Android RIVA Client: taRiva

[Fonceur]

The Tallus Android Riva application provides two way control over Charmed Quark Systems (CQC) home automation system via an Android phone or tablet (OS >= 1.6).

Information can be found on the wiki (http://talluscorp.com/mediawiki/index.php?title=TARiva_Application). The application can be downloaded here (http://dl.dropbox.com/u/408295/TARiva.apk).

Ideally, you would want a template matching the resolution of your device, but orientation change and swiping to see the rest of the template is supported.

http://talluscorp.com/mediawiki/images/5/50/TaRIVA_Template.jpg http://talluscorp.com/mediawiki/images/8/8f/TaRIVA_Background.jpg

(original post)

I fixed Broconne's code to send a proper logon request (unsigned byte issue and mixed types), which probably means that most of the steps involving talking TO the RIVA server are fine... Now I need to take care of the remaining 99% that involves reading FROM the RIVA server and doing something useful with that incoming data... ;)

[Dean Roddey]

Start off with a black background template and a piece of static text on it. Implement support for drawing of stuff incrementally as required. As long as you implement all of the push/pop stuff, you can easily just ignore any drawing commands you don't yet understand. They just won't display anything. So you don't have to do a go for broke type thing and try to implement it all in one fell swoop. So the area fill, push/pop stuff, and text drawing commands should get you enought to draw that simple type of interface, and prove it's basically working. Oh, and of course you will have to implement the sequence ack message or the server will stop and wait for you.

Just be careful to follow the asynchronous/synchronous processing guidelines in the protocol doc.

[Fonceur]

Broconne did a lot of the leg work already (he said about 80%), but he didn't have a riva license to actually test it, so there's still a lot of "connect the dots" to do... ;) Would probably have been faster if I had started from scratch, but it wouldn't have been as pretty internally... ;)

[Fonceur]

- I can send a login request.
- I get a login challenge back.
- I can create the MD5 password hash.

But as far as decrypting the challenge data and key, that's failing, as I'm getting 16 x 0 for each. So I guess I will need to print out the values from a challenge with the .NET RIVA client and figure out how to get the same result from the GNU Crypto Blowfish cipher... Probably another annoyance related to not having unsigned byte/short/integer in java. ;)

[Dean Roddey]

I wouldn't think so. The algorithm is very strict, so their implementation would have to deal with that or it wouldn't be of any use at all since it could never interoperate with any other working system. As long as you are passing them in a raw byte array or byte stream with the bytes, it should work out correctly.

[Fonceur]

OK, I guess I'm using the GNU Crypto implementation wrong, as I can't even encrypt/decrypt some data and get the original back... ;) I wish they hadn't disabled Blowfish in the default cipher of Android... ;)

[Fonceur]

I found BlowfishJ 2.15, which works on the Android, so now I can successfully log on... ;)

[Fonceur]

Currently, I can read the 59 messages associated with a logon and the template on the RIVA account that Batwater setup for testing...

So now it's time to figure out how to actually display the UI on Android... ;)

[Dean Roddey]

A mere implementation detail :-)

[Fonceur]

Would people be building new templates matching the size of there device (say 320x480) and only using the portrait (or landscape) orientation, or should I just go generic and expect people to reuse their 1024x768 (or whatever) templates and have some scrolling/swiping going on?

[potts.mike]

I dont use android but from my experiences with the iphone and ipad it would be best to allow scrolling of some sort.

on a side note did people see this:
http://www.engadget.com/2010/07/20/asus-eee-pad-ep101tc-opts-for-android-dumps-windows-embedded-co/
Could make for a good cqc device.

[personalt]

I would think that you would build the template to the size of the screen and use the swipe to swap templates. similar in how in the android OS you swipe to move between home screens.

[joeberg]

Personally, i plan to build new templates to fit the screen, so scrolling wouldn't be required. That doesn't mean scrolling wouldn't be beneficial to others though.

Joey

[sic0048]

I'm not sure which would be faster. Loading a larger template - something like 800x1920 and scrolling with a swipe, or loading a single smaller template at 800x480 and having 3 other templates that would load with a swipe.

I don't think I would want to scroll in two directions (up/down & left/right). I would probably stick with either up/down or left/right. But I think scrolling should be an option.

[Fonceur]

OK, so I'll get some scrolling/swiping going on. It seemed simple enough with a background image, I would imagine it will work similarly for some more complex canvas.

[znelbok]

I match the size to the device but I also create a page twice as wide and swip when needed

Mick

[Fonceur]

Getting there, the current version can display a simple template, with a background color and some text... I don't know if the font size will work as is or if I need to use some device independent format. I haven't seen how to select a font type, but it must be in there... ;)

[Dean Roddey]

The font criteria stuff provided is pretty simple, just intended to insure that you basically get the same look. I'm sure that OpenGL provides at least that level of font control. Most systems provide some sort of 'give me the best match for this' type of call. You might have to swizzle the values a bit if their way of describing fonts is a lot different.

[Fonceur]

Actually, I dropped the OpenGL idea, as that added overhead that wasn't really needed, not trying to get a 60 fps game here... ;) I am using the native Android canvas. I found where to set the font type and such, so it should work out fine. Only the font size might still be an issue, as it seems to only take point size, which I know does not scale with various resolutions.

http://talluscorp.com/mediawiki/images/5/50/TaRIVA_Template.jpg

[SamVimes2]

Looking good!

[Fonceur]

Version 0.11 alpha is available on the wiki:

http://talluscorp.com/mediawiki/index.php?title=TARiva_Application

It likely won't do more than display simple text/rectangle and choke on the more complicated stuff... ;)

[znelbok]

Just sent my phone away last night to have it checked for a fault so I can help with this until next week

Mick

[sic0048]

Wow! I guess I have to buy an Adroid phone now! Actually I am in the market and will probably get something in the next couple of weeks or month or so. I'd like to see what happens on the rooting front with the new DroidX before buying it. But I might be forced to pull the trigger early if Verizon really does announce the start of tiered data plans instead of the unlimited plan they have now.

[Fonceur]

I guess I have to buy an Adroid phone now!
Actually, I'm going on vacation Friday for more than a week, without a laptop or anything, so no need to hurry that much... ;)

[znelbok]

Before you get too far you should read some of the latest comments re security on the iPhone riva thread. The same issues apply here.

We need to be able to create a way of securly logging in without having a password pre entered but can be pre entered is desired.

Mick

[Fonceur]

Before you get too far you should read some of the latest comments re security on the iPhone riva thread. The same issues apply here.
Right, I went through the regular Riva thread, but haven't gone through that one yet.

EDIT: OK, I read the end of the thread... I'll be adding a "remember the password" checkbox or such, I'll be saving the password hash (instead of the password itself) and won't display the password...

[Fonceur]

The current version has some flicker issues, as I started the UI based on some gaming code (with continuous redraws), but I will be replacing that part by some update-driven code.

[sic0048]

Hmmmm, DroidX has been rooted! I'm that much closer to pulling the trigger. Now it just has to get back in stock!

[sic0048]

Is there a trial mode with tariva? If not, how do we purchase it? It doesn't appear on the wiki or your website.

I've loaded it on my new android phone and I get the CQC screen. I've changed the setting to point to my server, but nothing happens after that. It doesn't appear to even try to connect to the CQC server (according to logs).

Thanks,

[Fonceur]

Is there a trial mode with tariva?
Currently there are no limitations, this will come later... ;)

I've changed the setting to point to my server, but nothing happens after that. It doesn't appear to even try to connect to the CQC server (according to logs).
The best thing would be to check the Android log with Log collector or such.

Make sure ports 13516 and 13517 are opened in your firewall and to grab version .15 from the wiki.

[batwater]

Brian your other option is to turn on WiFi and have your phone talk to your server over your internal network. (assumption made that you were going OTA originally vs WiFi)

If you do not have it set up and you have a dynamic IP address assigned to your broadband device from your ISP, you will need to use a service like dyndns that maps via your router your dynamic IP address to a domain name that you create. The service is free.

-Ben

[Fonceur]

Any reason why I am getting:

New_template
Start_draw/End_draw block
Start_draw/End_draw block

Where the second Start_draw/End_draw block is an exact copy of the first one? It's doing that for 3 different templates...

[Dean Roddey]

What do you mean by exact copy? You mean the context between the two blocks is identical or the actual messages are identical? If it's the latter, and the sequence ids are starting over at 2 or 3 or whatever in the second ones, then you are losing the connection and logging back in again, which starts everything over including the sequence ids.

[Fonceur]

No the ID's are different, but it's the content that is identical, i.e. messageLoad(X) = messageLoad(X + 28))... I already have the full template by the message ID 31, so I don't see why it is sent a second time...

