Beiträge von Jonnie

Hi Tuur
Just some feedback for you.
It works great

I am on Linux using the Mozilla browser.

There is one small thing which may be a bug I am not sure.
The Look at numbers all work but the mouse numbers were static and did not change,
Then I tried G for the grid and the mouse numbers all started moving and seemed correct.
It seems for me I need to have G "on" for the mouse numbers to work.
I would not call this an issue, as long as I know then it works fine for me and all is good.
I just wanted to give the information so you have it.

Jonnie

Zitat von Tuur

Hi,

i think my ToolBox does exactly what you need.
https://krpano.com/forum/wbb/inde…74910#post74910

Press T to show hide.
use the 'numbers' section, put your mouse and or set your view and press P

Hope it helps,
Tuur

Hi Tuur
Thanks very much for taking the time to let me know - that is exactly what I am looking for - excellent, now it will take a little less time
I think.
Jonnie

Hi
I have had a number of years break from coding in KRPANO and so I am still trying to remember things.
Can anyone remind me of a simple hotspot tool that does not rely on droplets - I am working on LINUX so droplets will not work for me as they are based on OS specific code.

What I want to do is to be able to run a tour in my browser, click on a point in a scene and be given the numeric values for ath and atv ( horizontal and vertical coordinates) which I can then hand code myself into the XML. Alternatively it could just give me a live value for ath and atv and I could simply zoom my centre of vision in on the place where I want the hotspot to appear.

At the moment I am using educated guesses and then refining by repeated visits which is time consuming.
I do not want a tool that creates the tour or writes XML I just want the numeric values.

Hi,
I am working on Ubuntu 18.04.1 LTS after leaving the world of Windows some months ago.
After installing the current 64 bit linux version of Krpano I experienced the following with krpano tools.
The virtual tour editor tab shows a black screen with the error message "HTML5 browser with WebGL support required".
This seemed odd as this editor is documented as being there to avoid browser limitations.
I did not think this would be a graphics card issue but to be sure I ran both graphics cards on my hybrid machine ( DRI_PRIME=1 | ./'krpano Tools' ) which has always solved any graphics cards related issues in the past.

other tabs work its just the virtual tour editor that does not.

This is not a show stopper for me - returning to Krpano for a new project after a few years absence and so not as familar as I used to be,

I am exploring the word of WebVR and just wanted to use the virtual tour editor to set up a non-WebVR version of my tour in the hope that I could modify it for WebVR.
Instead I will go over to editing the code which is what I used to do when I was more familiar.

Problem solved!
Client needed to clear their cache.
I guess an older version of one of the SWF files was hanging around on their phone.

Zitat von ramirox3

Hi, Sony z3 compact, works fine, firefox, firefox beta, chrome, chrome beta
Ramiro

Many thanks Ramiro - kind regards Jon

Hi

Can anyone help offer suggestions on the following problem

I have a client for whom I made the following panorama ( of their campsite - aerial view )

http://www.raised-eye.com/production/pro…gronhoj_en.html

They say that they cannot view the panorama on Sony z3 phone and yet they can view a very similar panorama I made for another
campsite

http://www.raised-eye.com/production/project_raised/raised_en.html

The error on the Sony Z3 for the first panorama was a black screen with "ERROR: Adobe Flash Player 9/10 or higher needed" the second panorama apparently works fine.

I cannot physically go and look myself at the phone as I have relocated country since I did these panoramas and I do not have a Sony Z3.

The first thing I did was to make sure that both panoramas were using the same copy of the following files, I actually copied across from the panorama that is working. ( It is my working practice to rename krpano.js and krpano.swf to tour.js and tour.swf )

radar.swf
radar.js

tour.js
tour.swf

The version is a little dated but the important thing is that the four files mentioned above are the same in both panoramas and yet my client reports a difference in behaviour as described above.

All I could think of was to check that there were no referenced (swf) plugins that were of a different version so out of step with everything else but I could find none.

Can anyone comment on my basic line of reasoning? I am assuming that the difference in behaviour could only be caused by one of the panoramas referencing an JS/SWF plugin that is out of step with everything else? that somehow I have overlooked an JS/SWF plugin that is out of date compared with the others?

I am very sorry if this turns out to be a trivial issue, I have not done any Panorama work for a long time and I am now very rusty.

Obviously if anyone does have a Sony Z3 and can confirm that one panorama works and one does not I would be very grateful.

