Re: The new QWERTY device project
 

Completely agree ...
Too many devices have crap in them that are simply way overboard.

In some part...gerbick also hit on a part of the issue nowadays...
At least in the USA ...that is...
I read a report that desktop laptop and tablet sales the last few years were dropping away...
Yet smartphone sales were climbing...
With more and more people opting for handheld computing..camera.. and well ...pretty much everything...in a phone...
It makes for a more portable life ..in uncertain economic times I suppose.

Re: The new QWERTY device project
 

It certainly is! .... How does this relate to the keyboard phone? Will it be based on a SFOS port that has a ConsumerIR device?

Re: The new QWERTY device project
 

I'm not 100% certain on this but as far as I'm aware, TI are still making OMAP SoCs. It's true that they're no longer designing new chips but I believe they are still producing old ones.

The last SoC that TI put out was the OMAP5. It may not be the latest and greatest but it's several years newer than the OMAP3 from 2009 that you refer to and it is still a very capable chip. OMAP5 was chosen for the SoC to be used in the DragonBox Pyra.

I can think of a few reasons why OMAP is still relevant in 2017:

• It has far better mainline Linux support than any other chip used in smart phones.
• It does not bundle a modem on the SoC. Some may see this as a disadvantage for reasons such as cost or the physical board space taken up by separating modem but privacy advocates will tell you that it is better design to separate them. It's very easy to hide a backdoor on an SoC that bundles a modem that could give some external entity full control of the device.
• Its availability for smaller projects. The Pyra team evaluated several SoCs but it turned that OMAP was the only one they were able to use.

Re: The new QWERTY device project
 

The time you argue about architecture ... We just waiting our new... Precious...

Re: The new QWERTY device project
 

Yeah I'm with you here. I'd rather have a more modest screen resolution and much better battery life.

Also, with a modest resolution you don't need such a good graphics card/driver to give smooth performance - it's one of the things i'm slightly worried about with the Youyota tablet since people have said they had problems on the Jolla tablet.

Re: The new QWERTY device project
 

I'm in a plane. There's free Wifi. I'm too lazy to get up and fetch my laptop out of my carry-on bag. You can't imagine how much I'd like to have a keyboard on my Jolla right now. Please make it happen Chen, so that I can stay lazy and still enjoy the full internet experience forever.

[Edit] Regarding the last page, there is no optics/Mpix trade-off, except money wise. The reason optics can't be good on smartphones is not because of the Mpix race, that race is just a way to conceal the fact that a smartphone will never be better than decent. Yet, there are benefits for higher resolutions too, until a certain point. While the race has been ridiculous at times for cameras and other smartphones, we're not that far with Chen's device and the chosen resolution seems pretty reasonable by today's standard.

Same goes for the horsepower and the screen resolution. Pixels eat battery, and high resolution screens are not critical on small devices, but they are not useless and benefits are not unnoticeable. CPU power is useful when applications all require more of it than 5 years ago. Websites are heavier. Users' patience and expectations of responsiveness and smoothness have changed. CPU are also more power efficient than they used to be.

Maybe the only thing that is not a matter of opinions and differences among users, this phone needs to run the races it can or it won't sell, and won't exist. It's already skipping the cheese-slicing and mainstream races, there's not much room left for playing underground and be different if we want it to even happen. All these choices seem pretty reasonable, and again, how many of us like the Neo900 and look forward to it, but are still a bit concerned about whether the hardware will be enough when the phone ships (I can't wait to have mine and I understand the hardware choices, but I'd be lying if I'd say I wouldn't want more horsepower now that 4 years have passed)?

Re: The new QWERTY device project
 

Well. Airplane wifi is not a full internet experience as far as I have seen. It's usually pretty bad. What flight? The best thing with these flights are that no one tells you you can't watch video while departure or arrival :D

Chen, since you expect the device to be ready before new year, will the campaign start in sept or oct.

Re: The new QWERTY device project
 

I guess it's the purpose of this whole thing - so that people can't see the pixels. Because people believe that it's beautiful not to see the pixels. Because that's what Apple made with their Retina display. And whatever Apple does is, by definition, as an axiom or a dogma - beautiful and wonderful, and whatever you want to sell, shall you do as Apple does.

Re: The new QWERTY device project
 

When I used my VR headset on my old 5.5", FHD, android phone, I wished the screens resolution was higher.

