The long road to a final unit


*I* can see image change from resistive layers, *I* notice them, but most people probably will not
Well that's just it, can you really? Or do you just think that you can? A rigorous test would involve taking several different screens (resistive, capacitive, and no touch screen) and then asking you to guess, just by looking at it, which was which. In a perfect double blind test there will be no outward indication, all the LCDs will be exactly identical, displaying exactly the same image, the only difference will be in the coating put on top of them.This hearkens back to a lot of things in technology that people claim they experience but completely breaks down when investigated, people just form expectations. Are you sure you can actually tell the difference between resistive and capacitive and you aren't just seeing the difference between LCD screens or glass coated vs plastic?
 
That's a good point actually (and one Illuminerd alluded to in his last post to be fair). A glossy glass screen is very different to a matte plastic one, and the former might actually be better in out-of-doors environments, as while it will produce a lot of obscuring reflections if pointed so as to reflect the sun in your face, at any other angle it will pick up less ambient light than a matte screen illuminated by the sun from anywhere in front of it. So a glossy screen might actually be more usable out-of-doors than a matte one is (that said, I've had mobile phones with matte resistive screens that are perfectly visible out of doors, and ones with non-touchscreen glass fronts that disappear completely during the day outdoors, so perhaps it's not that decisive a difference to visibility after all).


As to which screen type you prefer indoors, matte versus glossy becomes more about taste than anything else (unless your room happens to be full of distracting reflections).
 
Can TI's rootfs for android be run from an SD card and booted manually? I would like that. Most of the time I want to use the regualr Linux though.
 
Why have a really nice screen and really nice keyboard and game controls, if you're then going to put an OS on it that assumes your only input method is ten sausages? Why do people so badly want to cover their screen with finger grease?
Actually, Android supports just about every input method you could imagine.  It even supports a mouse perfectly and just assumes a left click is the same as a finger tap on a screen spot, so I don't think it's fair to criticize Android in that regard.  It also supports full keyboard and HID interfaces, including a dedicated (and standard) game controller layout, so games can automatically detect what inputs are available and optimize based on that - by presenting onscreen controls if no physical ones are available, for example.
I know it supports a lot. But like I said, it assumes and it requires only fingers. So any Android app is required to have an interface that works with fingers only. That's also how the vast majority of Android users use their device; very few people actually connect game controllers or full keyboards to their smartphone or tablet. So typically, the UI of Android apps is designed for touch, with physical input only as an optional afterthought for advanced users. If you're out of luck you can't even get rid of the onscreen controls.

Also, lots of Android apps look best in the portrait screen orientation since that's the normal orientation for smartphones. Some games even refuse to rotate to landscape. That's also rather annoying when you're using it in a landscape clamshell.
 
Why have a really nice screen and really nice keyboard and game controls, if you're then going to put an OS on it that assumes your only input method is ten sausages? Why do people so badly want to cover their screen with finger grease?
Actually, Android supports just about every input method you could imagine.  It even supports a mouse perfectly and just assumes a left click is the same as a finger tap on a screen spot, so I don't think it's fair to criticize Android in that regard.  It also supports full keyboard and HID interfaces, including a dedicated (and standard) game controller layout, so games can automatically detect what inputs are available and optimize based on that - by presenting onscreen controls if no physical ones are available, for example.
I know it supports a lot. But like I said, it assumes and it requires only fingers. So any Android app is required to have an interface that works with fingers only. That's also how the vast majority of Android users use their device; very few people actually connect game controllers or full keyboards to their smartphone or tablet. So typically, the UI of Android apps is designed for touch, with physical input only as an optional afterthought for advanced users. If you're out of luck you can't even get rid of the onscreen controls.

Also, lots of Android apps look best in the portrait screen orientation since that's the normal orientation for smartphones. Some games even refuse to rotate to landscape. That's also rather annoying when you're using it in a landscape clamshell.
Excuse me but i believe you forget devices like the new nvidia shield tablet w/ controller and the older nvidia shield. ..Also there are many third party controllers for android tablets and very popular. Something is lacking with many games using touch so these devices suffice and are getting very plentiful.
 
The shield gets around the landscape/portrait problem in the vast vast majority of apps.


The menu system framework is designed around large buttons that are touch friendly by default, but you can change every aspect of that framework to include the size of those options. Also by default the dpad can navigate those lists and select one of the action buttons.


But because its open source many "theme" makers do change the size and style of that framework. When you change it once it changes it for every app that uses that framework. Moreover using the system GUI frameworks for menus isn't required for app developers, but its use is recommended and used often because it's main market is always going to be touchscreen. Games that target controllers usually have their own custom menu system.


