SmartAxiom / openhab

Automatically exported from code.google.com/p/openhab
0 stars 0 forks source link

Flexible HTML5-based UI for tablets #24

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago
The current WebApp.Net UI of openHAB has a row-based approach which works fine 
for smartphones, but does not use the screen of tablets efficiently.

The bigger tablet screens require a more flexible UI framework and 
configuration possibilities. SenchaTouch might be a good candidate to use here.

It should be tried to make use of the openHAB sitemap metamodel here as well, 
which possibly will have to be extended by some new properties and widget 
types. Ideally, currently existing sitemaps should render in a nice way on the 
new UI as well, so they can be used for both UIs at the same time.

Original issue reported on code.google.com by kai.openhab on 29 Mar 2011 at 7:55

GoogleCodeExporter commented 8 years ago
Yes, this seems nice.
This is what I think it should have as additional properties:

Sitemap object:
 - screen width
 - screen height
 - screen orientation
 - background image
 - background color
--------------------------

Widget object:
 - width
 - height
 - X position
 - Y position
 - border thickness
 - border color
 - theme color (set a color theme for widget's elements, for example the slider knob)
 - state images (set own images for ON and OFF states or other possible states, depending on the widget type)

In fact you could map many CSS attributes as openHAB sitemap meta properties.
--------------------------

New widget ideas:
 - Chart widget (it's coming to Sechna from it's big brother ExtJS 4.0)
 - Camera widget (simply stream JPEGs or maybe WebM video?)
--------------------------

There have to be a way for showing system messages, maybe in a popup window. A 
system message may be that a command can't be executed because the device is 
offline or doesn't respond.

The UI framework have to be as flexible as possible.

Another option to consider as good JS mobile framework is also jQuery Mobile.

Original comment by mishoboss on 29 Mar 2011 at 9:35

GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago
jQuery Mobile would have the advantage to be under the MIT license, not only 
GPL. This makes it easier to package it with the EPL licensed Equinox framework.

Original comment by kai.openhab on 29 Mar 2011 at 3:38

GoogleCodeExporter commented 8 years ago
The communication between openHAB and this UI should be done through a REST 
interface, see issue 25.

Original comment by kai.openhab on 29 Mar 2011 at 3:39

GoogleCodeExporter commented 8 years ago
Another plus for jQuery Mobile is that it is compatible with more mobile 
browsers. Sencha is best experienced on WebKit browsers (iOS, Android).

Original comment by mishoboss on 29 Mar 2011 at 6:35

GoogleCodeExporter commented 8 years ago
Hi, guys. I did some research for best technologies for building a tablet UI 
and/or desktop UI. jQuery Mobile is meant only for smartphones with small 
screens. It copies the iPhone user experience, i.e. nothing more than the 
current openHAB web UI. So, I don't think it is going to be useful if we want a 
decent tablet/desktop UI.

This is how I see it:

1. Using some good JavaScript framework with a lot of ready to use widgets 
(sliders, buttons, etc.), for example jQuery is perfect I think (ExtJS 4.0 is a 
perfect option too, but it is heavier and it's meant for building much more 
complex UI than we need).

2. Using HTML5 and CSS3. The default mobile browsers for iOS, Android OS and 
Windows Phone 7 have pretty good HTML5 support. However we don't care about the 
HTML5 support, because we are going to compile the webapp to native app (read 
about this later).

3. User defines his UI widget by widget in the config file. Then here comes two 
different approaches: 
- 3.1. Building the UI on the client device. The client requests a screen by 
REST and the server returns it in XML or even better - JSON. Then the 
JavaScript builds the UI on the client device.
- 3.2. Building the UI on the server. The client requests a screen by REST and 
the server builds it in ready to use HTML and returns it to the client.

4. Compile the webapp as native app for almost all mobile platforms using 
PhoneGap (http://www.phonegap.com/). It is a great open source mobile framework 
and compiler that compiles to the 6 most used mobile platforms. And now even 
Adobe added support for PhoneGap in Dreamweaver v5.5.

5. Using WebM for camera streaming. WebM is a new open video standard announced 
by Google and used in HTML5.  Mozilla Firefox 4, Opera 10.6, and Google Chrome 
already support it natively and there's a plugin for Internet Explorer. Also 
almost all mobile browsers support it. VLC seems to also support WebM, but I 
have to check if it can stream it or only play it.

