Related
Hi folks. Sorry to start a new thread but there's something bugging me around here. If by any chance you mods think this post is useless please close it and accept my apologies.
It's known to the most of you that the first version of drivers didn't worked on all roms. That issue forced the developers to do a workaround and slow down the process to get it working on those roms and created a new set of drivers that is not so fast as the first one.
So... i was thinking on this and decided to start a poll in order to give a hand to the developers in wich way they should follow. What do you prefer?
A Gold XDA based ROM with fastest drivers and less compatibility with all roms? (forcing all XDA members to flash their phones - not a big deal i presume)
Or a set of drivers compatible with most of the roms but with slower performance?
I know this is a controversial issue and the urge of the drivers working on our kaisers is huge but i think we would have more if we had a custom GOLD XDA rom
Just My 2 cents
Cheers,
Draco
kinda seems like a no-brainer honestly (unless you're petrified to flash or just despise 6.1). They will develop said drivers and you have a million chefs making a million roms from them...
Fared said:
kinda seems like a no-brainer honestly (unless you're petrified to flash or just despise 6.1). They will develop said drivers and you have a million chefs making a million roms from them...
Click to expand...
Click to collapse
I have no problems at all flashing my phone. I assume that there are some people who are not comfortable doing that but.... this is XDA. And if they are here that's because we all want to get the most out of our phones. And that is the tricky part of this poll. Developing the drivers based on one specific rom ( a ported one form Diamond or Raphael or an official kaiser one ) will be the more accurate and fastest way to get there.
Again... just my 2 cents
Draco
first:
flashing rom is a non-trivial task for a huge majority of users. leaving them out in the cold sounds like a very selfish position. If not for this reason, maybe avoiding a flood of "I bricked my tilt, help !!!!" threads may be a better one.
second:
Anyway, as i understood it what is slowing down FPS in the new version is the enabled VSync, which yields less tearing and improved smoothness (sounds like a shaver commercial, i'm afraid). I don't know what the refresh rate of a kaiser LCD is, but anything higher than 75fps (ie 15ms refresh rate) is a complete waste.
draxredd said:
second:
flashing rom is a non-trivial task for a huge majority of users. leaving them out in the cold sounds like a very selfish position.
Click to expand...
Click to collapse
Not at all draxredd. Although i'm a junior member this is the second time i'm on XDA. I'm here for almost 7 years. Believe me, this is not a selfish position. As i was able to get some help to start flashing all of the members here on XDA will get that help too. The first time i flashed my first phone i was scared to death but... it's all the same. It only hurts the first time.
As i stated before this is just my opinion. The more roms the developers must include in the drivers development process, the more difficult it will be and much more time will be wasted.
Draco
haste... makes waste.
And if only one rom was to be designated as a target for the driver, it has to be the official HTC one, for two reasons:
- enabling driver support for a huge majority of users
- sticking it to HTC, deep.
draxredd said:
And if only one rom was to be designated as a target for the driver, it has to be the official HTC one, for two reasons:
- enabling driver support for a huge majority of users
- sticking it to HTC, deep.
Click to expand...
Click to collapse
What about performance issues? Doesn't that count as well?
Anyway.... i've started the poll just to check your opinion against wide compatibility among all the roms versus one specific rom fully optimized. Let's not get far from the main question
By the way... my kaiser has the official 6.1 rom from HTC... for now
cheers,
Draco
I'm for Flashing myself, but the in the ideal world both sollutions would be best so no one comes off short. However the kind developers should be aiming at the most stable of the 2 first.
Looking at the driver progress it looks like theyre slowly getting there and getting the driver to work on the latest 6.1 roms
I voted for compatibility with the official ROM, simply because it's an identifiable standard that everyone recognises, and which everyone has on their phone by default. Cooked ROMs are great, but they're all individual, designed by individuals and used by individuals with the same tastes/needs - not one of them will suit everyone else, with or without drivers, whereas drivers released for the official ROM should work with pretty much everything else.
It's too early to say whether drivers developed for the standard ROM will always be slower than those for a cooked version, and I'm sure the very brilliant developer of those drivers will do their best to minimise any performance loss, assuming their is any in the longterm.
Vastly improved performance for the majority (and what that means for sales of the Touch Pro etc) has to take priority over a marginally improved performance in a niche ROM, aimed at an elite few, surely? At the end of the day it's entirely down to the developer what they do, irrespective of any poll here, but that's my take on it anyway.
Maybe i didn't expose correctly my idea.
There isn't any XDA Gold rom nor any kind of perfect ROM. The XDA Gold rom is just a concept and the first name that crossed my mind.
This is a big community and has some big brains around here. The exchange of knowledge bettween everyone is a plus and that opens a big wide world to explore.
So.... let's just imagine this scenario, ok?
The developers will pick a base rom. It really doesn't matter if it is a ported rom or a Official one. Preferably one that is fast and stable enough. The driver will be written over that rom. Users (those who want obviously) will test the rom for bugs and then chefs will try to cook out a 100% clean (note that by clean i mean a rom just as close as an official rom regarding to apps) fast and stable rom.
Of course this is just a crazy ideia but that i was thinking of. Create the first genuine XDA rom based on one device.
Note that there are many roms out there based on diferent releases. Make the drivers compatible with all of them will make the development process slow down and eventualy fail. That's the main reason i started this thread.
Now that the concept of the "XDA Gold rom" was properly explained (i hope) let's get on with the poll.
Draco
DracoLX said:
Now that the concept of the "XDA Gold rom" was properly explained (i hope) let's get on with the poll.
Click to expand...
