This is in reference to http://news.ycombinator.com/item?id=937032 for those who are confused.
* No easy way to search my personal web/reading history. The problem is not just collection, it's relevance. I want to go back and find a book recommendation on design I saw in a blog post sometime in the last week. I dunno where it is or if I bookmarked it. Solution? Google Desktop / WebMynd plus some radically different ideas about search.
* I can't use Amazon EC2 with Mosso CloudFiles quickly, cheaply, and at scale. Solution? Cloud peering.
The idea is to save enough data that you can have an algorithm that will automatically differentiate between stuff you closed within a couple of seconds and didn't really read and stuff that interested you and you actually read through. And the reason for doing this automatically is it seems very difficult to have the discipline take an action every time you see a page that you may want to save so that you can search through again (such as clicking a box for every such page).
Whatever mechanism it is, it needs to account for stuff read through readers and over multiple computers/devices too. I don't want to miss a search result just because I was reading it on an iPhone or in my Google Reader.
Or instead of caching locally, a search interface that would accept a long list of urls to look through.
Other things you'd take into account are:
* continuity in time - reading several pages on similar topics one after the other.
* analysis of the subgraph of the web graph that you visited
* you'd be more aggressive with query expansion because you work with much smaller corpus, e.g. you'd want to take hypernyms and hyponyms into account
* Too many notifications. I have to login to a half dozen different systems to keep up on what's happening. That's too much. There are notifications I just want to see, some I want to add to my todo list, some to my calendar, some I don't want to see again. A single-app solution for all of this would be brilliant. Email is too heavy: I want notifications, not correspondence. Bonus points if it's on my mobile.
* Some way to extract the most valuable points out of a meeting. A possible solution might be a 'real time digg' that lets people vote speakers up and down when they agree or disagree with their point. Correlate that with a copy of the transcript and emphasize sections with higher scores. Possibly two-dimensional voting for 'We need to remember this' and 'I agree with this'.
I'm surprised no one has done this?
...should be done by the weekend...
What sort of notifications? Please share some examples.
Class announcements (currently emailed to me)
Reminders from my calendar (currently email or rss'd)
New posts from blogs I follow (currently RSS)
Notification of shared document changes (currently login to each service to check)
Private messages on forums (currently emailed)
It can be a new service not built on Twitter, say on identi.ca or friendfeed.
There'd be an API, so parties sending out emails can build into the service with a pre-established userbase.
Do you want to build it?
(afterward ... with relationships established with businesses.. can follow through into my idea http://bit.ly/1MscB3)
Given the amount of reaction this comment thread got, I'm considering hacking out a prototype and seeing if it helps alleviate the problem. That being said, I've already got far too much on my plate. I don't think you'd ever come up with this idea if you were the kind of person who had time to kill checking their notifications.
I'm possibly willing to spend $4000 on this, or something like it, which I'll build the front-end - depending on my work schedule.
The other side of the coin is information you'd like from firms, but are not getting, rather than vice versa - hence the friendfeed for business information.
(edit: the other aspect is having the end user be able to respond to some type of company information, in a type of Google Wave, so another response can be made)
I built it with the intention of 'reducing' the amount of correspondance, and to only notify you of something that you exactly want.
Solution: site that lets you generate a list of sites to contact. With direct links to that site's contact us page.
2. New Season TV updates for basic cable, you know what I'm talking about you watch a TV channel and they keep repeating the same shows 20 times over.
Solution: site that let's you "follow" a show + channel, and get notifications when that channels starts airing a new season.
3. Where are they now? Where are all of those celebrities, politicians, musicians that have retired from the public eye? What are they doing now?
Solution: IMDB like site for retired personalities. Would show things like,
Name: John Doe
Famous For: Congressman from XX from 19xx-19xx
Latest TV appearance: Youtube link
Current Occupation: Lobbyist for ______
Name: James Doe
Famous For: Actor(Movie #1(amazon link) - Movie #2(amazon link)
Latest TV appearance: Youtube link
Current Occupation: retired in Florida, spends his spare time working on his bikes.You can be notified by email and Instant Messager (if you register using a gmail account).