Thanks in advance

Jon

PS: In case anyone is interested in the method for the aerial panoramas.

Self build steel tripod supporting purchased 10metre aluminium telescopic pole ( radio ham aerial )
Self build motorised head - nothing fancy ordinary 6V motors with a very, very low gearing and joystick to change polarity so
one speed forward and one speed backward.
Nikon camera with software for remote (laptop) control.
15 metre USB cable of high quality

I want to use the combo box plugin for scene selection but not on small screen devices.

The method I have been thinking of is a device dependent import for instance

<include url=\"real_combo.xml\" devices=\"desktop\" />
<include url=\"dummy_combo.xml\" devices=\"iphone\" />

The real_combo.xml would contain code for the regular flash/js plugin.

<plugin name="cb" url="combobox.swf" alturl="combobox.js" etc etc >

The dummy_combo.xml would as the name suggests just contain a minimal non-functional, non-visible plugin provided to keep the
code tidy something like

<plugin name="cb" url="" alturl="" visible="false" enabled="false" etc etc >

The code in my main file contains statements to update the combo box currently selected item if another method ( say a hotspot ) is
used to change the scene, otherwise the combo box would get out of step.

<action name="startscene">
plugin[cb].selectIdItem( item10 ); plugin[cb].closeList();" : "" ) .
</action>

In the above example plugin[cb] is the combo box and item10 is the position of the new scene in the combo box list.

If the startscene code is to be unconditional (that would be simplest) it would be nice to know how to create tidy code so that calling the methods
selectIdItem() and closeList() on the dummy plugin were not calls to functionality that did not exist - I want the dummy plugin
defined in dummy_combo.xml to have dummy methods or actions such that calling

plugin[cb].selectIdItem( item10 )

Would result in a clean successful call to a defined method/action but the function would merely be a "do nothing" dummy empty function, in otherwords how
to trivially define a dummy version of the krpano combobox plugin and methods for device dependent purpose.

This is leading to the question whether krpano actions can be associated or owned by a particular plugin so that we have
something that looks a little bit like object orientated code ( object (plugin) owns methods (actions) ).

The "methods" ( actions ) associated with my dummy plugin would not actually do anything they are there only to keep the code tidy and not to
have untidy loose ends that trigger error warnings.

Thanks Jon

Thanks Klaus

Thats the kind of thing I was looking for.

Kind regards
Jon

Hello

I often ask people I know to try out pano tours to see what usability issues might exist.

A recent participant said everything was easy and intuitive BUT....

"6. The only thing I didn't like was the "out of focus" at the start. I
think
it would be better if it didn't move until it was in focus
properly."

and then when questioned further on the matter.,..

"Maybe if the image didn't move until it were in focus? That
wouldn't be so
distracting. At first I thought there was something wrong
with it."

I have italicised the last sentence because this is the one I fear might cause the person to abort the attempt to view the panorama, my tracking system on my campsite
panorama tour suggests a small percentage of aborts and it is hard to tell why people exit within the first ten or fifteen seconds but obviously it is a concern if the reason
is that they think something is wrong during the loading stage.

This made me think it might be nice to have two predefined plugins...

< barmessage .... >
< loopingsmessage .... >

These plugins might be designed so as to automatically appear when the loopings or loading bar were displayed and automatically disapear
when loopings or loading bar disapeared.

The user could then define messages, perhaps as graphics perhaps as text along the lines of

"please wait the panorama is loading..."

It might be nice to have a positioning option that is relative to the loopings or loading bar position.

I thought even if usability issues dont exist this might be a nice touch and enhancement, perhaps it could be coded without an enhancement
but it would be nice for krpano to handle the decision when to display and when not to?

Jon

Hi

Those images look very good to me but I think many other people achieve the same quality.

Good Camera

Good lens

Work in Tif or RAW as much as possible in the workflow - dont go to 8bit jpegs until the last stages, I use a good editor that handles 16 bit tif, experiment
with saturation and contrast and also attention to detail on things like white balance, there is a lot in common with standard photography, the panorama
photographer has a slightly tougher time because 360 implies greater variation in lighting ( as a genalisation ).

I think the photographer had a very good headwind - the BBC building is iconic ( new and old ) and is lit up very effectively, equally well
the studio shot was taken under ideal professional lighting conditions and the newsreader was made up by professional make-up artists.

If you take a look at some of the ordinary daytime shots in this collection I think you will find they are of good quality but
a quality that many other experienced people achieve.

