On Mon, 2008-04-28 at 09:54 +0100, Matthew Paul Thomas wrote: > On Apr 27, 2008, at 7:30 PM, Dylan McCall wrote: > > ... > > I think meeting Firefox 3's notable features could be quite easy for > > Epiphany because its design is very flexible already. For example, I am > > working on a patch to add a "quick bookmark" button, which creates a > > bookmark with the page's name and a "Miscellaneous" tag on one click, > > (How would "Miscellaneous" be better than no tags at all?) Ideally, it shouldn't be, but it seems the easiest solution. My thought with the quick bookmarks is that they should be out of the user's way, not cluttering the more streamlined Bookmarks menu but still being searchable and known as "interesting" to the user. Another, tidier solution would probably be an arbitrary _hidden tag. > > > then pops up the bookmark editor on a second click. We already have > > most of the stuff in place to replicate the feature :) > > Even better than Firefox would be to automatically save the contents of > a page when you bookmark it (perhaps up to a maximum of 1 MB or so per > page). That way when you search your bookmarks, you can search their > contents, not just their titles (you're more likely to remember text > from the contents than to remember a title you might not have even > noticed). And if the page ever disappears, or when you're offline, > you'll still have a copy of the page to refer to. > Perhaps xesam could help us there? It would be rather out of scope for Epiphany to power a miniature search engine, although Beagle and Tracker seem to do all right with web pages. Something I think could be cool is bookmark tags being automatically picked based on page metadata. > > ... > > Oh, as for crazy features: How about the browser being able to watch > > changes for bookmarked pages? It could periodically check for changes, > > and when such occurs, it could promote the bookmark somehow. That could > > be done via RSS feeds and page scraping. > > ... > > Netscape versions 2~4 did this (minus the feeds part, of course), with > a "What's New?" menu item in the Bookmarks window. It worked okay in > the early '90s, when most Web pages were static. But now that most Web > pages are served dynamically, they often contain trivial changes (e.g. > a "Latest headlines" sidebar, or a copyright statement containing the > current year), and the last-modified date sent by the Web server is > more often wrong than right. So detecting whether a page has > significant changes is now an AI-class problem. > That is true. Still, there are some tricks out there, such as that thing with automatic bookmark names (argh, forget what it's called! That Firefox feature they demonstrated with ebay auctions, showing up to date status right in the bookmark title), and feeds, which don't necessarily need to be parsed, but can be expected to output in a sane manner. > Cheers
Attachment:
signature.asc
Description: This is a digitally signed message part