Again, when using a VR/Google daydream-headset, a high resolution screen is necessary, if you want to minimize/get rid of the visible pixels. For focusing there are lenses in front of the screen, which enlarge the pixels by a lot. To combat this, one just have to pump up the resolution.
A higher resolution logically means a higher load on the GPU, but there is a way to combat this. You don't have to run everything at the screens native resolution. One can browse the web just fine at 720p on their native 4K screen. Want to watch videos in its native resolution? Just change it yourself or let the OS change it automatically, based on what app you are running.
I think cyanogenmod is/was capable of that.
Samsung devices have that option too.

I agree. For everyday usage, it's more than enough. For special uses, a higher resolution screen is "necessary"/more than welcome.

Higher res. screens don't even need to be more expensive than lower resolution screens. The product that is available the most is the cheapest and the one that isn't, is more expensive.
Supply and demand.
Chens slider device is a good example of that. Chen has to pick a easily obtainable 5.5" fhd screen, because smaller screens are more expensive.


See, there is nothing wrong with high resolution screens.
The only thing wrong (currently) with them, is their implementation.
If Chen was to use a 4K screen in his 2nd slider device, you should be able to run it at 360p and save battery, but probably have to scale the OS/apps accordingly, to not get massive icons, buttons etc..

Re: The new QWERTY device project
 

Yeah, that's not how it works. yes running the 4K screen on the XZ Premium at 1080p does save (a noticeable amount of) battery, but any lower and you're looking at a few percent better, at best, which you wouldn't notice compared to it being in 1080p mode.

Re: The new QWERTY device project
 

I see. I didn't know that.

I wonder, how much more/less energy a 5.5" 4K display/device @1080/360p, consumes compared to a native 5.5" 1080/360p display/device.

If there is no noticeable difference, then the screen size is the factor that kills battery life. Resolution is "irrelevant" then.
Maybe, the GPU's simply can't clock slower and consume less energy, even if you go further down with the resolution.

Re: The new QWERTY device project
 

Regarding the screen resolution and pixel density...

My last long-term phone before my N900 was Treo 600. (I say long-term because I tried about three different phones that I did not like before I found the N900, none of them for longer than a few months.)

Treo 600 has a whopping 160x160 pixels, 2.5" square screen. And it was ABSOLUTELY BLOODY FANTASTIC! I cannot find the official spec but by my calculations, it works out at 90 pixels per inch. Yes, the pixels were clearly visible. No, you probably could not watch a video on it but it did not matter since the CPU was not up to the task anyway.

But the text was so fantastically readable! I only came to fully appreciate how good a low-res screen is when I switched to the N900 and suddenly I could not read my SMSes without modifying the style sheet and doubling the font size.

Which is to say, juiceme is absolutely right. A higher screen resolution does not have to mean smaller text and controls but it usually does. The temptation is too strong: we have all these pixels, it would be a shame not to use them. In the most crucial places that I use daily, such as SMS, where I have no alternative but to use the stock app that came with the device.

I have yet to see a counter argument. If anything, Jolla has made the text even less readable, with an even smaller font and an unfortunate choice of colours, only proving my point.

There is absolutely no reason for the pixel war other than to fool the gullible. If you can see a difference between 300 and 450 DPI then you must be a superman. The flip side, the tendency to use the same 12 pt font at a higher DPI, resulting in an unusable device, is just not worth it.

Re: The new QWERTY device project
 

These points are definitely subjective and depend on users' habits or even eyesight. I would never increase the font size in Sailfish Messages app. Even if it was easy (which it is, by the way, Sailfish comes with three font sizes configurable in Settings and applied system wide). I do understand that some would, but I would not, so I don't think downgrading the resolution from hardware would be satisfactory to everyone and should be the norm. Better to offer users capabilities that they can leave or take.

Re: The new QWERTY device project
 

This is one of those "back in days they made it better!"-things... :)

You can make screen with lower resolution readable if you set the font correctly for that screen. However, it will limit how much you can fit into that screen in a way that it still remains readable (not to mention pleasant to read) and this is where higher resolution screens usually work better. It's the same thing with projectors, low resolution image is usually mushier to read even from the distance you can't really notice individual pixels and throwing in a full HD projector increases readability especially on smaller items.

Even Master himself is for higher resolution... :)

Re: The new QWERTY device project
 

In a way, they did. Because they had to. They did not have octacores and 16GB RAM so they had to optimize. Jolla with 1GB struggles to cope with keeping more than 3 apps open, something that N900 does effortlessly with a quarter of that.

In terms of screen resolution, I appreciate that my post was a bit long, but please try to understand that I am not against a high resolution per se, only against poor optimization. Unfortunately, as is the case with pretty much everything in life, more resources are often interpreted as no need to optimize.