Syntax highlighting for natural language is a possible improvement. I have tried variations of this with no success so far: parts-of-speech tagging in color doesn't seem to help, for example. Automatic extraction and highlighting of named entities is probably most promising.
More generally, it's about ways to help humans understand electronic text better by providing them hints and interactions that were not possible (or were difficult) with paper.
An established example of this is searching within a page. There are different implementation of this, some better than others: search-as-you-type is good, and so is highlighting all results. I haven't yet seen implementations of query expansion like search engines do: correcting spelling mistakes, finding all word forms, finding synonyms. Finding several terms close to one another would also be helpful.
Solution: Make paper bags and sell those. Use less plastic. Spread awareness.
Solution: Make all the servers a network of bittorent peers.
This way load balancing, and bandwidth problems can be solved.
Downloads can be stopped and resumed.
After some time the server need not host the file at all.
There will be virtually no down time.
If web sites can be served using the bitorrent technology, awesome!
Solution: Some web service to exchange old books free of cost.In the Uk and europe (western europe anyway) supermarkets now either (a) don't let you have them at all, (b) make you pay for them, or (c) let you have them but make you ask the checkout operator instead of having a bunch that you can help yourself to.
Most supermarkets are also selling larger more sturdy plastic shopping bags. These are marketed as a 'green' iniative as using less of the other bags is supposedly good for the environment.
The Waitrose near me has makes a half-assed effort, whereby no matter how much stuff you have (literally, a heaped shopping cart), the lady will look at you and ask "Do you need a bag?" The idea is that if they ask, sometimes you'll say no, and use fewer bags. Usually I respond with something like "Nah, I'll just take the cart, thanks" or "It's fine, I'm a great juggler" or some other stupid quip.
Marks & Spencer also ask you if you want a bag. If you say yes, they charge you 10 pence PER bag. So if you do a lot of shopping at M&S, you really need to bring your own suitcase or something.
Tesco doesn't give a crap. They find baby seals and tie plastic Tesco bags on their heads.
* Video authoring: with the complexity of containers, codecs, DVD formats etc. it's still far from what the MP3 was even 6 or 7 years ago, despite iMovie's best efforts.
* Because of the way the Western 12-note musical system is devised, guitars have intonation problems that can never be fully overcome. I found the solution to this about 5 minutes ago: http://www.truetemperament.com/site/index.php Quite whacky.
2) Higher cab fares during night time so that drivers have incentives to operate during night, especially, on friday and saturday nights.
3) Lower medical/insurance costs
What? You're going to incentivize driving over taking a cab? At least in NYC, this is a huge mistake.
Incentivize biking for everyone and good public transit. If you're going to think about fixing transit, stop thinking about cars in the first place.
There's an idea - Bar Van, free rides home from the bar, completely funded by advertising from area businesses that want to support "don't drink and drive" initiatives.
"Oh crap, I have a late movie sitting here collecting fines. I'll just order another pizza so they'll pick it up..."
The idea is well intentioned, but what kind of liability do you think private companies would get from putting a bunch of drunk people in a crowded van at 2AM on a weekend? I think it would have to be a state sponsored thing for it to work.
Maybe there's no legal liability for Bob's Bakery in a situation like that, but it is certainly bad for business.
With Femtoo.com you can track events - you can specify what type of event you are interested in (conferences for example), enter some keywords (Java, Pottery etc) and when the event is schedules (3 months in advance), you will be notified.
You can be notified by email, Instant Message, or a URL callback.
When you've logged in, choose to create an 'Event Tracker'...enjoy!
(I sympathize, I built Femtoo to avoid missing concerts of my favourite bands!)
My solution, a voluntary agency to preapprove tv series to make sure they have adequate funding and buy in from the network to stay alive until the story reaches a conclusion, and that the story is coherent and makes sense. Viewers like me would only watch shows that get this approval.
better...belief persuasion. existence would be so much less traumatic if belief in some kind of afterlife or reincarnation could be instilled after childhood.
for serious this time: ...
nah, who am i kidding. i'm educated, certified, white, and working on my own startup. i live the most charmed existence imaginable (other than the no afterlife thing). i think it's better to scratch more worthwhile wounds.