4:09 a.m.
I have downloaded the latest test build r12932, and find that it
treats frames in an entirely different way to what it test builds did
less than 3 months ago.
Positioning the mouse pointer over any one frame and rotating the
scroll wheel used to scroll only the frame that it was underneath it;
now it scrolls the whole caboodle, and reveals a previously unseen
blank frame at the bottom. (??)
It also scrolls (using the slider bars) much less smoothly, and rather
more jerkily.
I am looking at:
Wild Flowers http://wildflowerfinder.org.uk/
--
Cheers
Roger
Burning the midnight candle at both ends of the barrel
4:24 a.m.
On 3 Oct 2011 Roger Darlington wrote:
I have downloaded the latest test build r12932, and find that it
treats frames in an entirely different way to what it test builds did
less than 3 months ago.
Positioning the mouse pointer over any one frame and rotating the
scroll wheel used to scroll only the frame that it was underneath it;
now it scrolls the whole caboodle, and reveals a previously unseen
blank frame at the bottom. (??)
It also scrolls (using the slider bars) much less smoothly, and
rather
more jerkily.
I am looking at:
Wild Flowers http://wildflowerfinder.org.uk/
<wishlist> It would be nice if it scrolled the frame first, and then
scrolled the whole frameset when you hit the top or bottom of the
frame. </wishlist>
Failing that, for this particular site I would prefer to be able to
scroll the frame with the scroll wheel and use the right hand scroll
bar for the whole caboodle. However, it's not a situation you come
across very often these days.
--
Richard Porter http://www.minijem.plus.com/
mailto:ricp@minijem.plus.com
I don't want a "user experience" - I just want stuff that works.
4:21 p.m.
On 3 Oct 2011, Richard Porter wrote:
On 3 Oct 2011 Roger Darlington wrote:
> I have downloaded the latest test build r12932, and find that it
> treats frames in an entirely different way to what it test builds did
> less than 3 months ago.
> Positioning the mouse pointer over any one frame and rotating
the
> scroll wheel used to scroll only the frame that it was underneath it;
> now it scrolls the whole caboodle, and reveals a previously unseen
> blank frame at the bottom. (??)
> It also scrolls (using the slider bars) much less smoothly, and
rather
> more jerkily.
> I am looking at:
> Wild Flowers http://wildflowerfinder.org.uk/
<wishlist> It would be nice if it scrolled the frame first, and
then
scrolled the whole frameset when you hit the top or bottom of the
frame. </wishlist>
Failing that, for this particular site I would prefer to be able to
scroll the frame with the scroll wheel and use the right hand scroll
bar for the whole caboodle. However, it's not a situation you come
across very often these days.
I cannot see why it should scroll the whole caboodle at all, certainly
scrolling it doesn't reveal anything of use to anyone.
I think the way Netsurf works with frames at the moment is definitely
4 steps backwards to the way it used to work.
Firefox now fares much better in this regard, but Netsurf used to be
much better than Firefox in this regard. Why? Because previously in
Netsurf you did not even have to click in the frame to scroll it with
the mouse wheel, merely hover the mouse over the selected frame. [With
Firefox you had to click in the selected frame before it would scroll
using the mouse wheel].
I will keep down-loading newer versions of Netsurf until I find one
that works as well as did the version of just 3 months ago.
--
Cheers
Roger
The only thing that moves is the wind.
9:31 p.m.
On Thu, October 6, 2011 8:21 am, Roger Darlington wrote:
I cannot see why it should scroll the whole caboodle at all,
certainly
scrolling it doesn't reveal anything of use to anyone.
I think the way Netsurf works with frames at the moment is definitely
4 steps backwards to the way it used to work.
I think it's worth pointing out that, at least AIUI, frames are currently
a work in progress.
When they "worked", they were provided by the RISC OS front-end and so we
were the only platform to have support. They are now provided by the
core: this makes them available to all platforms, and simplifies the RISC
OS code significantly (which as anyone who has looked at the code in
question will know, is no bad thing).
AFAIK, the internal APIs for things like scrolling are still being worked
on, so at present support is still less complete (for RISC OS users) than
it once was. Hopefully this should change over time, rather as the
bookmarks, history and cookies windows returned to normal following their
move into the core a while back.
--
Steve Fryatt - Leeds, England
http://www.stevefryatt.org.uk/
4:02 p.m.
On 6 Oct 2011, Steve Fryatt wrote:
On Thu, October 6, 2011 8:21 am, Roger Darlington wrote:
> I cannot see why it should scroll the whole caboodle at all,
certainly
> scrolling it doesn't reveal anything of use to anyone.
>
> I think the way Netsurf works with frames at the moment is definitely
> 4 steps backwards to the way it used to work.
I think it's worth pointing out that, at least AIUI, frames are
currently
a work in progress.
When they "worked", they were provided by the RISC OS
front-end and so we
were the only platform to have support. They are now provided by the
core: this makes them available to all platforms, and simplifies the RISC
OS code significantly (which as anyone who has looked at the code in
question will know, is no bad thing).