Header ID = 3, length = 35, code = NEW_TEMPLATE
Header ID = 4, length = 0, code = START_DRAW
Header ID = 5, length = 11, code = PUSH_CLIP_AREA
Header ID = 6, length = 11, code = FILL_AREA
Header ID = 7, length = 11, code = PUSH_CLIP_AREA
Header ID = 8, length = 0, code = PUSH_CONTEXT
Header ID = 9, length = 11, code = PUSH_CLIP_AREA
Header ID = 10, length = 0, code = PUSH_CONTEXT
Header ID = 11, length = 11, code = FILL_AREA
Header ID = 12, length = 0, code = POP_CONTEXT
Header ID = 13, length = 0, code = POP_CLIP_AREA
Header ID = 14, length = 0, code = POP_CONTEXT
Header ID = 15, length = 0, code = PUSH_CONTEXT
Header ID = 16, length = 11, code = PUSH_CLIP_AREA
Header ID = 17, length = 0, code = PUSH_CONTEXT
Header ID = 18, length = 11, code = FILL_AREA
Header ID = 19, length = 0, code = POP_CONTEXT
Header ID = 20, length = 0, code = POP_CLIP_AREA
Header ID = 21, length = 11, code = PUSH_CLIP_AREA
Header ID = 22, length = 11, code = PUSH_FONT
Header ID = 23, length = 1, code = SET_BACK_MIX_MODE
Header ID = 24, length = 4, code = SET_COLOR
Header ID = 25, length = 39, code = DRAW_MULTI_TEXT
Header ID = 26, length = 0, code = POP_FONT
Header ID = 27, length = 0, code = POP_CLIP_AREA
Header ID = 28, length = 0, code = POP_CONTEXT
Header ID = 29, length = 0, code = POP_CLIP_AREA
Header ID = 30, length = 0, code = POP_CLIP_AREA
Header ID = 31, length = 8, code = END_DRAW

Header ID = 32, length = 0, code = START_DRAW
Header ID = 33, length = 11, code = PUSH_CLIP_AREA
Header ID = 34, length = 11, code = FILL_AREA
Header ID = 35, length = 11, code = PUSH_CLIP_AREA
Header ID = 36, length = 0, code = PUSH_CONTEXT
Header ID = 37, length = 11, code = PUSH_CLIP_AREA
Header ID = 38, length = 0, code = PUSH_CONTEXT
Header ID = 39, length = 11, code = FILL_AREA
Header ID = 40, length = 0, code = POP_CONTEXT
Header ID = 41, length = 0, code = POP_CLIP_AREA
Header ID = 42, length = 0, code = POP_CONTEXT
Header ID = 43, length = 0, code = PUSH_CONTEXT
Header ID = 44, length = 11, code = PUSH_CLIP_AREA
Header ID = 45, length = 0, code = PUSH_CONTEXT
Header ID = 46, length = 11, code = FILL_AREA
Header ID = 47, length = 0, code = POP_CONTEXT
Header ID = 48, length = 0, code = POP_CLIP_AREA
Header ID = 49, length = 11, code = PUSH_CLIP_AREA
Header ID = 50, length = 11, code = PUSH_FONT
Header ID = 51, length = 1, code = SET_BACK_MIX_MODE
Header ID = 52, length = 4, code = SET_COLOR
Header ID = 53, length = 39, code = DRAW_MULTI_TEXT
Header ID = 54, length = 0, code = POP_FONT
Header ID = 55, length = 0, code = POP_CLIP_AREA
Header ID = 56, length = 0, code = POP_CONTEXT
Header ID = 57, length = 0, code = POP_CLIP_AREA
Header ID = 58, length = 0, code = POP_CLIP_AREA
Header ID = 59, length = 8, code = END_DRAW

[Dean Roddey]

Are there any onload actions or anything that might cause the screen to redraw again?

[Fonceur]

Are there any onload actions or anything that might cause the screen to redraw again?
I don't have the template myself, so I'll let Batwater reply... It's supposed to be a simple template with some text, but maybe he reused some older one...

[batwater]

The test template contains a single simple static text field all default settings.

-Ben

[Dean Roddey]

Oh well, I can't look into it at the moment. At the worst it's a little less than optimal in terms of optimization, but it won't hurt anything if it redraws again. I'll check it later.

[Fonceur]

At the worst it's a little less than optimal in terms of optimization, but it won't hurt anything if it redraws again.
If it happens for all the start/end draw blocks, that would double the CPU usage of the app, but yeah, it's not something major, just odd. ;)

[Dean Roddey]

It definitely shouldn't happen for all of them. I thought you meant it just happens on the initial load of the template, right? It's possible that there's some 'just in case we changed anything during init' type of redraw in there somewhere on the initial template load.

[batwater]

Just to confirm, this is the template being referenced (If I packaged it correctly) it is purely a static text field.
-Ben

[sic0048]

What could I be missing? I've loaded v.15 onto my DroidX. I've created a new DroidX user and it is set to load a simply template with one static text field. It has no background and is not transparent.

I've set the Droid Configuration to point to 192.168.1.2 (which is my CQC server address) and inputed the DroidX user name and password. Yet the screen simply sits at the white CQC initial screen and doesn't seem to do anything. I checked the Admin Interface, and there is no RIVA license being used, so I know it is not connected.

Anything else? I've connected other RIVA clients to this server before, so I know the firewall, etc is all set up to accept a user (actually the firewall is turned off).

EDIT - If I change the address to http://whs/ (the actual server name) the application will crash/close out. Perhaps that is better? At least it is a sign of life

[Dean Roddey]

Is the license screen still showing number of licenses used? Unless you are on an older version, that's bogus and should have been removed because there's no such info anymore in the latest version. If you aren't on the latest version, you probably should be since plenty of improvements and such to the RIVA stuff was done since then.

[Fonceur]

Well, I am using the RIVA version 2.0 protocol, so if you have an old version, that could be a problem... Otherwise, post or send me the Log collector results or such...

[Dean Roddey]

Yeh, the server is backwards compatible, but not forward compatible. Not that it would matter much in so far. The differences are almost all from the server to the client in the form of messages that the client would likely just ignore. However, it probably will reject a login with any protocol version higher than it understands.

[sic0048]

OK - I'm still on 3.3 I think, so I'll upgrade to the latest version and test again.

[Fonceur]

I thought you meant it just happens on the initial load of the template, right?
Actually, I have only been testing templates with mostly static text, so I'm not sure. ;)

Any idea what that CQC error means? ;)

Got packet sequence number 9, but expected 9

[Dean Roddey]

Oops, that error is probably being logged after the expected sequence number was incremented. So it probably really means got 9 and expected 8. I'll look at that. Or it could be the other way, it got 9 an expected 10, depending on which way the error went.

[Fonceur]

Actually, I think I got the other versions too, as I wasn't resetting the Sequence ID when the connection was dropped/reset. That most likely lead to some strange states... ;)

[Fonceur]

OK - I'm still on 3.3 I think, so I'll upgrade to the latest version and test again.
Also make sure to grab taRIVA .18 (or newer), as I've fixed a few issues.

[sic0048]

FYI - upgrading to v.18 didn't help, but moving to CQC beta 3.3.6 was the issue. Once I installed the latest beta, the Android RIVA client connects now.

Thanks!

[sic0048]

I am having trouble with both the RIVA client and SageTV client connecting from a non-local network. I've changed the address to read something like: http://myname.DNSforwarder.com

EDIT - I changed the address to myname.DNSforwarder.com and it works. Obviously that was the problem.

[sic0048]

OK - finally a constructive post!!!

It looks like the text widgets are working. I can see static text widgets and field text widgets (which show the correct field value).

The justification is not correct on some of the text however and images obviously don't work yet.

I can scroll around a template that is larger than the screen size. So that seems to work well. It is actually really easy to move the screen around.

Just an FYI - plus I am excited to get it working!!!

[Fonceur]

Good, as my main tester is going on vacation soon, I can use a new guinea pig... ;)

[batwater]

Brian, glad to hear you have progress, will start posting / comparing notes here!

My wife thinks I'm an alien for getting so stoked about being able to see our HVAC settings on my phone

-Ben

[Fonceur]

I can scroll around a template that is larger than the screen size.
Version .20 should work better with those.

[Fonceur]

I thought you meant it just happens on the initial load of the template, right? It's possible that there's some 'just in case we changed anything during init' type of redraw in there somewhere on the initial template load.
OK, testing a "clock template", indeed it's only the init that has the double messages, there are no other start/end block duplicate after. So I guess it's the New_Template that is followed by a start/end draw block, when it doesn't really need to.

[Dean Roddey]

Yeh, things can happen during load, and there's probably a just in case redraw at some point after everything is done.

[Fonceur]

As far as text-only templates go (including text buttons), version .23 is mostly working. It can handle any size or orientation of templates, though there's some issues with swiping (where almost everything will get blanked)... ;)

[sic0048]

As far as text-only templates go (including text buttons), version .23 is mostly working. It can handle any size or orientation of templates, though there's some issues with swiping (where almost everything will get blanked)... ;)

I just upgraded to this version. This is my exact experience. But I has better swiping on prior versions where the widgets didn't disappear. Now the swiping does make most widgets disappear.

[Fonceur]

But I has better swiping on prior versions where the widgets didn't disappear.
OK, try version .25, it should act better.

[Fonceur]

Looking at the image cache, the question is does everyone have an SD card to store the images, or should I store them in the main internal storage?

[sic0048]

OK, try version .25, it should act better.


Actually I cannot connect using version .25 using the same templates as before. I've attached a log for you, but I'm not sure it will tell you anything . It looks like it is choking on an image.

EDIT - I changed to load a simple template with only one widget - a single text widget - and it still won't connect (giving similar errors as log shows).

[sic0048]

Looking at the image cache, the question is does everyone have an SD card to store the images, or should I store them in the main internal storage?

I'd prefer it to use internal space, but that is because I have 8gb internal space. Most phones don't have that much. Would it be possible to give someone the option?

[Fonceur]

It looks like it is choking on an image.
Any chance you could copy that template over and remove the images? It's likely not sending a proper image query, as I haven't finished that code yet. ;)

[Fonceur]

EDIT - I changed to load a simple template with only one widget - a single text widget - and it still won't connect (giving similar errors as log shows).
Hmm, any chance you can reboot CQC to get out of that loop?

[Fonceur]

I'd prefer it to use internal space, but that is because I have 8gb internal space. Most phones don't have that much. Would it be possible to give someone the option?
I think that I can use the SD card if it's there, otherwise just stick a blob in the database using the internal storage... So you don't have a SD card, right?

