Re: Versioning (again, sorry!)

Steve Harris wrote:
> On 8 Oct 2009, at 05:39, Lee Feigenbaum wrote:
> 
>> Paul Gearon wrote:
> 
> [snip, good description of the problem]
> 
>> So that I understand whether we're in a position to re-open this topic 
>> or not, is this a theoretical "they don't get it" or an actual "they 
>> don't get it"?
> 
> In Garlik at least, it's an actual problem. I think only Luke and I 
> understand which versions of what part are 1.1, and which are 1.0.
> 
> The next WG will presumably have some 1.0s, some 1.1s, some 1.2s and 
> maybe some 2.0s. Nice.
> 
>>> If I had to explain it to someone (and I often do) then personally I'd
>>> like to say, "SPARQL 1 had two parts. The first part lets me do
>>> queries, and the second part describes how to connect to a SPARQL
>>> server and talk to it. SPARQL 2 also has those parts, expanding
>>> significantly on the capabilities of each. It also has a third part
>>> that lets me update data in a database."
>>
>> Yeah, this is pretty much how i've been doing it, myself, without any 
>> sort of confusion that I've noticed. (See slide 3 of 
>> http://www.slideshare.net/LeeFeigenbaum/sparql2-status .)
> 
> To me, that seems like it's adding to the confusion.

I agree, it's not a great situation.

I'm still hesitant to re-open the discussion, because while this is 
important, it's not clear to me that there is really new information here.

I took a look back at the original resolution & discussion around it at 
http://www.w3.org/2009/sparql/meeting/2009-05-06#resolution_3 and it 
doesn't really help refresh my recollection of who the biggest 
supporters of the numbering scheme are, other than remembering we went 
around the room and there was pretty clear consensus for numbering 
/Query as 1.1. I don't recall if we had a particularly long discussion 
about the fact that the new pieces would end up as /Foo 1.0.

Anyway, I'd like to hear if Steve and Paul's concerns are resonating 
with other working group participants. We can slot it in for a few 
minutes on the teleconference, since it's clear that Right Now is the 
time to get this right (if there is such a thing as "right").

Lee

> I was working on Sun/Solbourne OS's in the early 90's so I have a 
> natural concern about things like this happening. Encouraging this seems 
> a bit too much like it's heading down the old SunOS/Solaris parallel 
> versioning path. Trying to work out if software would run on your system 
> could take a team of engineers, and several diagrams - I'm maybe 
> exaggerating, but not by much :) ("yeah, but which 4.1")
> 
> What happens when a future working group wants to have an actual 
> SPARQL/Query 2.0? Then there will be huge confusion.
> 
> I think a neutral ISO-style "SPARQL 2010" type number (à la SQL), or 
> letters, or just about anything really, would be much clearer. If 
> internally it's called "SPARQL/Query 2010" (for instance), then the 
> peanut gallery referring to it as SPARQL2, is at least not going to be 
> confused with the "real" version number.
> 
> - Steve
>

Received on Thursday, 8 October 2009 12:26:52 UTC