Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Re: Dividing the viewport into areas with moveable borders

8 views
Skip to first unread message
Message has been deleted

Jukka K. Korpela

unread,
Oct 11, 2012, 8:41:01 AM10/11/12
to
2012-10-11 14:14, Tim Streater wrote:

> At present in my app I have a frameset of three frames arranged
> side-by-side. In the left-most, I subdivide horizontally but am using
> code for that. I took some time out to try a div-only test, the result
> of which can be seen at:
>
> <http://www.clothears.org.uk/temp/test.html>

That's an interesting exercise, but what do you actually win by moving
from frames (which are practically universally supported and do not
depend on CSS or JavaScript) to a much more complicated setup?

Apart from the "Frames killer" badge, that is.

Using frames, the borders are automatically moveable. In the good old
days, frames enthusiasts used to keep asking how to prevent that...

--
Yucca, http://www.cs.tut.fi/~jkorpela/
Message has been deleted

Eli the Bearded

unread,
Oct 11, 2012, 3:25:38 PM10/11/12
to
In comp.infosystems.www.authoring.html,
Tim Streater <timst...@greenbee.net> wrote:
> "Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:
> > 2012-10-11 14:14, Tim Streater wrote:
> > > <http://www.clothears.org.uk/temp/test.html>
> > That's an interesting exercise, but what do you actually win by moving
> > from frames (which are practically universally supported and do not
> > depend on CSS or JavaScript) to a much more complicated setup?

Frames suck big time in text mode browsers. Try using a framed site in
lynx or edbrowse. Here's Tim's page in edbrowse:

b http://www.clothears.org.uk/temp/test.html
14129
58
1,$p

Menu Bar

List


Preview


Activity Log


Actions


Tree

It has the same content as in Firefox. Frames don't work that way in
text mode. (edbrowse doesn't do much javascript, but it does do some.)

> > Using frames, the borders are automatically moveable. In the good old
> > days, frames enthusiasts used to keep asking how to prevent that...
> That is precisely my point. As far as I can tell, I gain nothing at all,
> for just the reasons you give. However, I'm interested to see if there
> is a smarter way to do it.

Resizable viewports that don't require frames have many potential uses.

Elijah
------
has been using edbrowse for twitter of late
Message has been deleted

Jukka K. Korpela

unread,
Oct 11, 2012, 4:09:41 PM10/11/12
to
2012-10-11 22:25, Eli the Bearded wrote:

> Frames suck big time in text mode browsers.

Text mode browsers have lost much of their importance since 1991 (or was
it 1992?), when I first used one. Even absolutely, and especially
relatively.

> Here's Tim's page in edbrowse:
>
> b http://www.clothears.org.uk/temp/test.html

That’s not a frames page. It’s a page that simulates frames in a sense,
as discussed.

A real frames site is viewable in Lynx at least, provided just that the
frames have meaningful names. Lynx effectively turns <frame> elements to
links. It also shows <noframes> content, but such content is usually
something between total cluelessness and insults (it *could* be
meaningful, too, but seldom is).

(Of course, a frames site could have *other* issues, and often has. But
let’s not blame frames for alt-less images or JavaScript-driven navigation.)

--
Yucca, http://www.cs.tut.fi/~jkorpela/

Eli the Bearded

unread,
Oct 11, 2012, 4:37:52 PM10/11/12
to
In comp.infosystems.www.authoring.html,
Jukka K. Korpela <jkor...@cs.tut.fi> wrote:
> 2012-10-11 22:25, Eli the Bearded wrote:
> > Frames suck big time in text mode browsers.
> Text mode browsers have lost much of their importance since 1991 (or was
> it 1992?), when I first used one. Even absolutely, and especially
> relatively.

I probably in a minority here, but I use them regularly.

a) "lynx -dump" is my preferred "html to rendered text" method.
b) I still care about and test for site usability for vision impaired
users, and have multiple different text mode browsers installed for
that purpose.
c) When my network connection gets swamped (bad wifi area, big download,
etc), non-graphical browsers work faster.