[sic0048]

Hmm, any chance you can reboot CQC to get out of that loop?

OK - that seemed to work. I was even able to go back to my original template. The widgets don't disappear when scrolling now. The scrolling is pretty choppy. I thought it was smoother before, but I could be completely wrong. Of course it might also have to do with environmental issues too. I am running on a remote wireless LAN and before perhaps I was on the local LAN network, I really don't remember.

I think that I can use the SD card if it's there, otherwise just stick a blob in the database using the internal storage... So you don't have a SD card, right?

No I have a 32gb SD card too :-) But that is getting filled up much quicker than the 8gb internal memory. Another benefit of having the new DroidX as most older Android phones have very limited internal storage.

[Fonceur]

The scrolling is pretty choppy. I thought it was smoother before, but I could be completely wrong.
Most likely my bad, I limited the refresh to every 250 ms during the scrolling, and forgot to take it out. Version .26 shouldn't be choppy and avoid the image download issue too.

No I have a 32gb SD card too :-)
OK, so I guess it would need to be user selectable then...

[sic0048]

Version 26 is has better scrolling. Nothing I would describe as smooth yet, but better than .25.

[Fonceur]

Version 26 is has better scrolling. Nothing I would describe as smooth yet, but better than .25.
Can you double check with the About that it's actually .26? Not sure if it has been uploaded yet... ;)

[sic0048]

Can you double check with the About that it's actually .26? Not sure if it has been uploaded yet... ;)

Darn you..... must have been a placebo effect.... It is still .25

[Fonceur]

Darn you..... must have been a placebo effect.... It is still .25
OK, try now, .26 is up.

[znelbok]

I got my phone back today so I can finally load this up

I think I have loaded it up OK, but when I try to run it I get an error

The application taRiva (process tallus.android.riva) has stopped unexpectedly. Please try again.

Did I do something wrong. I downlaoded the apk file directly from the web site and installed it from there (Galaxy S)

Mick

[Fonceur]

For debugging, the best thing is to install an app like Log Collector, then zip and attach the log. Did it crash after trying to load a complex template, or you never got to input any settings?

[sic0048]

OK, try now, .26 is up.

Much better!!! Thanks...

[znelbok]

For debugging, the best thing is to install an app like Log Collector, then zip and attach the log. Did it crash after trying to load a complex template, or you never got to input any settings?

The white screen with CQC sowed up and that is all, then it came up with the message.

Is there somewhere else for configuration such as server address etc, or is that after the white screen that I cant get past?

Mick

[Fonceur]

The white screen with CQC sowed up and that is all, then it came up with the message.
So it didn't like the default values, not sure what I have in there or why it would crash...

Is there somewhere else for configuration such as server address etc, or is that after the white screen that I cant get past?
You need to press the hardware Menu button to get to the configuration.

[znelbok]

So it didn't like the default values, not sure what I have in there or why it would crash...


You need to press the hardware Menu button to get to the configuration.
Can't find the harware button you are talking about. Can someone give me some clues as to where I find this.

[sic0048]

Can't find the harware button you are talking about. Can someone give me some clues as to where I find this.

It is the button you use to get to the menus or options for your phone. On some phones (like mine) it is one of the hard buttons. On other phones, it is not hard, but rather a touch button located at the bottom of most screens.

[Fonceur]

Can't find the harware button you are talking about.
As far as I know, except for some cheap Android tablets, all Android phones have some hardware Home, Menu and Back buttons...

[znelbok]

So what am I looking for because the only think I can find that referenced cqc is the icon in the applications area and that's where it crashes after I launch it

there is no specific option in the settings list for this app after I press the harware settngs button

Mick

[Fonceur]

You need taRIVA to be running in order to configure it... But as I said initially, if the application crashes on launch, you will need to install another application like "Log collector", so you can send me a report of the crash. Otherwise, there's nothing I can do.

[znelbok]

here are the two log files I have captured

Mick

[Fonceur]

here are the two log files I have captured
The second file has the start of the launch, but it cuts before the crash...

[znelbok]

I cant help other than I was asked to force quit and then I collected the log (both times).

i will try again a few time.

[znelbok]

Ahh a bit of luck now - I found the configuration screen. I was looking elsewhere instead of pressing the home button while the app was starting )somehing like the iPhone app where its in settings).

It still fails, but that is probably because it is not registered yet. How do I get a registration? The link on the website takes me to a page and reports "Unable to find Specified Product in Catalog."

Can we have a new thread started for this so that the link to where the app can be downloaded from can be easily found and it can also be kept up to date. Questions pertinant to this particular app can reside there as well (this thread started as a generic droid app thread)

Mick

[Fonceur]

It still fails, but that is probably because it is not registered yet.
Nah, the registration doesn't do anything just yet. I'm not sure how Log collector works, but it does look like you need to wait a bit for it to gather the logs, it doesn't seem to do it in real time...

Can we have a new thread started for this so that the link to where the app can be downloaded from can be easily found and it can also be kept up to date.
Yes, that did cross my mind before, and your wish has been granted. ;)

[Fonceur]

So I have this byte array (ba) which represents an image, but when I try to create a bitmap (bm) from it, all I get is null... Any ideas? Is there anything special about the byte array sent for ImgStart?

byte[] ba = message.GetByteArray();
Bitmap bm = BitmapFactory.decodeByteArray(ba, 0, ba.length);

[Dean Roddey]

It's not a bitmap. Well, in theory if it's cover art it could be a bitmap, but not very likely. You have to probe it and see what it is. It'll be either a PNG, JPEG, or bitmap. If it's from our repository, it's alway a PNG. If it's cover art, it could be any of the three. So you have to do a little poking around in the buffer to look for well known indicators of the file format and decode it appropriately. There may be some existing helper method on your platform that will do this for you, I dunno.

[znelbok]

latest log file.
taRIVA was started at 7:22 and forced to close at 7:23

Mick

[Fonceur]

latest log file.
Seems like you are touching the screen without being connected, and it doesn't like that or such...

[znelbok]

Well it just sits there and shows the CQC screen nothing else - didn't notice that touching was causing a problem.

The user account is set-up to show some simple text only as the default screen, so for some reason it is not loading that screen

[Fonceur]

Well it just sits there and shows the CQC screen nothing else - didn't notice that touching was causing a problem.
Yeah, it's not supposed to try to send anything when it's not connected, but something isn't working there... ;) As for why it's not connecting, you do have a RIVA server (> 3.3) running on 192.168.0.2 with port 13516/13517 opened in the firewall right? Could you check if the CQC log says anything?

[Fonceur]

So you have to do a little poking around in the buffer to look for well known indicators of the file format and decode it appropriately.
I was under the impression that it handled it automatically, but can't hurt to double check... ;)

[znelbok]

As for why it's not connecting, you do have a RIVA server (> 3.3)
Only running 3.0.3, otherwise everything else is there. I currently use the RIVA server with the iPhone.

Sorry for wasting your time. I should have read the page better. I assumed that because I was already using the RIVA server that this one would work as well.

Wont be upgrading for a while either as I dont have the money.

[Fonceur]

Wont be upgrading for a while either as I dont have the money.
I can look into using the 1.0 protocol too, I just need to figure out what to do without the initial New_Template message that was giving me the width and height... ;)

[Dean Roddey]

You could always let them configure a size that only's only overridden if you get a new template message.

[Fonceur]

Sounds like a plan, drop down list for the version, which hides/shows the dimensions...

[Fonceur]

So you have to do a little poking around in the buffer to look for well known indicators of the file format and decode it appropriately. There may be some existing helper method on your platform that will do this for you, I dunno.
Not having much luck on that side, so which method(s) are you using on the server side to create that byte array from an image? Maybe that will help me to do the reverse on the Android side... ;)

[batwater]

Greetings,

batwater and family has returned from a fun filled Caribbean vacation cruise and is now back from my brief hiatus (international cell rates are insane and no internet connection for a whole week, didn't die either :shock: )

On the SD card dialog, my vote is make it configurable. On my HTC Hero I will need to utilize the SD card as the main memory is a little more limited. My phone is already crabbing at me that it is nearly out of memory and I don't have a lot of stuff loaded. Amazing what 9 months can do to the android phone hardware, went from top of phone pile to lower middle in capability, sigh...

-Ben

[Fonceur]

On the SD card dialog, my vote is make it configurable.
Currently, I am using the application's cache directory, which uses the internal storage, but Android can clear it if it runs out of room. Adding an option to use the SD card should not be a problem.

[Dean Roddey]

The buffer is literally just the file contents of the graphics file. So, if it's a PNG file for instance, it's literally just the PNG file read into memory.

[Fonceur]

OK, so you are basically just doing something like:

byte[] byteArray = image.getBytes();

So once batwater sends me the PNG files, I can do the same on my side and should get the same byteArray, or at least figure out the differences between both...

[Dean Roddey]

Yeh, basically so. It should be just the raw image file contents.

[batwater]

For the images in template 4 I am using stock system images, specifically Buttons\Gel\Long Green, Long Purple for the pressed state, and Long Red. They are in the system image repo; I've sent the specific CQC files via email. Don't know if you can do anything with them...

-Ben

[Dean Roddey]

Those files are not the raw image files, so they won't tell you much. Those contain the raw image data, but they also contain other info.

[batwater]

Fonceur,

Text template is available on account 10 for your testing. I've included the field label # in the text on the template. Attached are screen shots of the 320x480 template and the Pallet Widget as well as the actual template.

Note, the static text boxes started out in order, that is the screen placement matched the Widget Pallet, as I added and shifted, the order was lost.