Kind regards
Jon

Hi

I am getting interested in the idea of having several sub-tours within one tour.

Example: Campsite - a tour of camping spaces, a tour of toilets and kitchens, a tour of a chalet - so three tours but presented as one.

I was looking in the forum for code structure ideas but I dont know what people call this type of tour - a tour with subtours? I tried seaching for "sub tour" and "subtour" but did not find anything.

The kind of structure I am thiking of is

Tour separated into 3 sub tours - combo box to choose between subtours - map is different for each subtour as are hotspots and thumbnails, in the simple scenario the way to hop between tours would be to use the combo box as the thumbnails, hotspots and map would be specific to each sub tour and would in general not represent a way of getting from one sub-tour to another.

Anyone who knows of postings on this subject or knows what term I should use to search for I would be grateful.

Kind regards
Jon

Hi

I just took a tour of Klaus's excellent tool for assessing how far each browser is along the route of implementing CSS3 transforms
sufficiently for krpano to work in html5 mode.

https://krpano.com/docu/html5/ - scroll down to Demonstration.

I wondered if other users of krpano might wish to do as I just did and send various browser companies/bodies a request to firefox
where the support is there but is buggy.

I used the following form...

http://input.mozilla.org/en-US/feedback - I am not sure of other language url's

I sent them a link to the above tool and suggested they might look into the issues.

Not sure IE would be moved by a similar request since they have not even got off the ground on this one but since Firefox is close maybe
its worth sending them the odd request?

Jon

Hi Klaus,

I read Nupsi's bug report.

I think I may have a related problem.

I took your standard thumbs code and added an extra plugin - this is simply to contain a graphic that says "click here to change scene", so I modified the actions buildthumbs, closethumbs and openthumbs in a fairly straightforward way.

the openthumbs and closethumbs functions merely get an extra line to turn off or turn on the visibility of the "click here to change scene" graphic, I dont want it displayed when the thumbs are open.

I am setting the url for the "click here" plugin using a style because I am working with six languages and hence six different graphics, one for each language - the outer xml wrapper is language specific and sets style urls for language dependent graphics.

The issue is that on ipad the "click here" graphic disapears completely after I have changed scene - the thumbs close as expected but the "click here" fails to re-appear
when its visibility field is set true by closethumbs.

I then experimented and changed the code in open / closethumbs to change the alpha of my "click here" plugin to 0.0 or 1.0 instead of using the visibilty option.
The behaviour is then that the "click here" does appear after a scene change but when I rotate the ipad it can get 'stuck' - for instance if I rotate from landscape
to portrait then the thumbs all correctly drop down to the bottom right corner but the "click here" remains higher up the screen at a position that corresponds
to its pixel height from the top of the screen when it appeared in landscape mode, it has not been correctly re-located to the bottom right after a screen rotation.

I am using the following code to set the url of the "click here" plugin....

<style name="changescenetip" url="pano_skin_guldager/change_scene_en.png" /> ( this is from the outer language specific xml wrapper - English in this case ).

I am using the following to set the url of my "click here" plugin within the build thumbs code.

addplugin(changescenetip);
plugin[changescenetip].loadstyle(changescenetip);

The rest of my additional code for the "click here" plugin is derived from the original thumbs code, for instance I have set the keep field true.

The interesting thing is that if I hard code the url instead of using the dynamic load style then all problems go away.

set(plugin[changescenetip].url, 'vtour/pano_skin_guldager/change_scene_en.png' );

I read your comments about a fix in the latest beta version but it seems there were some problems reported by others and I dont want to go off in that direction
just yet, Iam very close to a lot of client work and I dont want to make unnecessay changes at this crucial time without good reason.

Naturally I can create a cut down simplified version although this would be quite a lot of work so in the first instance I would like to ask whether you believe this
could be related to the problem that Nupsie reported?

Recap:

Ipad specific - no problems on desktop or android.
Simple extra plugin added to standard "thumbs" code - just an instructional graphic.
graphic url set by dynamic loading of a style.
plugin does not re-adjust position correctly when ipad is rotated.
It is desired that this plugin is not seen when thumbs expanded - if visible field is used then the graphic does not reappear when close_thumbs sets the field
true - however if the alpha field is used instead of the visibility field then when close_thumbs sets alpha 1.0 the graphic does re-appear.
The alpha method represents an acceptable workaround except for the additional problem of incorrect positioning after ipad is rotated.

