Thursday, December 22, 2011

A Gooey Browser

Let me talk to you for a bit about my dream for the future of internets.
This is the first part, as there are many things I want to say, and the future part(s) will have pictures (yay) hand-drawn by me(ohmai).
Let us start with one thing thing, that would in the end be two.

My biggest issue with StumbleUpon is that there is no in-between, between thumbs up and thumbs down. And then that there is no way to sort the things I liked.
I want an all-around net experience, either in website or browser form (though browser would probably be more effective and powerful for what all I want from my prime internetting experience.)

I want something that can divide my screen into... lets call them carrels. Customizable configuration and number.
I can do this a bit with Windows 7 and browser tabs. But I want something a bit more... gooey... than that.

Honestly, the only reason I haven't pursued my idea for this thing any further than just wanting it (even with my limited coding and such knowledge) is that I can't think of any way to build a better tagging engine. At least, not without more people on board with a Dublin Core or similar metadata format. Because if no one does the front-end work of tagging things, or classifying them in some system (Dewey, anyone?), at our present level of AI, no computer is going to be able to automatically sort everything and apply tags accordingly.
That's why things stay hidden on Google, I suppose.

Anyway, the vision, I guess.
I want StumbleUpon's functionality, pulling up random sites
(though it would be nice to somehow get things not suggested by users. Truly random websites. That goes back to not having the AI to accurately determine site content. Can't have the program trying random strings of letters/numbers/etc. without some way to determine what it is looking at. (Though, if it were to do this as a subfunction, not directly in the line of use... Forming it's own bank of websites as well as suggestions... Hmm.) Though it would be nice to allow NSFW as a filter so you could 'stumble' porn only, if you were so inclined)
liking and disliking, but I want a third option of 'seen'. Because there are times when something gets pulled up on Stumble that is mildly interesting, so I wouldn't mind seeing more things like it (so I don't want to dislike it) but not so interesting that I want to keep it (so I don't want to like it) but I don't want it to come up again (since I've seen it.) So a third option allowing me to tell the program that I saw the website, and don't want it added to my likes or dislikes would be good.
I would say that I want it to mark every landed page as "seen", but there are times when I skip things on Stumble because it isn't loading well, because I don't have time to look at it long enough yet, or because I think I would like it to come up again sometime, as a surprise, I guess. So I wouldn't want it marked as seen.
Though perhaps something like Google Reader's "Mark Unread" would work instead. Eh.

Anyway. That would be one functionality.
I would also like the option to on-the-fly turn on/off NSFW content, videos, animations and content filters.
I also want a "Challenge Me" filter. StumbleUpon keeps track of what you like and dislike, and formats your experience to you. I once wrote a draft of a post (that never got posted) that touched this. Here is the video I really wanted to show you that explains it a bit:
I want a filter I can turn on that will show me things that fall outside of those bubbles. Just because I never "liked" anything about Prop 8 being shot down from the conservative perspective doesn't mean I don't want to read it. Or that I should never have a chance to be exposed. It would still only show me new things (unseen), but it would ignore the filters of what I've liked/disliked.

Added to these in-use functions, I want post-use functions. This is where I want to attach bookmarking functions to the StumbleUpon functions (as modified above).
PearlTrees is my bookmarking application of choice. It allows me to organize my favourites into, well, trees.
The desire to set my Tree up in Dewey classification is a constant battle between end-use and front-end work. So far the desire for less front-end work has won out.
Regardless, this is part of why I want a 'seen' button on StumbleUpon, for easier conversion to other areas.
I can't find things I've liked on StumbleUpon. I want to be able to categorize them, or have them categorized for me(preferably). That goes back to the top and AI being unreliable.
Yet I don't know how far I've gotten in transferring my 'likes' into pearls as if I remove the like from the item on SU, it will come up again when I go stumbling. And I can't favourite directly from SU to PearlTrees.
It would be nice to have user-end classification, or tags as well. Because a specific site may mean something specific to me, so I want to give it a tag that I can find it by (say, "showmum"). This shouldn't affect anyone else's classification, just mine, within my "browser".


Now, I'm going to snip this here, mainly because I don't feel I'm making much sense and want to draw up some diagrams and junk to clear my head in order to explain better, but also because I have only touched on two of the many things I want to go into the carrels, and this post would get quite long if I didn't.

No comments:

Post a Comment