6. Using Highcharts (http://www.highcharts.com) for charts. It is a really nice 
jQuery plugin. If using ExtJS 4.0, there's a built-in charts engine there.

I would like to help building a really nice, really up to date and really 
flexible web UI for tablets and desktop (and everything with big screen, touch 
or not). My brother is a graphic designer and he could help too for icons and 
graphics.

Original comment by mishoboss on 4 May 2011 at 8:58

GoogleCodeExporter commented 8 years ago
Hi, thanks for sharing your thoughts :-) 
Let me try to comment them as a feedback:

Ok, so jQuery Mobile is no option then, agreed.

1. So you wouldn't go for any existing "touch framework" like Sencha Touch 
either? This sounds to me like a lot of work if you want to implement all such 
features yourself; I wouldn't dare trying this; but if you (and your brother 
;-)) are really motivated to do so, why not?

2. Not sure why you think HTML5 features are not of interest if we would use 
PhoneGap; could you elaborate on this?

3. I would go for 3.1. The REST interface could serve the sitemap as JSON and 
thus stay UI technology independent as this could be used by other UI 
implementations as well.

4. A cool option, but imho it should stay an option as it blows up the release 
cycle (getting the apps to the app stores etc.); the UI should also work nicely 
as a pure web application without being compiled to a native app.

5. Wouldn't this require the webcam to support WebM? Are there already any such 
webcams on the market? In general, I think we should simply use the HTML5 video 
element for webcams and leave the format to the webcam (h.264 should work well, 
too).

6. Highcharts looks pretty nice, would love to see my timeseries in it; but 
first, we will have to tackle feature 21, I'm afraid...

Original comment by kai.openhab on 4 May 2011 at 12:00

GoogleCodeExporter commented 8 years ago
Hi, Kai

1. Yes, every touch framework I see does pretty much the same thing - mimic the 
iPhone user experience. I think this is limiting for big screen apps. No, I 
don't mean to do everything by myself. There are great JS frameworks like 
jQuery that provides ready to use and highly configurable widgets. In fact we 
need only slider, button and switch (toggle) widgets, I think.

2. Maybe you misunderstood me. I mean if the webapp is compiled to native app 
we don't care about the HTML5 browser support.

3. Yes, I would too. It's a lot of work for a "thin client", but our days the 
thin clients have dual core CPU's and at least 512MB memory :)

4. Agreed!

5. In fact I don't have any experience with WebM and the video tag. I have to 
go deeper in that.

6. How is going the progress on feature 21? This feature is one of the things I 
abandoned OpenRemote ;)

Original comment by mishoboss on 4 May 2011 at 12:26

GoogleCodeExporter commented 8 years ago
1. Well, Sencha Touch does not stick to small iPhone screens - it is very well 
suited for tablets (see for example http://touchstyle.mobi/). And in contrast 
to a "pure" JS framework it brings many useful things for touch interfaces like 
toolbars, headers, menus, overlays, carousels etc. See the kitchensink for a 
feature overview: http://dev.sencha.com/deploy/touch/examples/kitchensink/

2. Yes, I indeed misunderstood you :-)

6. Not yet started; we actually had somebody very interested in contributing 
(to) this feature, timeline was to finish it during this year (hopefully 
earlier); but as I am also very much missing this feature, we should see to get 
it started soon.

Original comment by kai.openhab on 4 May 2011 at 12:52

GoogleCodeExporter commented 8 years ago
Sencha Touch works only with WebKit browsers. Like they say "Sencha Touch 
allows you to develop mobile web apps that look and feel native on iPhone, 
Android, and BlackBerry touch devices". 

iPhone, Android and BlackBerry's native browsers are WebKit based. However 
Windows Phone (IE), Symbian, Bada, etc. and every 3-rd party browser that is 
not WebKit (Opera, IE, Firefox) will not render the UI correctly.

So, I think Sencha is not the best option, though it is a really great 
framework.

Original comment by mishoboss on 4 May 2011 at 1:32

GoogleCodeExporter commented 8 years ago
I haven't seen any Windows Phone, Symbian, Bada tablets on the market so far... 
And for mobile phone use, we still have the current UI, which supports these 
platforms.