I would value even a guess as to whether this is the same problem as Nupsies - If I know its probably not an issue in my code then would prefer to think
about workarounds, I dont want to ship beta code at this late stage to my clients.

If its not related to Nupsies problem then I know its worth spending more hours closely scruitinising my code.

Thanks
Jon

Ammendment after initial posting - I misread Nupsie - he was talking of autorotation - whereas I am refering to ipad rotation ( portrait / landscape ) - sorry, probably less reason to suppose its related to Nupsies problem than I first thought.

Hello

Here is a link to an example that I made for illustration, it contains very little except what is needed to show the problem, I stripped out code that
was not required. The code was taken from the krpano example directory "simple tour" I then modified for a simple menu, removed thums and
changed the html so as to remove the overflow none css style statements which is required to show the problem.

( Issue is Ipad specific - Iphone not tested ).

http://www.buymydreamhotel.com/pano/simpletour.html

To reproduce - Ipad2 - safari - instructions are for LANDSCAPE view ( but problem may also be seen in portrait )

1. Load page and scroll page up or down until two "hellos" are visible at bottom between pano and bottom of screen.
2. Touch the fullscreen menu button.
3. If the pano appears fulscreen exactly as expected then go back to step 1 and with the pano scrolled to different positions up and down the screen, repeat until the pano appears full screen size but mis-aligned.

If the problem appears then you will probably see something like some lines of "hello" at the top of the screen and the further down the
screen the pano is wide enough for the screen but is mis-aligned ( since of course the hellos should not be visible ).

Now touch the screen in the vicinity of the "hellos" - you should be able to guide the pano into place, it will probably "Snap" into place.
Now it should look like a normal full screen except that the menu is now dead - it does work and indeed you can scroll the whole webpage up by pusing it up the screen, it may be necessary to reload the page so as to recover from full screen.

Behaviour varies for me but the above description is typical behaviour, sometimes the lower 25% of the pano becomes unresponsive sometimes less.

Kind regards
Jon

Update: I re-used code from the krpano examples directory, I just did a "copy-replace" text edit on the names for the image tile directory name in the xml. I had not noticed that the original example code was using mobile images for all devices, there is a comment in the original xml. Since I have tested that the pano does reproduce the problem I will not correct this oversight - its not relevent to the problem.

I hope the creator of this impressive giga pixel panorama will not mind me adding this link.

http://www.photoartkalmar.com/Photoart%20Kal…l/KHM.Wien.html

I quick search on the forum showed postings in German referring to the website on which this pano
is exhibited but I did not see any English language postings - I thought it was worth sharing.

Kind regards

Jon

Hi

I have been experimenting with the example webpage into which I embed my pano. Results so far indicate a high correlation between the problem and whether or not the webpage fits naturally in the ipad screen with no scrolling necessary.

My "trimmed" webpage that is carefully constructed so as not to exceed the screen height (or more accurately available area after the safari browser has taken up the top portion) of the ipad has not yet demonstrated the problem.

My "untrimmed" webpage that is longer than the available area exhibits the problem reliably.

My impression playing around leads me to think in terms of a "view port" and its equivalent tactile area - lets call it a "touch port" which should be aligned (so that visual events and touch events occur at the same screen cordinates) but have become mis-aligned, the degree or severity of the misalignment being in some way proportional to the amount by which the webpage height exceeds the available view area.

I dont know whether a "touch port" really exists within the ipad operating system but it is a useful device at a conceptual level to describe the observed behaviour.

For instance when I have a full screen case which exhibits what I would like to call "view port - vs - touch port mis-alignment" then I do find that the bottom portion of my pano which contains the menu strip is unresponsive to gestures that I intend to control the pano - for instance I cannot exit fullscreen mode using the button intended to do this since this lies at the bottom of the screen.

However this unresponsive lower portion of screen is actually responsive if I use gestures that are meaningful outside of the pano - for instance if I place my finger on the screen in this lower area and move it upwards then the correct response should be that the pano pans up.

Actually what happens is that the whole pano view port is moved up the screen - this is more appropriate behaviour for a point outside of the pano "touch port" but inside the "touch port" of the surounding webpage, ipad "thinks" I am touching a point outside of the pano when I am actually touching inside the pano.