> > Here's Tim's page in edbrowse:
> >
> > b http://www.clothears.org.uk/temp/test.html
>
> That's not a frames page. It's a page that simulates frames in a sense,
> as discussed.

Exactly, and it shows the whole page at once, unlike a frames site
where you can only see one frame at a time. Lynx works the same way,
showing you just a single frame at a time on a frames site. The frameless
frames are an improvement.

> (Of course, a frames site could have *other* issues, and often has. But
> let's not blame frames for alt-less images or JavaScript-driven navigation.)

edbrowse has javascript support (it uses the Spidermonkey libraries) to
deal with JavaScript-driven navigation. It has enough javascript support
that it shoots itself in the foot on one frames-based site I use.[*] That
site has javascript to redirect the browser to the frameset if you go to
a subpage. edbrowse cannot view any of the subpages. Lynx can view the
subpages, but then can't follow any of the javascript links on them.
Usability nightmare.

Elijah
------
[*] without naming it, let me state it is a game site that does not need JS

dorayme

unread,
Oct 11, 2012, 5:12:40 PM10/11/12
to
In article <k56eor$dlq$1...@dont-email.me>,
"Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:

OP did say "but it feels to me quite a lot of code to do
something provided for free by the browser if I use a frameset."

--
dorayme
Message has been deleted

Eli the Bearded

unread,
Oct 11, 2012, 6:16:30 PM10/11/12
to
In comp.infosystems.www.authoring.html,
Tim Streater <timst...@greenbee.net> wrote:
> Eli the Bearded <*@eli.users.panix.com> wrote:
> > Exactly, and it shows the whole page at once, unlike a frames site
> > where you can only see one frame at a time. Lynx works the same way,
> > showing you just a single frame at a time on a frames site. The frameless
> > frames are an improvement.
> Eh? My test simulated exactly the layout my app has with frames, all of
> which are visible all the time. I don't know what you are thinking about
> by saying "with frames you can only see one frame at a time".

I am referring to the way text mode browsers like lynx and edbrowse handle
framesets. Not to how graphical browsers do.

> What leads you to believe that only one of those is seen at a time?

When was the last time you used lynx?

$ lynx -dump frames.html

FRAME: [1]mainwindow
FRAME: [2]buttons
FRAME: [3]boxes

References

1. file://localhost/tmp/something.html
2. file://localhost/tmp/buttons.html
3. file://localhost/tmp/boxes.html
$ edbrowse
edbrowse ready
b /tmp/frames.html
243
49
1,$p
Frame {mainwindow}
Frame {buttons}
Frame {boxes}
q
$

Following a link will just show you a single frame at a time in those
browsers.

Elijah
------
the frameless frames show everything in lynx and edbrowse
Message has been deleted

Thomas 'PointedEars' Lahn

unread,
Oct 12, 2012, 3:07:13 PM10/12/12
to
Eli the Bearded wrote:

> In comp.infosystems.www.authoring.html,
> Tim Streater <timst...@greenbee.net> wrote:
>> "Jukka K. Korpela" <jkor...@cs.tut.fi> wrote:
>> > 2012-10-11 14:14, Tim Streater wrote:
>> > > <http://www.clothears.org.uk/temp/test.html>
>> > That's an interesting exercise, but what do you actually win by moving
>> > from frames (which are practically universally supported and do not
>> > depend on CSS or JavaScript) to a much more complicated setup?
>
> Frames suck big time in text mode browsers. Try using a framed site in
> lynx

BTDT. Both Lynx and (E)Links support framesets. There is a link list of
frames in Lynx. ELinks supports them in the common, graphic fashion; a
console wider than 80 columns is advised.