The focus should imho be iPad and Android-tablets, so I wouldn't rule out the 
use of Sencha Touch. Remember that if you try to do similar things as Sencha, 
you might have to restrict yourself to some rendering engine such as WebKit 
yourself as your JS code might otherwise become overly complex (or the UI very 
constraint as it has to support the different engines).

For users using a PC, it should be a fair assumption that they are also able to 
use a WebKit based browser (noting that the title of this issue is "for 
tablets", so PC-non-touch-browsers are not in the focus of this issue).

Original comment by kai.openhab on 4 May 2011 at 1:45

GoogleCodeExporter commented 8 years ago
Yes, I know this issue is only for tablet UI. However I have a little bit 
different understanding about the client apps. Instead of providing one web app 
for phones, another for tablets and another for PCs, I would go with just ONE 
web app that is capable of showing highly configurable UIs on every screen size.

I think it is going to be a hell to maintain several different apps, written in 
different ways. Also the user expirience is going to be the same on a phone, a 
tablet or PC. I think it is doable and it is the better approach.

Imagine that you can build your interface by defining widgets and their 
properties like X, Y, Width, Height, StateImageOFF, StateImageON, StateImage4 
(or any possibe state), BachgroundColor, Border, BorderColor, Label, etc. With 
these "tools" anyone can build any interface for any screen size. There just 
has to be one web app that can visualize it on every decent browser.

Of cource this is just my 2 cents :) 

Original comment by mishoboss on 4 May 2011 at 8:52

GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago

P.S. Besides that, what about HP TouchPad (WebOS), RIM PlayBook and the many 
Windows 7 tablets (and who knows what is coming next two years). Going with 
WebKit is a bad idea, in fact going with anything that isn't a standard is a 
bad idea. I like Sencha too, but... :( 

Original comment by mishoboss on 4 May 2011 at 9:35

GoogleCodeExporter commented 8 years ago
HP TouchPad: WebKit
RIM Playbook: WebKit
Windows 7 tablets: Just install a WebKit browser
Have a look at 
http://www.tested.com/news/hp-touchpad-vs-apple-ipad-vs-motorola-xoom-vs-blackbe
rry-playbook/1801/
So WebKit actually IS a de-facto-standard on tablet devices and I guess that 
any newcomers in the market will have to deal with that fact.

Original comment by kai.openhab on 4 May 2011 at 9:45

GoogleCodeExporter commented 8 years ago
I don't agree with the one-app-fits-all approach. I think Apple proved with the 
iOS devices that handhelds (and other touch devices) need totally different UI 
approaches - Windows mobile tried for years having a PC look&feel on a phone, 
but completely failed with it (luckily). And that you need different UIs for 
different screen sizes can be seen by the fact that iPads and Android tablets 
require dedicated apps (the apps designed for the phone screen are simply not 
adequate).
So you have to deal with three different UIs after all; if you do not restrict 
yourself to do it all-in-one, you gain flexibility and can really adapt to 
specific needs of the specific UI.
And don't get me wrong: It would of course be nice to use the same UI framework 
for phones and PCs as well; but the focus should be tablets and I don't see a 
problem in concentrating the efforts on the WebKit engine (see my previous 
comment).

Original comment by kai.openhab on 4 May 2011 at 9:59

GoogleCodeExporter commented 8 years ago
Discussions like this are something positive :) However we speak about little 
bit different things. Windows Mobile dictated the user interface - small 
buttons, small tree menus and so on. My idea is to let the user decide the look 
and feel of his interface. Let him build whatever he wants - define screen size 
and define widgets size, position and colors. This way if his target device is 
a smartphone with 320x240 px resolution, he could build a nice UI that best 
fits this size. Then he could build another UI for his 1024x768 px Windows 7 
tablet in the living room with bigger and/or more widgets on the screen. You 
don't dictate anything, it's all up to the user and his design skills.

All you have to provide is some decent toolset of properties like X, Y, Width, 
Height, Color, Label and so on. Then the web app that reads these properties 
and builds the UI is just ONE, not three. And you could have absolutely 
different UIs for the phone and for the tablet.

Original comment by mishoboss on 5 May 2011 at 8:14

GoogleCodeExporter commented 8 years ago
i think that approach is to low-level.

