Time for a new Twitter tag | Hacker News Reader

Time for a new Twitter tag

3 pointsrobchez8 comments

It was the original users of twitter who built the syntax we use today. The '@' symbol for referring to users, the 'RT' code for retweeting, the topic(?) '#' tag. The developers of twitter applications incorperated these into their applications, and eventually it became the standard, being adopted by Twitter.com themselves.

Since then Twitter has become a powerful method to distribute links to cool content. However, with the 160character limit, characters are a premium, so a whole bunch of services were born (bit.ly, goo.gl, etc.) to shorten URL's.

http://www.google.com.au/search?sourceid=chrome&ie=UTF-8&q=fluffy+bunnies (73 Characters or 45.63% of total tweet) became http://bit.ly/bRajwN (20 Characters or 12.5% of total tweet) a massive saving in tweet length. Why not take it a step further?

Why not remove the need for the "http://"? I mean do we really need to know that we are accessing the Hypertext Transfer Protocol? Google removed the http:// from Google Chrome back in April

By removing "http://" we will be saving seven characters from any tweet that has one link in it (or 4.38% of total tweet). TechCrunch recently reported that Twitter sees 90 million tweets per day with 25% containing links. That equates to 22.5 million links. All those links include "http://" that is 157.5million characters sent every day.

So if we remove it, how would my favorite Twitter software recognize a link? A new tag, hereby christened a linktag;

^

Why ^ ? Well @ and # are already tags. Moving along the keyboard, we can't use $ or % as I assume they are used quite a bit. But ^ is rarely used in text. It also looks like the head of an arrow pointing up, like a link out of the page.

The link tag will replace the 'http://' (7 characters or 4.38% of tweet) with '^' (1 character or .63% of tweet), this will mean that 135 million characters will be removed from tweets every day.

I dislike the ^ character in most contexts, including this one (as well as with Apple's "blocks" extension to C, where they are designed to resemble an asterisk -- but IMO, do a poor job of it).

Might I suggest the % symbol? It's full-height (like # and @), close on the keyboard, and not used at the beginning of words currently.

^ is often used by accounts that have multiple users. For example check out the way bank of america support reps "sign" each tweet with ^ and then initials:

http://twitter.com/bofa_help

I believe you can just leave out the http:// altogether and Twitter will still pick up the link.

No, it's time for a move back to hyperlinks with anchor text.

People who require tweets over SMS are welcome to stay in the past.

Interesting idea.

'^' doesn't feel right in the text flow, and as answerly noted, it's already used as an alt signing symbol.

I think a ':', '>', or even '//' would work well, still likely to be clear in context, suggestive of the original form. Examples:

:bit.ly/cYrNb3

>bit.ly/cYrNb3

//bit.ly/cYrNb3

The issue with all of those is that they are different than '@' and '#': they aren't full-height characters that fill up most of the space, they have a lot of white space. Other ones that seem to have that same (desired?) property are '$', '&', and '%' -- maybe one of those would be better.

However, that negates your point about resembling the original link, so I'm not sure if they are better.

I understand an aesthetic problem with vertical asymmetry and non-alignment in conveying this meaning, as with a '^'. It's off the main-line of text. It's pointing away from the content. It doesn't connect strongly to the substance.

I don't see the aesthetic advantage of 'fill up most of the space' -- as long as the glyph is easily recognizable and draws the eye to the substantive info.

Conveying motion to the substance (as with '>' or to a lesser-extent the leaning '//') fits the meaning of an outlink.

Providing a reminder of the elided part (as how ':' or '//' is suggestive of the missing 'http://) also fits the link usage.

I think '@' works for naming identities because it's evocative of email addresses and often read as 'at', suggesting a destination. Not sure why '#' won out; the familiarity of web users to '#anchors' in a page doesn't seem salient enough.

'$' and '&' have such strong meanings they're harder to overload for other purposes.

I think that '#' is a remnant of IRC channels, which used that 'tag' as well and are also focused on a specific subject.

(Just a personal theory, though, no evidence.)

Reply on news.ycombinator.com