Not enough people know this. If you're wary of a bit.ly link from a stranger, just add a "+" to the end. This will take you to a bit.ly page that has the full URL and a list of how many people have clicked it, who's tweeted it and when, etc. Also, if you're sharing a link, you could just add the + yourself to make it a bit less opaque. e.g. this link to Diaspora's Kickstarter page: http://bit.ly/dgpnNs+
I prefer to nag people to not use shorteners at all if you're not on Twitter. I don't know if it works, but I feel that it's a habit that needs to stop.
In your suggestion you fail to consider what prompted the 140 char restriction. The source of that restriction was that tweeter wanted the main source of tweets to be mobiles, using SMS. Now we can argue that they can allow people posting from the web to add some metadata but that would result in the inconsistency between the web and mobile protocols.
SMS has 140 bytes but not chars restrictions. When you use non-english letters you could not send 140 chars in one SMS. But you can use any 140 chars in Twitter.
So one to one mapping to SMS is not possible now.
When I enter SMS bigger then 140 bytes it's splitted transperently and the only consequence is increased payment for that SMS.
I think that nowdays the 140 char restriction is for keeping Twitter clean and readable, making it harder on the writer and easier on the reader - links are not hard to read and therefore there is no reason to restrict their length.
Also, users who actually use Twitter via SMS would probably not be capable of viewing the link anyhow so it could just say "(+lnk)" or something.
I used to manually type bit.ly links in from a tweet via SMS all the time. I didn't have a smartphone until a few weeks ago, and it was easier than opening twitter, typing in my password, finding the person's tweet, and clicking on it.
When I see a link I like and I want to pass it on to others, using bit.ly lets me track how many people follow it. If I just give out the link directly I have no idea how much effect my recommendation has had.
This is especially useful when recommending things for business. Already I've seent aht one way of recommending things is getting far more attention than another. Without some sort of tracking I would've known.
It's much more of an effort for me to put up some sort of referring system via other methods I can track - bit/ly provides me with this for free, both in terms of effort and money.
I don't deny any of that, I'm just saying that there are advantages.
For stuff I care about long term I put a reference on my own site, and I either cache the target page myself, of record enough of it that a Google search is likely to find something similar. That way I can fix it if it breaks.
For stuff that I don't care about beyond three months, why should I bother? OK, perhaps I should bother because of the potential for malware, etc, and I recognise that, but for me, the benefits outweigh the disadvantages.
If you think this is wrong, offer me a service that makes my life easier, otherwise all you're doing is saying "Don't do that."
And I'm not sure that I don't have at least some right to know how many people click on a link that I've provided, even if it's to someone else's domain. If I can get that information, and it's of use to me, why should I not do it?
Finally, I'm not talking about the links on my domains. For them I use long references. I'm talking about links provided in SMS, Facebook IM, twitter, and other similar media. I don't really see how providing short links there are "breaking the web" and would be interested in an clarification you'd care to offer.
(edited within minutes for typos and clarification.)
1) Most things on the internet don't last forever.
2) My idea of internet is that when I make impact I know about and can use that to make better impact. You are entitled to have your own idea of internet, but not to its superiority.
3) If you don't want me to know that you click on a link I sent you, don't click on it.
Shortened links are misleading, I might have not clicked your links if you wouldn't be hiding it. (already seen it, hate that specific blog, etc) - this makes your stats wrong cause many people might have clicked the link and closed it right away.
Last time I checked, bit.ly's statistics were counting all kinds of bots and did not allow filtering by useragent - they might have fixed this but in case they didn't then your stats are way wrong.
I'm aware of the counting limitations, and I believe I'm allowing for them in the rough and ready analysis I'm doing of the numbers I get. I have already seen that certain places get virtually no return, others get a lot. Adjusting my presences accordingly has increased my flow substantially, and while I'm aware that further tuning will require more detailed analysis, and more careful techniques, what I've done so far has been very useful as a first step.
I believe my points stand - bit.ly is useful for this context. I also believe you over-estimate the world's reactions to shortened links. My target is not the technically able, or even the technically aware. They see a link and click, unknowing and uncaring as to whether it's short, long or something in between.
I use the Greasemonkey userscript 'TinyURL Decoder' (http://userscripts.org/scripts/show/40582) which automatically decodes the tiny URLs on a page. Very useful on twitter (which the script has out-of-the-box support).
The vast majority of URL shorteners return the full URL with a HEAD request. Not practical for everyone, but the best way to do it if writing a service that might consume shortened URLs.