As said there is mouse, keyboard, controller, and stylus support in addition to touchscreen support in android. But the interface was designed to be optimized for touchscreen. Having used the Pandora and a Samsung note phone both with a stylus, I find the note to be more natural to use the stylus with because the interface is more touchscreen friendly, be it sausage fingers or precise pen doing it. "Needing" to use a stylus vs having the option to use the stylus.


There are also xserver projects already on android. I feel if someone put some work into it you could have a standard desktop manager in addition to the standard android interface and would open it up to a mountain of app port options.
 
Well Touchscreen aside, I don't like the bloat of Android nor would I like developing programs for it.
 
New Tweet from ED: "Working on the physical design of Pyras LCD Cable"

Bxz7zVtIQAAIsp3.jpg


Looks like a ribbon cable, I thought they went away from that (fragile) design and use a normal round standard cable to connect the screen with the mainboard?

https://twitter.com/EvilDragon1717/statuses/512561317497286656
 
Last edited by a moderator:
It's not a quick port - the full UI would need to be rewritten for Android... which would be the case for all apps using standard GTK, etc.

You can still run Android on the Pyra (TI offers a working rootfs for it), but it's not the main customer base and therefore not officially supported by us.
I don't want to run Android as my primary OS on the Pyra. I want to run two sessions of Android in independent Virtual Machines (work and play) on the Pyra in the background under Debian. Just like on an X86 VM environment I want the host (Debian Pyra) to handle the network bridging and resource allocations and a common clipboard to move data back and forth over environments. It's good to dream.

But Android as the primary OS on this device? No thanks.
 
I thought they went away from that (fragile) design and use a normal round standard cable to connect the screen with the mainboard?
I'm pretty sure flat ribbon cables are the standard. I don't think there's enough flexibility in the thicker wired kind, or maybe they just don't fit.
 
Looks like a ribbon cable, I thought they went away from that (fragile) design and use a normal round standard cable to connect the screen with the mainboard?

https://twitter.com/EvilDragon1717/statuses/512561317497286656
Ugh. Will this cable also use the "spiral of death" trick to go thru the hinge? I strongly suspect that the "spiral of death" is the cause of the LCD cable issues on pandora.
It's kinda hard to explain,but I think one edge of the cable ends up rubbing against another edge, causing a sort of band saw effect, cutting into the cable.
 
Last edited by a moderator:
I thought they went away from that (fragile) design and use a normal round standard cable to connect the screen with the mainboard?
I'm pretty sure flat ribbon cables are the standard. I don't think there's enough flexibility in the thicker wired kind, or maybe they just don't fit.
I'd just as soon they cut a slot in the top/bottom case parts and run a flat cable through it. Lenovo did this on most of it's Thinkpad line without complaints - and those have a LOT more wires. Simply get the cable out of the hinge. Then it can be made of thicker and better protected stuff.
 
After the traumatic CC Pandora ribbon cable deaths and countless hours of fixing these, I was sure ED went away from that design. He mentioned the other solution once, meaning a normal cable that goes from the mainboard to the LCD, supported by an extra circuit that also sits in the lid together with the screen.

Seems like the plans have changed. I find the ribbon cable solution "cheap" but let's hope ED knows what he is doing. I am still worried and scared every time I open the lid onto my Pandora and I don't want the same feeling when opening a Pyra. I'd like to see a strong, durable solution for the upcoming Handheld.
 
Last edited by a moderator:
I just watched the LCD cable replacement video again.  I thought the twist distributed the "folding" over more area of the cable, and was less destructive than without a twist.  Since I don't have as much time with my Pandora as many of you, please correct me if I am wrong.  Are there physical signs of the cable cutting into itself?  I have never been one to avoid opening my devices, and it seems many here are the same.  With all the horror stories related to the LCD cable, the screws, the mic, and anything I am forgetting, I am entertaining the idea of sending my Pandora home for the aluminum case change to avoid destroying it.

I'd just as soon they cut a slot in the top/bottom case parts and run a flat cable through it. Lenovo did this on most of it's Thinkpad line without complaints - and those have a LOT more wires. Simply get the cable out of the hinge. Then it can be made of thicker and better protected stuff.
If this would solve the problem I am ok with it.  As long as it isn't too fragile, I wouldn't mind being able to see some of what makes it work (clear case!).

I think no matter what method is used for the Pyra LCD cable we won't know what will happen to it without a lot of testing, or many units in users' hands.
 
I agree that we need another solution for the LCD connection. Hopefully with the MIPI interface (fewer electrical signals) and no speakers in the lid it should be easier to design something robust.

I already had to replace the ribbon in my rebirth Pandora when after two years of use it got simply cut into by the curved edge of the ribbon rubbing against the straight edge (exactly like Yoyobuae said).
 
Back
Top