But that is exactly the point! Howe much you can comfortably fit into the screen is not a function of the screen resolution, but the screen size. If you cram more things onto the same 4" screen only because the higher resolution allows you to, you will end up with an unusable interface.

Now turn it around. If you cannot squeeze more things into the screen despite having a higher resolution, then what is the point of the higher resolution?

Actually, what I would advocate is a constant pixel density, independent of the screen size.

Re: The new QWERTY device project
 

Agreed.

What we are seeing is that ...
since the always climbing increase in creation and production of devices with higher and higher and greater and greater maximums ..
The natural given is that there is a wastefulness ...
why optimize anything when cpu , ram, etc are always getting bigger and better?
forget optimization in fact...
do the the reverse in fact...
Bog the systems down as much as possible...
and that spurs faster obsolescence and keeps pace with the latest innovation, the latest increases in ability and the latest devices out.

Flat out..
we will not see honest and truly serious and innovative use and optimization of any device until they begin to hit "the Wall" of what is attainable...
THEN watch how fast the modern devices shuck the useless crap that bog them down.
it has happened before.. when tech has slowed in it's innovative output..
and it will happen again..

right now we are just going through a phase no different than many other areas...
the "Bigger is Better" philo always runs its cyclic course...
and then sanity kicks back in again...

Re: The new QWERTY device project
 

...and that also limited features (both functional and security ones).

Of course it is. As I mentioned previously, lower optimization is logical result from faster release cycle. If you are going to optimize in a way that squeezes everything out from limited set of resources (kind of what was done with Symbian), it will take resources (people, time, money) and one has to decide what's the best way to implement features X, Y and Z. Are you going to a) add in resources (money, people), b) add time/push back development of other features/optimizations or c) take advantage of additional processing power etc of the device. At the same time you have to take into consideration the competition which is pushing out new features at very quick pace. This was one thing that effectively killed Nokia mobile phones. Android back in those days was far from being optimized.

Point is, if you want to have both ultimate optimization AND the features to match competition, you will need considerable amount of money and people to do it. Otherwise you are screwed in no time. Jolla has been lamented about poor optimization (which is in many ways true of course), at the same time they have been lamented about lack of features X, Y, Z.... considering their resources those are conflicting demands, you cannot have both highly optimized code AND feature sets WITHOUT considerable financial and human resources OR time to implement it all. Sorry, it's just not doable.

I'd say it's combination of both screen size AND screen resolution, as smaller objects on higher resolution screens are more clear and pleasant to look than on lower resolution screen at the same size (in general, not talking about moving from something like Full HD to 4k on 5" display).

Re: The new QWERTY device project
 

Skimping on features or security due to limited resources is not "optimize" but "compromise".
What I mean is real optimization - the same features using less resources.

What we see instead is adding extra features that no one has asked for because the resources allow it. For example, the faded background in many stock Jolla applications (Email, Messages, Phone, People...). They add no value and more often than not, get in the way of usability. You may call them "features", I call them "bloat". You say, optimization costs money. I say that may be so, but so does bloat.

Agreed. But even you concede that there is a limit beyond which the usefulness is questionable. For me, the limit is at the point where you can no longer see the individual pixels. Even the N900 or Jolla screens are already at that point.

Re: The new QWERTY device project
 

There is a limit how far you can get with optimizing until you reach the point where implementing feature X with those limited resources becomes too heavy for HW to cope with (either it doesn't work, is unstable or it's so slow that nobody wants to use it any longer) if you don't run out of (other peoples) money before that. On the other hand, it's the same thing with screens and resolution - there is point where adding pixels won't do much good any longer. On features and optimization, in real life one has to choose the point where optimization is "good enough" instead of "optimal" because there usually are nobody to pay the price for "optimal".

I was going to say that IMO everyone should be just fine using only vi...

Most of time new features are something nobody asked for because they didn't have an idea such thing could or should exists. Same goes for all new products - they create the demand. I won't comment that much on mostly subjective things like faded backgrounds as those are more related to UX which needs to be more or less consistent and pleasing to the eye. Usually everyone has their own opinion about those.

I don't consider Jolla (the original one) screen to be on that point as I find it much more pleasant to use retina display. It doesn't mush smaller objects as much. But yes, there is limit in EVERYTHING beyond which usefulness/rationality can be questioned - optimization level included.

Re: The new QWERTY device project
 

The problem with the original Jolla sbj1 display is the dimness of it, not enough lux and contrast.
Resolutionwise it is pretty much optimal.