Although at a visual level the pano occupies the entire available area and therefore we cannot "see" the surounding web page - if we think in "touch space" rather than "visual space" then the pano "touch port" does not occupy the whole screen because it is mis-aligned with the pano view port - actually the "touch port" of the surounding webpage occupies the lower part of the window and so thinkking in "touch space" - it is natural that ipad thinks I want to scroll the pano viewport up the screen rather than thinking that I want to scroll the scene within the pano viewport.

I think that is an accurate rationalisation of the observed behaviour - only someone expert in ipad will know whether my talk of "touch port" mis-alignment is an accurate description of why the problem occurs or whether its just a useful device to describe a problem but no more meaningful than that.

Kind regards
Jon

Summary Of Problem in brief steps

1. Pano embedded as a div in a webpage - pano viewport smaller than available view area.
2. Touch full screen menu button.
3. Pano expands to an area suitable for fullscreen mode but the pano viewport is misaligned with the available view area so we still see some of the surounding webpage.
4. Touching surounding webpage I can move the webpage and the pano and help the pano into position, when it is close it "snaps" into place - visually things now look right, an ordinary fullscreen pano display.
5. The "touch port" however did not "snap" into place and is still misaligned, screen gestures performed with a viewport do not necessarily lie within the "touch port" corresponding to that "view port" - control of the pano is compromised and an non-graceful exit from fullscreen is necessary ( close browser ).
6. "Touch port" is just a concept I am using to explain a problem.

Hi

This is my second post on issues with full screen on ipad, it may later prove necessary to post code or alternatively to supply screen shots but first I will attempt a written description and see if that is enough to get any leads.

I have not exhausted the possibilities that this is a bug of my own making but I have now managed to reproduce the problem with a pano I made very early on when I relied almost exclusively on the droplets supplied with krpano examples and toolkit so it seems unlikely that the problem is in the xml as I am seeing it with code from different sources.

Case 1:

1. Embed pano in html file - pano div is sized 700 x 450 pixels so smaller than the screen on either ipad or my laptop.

2. Use menu buttons supplied in defaultskin.xml altered only so that the menu will appear on ipad - this is done simply because it is
necessary to be able to transition to full screen to show the problem (remove the device="desktop" statement for each button) I simply
need a fullscreen button to demonstrate the issue.

I included seven lines duplicated lines of html so that I had a bit of content above the pano.

hello<br> ( x 7 )
<div id="pano" style="width:700px; height:450px;">

This works fine - I cannot provoke the fault, I can alternate between full screen and 700x450 repeatedly using the fullscreen menu button, when the pano is fullscreen it "clicks" nicely into place on the screen and I cannot see any of the surounding webpage.

Case 2:

Increase the amount of content that appears above the pano, in my case I had 26 lines but I suspect it will depend on font size.

hello<br> ( x 26 )

<div id="pano" style="width:700px; height:450px;">

Now the problem occurs and is consistent, if I scroll the page so that the bottom edge of the pano div is visible then I scroll several lines of "hello" off the top of the screen.

Now when I tap the button for fullscreen a different and problematic behaviour occurs.

The pano div gets bigger but does not align itself properly with the available display area ( basically the full screen size minus the safari top menu bar ) I can still see quite a lot of the surounding webpage and also see lines of "hello", I am still seeing content from outside of the pano-div because the pano-div failed to align itself properly to "click into place". The effect is pretty much like a "zoom webpage so pano appears as big as view area but offset pano so that it does not fully occupy view area.

I can use my finger to now scroll the "missed" pano div so that it "locks into place" where it should have locked to give a correct JS viewer full screen effect. There is a definite "snap" behaviour - as soon as the pano-div is close to alignment it snaps into place decisively. Once in place it seems to be stable - after all there is now no control surface available outside the pano-div so actually there is no obvious method
of moving the pano out of place once it has snapped into place.

The problem now is that an area of the bottom of the pano div is now unresponsive, the menu does not work and I cannot pan the panorama using my finger if I place my finger in the bottom portion of the pano div. I can only get a response if I use the upper portion of the div.

I am not sure but I think that the height of this unresponsive area may be related to the amount of mis-alignment when the "fullscreen div" failed to align itself properly - this is based only on subjective observation but it does seem that in cases where the mis-alignment is greatest and the pano div is too low in the screen area ( so that more webpage and more lines of "hello" still appear at the top ) then the unresponsive area at the bottom seems to be higher ( this can be tested once the pano has been moved on the screen and "snapped" into place ).