If we want to reach a broad set of users we _should_ restrict the possibilities 
of UI design or in other words we _should_ provide simple templates to easily 
create new interfaces. Even me (somehow experienced) are not interested in 
building every new page of my visualization from scratch (position the button, 
size, style, tweak it) and after all the UI looks ugly again because my 
experience is not in UI design.

I think using a framework from a _very_ experienced UI crew should be the first 
approach. After that somebody could develop another UI so meet your 
requirenments. By the way: do you know commandfusion? It is not HTML5 based, 
but it looks very much like what you suggested.

No

Original comment by teichsta on 5 May 2011 at 8:34

GoogleCodeExporter commented 8 years ago
Hi, teichsta. Yes, maybe it is not very easy to build a pretty and useful 
interface, but it worth the time spent. It is extremely important for a 
domotics system to have highly customizeble interface engine. Once you get that 
done, you can always provide some predefined templates or create simple wizard 
with NEXT, NEXT, FINISH that generates some basic UI, that could be extended by 
modifying the generated code later.

It is not so difficult to be done, but I think it is essential for every 
domotics project.

Original comment by mishoboss on 5 May 2011 at 9:14

GoogleCodeExporter commented 8 years ago
I do not see why your requirements are specific to domotics; to me it simply 
describes a very flexible and powerful UI framework; so yes, for a framework 
user, all what you say makes sense. But I agree with teichsta that for an 
openHAB user, this is far too much complexity and it should instead work with 
little efforts from the user (just as the current web UI does). So for openHAB 
use, you would definitely have to design "restrictive" pre-defined templates.

Having said that, I think if you really succeed in building such a flexible 
framework, it will be useful for much more than merely serving as an openHAB 
UI. I'd therefore suggest that you develop it independently of openHAB as a 
separate project and we then do the linkage through the predefined templates.

This would allow openHAB to concentrate on its core values: Doing integration 
of different systems (and offering flexible interfaces to connect different UI 
frameworks on top) and providing automation functionality.

Original comment by kai.openhab on 5 May 2011 at 9:44

GoogleCodeExporter commented 8 years ago
Well, now you have to design your openHAB interface too. It just has very 
little possibilities to make sophisticated design, because it is meant for 
phones, not tablets.

How you imagine the tablet interface? What is going to be configurable?

Original comment by mishoboss on 5 May 2011 at 11:01

GoogleCodeExporter commented 8 years ago
Here's a concept for a tablet interface: 
http://knx-user-forum.de/attachments/knx-eib-forum/8490d1269090390-misterhouse-i
pad-native-application-eibpad-concept.pdf

This could imho be a good starting point and it is close to the look&feel of 
other tablet apps.

Original comment by kai.openhab on 5 May 2011 at 4:45

GoogleCodeExporter commented 8 years ago
Well, if you ask me, I would go with the OpenRemote/Elve approach. I like to 
have full control over how things look like. I like to choose colors, button 
images and so on. I want to make it entirely mine.

However this concept looks nice too. They talk about grid size and widget's 
position (and I hope widget size too, at least for the camera and the chart 
widgets). It seems to give good amount of customization options.

Take a look at this interface just for inspiration, though it's not a perfect - 
http://www.mavromatic.com/images/grouse_interface_preview_2.JPG

Original comment by mishoboss on 5 May 2011 at 5:32

GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago
You may check Dojo Mobile too when considering the right mobile framework for 
the tablet UI. In my opinion it's no worse than Sencha and jQuery Mobile. 
Unlike Sencha, Dojo Mobile is truly multi-platform and uses only strict HTML5 
and CSS3 (not WebKit specific functions). I tested it with my iPhone 1-st gen 
with iOS 3.0 and I must say the animations and transitions run pretty sleek (I 
can't say this for Sencha or jQuery Mobile). It runs almost perfect on a PC on 
Opera 11 and Firefox 4 too.

In the upcoming Dojo 1.7 there will be charts out of the box. In Sencha they 
will be paid as far as I know.

