Re: Concurrency in an RDB

On Dec 20, 3:13 pm, monaro..._at_gmail.com wrote:
> Marshall wrote:
>
> > > > > Real time, interactive editing of replicated data is very well suited
> > > > > to OT.
>
> > > > This is nothing but more description; you have already supplied
> > > > plenty of that. My interest is in an example.
> > > The shared jigsaw, text and white board components were real
> > > examples that I gave.
>
> > No they weren't. They were handwavy descriptions of something
> > one could build that you claim used some of the techniques
> > you're proposing. *I* haven't seen anything yet except vague
> > descriptions.
>
> David wrote, "I have built a working system and the results
> are very impressive. In our office we've collaborated on a large
> virtual world of components including text documents, jigsaws,
> white boards etc. "

So he wrote a couple of apps. Woo hoo. And he's told us, in the most excruciatingly high level, handwavy manner possible, that they have this and that awesome property. But he hasn't told us anything about specifics. The system "achieve[s] intention preservation, causality preservation and of course convergence at quiescence." Also, "remote edits are incorporated asynchronously after undergoing transformation." The actual details of how any of this is accomplished are somehow omitted.

I assume his applications are also "mission critical" as well as being "enterprise-level."

> Something one *could* build.
>
> Is English your second language?

Wow, an insult in which you imply that I don't speak English very well. Isn't *that* original, and oh so relevant?

Or perhaps you intend to imply that one could *not* build such apps. But I assert that one could.

> > You do know this *is* a theory group, right? Do you actually
> > have any theory? Any papers? Any math? Computational
> > models? Equations? Examples, even?
>
> > Okay, I think you're in the wrong newsgroup. This is a database
> > theory newsgroup. Those with an application framework du jour
> > are directed to comp.object. OT is OT here.
>
> It seems to me that you have led David away from discussing
> database theory.

I asked David several times to give us some low level details about what he was talking about, a question he seems not to have understood. At no time has he broached the subject of database theory, except to discount parts of it.

Again, any theory? Papers? Math? Computational models? Further handwaving may be safely omitted; we have already had plenty.

> OT may be OT here, but you're the one that wanted to talk
> about it. Perhaps you're in the wrong news group?

No, I know right where I am. And I'm happy to pit my history of almost 2000 messages in this group alone against your sock-puppet, 8 days old profile any time.

> In order to answer Davids original question, we need
> to know what assumptions are being made, and hence,
> we do need to know the implications of the OT approach.

What original question is that? His original post didn't ask anything. Mostly he just seems to be here to amaze us with the fact that he's written a shared whiteboard app and that he has some ideas about grossly restricting what applications can do and what they can count on to make his programming model more applicable.

Eventually he asked doubtingly about the existence of updates that take a long time. Is that the "original question" you mean?

> A discussion of OT in comp.object would not talk about deadlock,
> transactions, and all the other database theory related topics raised
> in this thread.

Yeah, they are pretty light on hard technical matters, I admit. More about the frameworks and the "extreme" and the "paradigms" and all. Still, deadlock, atomicity, etc. are all on-topic for object systems as much as database systems, and their fast-and-loose approach is probably a better fit.

> It's these issues that David seems interested in, not
> a discussion about OT.

You seem to know a lot about what he's interested in and all, for someone who only sprang into existence eight days ago, and has only ever commented on just this one thread in all that time. Hmmm; perhaps you *do* know just what he's interested in.

Marshall Received on Thu Dec 21 2006 - 02:29:30 CET