John J Foerch <[EMAIL PROTECTED]> writes: > John J Foerch <[EMAIL PROTECTED]> writes: >> Yes, there is some overlap between the things for which browser >> objects are best vs. completions. In the case of non-hyperlinked URLs, >> I have the impression, correct me if I'm wrong, that positioning the >> hint elements is not quite possible, making completions the logical >> choice. > > ...Or perhaps the idea of switching completion-sets during completion is > redundant of browser objects, and we should not go that way. Though I > do like the idea of "narrowing completions" very much.
Sorry I keep following up on my own posts here, but I have been thinking about this problem throughout the morning. Let us make new browser-object-classes for history, bookmarks, and webjumps, and change find-url to be a browser-object command for which the default browser-object-class would be some kind of composition of bookmarks plus webjumps. We will need to brainstorm about how this "browser-object-class composition" would work, but I think we can come up with something. This brings the completions system into the browser-object system, and if we further find a means to switch browser-object-classes during an ongoing interaction, that will cleanly solve the issue that David raised, plus switching between any of the existing browser-object-classes. For my part, I'm not convinced of the need for this browser-object-class switching, because the browser-object system is built on a semantics of prefix commands. However, I'm willing to suspend judgement and see where this line of development leads. -- John Foerch _______________________________________________ Conkeror mailing list [email protected] https://www.mozdev.org/mailman/listinfo/conkeror