Regarding the idea presented before that a modern (too-)high-resolution display could be used by cutting down the resolution by for example using 4x4 groups of pixels; I am not totally confident that it will cut down power consumption as much as can be hoped;
You will still need a high-end display driver for it I guess, can you even interface it to a more energy-efficient chipset?

Re: The new QWERTY device project
 

AFAIK, the power consumption of typical LCD screen is about 75% for back-light. As for pixels, it's not that much about the number of pixels but the area of screen they have to cover. One big pixel can use roughly same amount of juice as several smaller pixels covering the same area. There is some power "wasted" on circuitry needed for handling extra pixels but AFAIK it shouldn't be significant compared to other things consuming power.

Bigger resolutions do need more capable SoC/GPU, which may or may not lead to increased power consumption. It depends on how much newer SoC/GPU has improved on performance AND how that has affected power efficiency on those chips. Moving to smaller manufacturing process helps on this regard, as well as possibility to cram more transistors to smaller die (of course the design/architecture has to be good). So, I do question the idea that older, low powered chip coupled with low resolution screen automatically results in more power efficient device compared to newer chip with relatively high resolution screen if the devices are similar in size.

Re: The new QWERTY device project
 

Sorry guys, do you really think this discussion has still anything to do with the initial thread? Wouldn't it make sense to open a new thread for this screen discussion?

The screen is already chosen it is 5.5" IPS and 1080p.

Re: The new QWERTY device project
 

y'know....
I have been pondering this same question for a while now...long while...

I think as a serious topic as it is...there should be a separate thread for it...

Also I think there should be a thread in off topic...
Called the "Waaaayyyy Off Topic Hijack Thread" ..
whose sole purpose is to deviate as much as humanly possible,
from the pertinent topic being discussed at the time in it.

I am sure I will enjoy spending some time there...
:D:D:D

Re: The new QWERTY device project
 

Yeah, I have been quiet lately due to off topic discussion. But I'm sure I'm some how blamed anyway. As always ;)

On the other ...side/hand/frontier? What else is there to discuss in this thread? We simply have to wait for Chen to announce the sailingChen campaign.

Re: The new QWERTY device project
 

Ahhhhhh Dave is involved, that explains anything.
It couldn't have ended differently, than running off topic...
:p

Re: The new QWERTY device project
 

hahahahaaha...
ah..
well to commemorate

I will indeed make the said thread as a homage to the master ...
Mr. Dave.

once I get home ...

I have a fitting topic start too.

Re: The new QWERTY device project
 

Back to topic:
@chenliangchen if you consider to offer this device also with android. Wouldn't it make sense to offer the device with a dual boot system right from the start? Maybe that would save some people from the dark side by just playing with SFOS.

And something different, how long do you think will it take to port SFOS to the livermorium device. To be honest I'm a little bit concerned about that, because Jolla is releasing Sailfish X unfinished.

Re: The new QWERTY device project
 

Hehe. It always works better to move Off topic here as off topic is the only true way to be on topic for threads over 2-3 pages. Becouse of the fact that all valid discussions has already be been discussed from the info we got. But to keep a thread/project alive over time, Off topic most be allowed. Why not continue here until Chen ship the Moto mod and start building a qwerty slider. We can have a lot of fun in this Thread over the years to come until something might be ready.

Like having a contest when we think it is ready for customers:
Chen thought/hoped before end of this year.
My bet was in about 2 to 3 years.
Anyone else?

Re: The new QWERTY device project
 

TL;DR: The fact that Sailfish X isn't 100% finished has no impact on SFOS on livermorium. In fact, when you look at it, the remaining work to get Sailfish X being perfect on Sony Experia X is about all the work that would be needed to get it on Livermorium too.

-----

When you look into the detail, Jolla has just finised the long and exhausting part of porting the whole of Sailfish OS from 32 bits to 64 bits.

Up until now we had 32bits ARMv8 used by nearly every smartphone port (and 32bits 486-like Atom used by the tablets). Sony Experia X is the first ever 64 bits CPU (an ARM64 one).

In theory that requires "just" recompiling everything for the new architecture. In practice, the devil is in the detail and it's a long process of chasing all the minute small unexpected bugs.

So, now with Sailfish X all the hard work (or at least tedious work) has been done.

The things that are missing on the Sony Experia X are very specific to *that* platform (getting the drivers to work and plugging them into the users space to work - they "just" need to get BlueZ5 working).


So for a potential Sailfish OS port on this upcoming livermorium, it shouldn't take as much time :
Sailfish X is already working on 64bits architecture.