For font placement, H= horizontal, V=vertical, C=Center, R=Right, L, T, B, etc. There are other attributes to set but this is a good start, once these are knocked out, we can modify to include the fancier stuff.

edit: template and screen shot updated to fix typos, sorry. :oops:

-Ben

[Fonceur]

For 05/06 you have a mismatch left justify/H:R... ;)

I've fixed the bold text (misread how it worked). The italic isn't working, might be specific to the font used... I've implemented some vertical justification that seems to work fine.

As for the horizontal justification, there's something weird going on, like it's getting an offset somewhere.

[batwater]

Doh, copy and paste, foiled again, missed 07 also, fixed on test template. .28 verified on Hero, bold and vertical justification behaving as expected. Original attachments updated.

On the horizontal text offset, left, right, center, seems like the left origin of the text box might not be taken into account, don't know how it works, just thinking out loud...

-Ben

[Fonceur]

OK, so 1/5/6/7/8 are fine, though the fonts on 7 seems a bit bigger...

From what I can see, it writes the text left, right or centered on the initial point... Not quite what I was expecting, but easy enough to adapt... ;)

[Dean Roddey]

The justification stuff get's pretty complex when blurs and such are involved. I can send you my code that does those calculations if you want. Just e-mail me.

[batwater]

.29 is a vast improvement for proper text placement!

-Ben

[Fonceur]

Only running 3.0.3, otherwise everything else is there.
OK, you can try version .30 (once it's up on the wiki), by default it now uses the RIVA protocol 1.0. You will need to specify the width and height of the template, as that's only available in 2.0.

[batwater]

Hi Fonceur,

Tested a template pop-up with .30, mostly works, text more improved, great progress, thanks for your hard work!!! I've updated template 10 Times NR to be italic, it shows up italics on my Hero. Must be Ariel on droid for some reason.

Dean are fonts downloaded to the client via RIVA protocol or is it dependent on what is installed on the device?

-Ben

[Fonceur]

Dean are fonts downloaded to the client via RIVA protocol or is it dependent on what is installed on the device?
The iPhone RIVA client will download some missing fonts, but that's not part of the protocol. ;) All I get is the name of the font and other parameters to use with the text.

[Dean Roddey]

Yeh, it would be pretty crazy to try to do that. Ultimately the clients have to just try to be clever about choosing fonts that are close to the ones chosen. Given all the font formats and all that, trying to download them would get into a lot of complexity that I would rather avoid. If more font metrics are necessary to allow good selection, then that's something we can provide.

[Fonceur]

Yeh, basically so. It should be just the raw image file contents.
OK, I found the issue. I wasn't reading the unsigned integer left over from the length of error text... Did I mention how much I hate that optional "length of text" + text not part of the structure? ;)

[znelbok]

OK, you can try version .30 (once it's up on the wiki), by default it now uses the RIVA protocol 1.0. You will need to specify the width and height of the template, as that's only available in 2.0.

Excellent

Installed it this morning and it works with caveats.

The interface was build in portrait but I had the phone in Landscape when I started the app. it showed the screen fine (top portion) but when I rotated the phone it rotated the image, but did not display the lower portion, it was just blank.

I then added a command button to the page and the text displayed and then the image displayed after I pressed it. It executed the comamnd fine (load a new page).

I have just been loading my iphone templates and had these issues.

One of my [iPhone] templates is twice the width of the screen so I can scroll left and right, this would not load up at all (proabably a CQC version thing) and the app would fail and request a force close.

The background was just black, and not the grey / white belnd that I use (I dont use an image becuase I want it to be fast).

Images on command buttons would just show after being pressed - it could be just me being impaitent and they are not loading fast enough.

Comments/questions
Can we hide the password in the config page?
Can we add sounds so that a click can be heard when a button is pressed?
Not trying to make this an iPhone app, but how many of the option in the iPhone app are do-able here in the taRiva app?

As it stands right now its a working example that I could get by with. Might not be the prettiest - so well done on this.

Keep up the great work

Mick

[Fonceur]

The interface was build in portrait but I had the phone in Landscape when I started the app. it showed the screen fine (top portion) but when I rotated the phone it rotated the image, but did not display the lower portion, it was just blank.
And you could not swipe to reach the extra area?

One of my [iPhone] templates is twice the width of the screen so I can scroll left and right, this would not load up at all (proabably a CQC version thing) and the app would fail and request a force close.
Did you update the width and height on the configuration, since you are using the RIVA 1.0 protocol which doesn't provide those info... I guess the iPhone app might be rescaling on the initial FillArea with black or such... Also I haven't implemented all the commands yet, so it's very possible that the template made use of something undefined yet, which crashed the app... ;)

Images on command buttons would just show after being pressed - it could be just me being impaitent and they are not loading fast enough.
Actually, if the images haven't been downloaded by the time the drawing is going on, they just won't show until something forces a refresh, like that press. I guess I could delay the UI update until the images are downloaded and ready to display or such... OTOH, once they are in the cache, it should be fine for the next time. ;)

Can we hide the password in the config page?
Yeah, I probably should not populate that field, it just makes the current testing easier. ;)

Can we add sounds so that a click can be heard when a button is pressed?
There is no buttons defined as such, which means every press, anywhere on the screen would do that, making it pointless...

Not trying to make this an iPhone app, but how many of the option in the iPhone app are do-able here in the taRiva app?
Well, I don't have the iPhone app, so not sure what it all does, though I guess that now that the basic is working, I could snoop a bit... ;)

[znelbok]

And you could not swipe to reach the extra area?

No


Did you update the width and height on the configuration, since you are using the RIVA 1.0 protocol which doesn't provide those info... I guess the iPhone app might be rescaling on the initial FillArea with black or such... Also I haven't implemented all the commands yet, so it's very possible that the template made use of something undefined yet, which crashed the app... ;)

No, Just had it at the standard width. This is all based on 5 mins of testing so I will play some more tonight.


There is no buttons defined as such, which means every press, anywhere on the screen would do that, making it pointless...

? - I am not sure what you mean, but what I was asking for was the ability to add a command to an action that makes a click sound etc. so taht when the user presses a "button" aka check box, ccommand widget etc, a sound is made.


Well, I don't have the iPhone app, so not sure what it all does, though I guess that now that the basic is working, I could snoop a bit... ;)
A lot of feedback has been given for the iPhone app so it makes a great place to start for this app (and there is no need to re-invent the wheel).
Once this app has caught up, it would be great if they could prgress together with the same features.
I do understand that there may be some things specific to each OS (guestures etc)

[Fonceur]

? - I am not sure what you mean, but what I was asking for was the ability to add a command to an action that makes a click sound etc. so taht when the user presses a "button" aka check box, ccommand widget etc, a sound is made.
Right, and what I'm saying is that a round thing on the screenshot of the UI could be a button or the nose of a clown, there is no way to know it... ;) The RIVA protocol does not position buttons on screen, it just draws a set of shapes and images, when you press somewhere on the screen, it's the RIVA server that will decide whether there was something active like a button there.

Now there is some kind of "play wave" action that was added to the template side, which can send a play command over the RIVA protocol, and could be associated to do that, but...

[SamVimes2]

The way the iphone app does this is by reacting to the PlayWav command with some built-in iphone sounds.

So if a button has the action System::PlayWav(Beep1) iRIVA plays one of the iPhone's beeps.

More broadly I am following this pretty closely and am happy to provide any help gleaned from developing the iPhone app.

[Fonceur]

The way the iphone app does this is by reacting to the PlayWav command with some built-in iphone sounds.
Right, that's how I figured it might work.

[SamVimes2]

Hi Fonceur,

I installed this tonight on my wife's android phone (HTC Desire 800x480). I have a series of ipod (320x480) and iphone4 (640x960) templates to test with, but they are very image heavy so it's hard at the moment.

If i have text over an image background, should I be seeing that?

Some comments off the bat, based on my own heavy usage of iRIVA and what people have asked for throughout its development:
-some kind of network traffic indicator (on i* this is a light spinny circle in the upper left of the screen)
-some kind of mapping of gestures/orientation to hotkeys. Here (http://www.mountford.net/cqcclient/usersguide.html)is what we do on iRIVA; I think it would be great if we could keep it parallel.
-some kind of orientation locking options
-some kind of profiling system, with autoconnection features based on connection type

All in all keep up the great work! I will continue to play and post comments.

[Fonceur]

If i have text over an image background, should I be seeing that?
Yes, which part aren't you seeing? You might have to wait for the images to download and reconnect to actually see them. Only got the images to work in the last 48 hours, so there isn't much logic around them... ;)

-some kind of network traffic indicator (on i* this is a light spinny circle in the upper left of the screen)
Currently, the app itself hardly knows it's connection state... ;) I'd probably hide that on the information/configuration pages, as the screen real estate is at a premium... Otherwise, I'd probably just un-hide the status bar, reducing the available template space by as much, to use the notification area...

-some kind of mapping of gestures/orientation to hotkeys.
Not sure I see the point, or how you'd define hotkeys... A typical Android phone has like 5 hard button already used... ;)

-some kind of orientation locking options
Don't really see the point of that either...

-some kind of profiling system, with autoconnection features based on connection type
Not quite sure how that's supposed to be helpful...

[znelbok]

Don't really see the point of that either...


I tought the same as well at first, but these phone are to ready to swich from landscape to portrait (does it to me all the time when I am lying on the couch) and its a pain if you have only written your templates in one format. Locking it would allow the phone to be swung around without it rotating the screen. I started out building two templates, but gave up and now lock the screen to portrait and only use templates of one orientation.

