A date with the Pyra


OK my andriod opinion: I don't give a f... to have it. Why?

I want the pyra BECAUSE it is not an andriod. Basicly andriod is for me: Guys selling you a device, which "seams" to be open because it's"Linux", but not givingyou openess by that but spying on you, forcing you to have an account to get free sonftware (my calling "store" what should be an apt-get command).

For years now I wonder why it is claimed that you can't use a modern smartphone as a desktop replacement, while I read that anouncement on my 15 year old thinkpad, which has what 10% of the hardware power???

So for me the pyra will be "the ultimate tiny laptop" (finally). Beeing more open than any other machine available (except bunnies novena I guess) running my faviored OS, needing noe could service to have my calandar working, beeing able to type on a keyboard (not a crapy ugly non-tacktile responsive touchscreen shit), not beeing forced to hack it forst to have under controll ... I think you get my point.

And yes: I don't have any andriod device at all!

I personally see no reason at all to have andriod running on the pyra for these reasons. I don't mind if somebody want to have it (what i will never understand), but as a major course? no.

As for the 90% guys mentioned: I have the impression we are more likely the other 10% here.

Pyra will be the best power/size/openess combination that was there ever! That's why i want to have it and be willing to pay for, not to have an expensive "other" andrion device.
 
I am curious: for 90% + of perspective Pyra users,...

Call me stupid but I really don't understant your stand?
I don't think Steven Craft is especially interested in a Pyra that sports Android on Day 1, he just wanted to ask a (in my opinion) legitimate question. All those that speak out against Android either at all or want it just as an "afterthought" are the 10% percent he is implying. But what about the silent majority ? In my opinion 90% is an enourmous exxageration, but there sure is a big percentage of users, that want the Pyra just for gaming and they probably don't care about a "full" Linux, they probably are in the situation that they already know Android and the software library it offers - that may change later on, when they discover that Android is actually limiting them on the Pyra, or maybe not.

My personal opionen is the minority one, I don't frickin care about Android as it offers no value for me, and ED should (and I am pretty sure he is) concentrating on the UIP the Pyra has, beeing a portable and open(!) Linux MiniPC
 
Very interesting article. It reads that ALSA is a source of lag. Would that be the case if Pyra uses it ? What about PulseAudio ?

*Everything* runs through ALSA, it's ALSA that provides all audio drivers and the interfaces to feed them with audio data - PulseAudio is just an additional front-end that runs on top of ALSA.

ALSA is very modular and configurable, you can just use it to directly access an audio device (which blocks access for every other program on hardware without a hardware mixer - pretty much every chip out there), or you can configure it to be a full fledged audio framework doing all the mixing and whatever complex stuff you can come up with. By default PulseAudio grabs the hardware through ALSA exclusively, does the mixing itself and configures ALSA that any audio data being fed directly to ALSA's default device is being redirected to PulseAudio (ALSA -> PA -> ALSA -> HW).

The basic latency introduced by ALSA is something you can hardly avoid, it's the other stuff you're doing with the audio data that should have the largest impact on latency. JACK is often being used for low latency audio with studio applications and it runs on top of ALSA as well - but it's a really bad idea using it for your everyday needs.
 
If you just want a decent android gaming device there's plenty to choose from the likes of JXD or NVidia.  The only thing those are missing is a keyboard, and the only reason you'd want one of those if you're only interested in emulation or gaming is for playing old home computers (I believe the reason the Pandora got a keyboard was because despite it being an unofficial followup to the GP2X, somebody pointed out it'd likely be able to emulate Amiga).

For me at least, the Pandora has turned out to be something much more capable and usable, and I'm glad the Pyra is building on that personally.  There may be a significant number of people interested in a purely Android device for gaming, but those are already well served IMO.  There's a handful who really really want to emulate Amiga or something older *a lot*, and who therefore want a keyboard, and those will have to decide between a Pyra or an Android device with a plug-in keyboard most likely (or a tablet if you're less bothered about pocketability).
 
Pyra is like a tiny laptop. Would you run android on your laptop? I'm sure some do, but for some good reason most people don't.

There are quite a few Android based Tablet / Laptop hybrids as well as a selection of dedicated clip on keyboards.

This suggests to me that there is a subset of the worlds 1 Billion plus Android users that do value a hardware keyboard, I'm going to go out on a limb here and suggest this might be because they do a lot of typing (email, office apps etc.).

This subset is rather small, and the percentage that would buy a Pyra due to it running Android & having a hardware keyboard is probably a small fraction of Android + Keyboard users. However, even if just 1000 Android users could be persuaded to buy a Pyra, because they had the OPTION of running Android, then that would be a worthy boost to Sales IMHO.

Due to the relatively high cost of Pyra hardware I think that a major aspect that will impact on the sales figures will be an ability to clearly stress the unique / unusual aspects of the Pyra project. One unusual feature that IMHO would appeal to potential users is the OPTIONAL ability to easily install and use multiple OS's.
 
[...]


This subset is rather small, and the percentage that would buy a Pyra due to it running Android & having a hardware keyboard is probably a small fraction of Android + Keyboard users. However, even if just 1000 Android users could be persuaded to buy a Pyra, because they had the OPTION of running Android, then that would be a worthy boost to Sales IMHO.

Due to the relatively high cost of Pyra hardware I think that a major aspect that will impact on the sales figures will be an ability to clearly stress the unique / unusual aspects of the Pyra project. One unusual feature that IMHO would appeal to potential users is the OPTIONAL ability to easily install and use multiple OS's.
The option of Android is there, nobody is taking away the freedom to use it, but it doesn't make sense to invest already scarce resources into Android when even the default "official" OS is not completely there. And as has been pointed out, even focusing on Pyra OS, not all is lost regarding Android, since you WILL need a complete kernel anyway and this is (obviously) one of the main focus points right now. Android support WILL come in one form or another further down the road, and I'm glad of that since I have quite a few games in my humble bundle account which I would like to play, but I think we can agree that, even though Android could be a selling point, getting out a product as polished as possible is better for business, right?
 