The main problem is going to be drivers :
- getting the binary Android linux kernel fork provided(*) by Qualcomm to run
- using libhybris to have all the binary proprietary drivers (eg: EGL drivers) to run on GNU/Linux instead of Android.
- plugging all the other userspace component.

Should be work within the "months" range.
Somewhere between 1 month (we're terribly lucking and everything works on the first attemps) and 6 months at worst.

(Again, that's incredibly fast, because the tedious work is already been done)

----

(*) : Or if we're extremely lucky, maybe that the PCB that chen has decided to use as a base happens to be useable with the vanilla upstream linux kernel and freedreno drivers.

Re: The new QWERTY device project
 

Another option to get faster, would be to leverage the community porting (in progress) for the Motorola Z Play, using the same MSM8953 chip as seen here :
https://talk.maemo.org/showthread.php?t=99669

Seeing as Chen said in the linked thread he had sent a private message to the porter, and that he is working closely with Motorola for the keyboard Moto Mod, I wouldn't be suprised if there is some truth behind this. ;)

Re: The new QWERTY device project
 

In the community we already did the work for 64bit arm devices, this was before the work on Sailfish X had even started and was mainly pioneered by Ghosalmartin for the Nexus 5X (bullhead).

The userspace will still be 32bit and indeed that is a feature because there are no aarch64 compiled apps around for sailfish at all.


Bluez is fine, its the kernel HCI driver with the broadcom based BT chips which I already warned Chen against.

The main problem is *always* the drivers, and also the specific android base used to interface with those drivers which Chen probably won't get a huge amount of choice on anyway.

Re: The new QWERTY device project
 

I've completely missed this port.
Thank you for enriching my knowledge.

Well, that's sad.
I was hoping for a full 64bits support.

- Lots of 3rd party apps are "noarch" anyway (QML + Javascript, mostly - same package regardless of the CPU arch).

- Supporting Aarch64 is basically just adding 1 more available target on their OpenBuildSystem compile farm.

- We need multiple archs due to Atom's 486/IA32 anyway

- There aren't that many 3rd party apps with binary code anyway, compared to Android or iOS. Lots of them are opensource and even if the original author is unavailable, someone else could easily fork the repo and recompile.
Only the few closed-source apps, with a binary component, whose dev is unavailable, and that will need a Aarch64 recompile would be affacted by the switch.



Let's hope that by the time of Livermorium, we'll manage to find a way to get it working.

I haven't been paying much attention, is the bluetooth core inside Experia X anyway related to the core in chen's selected PCB ?
i.e.: will the work that Jolla will need to put into Sailfish X's bluetooth driver benefit Livermorium in anyway ?

Re: The new QWERTY device project
 

Anyone know when moto mod is estimated to be completely delivered?

Re: The new QWERTY device project
 

According to the IGG page

Re: The new QWERTY device project
 

There's practically zero apps that install on the Jolla Tablet for this reason, nobody bothered to build for it. Indeed at some point Jolla is going to have to let the community support unmaintained apps on Harbour because sooner or later a Qt upgrade is going to break them. and there's also Alien-Dalvik to consider in this equation.

But so far Jolla hasn't said anything about building or supporting fully 64bit sailfish, and I would expect that they would let us community porters have it first for testing before rolling it out as a fully supported thing.

No idea. My point from mentioning it was basically to say that its a hardware specific problem and mostly likely irrelevant for the Chenphone.

Re: The new QWERTY device project
 

Soon?

(Sorry OT and bad joke...)

Re: The new QWERTY device project
 

When I was discussing the GCC update during one of the Mer meetings, Jolla seemed most concerned about those closed-source applications.

Re: The new QWERTY device project
 

@Chen: Are you planning to update your signature with this here project any time soon, or perhaps are you waiting for your Moto mod to be done with ?

There *might* be people that read your posts on other threads and don't know about the "Chen[s]phone".

The more the merrier.
Strength in numbers.
All that...

Re: The new QWERTY device project
 

Yes thank you! I will need to deliver this Keyboard Mod first.

I have decided to take the time longer to make the product as good as possible, instead of releasing in a hurry. It's about 85-90% of my "ideal product" at the moment but I want to make that 100%...

Once Motorola formally launches this Mod, and I start fulfilling the supporters in IGG, I will close the Keyboard Mod campaign and formally move to this one.

But because of running late with keyboard mod I don't think we can make the hardware end of year anymore. But won't be too long. Building this device (not inc. software) isn't more difficult than Moto mod and I know exactly what to do...

Re: The new QWERTY device project
 

Be careful on that! The last 20% usually need more effort than the other 80%. It might take ages or even kill your project as a whole. Especially if your aim is an "ideal product".