AFAIK, the internal APIs for things like scrolling are still being
worked
on, so at present support is still less complete (for RISC OS users) than
it once was. Hopefully this should change over time, rather as the
bookmarks, history and cookies windows returned to normal following their
move into the core a while back.
OK, many thanks Steve. I will keep downloading new ones...
--
Cheers
Roger
Sailing on the Ship of Theseus across Heraclitus' River
9:10 p.m.
New subject: frames [on RISC OS]
On 3 Oct, Roger Darlington wrote in message
<02a8841c52.rogerarm(a)rogerarm.freeuk.com>:
I have downloaded the latest test build r12932, and find that it
treats
frames in an entirely different way to what it test builds did less than 3
months ago.
Positioning the mouse pointer over any one frame and rotating the scroll
wheel used to scroll only the frame that it was underneath it; now it
scrolls the whole caboodle, and reveals a previously unseen blank frame at
the bottom. (??)
Some of these issues have now been sorted in the RISC OS front-end, as of
r13292.
Frames can be scrolled using the scrollbars and by dragging, as has always
be the case.
In addition, they can now be scrolled using the keyboard. The browser
window must have focus, and the caret must not be in the URL bar (as many of
the scroll keys are also used by URL completion); a click somewhere in the
main part of browser window will achieve this. After that, the cursor keys,
Page Up/ Down, Home and End will have the usual effects: either on the whole
page (for unframed sites) or for the frame under the mouse pointer.
The scroll wheel is more tricky. In theory it will behave as the keyboard
does, and scroll whatever is under the pointer. However, the problem is
that there's no standard API for passing on scrollwheel events under
RISC OS, so there may well be incompatibilities. Operation on Select and
RISC OS 6 remains untested.
* The scroll wheel won't work under RISC OS 4.02 or earlier.
* Later versions of RISC OS 4 and 6, which include ROL's scroll wheel API,
should work "out of the box". If you have it configured to "Focus"
or
"Focus or Pointer" behaviour, however, remember that NetSurf still needs the
pointer to be over the window in question so that it knows which object
you're trying to scroll. Also remember that support on these OSs is
untested.
* RISC OS 5 without HID should work as standard. However, the OS seems to
send scroll requests in clumps of five to the browser, so it may scroll more
that you would expect.
* RISC OS 5 with HID should also work, but NetSurf needs to get scroll
information from HID in "Scroll Request" mode. If this is your default
setting for HID then things will just work; if not (and I think that HID
defaults to a different default, IYSWIM) then you need to add an exception.
Select 'Scroll wheel' from HID's iconbar menu; click Menu on the list that
appears and choose 'Edit->Add new line at bottom'. Select the new line,
then choose 'Edit->Edit line...' from the menu. In the dialogue, set
'Application' to "NetSurf" (via the pop-up menu, if NetSurf is running),
and
set 'Vertical method' and 'Horizontal method' to "Scroll
Request". 'Step'
should be "1" for both. Click 'OK'; you'll then need to save the
settings
(which HID might prompt for when you shut down).
* If you have a scroll wheel setup that scrolls by sending Open Window
events to NetSurf (which is how HID works in some other modes, and how I
think STD's old PS2Mouse system worked pre-ROL), then things will still work
on unframed sites. With frames, however, you'll see the effect described by
Roger above: the whole site will simply scroll out of view and leave a blank
work area in its place.
--
Steve Fryatt - Leeds, England Wakefield Acorn & RISC OS Show
Saturday 28 April 2012
http://www.stevefryatt.org.uk/ http://www.wakefieldshow.org.uk/
9:02 a.m.
New subject: frames [on RISC OS]
On 18 Dec, Steve Fryatt wrote:
RISC OS 5 with HID should also work, but NetSurf needs to get scroll
information from HID in "Scroll Request" mode. If this is your default
setting for HID then things will just work; if not (and I think that
HID defaults to a different default, IYSWIM) then you need to add an
exception. Select 'Scroll wheel' from HID's iconbar menu; click Menu
on the list that appears and choose 'Edit->Add new line at bottom'.
Select the new line, then choose 'Edit->Edit line...' from the menu.
In the dialogue, set 'Application' to "NetSurf" (via the pop-up menu,
if NetSurf is running), and set 'Vertical method' and 'Horizontal
method' to "Scroll Request". 'Step' should be "1" for
both. Click
'OK'; you'll then need to save the settings (which HID might prompt
for when you shut down)
Since following this advice, long menus no longer scroll with the
wheel. RO 5.16 and r13295.
Example site - http://www.tvguide.co.uk/ - scroll to the bottom of the
page and click on the "Add Channel" menu - you have to use the scroll
bar to scroll the menu.
--
Brian Howlett
---------------------------------------------
"What's the matter Colonel Sanders? Chicken?"
10:42 p.m.
New subject: frames [on RISC OS]
On Mon, December 19, 2011 12:02 am, Brian Howlett wrote:
Since following this advice, long menus no longer scroll with the
wheel. RO 5.16 and r13295.
