which a number of services (websites)
>> might run.
>
> You're right, it is.
>
>> Software might be written in a way that facilitate OSSD
>> conformance of an instance, or even makes a new installation OSSD
>> conformant by default. Conformant instances, and software that by
>> default creates conformant instances, all great, of course!
>
> Yes, that's what they try to do :)

Great. I could see some kind of activity, from Open Definition or
other, to encourage such out-of-the-box conformance from developers.

Mike