> or edbrowse

<http://the-brannons.com/edbrowse/>

I have never heard of that. The latest snapshot is dated 2011-11-30, and
requires SpiderMonkey 1.8.5 ("and later", but given Mozilla's pace that is
probably irrelevant now). No accessibility features. No distribution
packages. Probably irrelevant.

Screen reader software reads lynx or w3m output, or even that of graphic
browsers like Firefox (e. g., with Orca), these days.

But: Frame*sets* should not be used where they can be *reasonably* avoided.
However, for some use-cases, like indexes (e. g. of API documentation) they
are very useful and backwards-compatible, much more backwards-compatible for
example than any CSS+ES/JS+XHR approach. So they can *at least* be a viable
fallback for that (not all ELinks builds come with ES/JS support, AFAIK none
has XHR support).


PointedEars
--
realism: HTML 4.01 Strict
evangelism: XHTML 1.0 Strict
madness: XHTML 1.1 as application/xhtml+xml
-- Bjoern Hoehrmann

tlvp

unread,
Oct 13, 2012, 6:22:18 PM10/13/12
to
On Thu, 11 Oct 2012 12:14:45 +0100, Tim Streater wrote:

> ...
> At present in my app I have a frameset of three frames arranged
> side-by-side. In the left-most, I subdivide horizontally but am using
> code for that. I took some time out to try a div-only test, the result
> of which can be seen at:
>
> <http://www.clothears.org.uk/temp/test.html>
>
> The salmon-coloured dividers can be moved around with the mouse, quite
> smoothly it seems, but it feels to me quite a lot of code to do
> something provided for free by the browser if I use a frameset.

What an astonishing tour de force, Tim! I had no idea one could do that.

Cheers, -- tlvp
--
Avant de repondre, jeter la poubelle, SVP.
Message has been deleted

tlvp

unread,
Oct 14, 2012, 4:41:27 PM10/14/12
to
On Sun, 14 Oct 2012 12:39:46 +0100, Tim Streater wrote:

> ... some folks are too quick to say "frames are evil" ...

I'll say only frame-based sites annoy me -- if there's anything there worth
bookmarking, simply asking a browser to make a bookmark captures only the
site's entrance URL, not the particular page you've navigated to (which may
well be a long, intricate, and nearly irreproducible click-stream in). So
you have to have the presence of mind to right-click in the frame of
interest and bookmark that frame ... if you realize it's a framed site :-).

That's not enough to qualify as "evil" -- but it is enough to keep me from
ever wanting to inflict a framed site on the world :-) .

Oh ... and as for that other thing -- you're welcome. Cheers, -- tlvp
Message has been deleted

tlvp

unread,
Oct 15, 2012, 2:25:19 AM10/15/12
to
On Sun, 14 Oct 2012 22:41:33 +0100, Tim Streater wrote, quoting:

>> ... -- but it is enough to keep me from
>> ever wanting to inflict a framed site on the world :-) .
>
> Well, I take the point generally, but in my case it's not a website,
> it's an application which is not intended to be bookmarked and will be
> unreachable by crawlers.

Fair enough, then -- you're not "inflicting [it] on the world" :-) .

Cheers, -- trlvp
Message has been deleted

Eli the Bearded

unread,
Oct 23, 2012, 4:00:24 PM10/23/12
to
Thomas 'PointedEars' Lahn <dci...@PointedEars.de> wrote:
> Eli the Bearded wrote:
> > or edbrowse
> <http://the-brannons.com/edbrowse/>
...
> Screen reader software reads lynx or w3m output, or even that of graphic
> browsers like Firefox (e. g., with Orca), these days.

I gather that edbrowse is written with a braille reader in mind. The
ed(1)-style interface, developed for line-at-a-time hardcopy terminals
is well-suited for other one-line displays, like most braille terminals.
(Braille wheels displays seem more akin to an endless ticker-tape than
a fixed-size line. I'm not sure they ever reached commercial production.)

Elijah
------
back from vacation and reading the group again
0 new messages