Mick

[Fonceur]

I tought the same as well at first, but these phone are to ready to swich from landscape to portrait
So you just rotate it back, if that wasn't on purpose...

and its a pain if you have only written your templates in one format.
Why would it matter? You are still displaying the same template, a rotation doesn't call up a new template...

Locking it would allow the phone to be swung around without it rotating the screen.
Seems silly to read your text at a 90 degrees angle...

[znelbok]

you sound like you have not experienced this yet. I find it happens a lot and it annoying and takes time (only a second but still).

It is silly to read at a 90 degree angle (relative to head), and hence the lock orientation is good for that, I can lie down and hold the phone horizontal (butin line with my body/head) and it does not rotate on me (so I am then reading text at 90 deg to the normal).

Ultimately your choice, but I am voting for locking the orientation as an option to be added.

[Fonceur]

you sound like you have not experienced this yet.
I don't have a cell phone or a tablet, so no.. ;)

I can lie down and hold the phone horizontal (but in line with my body/head) and it does not rotate on me
Ah, that's what I was missing! From watching videos on my iPod touch, either in bed or on the couch, I can see the point of locking the orientation.

[sic0048]

I don't know what it would take to impliment, but I can see the desire to lock the screen from rotating too. I'd like to see it as a user defined option.

[Fonceur]

I got the screen orientation lock to work. So now I guess you're going to want it to be persistent between restart... ;)

EDIT: Turns out it was easy enough to make it persistent... ;)

[znelbok]

easier than convincing you

[Fonceur]

easier than convincing you
Indeed! ;) I was sure that I would have had to do some complex stuff but it turned out it was just a question of setting a value, and Android take cares of everything else... ;)

Since the forum doesn't allow files the size of the .apk, and since it takes time for them to show up on the wiki, you can grab them from DropBox. If you don't already have a DropBox account, use the referral (we both get an extra 250Mb or so):

http://www.dropbox.com/referrals/NTQwODI5NTk

Then send me a pm with the email address you used, so I can share a folder with you. You can even install an Android DropBox app to get direct access through your phone.

[Fonceur]

As far as I can tell, the issue with the images is due to trying to read more than 2k bytes, when they are not already in the buffer. With some loop and possibly some delay, I can hopefully get a work around.

http://talluscorp.com/mediawiki/images/a/aa/TaRIVA_Template_4.jpg

[sic0048]

Just an FYI: On her 38, I have a a single button on my template. The graphic doesn't appear until I touch where the text is. When I do that the button graphic appears.

[Fonceur]

Just an FYI: On her 38, I have a a single button on my template. The graphic doesn't appear until I touch where the text is. When I do that the button graphic appears.
Right, the download of the images happens at the same time as the drawing of the UI, so you pretty much need to pre-fill the image cache, unless the images are small enough.

I tried delaying the UI drawing when an image is still downloading, but it tends to crash due to simultaneous access of the database (which can also happen the first time the template with images is loaded). Hopefully I can modify my database access and resolve the issue.

[Fonceur]

Version .39 has a slight improvement for the image saving and is less likely to crash during loading, due to the database access...

[Fonceur]

So Broconne is the one that did most of the inputstream reading routine, dealing with the big/small endian stuff and all. So that when you read an unsign short, which is coming as 2 bytes, you basically get:

AC 00 --> 00AC = 172
AC 01 --> 01AC = 428

So if now you want to read a short coming as 2 bytes, could someone fill in the ???:

AC 00 --> 00AC = 172
AC 01 --> ??? = ???

I would guess it's -172 or maybe even (172 - 256) = -84, but... ;)

EDIT: Er, I guess it's more complex than that... ;)

EDIT2: While I'm still not sure how a negative short would be handled, "AC 01" would be the same for signed/unsigned short, the problem I thought I had is somewhere else... ;) I had an error in my log printing, so shorts looked odd.

[Fonceur]

Ultimately the clients have to just try to be clever about choosing fonts that are close to the ones chosen. Given all the font formats and all that, trying to download them would get into a lot of complexity that I would rather avoid. If more font metrics are necessary to allow good selection, then that's something we can provide.
As far as I can tell, the Android devices only come with the Droid fonts (http://www.droidfonts.com), in the following flavors:

* Droid Sans (regular, bold)
* Droid Sans Mono (regular, bold)
* Droid Serif (regular, italic, bold, bold italic)

Everything else gets replaced by one of those.

Eventually, I might be able to include some kind of extra fonts support through the SD card or such, but don't hold your breath, and just design your templates in consequence... ;)

[batwater]

I've located a source for the fonts at this site (http://damieng.com/blog/2007/11/14/droid-font-family-courtesy-of-google-ascender) direct download at this URL (http://android.git.kernel.org/?p=platform/frameworks/base.git;a=snapshot;h=39b04dd27e6d20809f8ff26920d1 e761a0005252;sf=tgz).

Note that I have not installed yet.

-Ben

[Fonceur]

I've located a source for the fonts
You probably don't need the Japanese or Ahem fonts in there. ;) But using the Droid fonts when designing your RIVA templates should ensure a proper match in taRIVA...

[Fonceur]

I have updated the wiki (http://talluscorp.com/mediawiki/index.php?title=TARiva_Application) with more images of test templates.

[SamVimes2]

I upgraded my device (HT Desire) to Android 2.2. I get some images now, but the screen gets all "jumpy" when i touch. Is anyone having success with Froyo?

Thanks!

[Fonceur]

I get some images now
You should get them all the next time, as the cache will have been pre-filled...

but the screen gets all "jumpy" when i touch.
A few versions had such issues, but .43 should be fine... So what were you trying to do?

[SamVimes2]

All the links I can find still link to version .42, which definitely has the image freak-out problem...reminds me of the 'advil' with the funny stamps on it I used to get at the parties!

Can you point me to where I can find .43? Also, can you maybe version your .apk files so I don't make the same mistake in the future?

In addition, I'd like to go ahead and pay for the app, but I can't find it on the website, can you point me in the right direction for that too?

[Fonceur]

Can you point me to where I can find .43?
Post 134 has the info for the best way to get it... ;) Basically that's where the Tallus owner picks it anyway.

Also, can you maybe version your .apk files so I don't make the same mistake in the future?
Kinda defeat the point of linking to a single file... ;) But dropbox has a text file with the version.

In addition, I'd like to go ahead and pay for the app, but I can't find it on the website, can you point me in the right direction for that too?
I'd kinda like to polish it up some more before that... ;)

[SamVimes2]

Ok...I must have misunderstood the wiki:

The most current version is always here or from the About screen: Direct download for Android web connected devices


I already have a dropbox account so I can't refer you. I will email you.

[Fonceur]

Ok...I must have misunderstood the wiki:
No, but the guy that wrote that is also the one that is not keeping up... ;) Which is why on the SageTV forum I posts the files directly in the thread. I would do the same here, but the .apk size is bigger than what's allowed on this forum. Giving people access to DropBox was the best work around I could find.

I already have a dropbox account so I can't refer you.
That's fine, I'm already up to 3 Gb and don't even use 25% with my various shared folders. ;)

[batwater]

Sam,

The DropBox method is by far the easiest. If you don't have the free android client you should grab it. From the shared Tallus_Android DropBox folder you simply select the .apk file and it installs, no muss, no fuss.

-Ben

[Fonceur]

Version 0.44

- Added support for multiple Username/Hostname combos.

Obviously you can use different passwords for each... The protocol version is assumed to be the same, though if you are using protocol 1.0, you can pick individual resolutions.

[Fonceur]

Version 0.45

- Fix: On launch always used the first Profile.
- Fix: Would crash switching profiles if the first was not connected.

[SamVimes2]

Hi Fonceur,

I got the drop box set up; you're right that its very easy that way.

After installing .45, I see that it still has the "shakes," even on the opening splash screen. Is that intended?

[Fonceur]

Version 0.46

- Fix: Was sending the wrong X coordinate with Press/Release.


Basically if you had a template that was wider than your phone, any tap passed the initial view was off...

[SamVimes2]

Still does crazy shakes on the opening splash screen. (But I'm super impressed at turnaround time!)

[Fonceur]

After installing .45, I see that it still has the "shakes," even on the opening splash screen. Is that intended?
I'm still not sure of what you are describing... It sounds like you are swiping (thus updating the screen) while you are still receiving data to update the screen, which indeed is a bit of a problem...

[SamVimes2]

1) launch the app
2) see cqc image splash screen
3) touch the screen
4) see it shake wildly and distort
5) pick a profile and connect
6) touch the screen
7) see it shake wildly and distort

[SamVimes2]

Also there are no images in this one (they had been working in .42).

I am testing with my iPod templates which are 320x480 but on my information screen it says that the resolution is "320x533 (320x480)". The device is 480x800, yet the templates seem to be stretching to fill the screen. Is that intended?

[Fonceur]

1) launch the app
2) see cqc image splash screen
3) touch the screen
4) see it shake wildly and distort
If the last profile is valid, you are already in the process of updating the screen, and then force more redraws as you touch the screen... It's kinda like if you have a printer that's printing something and you decide to pull the paper because it's not going fast enough... ;)

5) pick a profile and connect
6) touch the screen
7) see it shake wildly and distort
Again, once you've picked a Profile, you are loading it, so by touching the screen you are performing more redraws...

Of course if there was an indication that something is going on, you might not be tempted/abled to do other things at the same time... ;)

[Fonceur]

Also there are no images in this one (they had been working in .42).
Probably just a caching issue, they should be there if you reload.