Just my 2 cents :)

Original comment by mishoboss on 21 Jun 2011 at 4:04

GoogleCodeExporter commented 8 years ago
Another option is Adobe Flex 4.5.1 that came out 6 days ago. It builds native 
cross-platform applications for iOS devices, Android OS and Blackberry PlayBook 
OS (and I bet the next step is Windows Phone 7 support). Seems like the 
performance is way better than HTML5. Take a look at this video for a real-life 
example - 
http://tv.adobe.com/watch/adc-presents/financial-trader-app-for-ios-android-and-
blackberry-playbook/

Original comment by mishoboss on 27 Jun 2011 at 11:24

GoogleCodeExporter commented 8 years ago
Using Dojo Mobile is a great idea. I've used the Dojo Toolkit for some 
Enterprise projects. Moreover openHab is a fantastic project. I'm using it for 
a couple of weeks now and I'm very intrested to bring this project a step 
forward. If you like I can start the developing of a Dojo Mobile based UI. Also 
I can start working on a REST Interface as communication Layer and on a 
persistance storeage to provide historical data e.g. for charts.

You can find some amazing examples here:
http://chrism.dojotoolkit.org/mobile-0.2/make_samples/dojo-samples/demos/mobile-
gallery/demo.html

Original comment by oliver.m...@gmail.com on 29 Jun 2011 at 12:55

GoogleCodeExporter commented 8 years ago
Hi Oliver,

welcome to openHAB and glad to see that you are going to contribute some very 
promising things!

AFAIK Kai is already working on the REST-issue (see 
http://code.google.com/p/openhab/issues/detail?id=25), at least there are some 
thoughts on how to develop such an interface. Especially the URL-Design would 
probably be a time-consuming task because neither kai nor me are experts in 
REST. Are you probably?

Dojo seems to be an interesting framework but keep in mind, the most 
interesting device to serve currently would be a tablet (especially an iPad). 
The working iPhone UI works pretty well and we should not spend resources on 
developing the second mobile device UI, shouldn't we?

The persistence storage would be interesting as well! If you don't mind, just 
open a new issue and we can discuss all details there!

Looking forward to working with you on the coolest opensource project since 
sliced bread ;-)

Cheers,

Thomas E.-E.

P.S.: kai is on vacation currently so he won't be able to comment things in 
more detail.

Original comment by teichsta on 29 Jun 2011 at 2:28

GoogleCodeExporter commented 8 years ago
Hi teichsta,
I would recommend a pure HTML / JS WebUI with no jsp parts, so the UI is 
completly separated from the server side. On the client side there should be a 
page controller which decides how to render the page and which widgets or 
portlets to use e.g. different layout for tablets, mobiles and desktops. All 
this is possible with dojo and moreover it is PhoneGap compatible, so one can 
build native apps for a store. 

The REST part could be tricky. I've some experiance with Jersey but I never 
used it in an OSGI environment. Since Version 1.2 Jersey added basic OSGI 
support. Jersy has some really nice features like Declarative Hyperlinking 
(http://jersey.java.net/nonav/documentation/latest/linking.html). I don't know 
which REST Framework Kai will use. In Issue 25 Restlet was mentioned, bu far as 
I know Restlet is missing some features wich are required for a REST based 
WebUI e.g. websocket / comet support. In the latest milstone they added a NIO 
extension, which can be used to implement these features. On the other side one 
can use Jetty for websocket / Comet support.
There is a great publish / subscribe framework called Project Atmosphere 
(http://atmosphere.java.net/) it offers AjaxPush/Comet and websocket support 
and it is very easy to use. For the REST part it uses Jersey. 
At the weekend I will have a deeper look into the openHab architecture and will 
think about possible integration solutions.
Maybe you can point me to some of the things you've allready thought about.

Regards
Oliver

links to 

Original comment by oliver.m...@gmail.com on 1 Jul 2011 at 10:18

GoogleCodeExporter commented 8 years ago
Hi Oliver,

thanks for your input, but let's discuss the REST details in the corresponding 
issue they are somewhat off topic here :-).

Regards,

Thomas

Original comment by teichsta on 2 Jul 2011 at 5:31

GoogleCodeExporter commented 8 years ago
Hi Oliver,

> I don't know which REST Framework Kai will use

Neither do I! Thanks a lot for your opinion on the REST frameworks, this is 
indeed very helpful.
As Thomas suggested, I will continue the discussion on the REST issue 
(http://code.google.com/p/openhab/issues/detail?id=25).