Example site - http://www.tvguide.co.uk/ - scroll to the bottom of the
page and click on the "Add Channel" menu - you have to use the scroll
bar to scroll the menu.
Does that only applies to long menus in NetSurf?
If so, then presumably NetSurf's menus look just like NetSurf windows and
HID now sends them scroll requests as well. Unless those requests end up
being sent to NetSurf so that it can scroll the menus manually in the way
that HID would have done before (I'm not sure, and I'm not near enough to
a RISC OS box to check), then there's no way around the problem. Even if
they do, it will be messy and I suspect we're straying into "undocumented
areas" of the Wimp.
This could be more of an HID feature than a NetSurf one: I'll investigate
when I get the chance. As I said in my first post, scroll wheel support
on RISC OS is something of a minefield... :-(
--
Steve Fryatt - Leeds, England Wakefield Acorn & RISC OS Show
Saturday 28 April 2012
http://www.stevefryatt.org.uk/ http://www.wakefieldshow.org.uk/
1:50 a.m.
New subject: frames [on RISC OS]
On 19 Dec, Steve Fryatt wrote:
This could be more of an HID feature than a NetSurf one: I'll
investigate when I get the chance. As I said in my first post, scroll
wheel support on RISC OS is something of a minefield...
If I set the scroll wheel settings in HID back to the way I had it
before (Fixed distance/200 units) the menu scrolls OK, but obviously
the frames issue would still be a problem.
However, as so very few sites use frames these days (that I use,
anyway!) I can live with it.
--
Brian Howlett
---------------------------------------------------------------
Every 4 seconds a woman has a baby. Our problem is to find this
woman and stop her.
8:31 a.m.
New subject: frames [on RISC OS]
On 19 Dec, Brian Howlett wrote in message
<065e1f4452.Brian(a)bhowlett.plus.net>:
On 19 Dec, Steve Fryatt wrote:
> This could be more of an HID feature than a NetSurf one: I'll
> investigate when I get the chance. As I said in my first post, scroll
> wheel support on RISC OS is something of a minefield...
If I set the scroll wheel settings in HID back to the way I had it before
(Fixed distance/200 units) the menu scrolls OK, but obviously the frames
issue would still be a problem.
However, as so very few sites use frames these days (that I use, anyway!)
I can live with it.
Try r13300. It's still not perfect, but I'm not sure it's handling of (say)
the scroll wheel in the Themes config dialogue is any more broken than what
happens normally (with HID; still untested with Select).
--
Steve Fryatt - Leeds, England Wakefield Acorn & RISC OS Show
Saturday 28 April 2012
http://www.stevefryatt.org.uk/ http://www.wakefieldshow.org.uk/
8:59 a.m.
New subject: frames [on RISC OS]
On 19 Dec, Steve Fryatt wrote:
Try r13300. It's still not perfect, but I'm not sure it's
handling of
(say) the scroll wheel in the Themes config dialogue is any more
broken than what happens normally (with HID; still untested with
Select).
That's sorted it for me, anyway.
Quick work ;)
--
Brian Howlett
---------------------------------------------------------------
I came home from work the other day and found that someone had
stolen all my furniture, and replaced it with exact replicas...
10:34 p.m.
New subject: frames [on RISC OS]
On 18 Dec 2011, Steve Fryatt wrote:
On 3 Oct, Roger Darlington wrote in message
<02a8841c52.rogerarm(a)rogerarm.freeuk.com>:
> I have downloaded the latest test build r12932, and find that it
treats
> frames in an entirely different way to what it test builds did less than 3
> months ago.
>
> Positioning the mouse pointer over any one frame and rotating the scroll
> wheel used to scroll only the frame that it was underneath it; now it
> scrolls the whole caboodle, and reveals a previously unseen blank frame at
> the bottom. (??)
Some of these issues have now been sorted in the RISC OS front-end,
as of
r13292.
Indeed so Steve:-)
Frames can be scrolled using the scrollbars and by dragging, as has
always
be the case.
Yes, it now works properly as before (VA-RPC SA RO4.39). (NS r13437)
When using the scroll bars to scroll, it is still noticeably slower
than versions of Netsurf of 9? months previous, perhaps half as slow
(or twice as jerky). Something has happened to the update speed.
Can this ever be restored to its former responsiveness, or is it some
consequence of the way NS now operates?
In addition, they can now be scrolled using the keyboard. With
frames, however, you'll see the effect described by
Roger above: the whole site will simply scroll out of view and leave a blank
work area in its place.
No, it will now scroll without ever showing a blank work area:-))
I notice, however, that it doesn't clear the keyboard buffer when you
release the Arrow Keys, but continues scrolling for some short period
depending upon stored keys in the keyboard buffer. But it is not a big
deal.
Well done. Sterling work :-)
--
Cheers
Roger
If I knew where I was I wouldn't be here
4244
days inactive
4372
days old
netsurf-users@netsurf-browser.org
11 comments
4 participants
participants (4)
-
Brian Howlett -
Richard Porter -
Roger Darlington -
Steve Fryatt