I am testing with my iPod templates which are 320x480 but on my information screen it says that the resolution is "320x533 (320x480)". The device is 480x800, yet the templates seem to be stretching to fill the screen. Is that intended?
Hmm, the template is correctly reported as (320x480). Now if it reports 320x533, and rescale to 480x800, that might explain the shake and stretch, as it seem the resolution is not always reported as the same, or it somehow switches to a different graphic mode or something...

[SamVimes2]

In the information screen, is the "X x Y (Z x W)" supposed to represent "device res (template res)"? If so my device res is not being read correctly.

[Fonceur]

In the information screen, is the "X x Y (Z x W)" supposed to represent "device res (template res)"? If so my device res is not being read correctly.
Indeed, which is what I'm saying... ;) If you rotate the phone and check the resolution again, what does it say?

[SamVimes2]

It says 533x320. I'm also not getting images, no matter how long I let it sit or how many times I close and reopen (though I did have working images in .42)

[Fonceur]

OK, going to need some log collector report(s) then...

[sic0048]

I see the shaking issue when in landscape mode, but not in normal orientation. I am using a 480x854 template.

I also don't get any images and cannot get a cover art browser to show any cover art.

[Fonceur]

I see the shaking issue when in landscape mode, but not in normal orientation. I am using a 480x854 template.
What is the reported resolution on the Information screen, and what is the actual resolution of your phone?

I also don't get any images and cannot get a cover art browser to show any cover art.
Are you using the small images for the browser? Do you at least get the navigation images? If you reload the app with the same page, do you see those images?

If you could send me the result from Log Collector, I can have a look...

[batwater]

I get images when my signal strength is really good but so so when I have poor reception. I'm on Sprint which is very tolerant of low signal strength and still having data flow with other apps. I'm wondering whether this is caused by file I/O time outs?

I have a MyMovies repo template, system generated at 320x480, that I can browse, however as it stands, you have to page forward to the next page (next page is also blank) and then back,then the images show up. This is a known issue.

What type of connection are you using for your testing, cellular or WiFi?

-Ben

[Fonceur]

Version 0.47

- Wait up to 1 second per image download.
- Added a "Loading. Please wait..." between the Start/End operations.


So now you should be able to see the images from a Media Browser, assuming you are using the small images option and have a decent speed. The actual connection will not display the wait dialog, only once it starts processing the Start_Draw...

[Fonceur]

Currently, if you want to swipe across, there is a Press and Release sent to the server... So if you happen to hit a button, that can be bad. ;)

What I propose is to only send the Press/Release for a "long press", assuming I get access to that somewhere...

[batwater]

Was there an issue with .47, drop box has .46?
-Ben

[batwater]

Currently, if you want to swipe across, there is a Press and Release sent to the server... So if you happen to hit a button, that can be bad. ;)

What I propose is to only send the Press/Release for a "long press", assuming I get access to that somewhere...

What about multi touch, eg 2 fingers for a swipe message vs. long press? This is distinctly different from a press. Although a long press more likely ensures that the button press was intentional.

Might want to reserve the multi touch option(s) for special triggers...

-Ben

[Fonceur]

Was there an issue with .47, drop box has .46?
Er, what was wrong is that it was late, and I don't let DropBox running by default... ;) It's there now.

[batwater]

Version 0.47

- Wait up to 1 second per image download.
- Added a "Loading. Please wait..." between the Start/End operations.

So now you should be able to see the images from a Media Browser, assuming you are using the small images option and have a decent speed. The actual connection will not display the wait dialog, only once it starts processing the Start_Draw...
Media browser images now show up, yay!

The Loading. Please wait.. is appreciated but extremely annoying as it is showing up constantly and is actually flashing fairly regularly if briefly in the background. For example test template #8 with all of the alarm status images, this flashes almost constantly. For the media browser it shows up for every image loaded.

Perhaps as a compromise there could be a very small indicator in a corner that indicates I/O life?

Thanks!!!
-Ben

[Fonceur]

The Loading. Please wait.. is appreciated but extremely annoying as it is showing up constantly and is actually flashing fairly regularly if briefly in the background.
Right, as I said, every Start/End blocks are generating it, so if something on a template changes rapidly (like the time) you do get a lot of redraws. It was mostly so people could get a feel for what was going on, instead of having people try to connect to the server when there's already a connection going on...

I do want something less intrusive and more accurate, but that's what I have found so far... ;)

[batwater]

Agree completely,it is very helpful to see what is happening. I do like the status indicator! Hopefully you can find a little blinky of some sort.

The other issue that is happening is on #8 the scroll list index is getting reset back to the beginning when trying to scroll now.

-Ben

[Fonceur]

The other issue that is happening is on #8 the scroll list index is getting reset back to the beginning when trying to scroll now.
Mind elaborating? Everything seems to be working fine on it.. I click the up/down/top/bottom arrows and they work as they should, or do you mean something else by "scroll"?

[batwater]

Sorry, I mean the up and down arrows. The list is getting reset back to the beginning, not every time, but fairly consistently when the buttons are pressed.
-Ben

Edit,

Appears to happen when I press when the please wait dialog is on screen

[Fonceur]

Sorry, I mean the up and down arrows. The list is getting reset back to the beginning, not every time, but fairly consistently when the buttons are pressed.
Still not seeing it... Maybe some data are being added on the list which resets it?

[Fonceur]

Appears to happen when I press when the please wait dialog is on screen
Well, if the dialog is on screen, that means it's already redrawing, possibly because the list has changed...

[batwater]

Still not seeing it... Maybe some data are being added on the list which resets it?

I think I have figured out what was happening but Dean I need you to validate text list box behavior.

If I am scrolling a text list and that list is updated will the list reset itself to the beginning if there is an active scroll request in progress (in this case via RIVA)?

I believe what was happening when I was testing yesterday was 2 things in combination that was causing what I was seeing: 1) slower cellular network connection, 2) alarm panel reporting back zone messages while I was trying to scroll the zone message list box.

Thanks
-Ben

[Dean Roddey]

Yeh, the list will reset itself. It's not a list in the sense that the typical local list box in a program is. It's just displaying the contents of a field. If the field value changes, it doesn't know if it changed because a single item was added or if the whole thing was changed. So it just reloads and goes back to the start.

[jkish]

Yeh, the list will reset itself. It's not a list in the sense that the typical local list box in a program is. It's just displaying the contents of a field. If the field value changes, it doesn't know if it changed because a single item was added or if the whole thing was changed. So it just reloads and goes back to the start.

Maybe it could be a little smart...

Like if the field source is the same and the number of items is still >= the old view index + list size, then don't change the view index.

[jrlewis]

Any way to signal so you could use a separate fieldwrite for the stringlist? One that signals the widget should be "reset" and set its index to 0 and one that just updates the field, but doesn't cause any linked widget to change its current index.

[Fonceur]

I'm trying to replicate the effect of the progress bar, but using a linear gradient between light gray and red, I sure don't get the same effect... ;)

Based on the IV, it's like color 2 (red) is used for 60% and there's a gradient for the initial 40%, but even then it's more like orange, there's no gray in there...

The best compromise I got was by using the linear gradient on 50% of the image...

http://talluscorp.com/mediawiki/images/9/99/TaRIVA_Template_weather.jpg

[Dean Roddey]

How are you going about doing all the image processing and whatnot?

[Fonceur]

Actually, I think the fill from bottom to top and right to left are busted in the IV itself, as trying the 2 other orientations do give a proper gradient in the IV...

EDIT: Actually, even the other orders are not quite right...

If you do top (white) to bottom (red) at 65%, the gradient will be spread from 0 to 100%. So it stops pinkish...

If you do bottom (white) to top (red) at 65%, the gradient will be spread from 0 to 20% at the bottom, using the 80-100% red, with the 20 - 65% using the 100% red.

The horizontal results are similar...

[Fonceur]

How are you going about doing all the image processing and whatnot?
It's a simple alpha level extraction, refilled using a linear gradient for the painting brush.

Shader shader = new LinearGradient(0, height, 0, 0, c1, c2, mode);
Paint paint = new Paint();
paint.setShader(shader);
final Bitmap extracted = bitmap.extractAlpha();
canvas.drawBitmap(extracted, src, dst, paint);

[Fonceur]

Version 0.49

- Implemented the remaining fill orders of the Progress bar.


Version 0.48

- Fix: Progress bar will no longer crash the application.
- Implemented the vertical progress bar, filling from bottom to top.

[Dean Roddey]

I think that maybe the issue on the fill is that perhaps you are doing the gradient for the whole size, instead of just the size of the actual fill bar image?

[Fonceur]

I think that maybe the issue on the fill is that perhaps you are doing the gradient for the whole size, instead of just the size of the actual fill bar image?
As I said, the problem isn't on my side, it's just that the IV isn't showing it properly, which is why I wasn't able to replicate it... ;) If you want to convince yourself, just use a progress bar with a gradient from white to red, then try the 4 orders. You would expect the top to bottom (left to right) to be mirrored of bottom to top (right to left), but they're not in the CQC IV... ;)

[Dean Roddey]

Oh, OK. I'll check that. I did test them at first, but I've only used left to right or bottom to top since then myself, so I might not have noticed it. I need to do a test template for the progress bar really.

[Fonceur]

Version 0.50

- Fix: Make sure not to use a deleted profile.
- Re-enabled DrawLine.
- Replaced some paint.setARGB by paint.setColor.

[batwater]

This is a functional design question on the configuration settings:

I set up a new account, neglected to select v2 on the RIVA protocol setting. I can change the setting and apply and the template loads properly using v2 but when I exit and reenter the protocol config change has not been applied to that account.

I think it makes sense to save whatever the current client settings are when the apply button is pressed. Either that or an additional update button will need to be added otherwise users will be forced to add a new account and then delete the old one in order change or correct settings.