Regards,
Kai

Original comment by kai.openhab on 11 Jul 2011 at 8:31

GoogleCodeExporter commented 8 years ago
> I would recommend a pure HTML / JS WebUI with no jsp parts, so the UI is 
completly
> separated from the server side. On the client side there should be a page 
controller 
> which decides how to render the page and which widgets or portlets to use 
e.g. 
> different layout for tablets, mobiles and desktops. All this is possible with 
dojo 
> and moreover it is PhoneGap compatible, so one can build native apps for a 
store.

This is exactly what I've suggested earlier in this topic :)

> The working iPhone UI works pretty well and we should not spend resources on 
developing the second mobile device UI, shouldn't we?

If it was up to me I would go with just ONE app for everything - touchscreen 
phones, tablets, PCs, etc. Or at least the app for phones and tablets could be 
the same. Every JS framework I see provides tools to deal with different screen 
resolutions and orientations.

Also the current Mobile UI isn't perfect, mainly due to the used framework. See 
my issue here for more info - 
http://code.google.com/p/openhab/issues/detail?id=32
This issue is going to be solved with the REST API you talk about.

And another plus for ONE app is that you maintain only ONE app ;)

Original comment by mishoboss on 12 Jul 2011 at 8:43

GoogleCodeExporter commented 8 years ago
I have created a static prototype with Sencha Touch - please find attached a 
few screenshots. I will soon also provide a url, where this can be tested 
directly on an iPad :-)
Feedback welcome!

Original comment by xthirtyn...@gmail.com on 16 Sep 2011 at 10:40

Attachments:

GoogleCodeExporter commented 8 years ago
[deleted comment]
GoogleCodeExporter commented 8 years ago
The prototype is now available at http://goo.gl/GpZ4H !
So go and use your webkit-based browser or better directly your tablet to try 
it out.

Please note that this is a static prototype, so nothing happens if you operate 
any switches or sliders (not even the icon updates to the new status). 
Furthermore there is a rendering bug on the "Rooms" tab, you need to change 
your tablet orientation once to make it render correctly.

This prototype is just meant to give a first impression of how an openHAB 
tablet UI could look like based on the Sencha Touch framework. Nothing is 
carved in stone yet and the exact features and UI guidelines (vertical 
scrolling as on the "Rooms" tab or rather horizontal scrolling as on the 
"Functions" tab?) still have to be worked out.

Nevertheless, I think it is a good starting point and I welcome your feedback 
and ideas!
After working on this prototype, I have the impression that Sencha Touch is a 
good choice for tablet UIs (while it is not well suited if you want to design a 
desktop UI), so I think I go further this road.

Kai

Original comment by kai.openhab on 17 Sep 2011 at 10:18

GoogleCodeExporter commented 8 years ago
Hi Kai, nice work¡ 

We have played with your prototype in chrome, and it looks fine. But i can't 
see it in my iPhone (although i can see sencha touch demos, for example).

Any idea?

Original comment by dmora...@gpm.es on 19 Sep 2011 at 8:58

GoogleCodeExporter commented 8 years ago
This UI is really only designed for tablets - I haven't tried it on the iPhone 
at all and would need to do many changes in the code to support it (as it would 
require a very different screen and menu layout).

Regards,
Kai

Original comment by kai.openhab on 19 Sep 2011 at 9:24

GoogleCodeExporter commented 8 years ago
Thanks for the info, Kai.

It looks truly awesome, specially the alerts and graphs (talking about the 
persistence module) as they are new components to the UI.

But, do you think sencha is a good choice?, do you think it will be easy to use 
the rest api with this UI? 

Regards

Original comment by dmora...@gpm.es on 19 Sep 2011 at 9:51

GoogleCodeExporter commented 8 years ago
Yes, Sencha is really made for getting information via AJAX/REST, so I think it 
is well suited for this. Nonetheless, it will mean quite some effort to adapt 
the current static prototype to a fully working UI that is fed through REST...

Original comment by kai.openhab on 19 Sep 2011 at 11:21

GoogleCodeExporter commented 8 years ago
Sencha is the best choice up to date. I follow it pretty close, as well as Dojo 
Mobile and jQuery Mobile (not suitable at all). I think to build Sencha 
interface for openHAB, but I wait for Sencha 2.0 as there are a lot of API 
changes there.

