Not much text here. Just a video.

[matzesu]

Not realy, this "Pyra Phone" was just a Updatet PCB for the "Openmoko" Phone, Nikolaus made allready a new Board for the Openmoko who was the GTA 05

 

[CommanderB]

Not realy, this "Pyra Phone" was just a Updatet PCB for the "Openmoko" Phone, Nikolaus made allready a new Board for the Openmoko who was the GTA 05

No, if I understood correctly, the PyraPhone is going to be a completely new design and not based on the GTA04.

I wonder, whether ED is going to create a plastic case for it. I can imagine that it is non-trivial to design the necessary molds for it, if one takes the case problems of the Pyra as a reference

 

[TrashyMG]

I wonder, whether ED is going to create a plastic case for it. I can imagine that it is non-trivial to design the necessary molds for it, if one takes the case problems of the Pyra as a reference

Well ED is dealing with a completely different company now for the cases, the new people seem to actually enjoy doing the work.

But I suppose case manufacturing on this phone will depend on how many people actually want the PyraPhone.

 

[Coupo]

That's nice!

Personally I don't need videos. Text is just fine.

Keep up the great work

 

[Linux-SWAT]

Have you been following the preorder thread? People have been cancelling their preorders. I think outside the truely hardcore, yes, people do need convincing that the wait is almost over.

http://p.rdw.se/pyra-preorder.html

 

[PowerGod]

Videos are really needed only for HELLOLO tests !!

 

[pmprog]

http://p.rdw.se/pyra-preorder.html

So, yes, your are aware than more people are cancelling than signing up, and the trend is going the wrong direction.
I don't want to sound negative, but I'd hate to see all this time and money be for nothing if releases don't start happening before the cancellations get more and more.

 

[TrashyMG]

Since EvilDragon is handling the cancellations, he should be fully aware of it.

 

[EvilDragon]

Have you been following the preorder thread? People have been cancelling their preorders. I think outside the truely hardcore, yes, people do need convincing that the wait is almost over.

Well, most people had reasons like urgently needing the money, some even asked if they can revive their preorder later on when they have more money back. It's actually not that many who are cancelling, which is good

 

[pmprog]

Well, most people had reasons like urgently needing the money, some even asked if they can revive their preorder later on when they have more money back. It's actually not that many who are cancelling, which is good

Glad to hear!

 

[EvilDragon]

Some more info about the tests:

As mentioned the failing TWL and 4G/LTE test is indeed a problem in the software setup, as it fails on all tested units.

Two of them won't boot (one has memory issues, one can't read from the SD Card), which could be a missing capacitor or other simple thing, but we'll check that.
The non-working keys seems to be the main issue right now: 50% of the tested units have at least one non-working key. However, as bad as this may sound, it's probably just a parameter that needs to be fixed during production:
Most of the times, it's the "," key that fails, so we have similarities. And more importantly: Only ONE key fails - even though we're using a matrix. If a row or column of a Matrix fails, this would lead to ALL of the keys connected in that row or column to not work - but this is not the case. So the connection of the key to the keyboard controller is fine.
All keys have an anti-shadow diode though - and if that one is missing or polarity reversed, that one key will not work.

This means: Most likely, the place for the population of the diodes isn't 100% accurate in the machine program right now, so sometimes a diode is misplaced, especially the "," one.
This also means: all we need to do on the prototype boards to make them fully work is fix the diode.

This has to be confirmed - but there's nothing else that could possibly cause this. And this is exactly what such a prototype run is for: Find out which parts cause issues in population. It's not always the complex parts like the SoC, but sometimes even simple parts like the diode.

I'll keep you updated

 

[inkling]

text is my preference, but videos are nice.

 

[Linux-SWAT]

We don't need the ",".

 

[EvilDragon]

We don't need the ",".

Oh, you really do!

The "/" is on there as well, which renders the terminal pretty much useless

 

[Djoga'Ro]

We don't need the ",".

You only programm in Whitespace, or something? ;P

 

[zoranc]

text is my preference, but videos are nice.

Same here, so do whatever suits you better.

Quarter of a content, but twice as often would be also preferable for me.

 

[CommanderB]

Some more info about the tests:

As mentioned the failing TWL and 4G/LTE test is indeed a problem in the software setup, as it fails on all tested units.

Two of them won't boot (one has memory issues, one can't read from the SD Card), which could be a missing capacitor or other simple thing, but we'll check that.
The non-working keys seems to be the main issue right now: 50% of the tested units have at least one non-working key. However, as bad as this may sound, it's probably just a parameter that needs to be fixed during production:
Most of the times, it's the "," key that fails, so we have similarities. And more importantly: Only ONE key fails - even though we're using a matrix. If a row or column of a Matrix fails, this would lead to ALL of the keys connected in that row or column to not work - but this is not the case. So the connection of the key to the keyboard controller is fine.
All keys have an anti-shadow diode though - and if that one is missing or polarity reversed, that one key will not work.

This means: Most likely, the place for the population of the diodes isn't 100% accurate in the machine program right now, so sometimes a diode is misplaced, especially the "," one.
This also means: all we need to do on the prototype boards to make them fully work is fix the diode.

This has to be confirmed - but there's nothing else that could possibly cause this. And this is exactly what such a prototype run is for: Find out which parts cause issues in population. It's not always the complex parts like the SoC, but sometimes even simple parts like the diode.

I'll keep you updated

Is the test-tool open source? If yes, the community could provide some patches to help you speed up the development

 

[docbroke]

Well, most people had reasons like urgently needing the money, some even asked if they can revive their preorder later on when they have more money back. It's actually not that many who are cancelling, which is good

That's some optimism. This trend will change to positve again once pyra will released, but till that it may even get worse.

 

[CommanderB]

That's some optimism. This trend will change to positve again once pyra will released, but till that it may even get worse.

I am not even sure, if ED wants more pre-orders at this point. At the moment, only 500 Pyras can be produced per batch. So before the production can be scaled for many pre-orders the initial ones have to be fully paid and delivered.

Also more pre-orders would increase pressure on ED which is counterproductive in the current development state with its possible set backs.

 

[EvilDragon]

Is the test-tool open source? If yes, the community could provide some patches to help you speed up the development

Is that a question?
Of course it is

https://dev.pyra-handheld.com/packages/pyra-scripts/tree/master/scripts

The init-hwtest.sh is the one that's being run, mostly using hw-test inside the Letux-Directory (which is from hns).