Thanks!
-Ben

[batwater]

I have noticed when I have a slower cellular connection and I go into the config menu the please wait message box takes over focus and the config dialog is no longer accessible. This primarily occurs on templates with a lot of fields like my alarm panel status. The other thing that occurs on this same type of template is that the please wait message flashes in and out rapidly (I know this is for debugging and to indicate that there is I/O happening) making it very challenging to make any config changes.

This is a philosophical design question:

It is desirable for the template that is loaded to continue interacting with the server but I am wondering (presuming this is possible) if it makes more sense to quiesce I/O with the server when making config changes? Either that or not allow the config dialog to loose focus. It is possible that the user may just be going into the config to check something but not really making any changes.

-Ben

[Fonceur]

I set up a new account, neglected to select v2 on the RIVA protocol setting.
Technically, the RIVA version applies to all, it's not on a per Profile basis. One possibility is that with the recent changes, I might not be loading/saving it properly, so I'll double check.

[Fonceur]

I have noticed when I have a slower cellular connection and I go into the config menu the please wait message box takes over focus and the config dialog is no longer accessible.
Right, I had to use some work around so it wouldn't crash in those cases, which has side effects. Having the progress bar and a dialog at the same time is a bad idea.

As for your big template with its 50 dynamic fields, I'd suggest breaking it down into multiple smaller ones, with a navigation layer... ;)

It is desirable for the template that is loaded to continue interacting with the server but I am wondering
I see a few possibilities:

1) Replace the "Please wait" by something less intrusive.

2) Block the update queue while a dialog (i.e. config) is up.

3) Replace the dialog (as overlay) by their own individual page.

[Fonceur]

Version 0.51

- Fix: Protocol version was never updated...

[batwater]

Benefit of having dropbox loaded on my laptop, I knew a new version had just been dropped :-)

Thanks!

[batwater]

Right, I had to use some work around so it wouldn't crash in those cases, which has side effects. Having the progress bar and a dialog at the same time is a bad idea.

As for your big template with its 50 dynamic fields, I'd suggest breaking it down into multiple smaller ones, with a navigation layer... ;)
Already working on, testing extremes with the current one ;-)
I see a few possibilities:

1) Replace the "Please wait" by something less intrusive.

2) Block the update queue while a dialog (i.e. config) is up.

3) Replace the dialog (as overlay) by their own individual page.
Hmmm, option 3 might be the cleanest, anyone else testing the client have thoughts on this? I also would like to see a less intrusive method of indicating I/O activity but I also see that as a lower priority.

-Ben

[Fonceur]

Version 0.52

- Keep the current's templates images in memory.
- Close the Info screen after pressing the Clear cache button.
- Open/close the TCP connection for each individual image.

[Fonceur]

Version 0.53

- Fix: Was not handling the Pop...
- Re-enabled all clipping.

[sic0048]

What type of images should we see now?

Here is a screen shot of your new mobile templates. None of the images show up and the text appears slightly fuzzy.

http://i185.photobucket.com/albums/x46/sic0048/CAP201009131709.jpg

[Fonceur]

What type of images should we see now?
This is using taRIVA .53? I guess the images were too big for your connection speed. If you switch to a different day and back, you should see them...

For the ???, it just means you didn't load/setup the variable driver. As for the fuzzy text, I guess Android doesn't scale fonts properly for all sizes...

[Fonceur]

I think most of the elements are in, so it is now possible to buy a registration key for $19.99. Of course there are still more iterations to come...

Just a heads up, some of the features will be moved to the "registered only" version...

[sic0048]

Purchased!

On another note, I deleted the user profile that I had stored in the RIVA client and now the application force closes anytime I try to open it. I assume it is because it is trying to connect, yet there is no server information stored.

I deleted the profile because I lost the ability to connect. I tried both using a LAN setup as well as over 3g and nothing would work. The odd thing is that nothing changed. I didn't change versions and I tried using both your templates and my old template without luck. The screen would just it there on the CQC initial page. There was a thread error in the CQC log, but no details.

Checking the phone log I see an entry that says "W/System.err(10987):java.net.SocketException: Broken Pipe as well as some other stuff. I'm going to send you an email of the log.

[SamVimes2]

Hi Fonceur,

Any thoughts on the issue above where some devices' resolutions are being misreported? I'd love to buy and use the app!

Thanks.

[Fonceur]

On another note, I deleted the user profile that I had stored in the RIVA client and now the application force closes anytime I try to open it.
Actually, if you delete the first item I can see a problem... Fixing that for .54.

I tried both using a LAN setup as well as over 3g and nothing would work.
The last time something similar was reported, it was due to the user turning off wifi or such...

[Fonceur]

Any thoughts on the issue above where some devices' resolutions are being misreported?
I will need to add some more debugging text and use multiple ways of querying the resolution... In the mean time, what is the maker/version of your phone?

[sic0048]

The last time something similar was reported, it was due to the user turning off wifi or such...

I've tried it both ways. It is strange that it was working fine at one point yesterday and I was able to take the screen shots above, and then later that day, it won't connect. I tried both the "Home" LAN setup over the wireless network, and the "Away" setup which points to a dynamic DNS server which works over 3G or wireless. I'm hoping the logs show something because the CQC log shows the connection trying to be made, but some error causes it to stop.

I haven't tried the windows version of the RIVA client in a week or so, but I was having issues with it connecting too. Oddly enough, my iPaqs connect just fine without any issues. I'll try the Windows version again ASAP, but it might be a couple days.

[Fonceur]

Any thoughts on the issue above where some devices' resolutions are being misreported?
As a data point, sic0048's DroidX which is advertised as 480 x 854 is being reported as 320 X 569 by my detection method...

[sic0048]

Perhaps that explains why I get the jittery screen when I scroll.

[SamVimes2]

It's an HTC Desire with 480x800 resolution. taRIVA reports it as 320x533.

[SamVimes2]

As for the fuzzy text, I guess Android doesn't scale fonts properly for all sizes...

I think this is related to the misreporting and resulting scaling of his resolution - I see something similar on mine.

[Fonceur]

Version 0.54

- Fix: Deleting the first Profile was bad.
- The orientation lock and multiple Profiles now require a valid Registration.
- Try to discover the screen resolution through DisplayMetrics.


On the Information screen, you will now se a third resolution displayed in [], see if that method is more accurate. It is not actually used in the app, until I figure one more accurate than my current solution... ;)

If you have multiple Profiles already defined on an unregistered copy, it won't delete them, but you won't be able to add more. While you can display them and possibly edit them, it will still only connect to the first profile.

[Fonceur]

I think this is related to the misreporting and resulting scaling of his resolution - I see something similar on mine.
Right, that did cross my mind after I posted, though I wasn't sure if he had misreporting issues, which indeed he does. ;)

[sic0048]

Loaded .54. Still cannot connect. I've sent another log. Here is the CQC log where you can see that the connection is attempted, but errors out for some reason.
http://i185.photobucket.com/albums/x46/sic0048/RIVAerror1.png

Also, the directions from the receipt said to download the app from the web site. I don't get a good file when I do this (ie it cannot be opened and installed).

[SamVimes2]

.54 shows [320x533] ie that method has the same problem.

I am also seeing lots of errors in the log, though it does connect:
http://img709.imageshack.us/img709/2972/tarivaerrors.png (http://img709.imageshack.us/i/tarivaerrors.png/)

[Fonceur]

And the window/iphone RIVA clients don't generate those errors?

EDIT: Actually, it might just be that I'm sending a LOGOFF message to it on disconnect and the image thread doesn't like it...

[SamVimes2]

Quick google found this, looks like you need to explicitly support multiple resolutions , or Android will get smart and try and scale things for you:

http://developer.android.com/guide/practices/screens_support.html
Auto-scaling of pixel dimensions and coordinates
For instance, suppose a given device is using a WVGA high-denisty screen, which is 480x800 and about the same size as a traditional HVGA screen, but it's running an app that states that it does not support multiple densities. In this case, the system will "lie" to the application when it queries for screen dimensions, and report 320x533. Then, when the app does drawing operations, such as invalidating the rectangle from (10,10) to (100, 100), the system will likewise automatically transform the coordinates by scaling them the appropriate amount, and actually invalidate the region (15,15) to (150, 150). The same thing happens in the other direction, if the application is running on a lower-density screen, coordinates are scaled down.

[Fonceur]

Interesting, so the hardware is lying to the software... ;) The only thing is that I was targeting Android 1.5 so it would work on the cheap Chinese tablets, while multiple resolutions was only introduced in 1.6.

So I will bump the Android OS requirement to 1.6, unless someone really wants/needs 1.5...

[Fonceur]

Version 0.55

- Bumped the requirement to Android OS 1.6.
- Improved the support for higher resolution screens.

[Fonceur]

Version 0.56

- Fix: Removed the logoff message from the image download thread.
- Added a higher resolution default image.

[batwater]

First, Fonceur, excellent work on the Android client for CQC!!!

For those that haven't done so, DropBox is by far the easiest way to get updates to the taRIVA and taSageTV clients on your phone!

Patiently waiting for key to arrive... :-)

Again, thank you for your hard work!

-Ben

[Fonceur]

Version 0.57

- Fix: Did not properly handle the position when templates were shorter than the screen.

[Fonceur]

Version 0.58

- Can select the delay to wait for images from the Information screen (Registered user).


By default, each image to be drawn is given one second to finish downloading (checked every 100 ms), from the previous drawing command, then the next command is processed. With small images and a good signal, this is plenty and of course once the images are in the cache, then it's not an issue. On a bad signal, it might not be enough, so you would end up with missing images (they do get fully downloaded, just not in time to be displayed). Now Registered users can adjust that delay between 0 and 10 seconds... So if you want to use large images for your media art browser and don't mind waiting 10 seconds per image, now you can... ;)

[Fonceur]

Version 0.59

- Can select the Touch delay from the Information screen (Registered user).