JACKis often being used for low latency audio with studio applications and it runs on top of ALSA as well - but it's a really bad idea using it for your everyday needs.

Care to elaborate? I guess the issues would be the same as with PA more or less. On the other hand maybe having just ALSA would be the best option to keep latency low and keep things simpler if compatibility is manageable. Anyway just curious regarding that.
 
Last edited by a moderator:
Pyra is like a tiny laptop. Would you run android on your laptop? I'm sure some do, but for some good reason most people don't.

There are quite a few Android based Tablet / Laptop hybrids as well as a selection of dedicated clip on keyboards.
Yes, hence the "I'm sure some do". For those that actually only need what android can provide and are certain they never need anything else it may well be a good option.

This suggests to me that there is a subset of the worlds 1 Billion plus Android users that do value a hardware keyboard, I'm going to go out on a limb here and suggest this might be because they do a lot of typing (email, office apps etc.).

This subset is rather small, and the percentage that would buy a Pyra due to it running Android & having a hardware keyboard is probably a small fraction of Android + Keyboard users. However, even if just 1000 Android users could be persuaded to buy a Pyra, because they had the OPTION of running Android, then that would be a worthy boost to Sales IMHO.

Due to the relatively high cost of Pyra hardware I think that a major aspect that will impact on the sales figures will be an ability to clearly stress the unique / unusual aspects of the Pyra project. One unusual feature that IMHO would appeal to potential users is the OPTIONAL ability to easily install and use multiple OS's.
Well, you have to weigh the upsides against the downsides. How much developer work does it take to accomplish? Maintain? The maintenance part is one that's often overlooked, but is typically the most work. Also you have to weigh in that giving a prototype to someone for Android work means someone not getting one for Linux work. Is it still worth the downsides?

IMHO Pyra should stay the course for now and focus on getting one OS working beautifully before branching out. Let's focus our efforts until we have a leg to stand upon. Android can be ported when Linux works.
 
Last edited by a moderator:
There's going to be a significant amount of overlap between Linux, Android, and Android-in-VM.  Linux is a sure thing, there's people eager enough for an Android-in-VM that I'm sure that'll be maintained by the community, the remaining parts for a base install Android probably aren't going to be much more beyond that.  It may even be easier the other way around: base install Linux and base install Android makes Android-in-VM trivial. We'll have to wait and see who starts doing what and if there are any places resources can be applied.
 
I had an iMac a few years ago with BootCamp installed that could boot to Mac OS X or Windows. I could also launch the Windows partition inside the Mac OS using VMWare (http://kb.vmware.com/selfservice/search.do?cmd=displayKC&docType=kc&docTypeID=DT_KB_1_1&externalId=1014618).  This was a nice set up in that most of the time, I could just use the vm to dip into Windows but I could also fully boot into Windows as needed.

This approach could be something to consider when developing Android for Pyra.
 
Some sort of ChromeOS/Android convergence is likely (maybe some more clues with the upcoming Pixel C, maybe later). 
 
I guess the issues would be the same as with PA more or less.
Probably worse - it solely focuses on low latency (using tricks like realtime scheduling) and flexible rerouting of the audio data flows, high CPU load is kinda intentional - but I'd guess PA would have a rather bad latency. As far as you only care about a single program, sending audio data directly to the ALSA HW device will get you the lowest latency possible without any extra CPU load (which is what the JACK guys actually recommend) - but that also means no other program may output audio data at the same time, sound cards that can do the mixing in hardware are rare. Obviously, this doesn't work as long as there's some audio daemon running that wants to do exactly that as well, which is probably one of the reasons why Android's latency is that bad.
 
How much of a performance hit would an android emulation layer (Android in a VM) be compared to booting android separately?  if it's mostly small/unnoticeable, i agree with Grench that we should definitely concentrate efforts on the Virtual Android rather than Booted Android.
 
(Someone please correct me if/where I am wrong).

Since Android is/was based on Dalvik, a Java-like virtual machine, is it not possible to just run Dalvik in a CHROOT or some such thing? I am aware that there is a separate windowing-system for Android, but would it be possible to use a Xephyr-like program for that?
 
(Someone please correct me if/where I am wrong).

Since Android is/was based on Dalvik, a Java-like virtual machine, is it not possible to just run Dalvik in a CHROOT or some such thing? I am aware that there is a separate windowing-system for Android, but would it be possible to use a Xephyr-like program for that?

This is pretty much what Alien Dalvik does, but it's not quite as simple as that as you also need to implement the various system services and interfaces for applications to work right. Users of Jolla such as myself have some experience in this :)
 
Honestly, I don't think we're even in the position to worry about Android.  From what I can gather from comments from EvilDragon, Notaz, etc., there's an enormous amount of work that needs to be done on the OS side just to get the standard OS at an acceptable state for us.

-God Ginrai
 
Given that ED has already stated that he's willing to put money into getting Android working, (after the Pyra is released and is selling well) I'm sure we'll end up with a solution that gives us the best of both worlds.  It just won't be available on launch.

-Neelix
 
Given that ED has already stated that he's willing to put money into getting Android working, (after the Pyra is released and is selling well) I'm sure we'll end up with a solution that gives us the best of both worlds.  It just won't be available on launch.

-Neelix

Sounds entirely reasonable, from a personal perspective I'm in no rush to have a fully functional version of Android on the Pyra, as long as it comes one day I'll be happy. However I do think that having the option of Android available at launch would prove favourable with reviewers and would therefore help to boost sales.
 
 
Back
Top