Closed GoogleCodeExporter closed 9 years ago
This is typically a temporary problem caused by long tweets like "Oh
nooooooooooooooooooooooooooo!". It's something that could probably be "fixed"
in such a
way that makes it worse for others. Any fix should be careful not to stop long
URLs
from working.
Original comment by david.carrington
on 11 Aug 2009 at 9:31
Issue 23 has been merged into this issue.
Original comment by david.carrington
on 11 Aug 2009 at 9:38
Been looking in to this - it's especially bad when you set URLs to autoexpand.
HTML doesn't support wrapping inside tables (used in Dabr) and it's not
supported via
CSS until CSS3 (rules out lots of mobile phones).
So, I'm going to see which mobile browsers support the <wbr> tag
(http://net.tutsplus.com/articles/web-roundups/10-rare-html-tags-you-really-shou
ld-know/
and http://www.quirksmode.org/oddsandends/wbr.html for details)
I reckon on placing it in any word >=20 characters. That sound about right?
I figure, worse case scenario the unsupported browser will just ignore it.
Original comment by terence.eden
on 24 Dec 2009 at 10:47
We can accomplish this using PHP's wordwrap function -
http://www.php.net/manual/en/function.wordwrap.php
The question is, how many characters should this be cut off at? I'd say that 32
characters is long enough not to break most English words - but it may break
apart
some German and Japanese words.
Without knowing the width of the phone's screen (and font size)* etc it's
impracticable to calculate the "best" word wrap length.
So, for now, I'm suggesting a word wrap of 32 characters for URLs only.
Any objections / questions?
T
*Well, you could using DeviceAtlas or WURFL - but that would involve installing
extra
software and could mean spending time calculating a width which may be
incorrect.
Original comment by terence.eden
on 9 Feb 2010 at 1:52
While this benefits mobile users, it's also slightly detrimental to PC users -
approx
35%. Stupidly - I had not previously considered that this could be handled by a
theme() function and ignored appropriately on desktops.
The majority of mobile traffic comes through Opera Mini, where it is impossible
to
judge screen size.
Original comment by david.carrington
on 9 Feb 2010 at 2:04
[deleted comment]
In URLs it's not that noticeable. A random URL when rendered looks like
http://nlabnetworks.typepad.com/_transliteracy/2010/02/thoughts-a_t-half-time-at
-the-transliteracy_-conference-.html
Funnily enough, I use the Touch Mode when I use via the desktop. I don't
suppose
you've got any way of knowing which theme is used by the majority of desktop
browsers?
Original comment by terence.eden
on 9 Feb 2010 at 2:17
I don't have any storage or stats on what themes people are using.
Original comment by david.carrington
on 9 Feb 2010 at 2:18
curious as to why, with expanded URLs, the entire URL needs to be _shown_
instead of just the domain the link is
coming from
Original comment by ldoug...@gmail.com
on 11 Feb 2010 at 12:42
@lgouglas I agree that you could show just the domain - like Slashdot. But,
for me,
that's only half the point. Not only do I want to know that the link I'm going
to is
on a safe domain, I also want to know whether it's a story I've read before.
So, seeing "@edent check out [example.com]" isn't as useful as "@edent check out
example.com/blog/story-about-dabr"
I could set it so that you see the short URL but when you hover you see the
full URL
- but that breaks the tracking built in to some URL shorteners. Which,
admittedly,
isn't our concern.
Original comment by terence.eden
on 11 Feb 2010 at 8:27
ahh - I didn't think about url shortener tracking. Plus i agree the hover
wouldn't be
as useful on a mobile. Not sure how you could handle it other than showing the
short
link as well as the long link - maybe below? maybe truncated?
Original comment by ldoug...@gmail.com
on 13 Feb 2010 at 6:42
Since removing TABLEs and switching to DIVs the affects of really
looooooooooooooooooooooooooooong words is negated in that only that tweet
scrolls over, whilst the rest remain in the viewport.
Original comment by ryancul...@gmail.com
on 3 May 2011 at 10:12
Original issue reported on code.google.com by
Maxthon...@gmail.com
on 21 Jun 2009 at 1:02