Basically this parameter determines if a touch should be sent to the CQC RIVA server or if it was meant for taRIVA. It also prevents accidental button click when swiping.

By default it is set for 300 ms, so if the initial touch last for less than 300 ms, it is assumed that you either didn't really mean it or if it is followed by a move, that you wanted to perform a swipe.

If you press for more than 300 ms, on release the Press/Release will be sent, or if you started to move after that initial 300 ms, the Move command will be sent to the CQC RIVA Server (i.e. using a slider).

[SamVimes2]

Hi Fonceur,

I paid to register on the Tallus site; how do I get my key?

Using .59 (i guess in unregistered mode?) I no longer have the shakes, but smaller templates are being 'shrunk'. Also I still do not get all images and have tons of exceptions in the image download thread in the CQC logs.

[Fonceur]

I paid to register on the Tallus site; how do I get my key?
If you included the Phone's Android ID, you should be receiving it by email tonight (Tallus' owner seems busy with work lately).

but smaller templates are being 'shrunk'.
Are they really shrunk or just using the proper portion of the screen? That is, a 320x480 template won't fill a 480x800 screen...

Also I still do not get all images
Are those big images, which actually show up the next time the template is loaded?

and have tons of exceptions in the image download thread in the CQC logs.
Right, probably because I was sending a command saying I'd request more images, and then closed the connection... ;)

[SamVimes2]

-Where was I to include the Android ID? I don't believe I did.
-They are really shrunk. They start out the right proportion of the screen, then shrink on first touch.
-They are not necessarily big images, and they do not show up next time.
-Are you suggesting this behavior is by design?

[Fonceur]

-Where was I to include the Android ID? I don't believe I did.
From the wiki: "Purchase TARiva Here and make sure to email your Android's ID (after you install the application on the Android device and go to the Registration window) for licensing. " I think you are supposed to include it in the "Special Instructions" section.

-They are really shrunk. They start out the right proportion of the screen, then shrink on first touch.
Hmm, might not be using the right size somewhere.

-They are not necessarily big images, and they do not show up next time.
I guess it might have been due to the image thread issue...

-Are you suggesting this behavior is by design?
Which part?

[Fonceur]

Version 0.60

- Fix: The image connection wrongly claimed more requests would be made.
- Added a sound when a Press last longer than the "Touch delay".

[batwater]

Hi Fonceur,

First I really like the way the latest changes are heading, thank you! ;)

Button press vs swipe vs slider:
Here is the challenge I am having, with the latest changes it has become even more difficult to register a button press. What I believe is happening is that a "swipe" movement is being picked up by the client even though I am not trying to swipe but simply push a button thus a move (slider) command is being sent back to the server. So the challenge as I see it is how to clearly identify a swipe vs a button press vs a slide . As it stands I am getting maybe 5% response on button presses. I understand the time delay behavior its just not translating on my phone. Perhaps for slide the action can be a double tap and then a slide; thinking out loud here...

I have not found anywhere where I can adjust swipe sensitivity, nor am I inclined to change it as everything else on the phone works without any issue.

Audio feedback:
When I first exceeded the press threshold the beep sound that issued from my phone sounded like my alarm was having a fault, it was kinda funny, didn't make the connection right away. Later I had set the time delay to 0 as a test for the issue discussed above so every screen tap caused the same high pitched beep beep. As I said it was kind of funny up until my wife explained to me that she would hurt me if I didn't stop whatever was causing the noise. Guess it irritated her. :shock: (I turned the media volume down rather than risk her ire) Can you make Haptic feedback an option instead of sound, much less intrusive?

Thanks!
-Ben

[Fonceur]

Here is the challenge I am having, with the latest changes it has become even more difficult to register a button press.
Hmm, maybe you're shaking too much and should cut down on the coffee? ;) I guess I need to ignore small initial movements...

Audio feedback
I just grabbed some "free Android sound", I couldn't find a click one... ;)

As I said it was kind of funny up until my wife explained to me that she would hurt me if I didn't stop whatever was causing the noise.
Maybe you have your sounds too high... ;)

[Fonceur]

Version 0.61

- Replaced the Press sound by the Haptic feedback of a keyboard tap.
- Initial movements of less than 10 pixels are ignored.


While you can fine tune the end of a movement, if all you want is a small movement, you might have to go passed the point and back...

[Fonceur]

Version 0.62

- Can select the Touch sensitivity from the Information screen (Registered user).
- Removed the debugging information about incoming/outgoing messages.

[batwater]

Hi Fonceur,

Hmm, maybe you're shaking too much and should cut down on the coffee? ;) I guess I need to ignore small initial movements...
No can do on the coffee, actually there was blood in my coffee stream, fixed that later... Thanks for the change!

Maybe you have your sounds too high... ;)
Yes it was

Haptic:
I am not able to trigger a haptic response at all. (it's turned on for the phone)

Button behavior:
After the 1st button press and subsequent expectged overlay load When I press a button for a different overlay I get the same overlay again (please wait dialog shows.) A 2nd press will the overlay associated with that button will load.

Button press delay:
What affects the response time for a button press to register as a button press. From a UI responsiveness perspective it seems far longer then how Android behaves directly in the UI; perhaps others could weigh in on their experience with this? I know we must find a happy balance or can this too be adjustable?

Wiki has definitions of each of the parameters!

Thanks!!!
-Ben

[Fonceur]

I am not able to trigger a haptic response at all.
I guess I'll need to make a version where you can test the various one...

After the 1st button press and subsequent expectged overlay load When I press a button for a different overlay I get the same overlay again (please wait dialog shows.) A 2nd press will the overlay associated with that button will load.
Maybe there's a problem and it uses the previous one or something odd like that.

What affects the response time for a button press to register as a button press.
That's the Touch delay parameter...

[sic0048]

Alright, I've gotten my RIVA issues sorted to where I can connect again via the android client, but my templates don't seem to work as others as.

Here is what the template looks like when I first load it (Fonceur's templates). There are not images being loaded and this is over my LAN connection at home, so speed is not an issue here.

http://i185.photobucket.com/albums/x46/sic0048/Riva_64_1.jpg

If I press on the screen, it gets even worse with the screen shrinking even smaller.....

http://i185.photobucket.com/albums/x46/sic0048/Riva_64_2.jpg

This is on a DroidX running Froyo (2.2). Under the information screen, it shows this.
http://i185.photobucket.com/albums/x46/sic0048/Riva_64_3.jpg

[Fonceur]

Can you restart the app to get on the first image, hit menu and information, hopefully that won't be rescaling just yet... Also if you could send me a log please. The emulator won't allow me to try those higher resolution, as it has hardcoded limited available memory for the video buffer.

[sic0048]

Can you restart the app to get on the first image, hit menu and information, hopefully that won't be rescaling just yet....

It says resolution = 480x854
[480x854]
(384x544)

[Fonceur]

Version 0.63a

- Fix: Haptic feedback didn't work right...

[Fonceur]

It says resolution = 480x854
Hmm, not sure what is happening then, so I'll really need that log...

Go put a wrong username, so you only get the default image. Close the app, restart it and see if it behaves normally unconnected, or you still get that resizing.

[batwater]

Haptic feedback now working, thanks.

I have started to formalize my alarm overlay template (#16, alarm button) I am getting a line between the 2 adjacent, butted up next to each other, text widgets that I do not get on the IV or on iRIVA. On both the 2 show up as a single green rectangle.

-Ben

[sic0048]

Hmm, not sure what is happening then, so I'll really need that log...

Go put a wrong username, so you only get the default image. Close the app, restart it and see if it behaves normally unconnected, or you still get that resizing.

Did this and the last resolution shows as (697x414).

I included a couple more screen shots. This is how the initial page has always looked. It doesn't fill the screen properly in portrait mode and is a little smaller than the screen in landscape mode.

http://i185.photobucket.com/albums/x46/sic0048/LoadScreen1.jpg

It does get resized oddly if you press the screen.
http://i185.photobucket.com/albums/x46/sic0048/LoadScreen3.jpg

Log sent too.

[Fonceur]

Haptic feedback now working, thanks.
Actually, I did a double tap on that one... There's the Haptic feedback for a long press and there's a 100 ms vibrate, can you feel both? ;)

On both the 2 show up as a single green rectangle.
I might have cut some things to 1 less pixel than they should, has it seemed to end up with a size +1...

[batwater]

Actually, I did a double tap on that one... There's the Haptic feedback for a long press and there's a 100 ms vibrate, can you feel both? ;)


I might have cut some things to 1 less pixel than they should, has it seemed to end up with a size +1...

Im not feeling the 100 ms press.

-Ben

[Fonceur]

It does get resized oddly if you press the screen.
OK, that is something I can test with the emulator, but it's not acting like that, so it's definitely something related to the hardware. It must be switching from zoomed 320x544 to 480x800 at some point when you press.

[batwater]

Im not feeling the 100 ms press.


.63b corrected the 100ms feedback. This method works very well to provide notification that a button has been pressed. Excellent!

-Ben

[znelbok]

Sorry to ask a question that will be answered here already, but there are too many posts.

I installed the lastest this morning (an upgrade) and it does not work. I only get the splash screen.

I did briefly see something about deleting a profile or something to that effect. Can someone please elaborate on what I need to do

Thanks

Mick

[Fonceur]

Can someone please elaborate on what I need to do
Go back to the configuration and make sure it is still valid...