libera/#ecl - IRC Chatlog
Search
20:32:18
paulapatience
I'm not a huge fan of ggplot's short names, especially in Lisp, but it might be worth reconsidering it
20:34:17
paulapatience
I just spent several hours finetuning print-object so that the object hierarchy of a composed plot is easy to read, and also possibly readable
2:07:14
paulapatience
Well, it seems that "caption" is a synonym for "title". (captions :x "time" :y "speed") might be ok.
2:33:06
paulapatience
While we're on the topic of naming, jackdaniel, what do you think of the name "trace" for something similar to ggplot layers?
2:34:27
paulapatience
It's not exactly the same thing as a ggplot layer, basically it's a collection of data, mapping and geom. (I haven't decided where to put transforms, so for now it is a subset of a ggplot layer.)
2:34:53
paulapatience
But the reason I didn't want to use layer is because I wanted to have the layered-trace class, which contains a list of traces.
2:36:13
paulapatience
It was either trace, layered-trace, or layer and layers, but in the latter case there is a different of only one character in the names, and also, in my figure class, I wanted to accept a trace or a layered-trace, and with the singular term I can just have an initarg :trace, whereas with the plural it would be awkward, i.e., :layer layer, :layer layers, or :layers layer, :layers layers.
5:42:53
jackdaniel
I think that layers is a fine name, and if you have many layers then they can be, by definition, layered :)
5:43:16
jackdaniel
also the farther you move from the original wording of ggplot, harder it is to comprehend meaning for outsiders who know ggplot
5:43:59
jackdaniel
and layers may be specified to accept one-or-more layers (and in the constructor you could coerce it to a sequence)