[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]

Re: What's a quark?



Tuomas Lukka wrote:
> > In your new merge terminology, how exactly is a quark defined? Consider
> > cell neg and cell pos, which are connected on dimension dim. Are neg's
> > posward and pos' negward connection different quarks-- or is the whole
> > connection, from neg to pos, a single quark? In the Design Problems doc,
> > it sounds like you mean the former, but in the change list code, it
> > sounds like the latter.
> 
> I think of it more as the latter. But the distinction does not make
> it into practice since quarks immdiately bond to form atoms
> (physicists talk of "confinement": unlike gravity, the force does
> not decay).

If that were so, then the concept of a "quark" would not make it into
practice at all, for the same reason, and we could drop it. But as
terminology is for programmers, too, and we want to understand each
other as well as each other's code, it would be good if we would settle
on a definition. Like if I read your Design Problems definition and
understand the former, then go to SimpleDimChangeList, I will have
problems understanding what the data structure in the code is. If you
define it as the latter in Design Problems, then I will understand
SimpleDimChangeList more easily. If you use constantly changing
definitions, and I don't realize that, then we'll have serious
communication problems. ;)

BTW, I'll comment on the Mediaserver doc and rewrite the Semantics doc
this weekend.

-b.