If this explanation is unclear I can try and supply pics but it will probably mean taking photos.

Any ideas?

kind regards
Jon

Hi

I have to apologise in advance if there is naivety here, I have only just purchased an Ipad2 for testing after getting tired of doing my "debug sessions" in local shops.

My preferred implementation is to embed the tour into a div in an existing webpage so that a small version of the pano is visible as soon as the webpage is loaded, for example a 600 x 450 pixel div - the contents of the webpage are visible around the div. I use an extra scene which acts as a "splash" screen - the static image being provided by a plugin - this spash screen gives a ghosted impression of the first scene of the tour and has a "press play" button. In the following screen shot I have tried to show the webpage into which the pano is embeded ( yellow menu bar, background sea image and so on ), the pano is actually showing the last scene which does not have any <image> but does have plugins to give a start button and show a static image so that the visitor gets a taste of the subject matter of the tour.

If the user does press play they will not see a full screen pano - the pano will stay the same size unless they choose to go fullscreen using the menu.

Some implementers seem to prefer a link that takes the viewer to a full screen pano so they do not deal with panos occupying less than 100% of the screen and may not have encountered the same problems.

My debug sessions down the local shops had shown some anomalous behaviour, for instance when I hit the "fullscreen button" the pano increased in size but did not "click" into place on the screen so the effect was as if I had zoomed in on the whole webpage and the pano was partially visible and displaced so that although it might the right size to fill the screen*** it was not aligned with the screen boundaries
but floating around to the left, right, up and/or down directions (***ok I know the browser takes the very top in the JS viewer but it will get too complicated if I keep on adding that qualification).

Sometimes the lower half of the pano becomes unresponsive and I can only get items in the upper part of the pano to respond - the pano's own menu goes dead and does not respond. I am not sure if I have any bugs in my code but playing around with the ipad I can see other problems and I wanted to get comment on these first as I have a hunch that exploring the simplest issue may shed light on these other issues.

Start Of 'Simple Issue' Description

On ipad I look at a pano embedded on a webpage like above.

I use finger movements to move around and the two finger gesture to zoom in.

The immediate problem is that I can zoom in over the pano sufficiently that I can no longer see any of the surounding web page and indeed so that I cannot see the pano menu.

I can put myself into a position where all finger gestures are interpretted as control movements in the context of the pano, a single finger gesture is interpretted as a "move pano" movement and a two finger gesture means I zoom the pano and not the webpage - this is of course perfectly sensible and as expected except that I now appear to have cut myself off from any interaction with the surounding webpage - I cannot move the webpage around so that I can get control back on the webpage "surface" and I cannot even pan the pano div around the screen - all "pan" gestures are interpretted as "pan the scene inside the pano" - AND NOT - "pan the pano div around the screen".

As mentioned earlier I can zoom and pan myself to a position where I cannot even see the pano menu.

I appear to have no method of escape other than hitting the browsers "back" button or similar so that I visit the page I was on BEFORE I looked at the page containing the pano ( and doubtless other non graceful ways of getting out ).

The examples I see on the krpano website - for instance the Kuchelerhaus Tour http://www.krpano/iphone do not seem to suffer this problem because the visitor never sees the pano embedded as a small div in the webpage - they simply see a text link and then the fullscreen pano - hitting the "back" button on the browser then seems an intuitive and graceful thing to do to "escape". However this is not the case in my implementation, it no longer seems intuitive or graceful.

To reproduce this problem all I have to do is to use the two finger gesture over the area of the surounding webpage to zoom my pano div until it is bigger than the screen but does not yet fully cover it. I then use a single pan gesture to centre the overlarge pano on the screen and now I have no "control surface" left which is not "pano control surface" - all subsequent gestures are interpretted as pano gestures and not "get me out of here" gestures - my pano menu is no use because I zoomed it out of the screen area.

Does anyone have experience of dealing with this kind of issue on ipad? I feel I should deal with this issue first before going on to try and understand the other anamolous behaviour I have been seeing. I am doing some work for paying customers so I have to try and stop the user getting themselves stuck.

Kind regards

Jon.

Hi Klaus

I used your code and it worked just fine - thanks.

The only thing I added in addition was to add the line

if(stagewidth GT stageheight, landscape(), portrait());

into the krpano startup action.

I figured it was simply easier to add this line rather than to risk the possibility that devices can start up in either
portrait or landscape without the resize event handler being called.

Thanks again

Kind regards

Jonnie