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

Re: [zzdev] Many-to-many: bad name

On Sat, 19 Aug 2000, Benjamin Fallenstein wrote:

> I think the many-to-many raster has a bad name. Sure, showing
> intersections is one thing it does, but the name doesn't describe what
> it does well.

Well, it describes what it was initially planned to do well. The other
good stuff were a surprise. It came to be because we wanted to show the
financier how zz can be used as a frontend for SQL databases.
> Basicly, in its current form, it is a brick-style hard raster with
> floating headcells. 

Yes, that's about it. Except for the important part that it looks at the
intersections of the headcells, not directly connected cells. This is very
different from usual rasters.

>And this hints on a possible extension of it: as a
> grid or "hard" raster in general. Letting vanishing do the job just
> makes stuff more complicated, really.

Surprising: we were just discussing this last morning after the demo.

> floating headcells on/off
> show same cell more than once on/off
> use H or I or E or M or 3 or W or whatever raster (H and I are 'nuff initially)
> brick mode on/off (no, wait, we have the gap structparam for that, don't we)

Also, whether 

> Also, the headcells should have a different flob path. Why is that
> commented out? (The animations look very awkward currently.)

Hmm.. I'm not sure. We were a bit tired after working almost all night to
make that demo.

> Oh, the name. "HardRaster" would make sense to us who are accustomed to
> calling the non-vanishing stuff hard; but I thing GridRaster would
> suggest the actual meaning much better.
> All this is not urgent, but should be kept in mind -- it would be nice
> to have working hard rasters with floating headcells.

Good points - will have to consider. We're probably going to rename almost
everything while Ted's here...