> This UI is really only designed for tablets - I haven't tried it on the 
iPhone at all and would need to do many changes in the code to support it (as 
it would require a very different screen and menu layout).

Sencha provides really flexible layouts that could visualize the content 
differently regarding the screen size. I really think that this interface MUST 
be suitable for wide range of touchscreen devices, not only for tablets. And 
where ends the phones and starts the tablets, when you have the Samsung Nexus 
Prime with 4.65" 1280x720 pixels display (a phone) and the Archos 43 with 4.3" 
480x854 pixels display (a tablet)? :-)

Original comment by mishoboss on 31 Oct 2011 at 12:18

GoogleCodeExporter commented 8 years ago
looks great on chrome ! I'm already using openhab and old webUI of it, but will 
have a motorolla Xoom in 2 days and want to try it there with new UI. Is there 
a possibility to have it for testing ? I can help with some development too. 

Original comment by ramazany...@gmail.com on 9 Dec 2011 at 12:37

GoogleCodeExporter commented 8 years ago
For testing, you can simply point your tablet browser as well to 
http://goo.gl/GpZ4H - it should be immediately available.
I had myself the chance to test the prototype on a Galaxy Tab 10.1 this week - 
unfortunately, the result was rather disappointing. While it works quite smooth 
on the iPad2 (with the exception of initial loading and orientation changes), 
it was dreadfully slow on the Galaxy tab - and when scrolling, the buttons 
moved at a different speed than the page background :-(
So not really convincing at all, which made me think that I should reconsider 
the choice of the framework again. Unless Sencha comes with updates that solve 
these performance issues.
Another interesting framework would be the Vaadin TouchKit, see 
https://vaadin.com/add-ons/touchkit. But I doubt that I will find the time to 
do another prototype at the moment...

Original comment by kai.openhab on 9 Dec 2011 at 12:52

GoogleCodeExporter commented 8 years ago
do you have your prototype sources in the mercurial? 

Original comment by ramazany...@gmail.com on 9 Dec 2011 at 6:52

GoogleCodeExporter commented 8 years ago
regarding performance I think it is better to use Sencha 2.0: "The second big 
improvement we’ve made in Sencha Touch 2.0 is Android performance, 
particularly when it comes to scrolling and animation"

Original comment by dina.gal...@gmail.com on 9 Dec 2011 at 8:43

GoogleCodeExporter commented 8 years ago
Would be nice if performance would improve on Android. Really lags with Dolphin 
Browser, just a bit better with the stock browser

Original comment by honkton...@gmail.com on 9 Dec 2011 at 8:49

GoogleCodeExporter commented 8 years ago
> do you have your prototype sources in the mercurial? 
Yes, in the wiki repo: 
http://code.google.com/p/openhab/source/browse/prototype/tablet/?repo=wiki

> regarding performance I think it is better to use Sencha 2.0
Sounds promising, I will have a look how easy it is to migrate to version 2.0...

Original comment by kai.openhab on 9 Dec 2011 at 10:30

GoogleCodeExporter commented 8 years ago
I work on a UI based on Sencha 2.0. It runs relatively smooth on iPhone 1-st 
gen with iOS 3.0 and great on Samsung Galaxy S with Android 2.3.4. I have a 
little more work to do (I hope to be done in a day or two) and I will post here 
my work for testing with different devices. Don't expect ready to use 
application, it's just a proof of concept.

Original comment by mishoboss on 16 Dec 2011 at 8:10

GoogleCodeExporter commented 8 years ago
Note: This Sencha Touch UI has been released with openHAB 0.9.0 as a preview, 
see http://demo.openhab.org:8080/touch/ and 
http://code.google.com/p/openhab/source/browse/#hg%2Fdistribution%2Fopenhabhome%
2Fwebapps%2Ftouch.

Original comment by kai.openhab on 2 Feb 2012 at 9:30

GoogleCodeExporter commented 8 years ago
Wanted to try it out on my iPad2 but after asking for sitemap it stays in 
"loading..." state forever.
Should the demo work with openHAB 0.9.1?

Original comment by h...@harleyman.lu on 27 Apr 2012 at 5:01