Click to collapse
As long as "XDA Gold" = HTC Latest Official version I cant see a problem. It's when drivers are developed for anything other than the official ROM that we'll encounter problems. All the chefs can base their designs starting with official ROMs and work out from there. In the end it will depend on whether those developing the drivers want to satisfy the larger number using official ROMs or those using Custom ROMs where the chef has not saught to maintain driver compatibility. I've therefore not voted because it depends whether "XDA Gold" does in fact = HTC Latest Official version - we weren't told.
Are you even sure that theifference is that huge?
They might be able to provide two sets to cover all ends without much additional work.
DracoLX said:
Note that there are many roms out there based on diferent releases. Make the drivers compatible with all of them will make the development process slow down and eventualy fail. That's the main reason i started this thread.
Now that the concept of the "XDA Gold rom" was properly explained (i hope) let's get on with the poll.
Draco
Click to expand...
Click to collapse
IMO, The drivers only need to be compatible with an official based rom. It's up to the chefs to make the drivers work with their rom. Now, if the root of the compatibility issue can be resolved, I think things can progress forward at a faster rate.
DracoLX said:
It's known to the most of you that the first version of drivers didn't worked on all roms. That issue forced the developers to do a workaround and slow down the process to get it working on those roms and created a new set of drivers that is not so fast as the first one.
Click to expand...
Click to collapse
FALSE. Two separate issues.
First, is dlls load differently between stock or hacked/cooked roms due to memory mangling by G'realoc going into cooked roms. An attempt to fix that was in the updated release.
Two, the screen update style in the first release was really just a hack that hijacked DDI's screen refresh routines. The 2nd release finally got the 'official' way to work so the hack was removed.
I voted faster...there isn't any other reason I came to XDA other than customizing.
Anyone who doesn't want a faster cellphone won't appreciate the new D3D drivers anyway.
There are may reasons the XDA Gold Rom (HTC Original 6.1) may be slower. Cooks reduce the software on the rom and use newer software versions, these may be faster than the previous ones.
Also this software is not from the Kaiser sometimes these may be causing the incompatiblity problems with the drivers.
The HTC original rom is slower to start than most cooked roms anyway, so way wouldn't the 3d drivers also run slower.
I decided not to vote, I think speed is a non issue as long as the drivers work on the HTC original rom, and are stable. You will always get rom's that perform faster, that's what cooks do round here.
I also don't think you would get a big jump in performance writing for just one rom, the hardware is the same in all the HTC Kaiser's handsets. Software can only make small difference to the performance of hardware if the drivers are written correctly.
personally i would love the drivers in a rom like dutty's diamond v1, but i think building them into a standard(ish) htc rom is best for the comunity; it sticks it to htc best as it shows u guys can do what they couldn't(or wouldn't), and makes flashing a rom feel safer to noobs.
and anyway we all know within hours of a rom with the drivers being released there will be dozens of cooked versions for all tastes with the drivers included too.
So I reserved my Hero from Sprint but going through the forums so far I'm a bit hesitant now. There doesn't appear to be that many mods (only a few really) so far for the Hero. Is that because the device is so new and I should just give it time or is it that the device really isn't that popular of an Android phone so I shouldn't hold my breath for more customizable features? Also of all the hacks/mods they are so far for the European version of the phone and there is no telling if any of them will work on the Sprint or American version so that even further lowers my chances of people releasing good Quality hacks/mods for it. The main reason I wanted an Android phone was because of the customizable interface and the theming and skinning possibilities; I just want to make sure I'm not shooting my self in the foot by making the Hero my Android phone of choice.... Any suggestions?
-Michael
The only real "issue" with regard to the European/Worldwide Hero at present is the lack of kernel source, which makes certain modifications difficult. I'd expect this to change very quickly when the source is finally released.
However, this doesn't really help you much at present as the Sprint Hero is very much an unknown quantity. Until such time as some US devs get their hands on one, we won't know what is or isn't possible to do with the Sprint Hero.
Given your concerns, I'd probably be inclined to wait at least a month before getting one so that you can see what the state of play is with regard to rooting and custom ROMs for the Sprint Hero.
Regards,
Dave
I feel the same except I did go out and upgrade from a customisable G1 to a stale Hero. However I playing the long game tbh, think about it.... the hero as shipped by the bucket load plus at the moment its very hard to cook for a fresh because of the lack of kernel source, once thats release its going to be open season! And remember the hero is a great phone in it own.
How do you like your Hero? I"m currently using an iPhone and while I'm not getting rid of the iPhone because I really do enjoy the phone, I'm really looking forward to the customizability of the Hero and Sense UI even without the hacks.
ratcom said:
plus at the moment its very hard to cook for a fresh because of the lack of kernel source, once thats release its going to be open season!
Click to expand...
Click to collapse
How long do you think it'll be before the source is released? What's the precedent with other Android phones? Do you think because it's Sprint's first Android phone that they would take a twice as much time or even longer before releasing the code?
I believe that HTC have to release kernel source at some point which will enable most of us to cook its all just a matter of time. Plus remember how many hero's there are out there, I would think theres alot more than say G1's?
short notice: the hero is one awesome phone even not rooted. i didn't play around with the stock rom (1.76 for t-mobile germany) for long, but even that wasn't bad. MCR2.3 makes it just more awesome. i would buy it again, even if it wasn't rootable at the moment.
I may be the only one interested in this, and I know its set for froyo 2.2 and not sure if its possible but is there no way to port this back one version? Maybe make it work for us?
I really like the idea and would be disappointed to have to wait the loooooooooong time before 2.2 is ported to G1.
no it uses the new pushy mc jig from froyo
probably could back port the whole framework, but seems like a wast ogf time to me
atantekeier said:
I may be the only one interested in this, and I know its set for froyo 2.2 and not sure if its possible but is there no way to port this back one version? Maybe make it work for us?
I really like the idea and would be disappointed to have to wait the loooooooooong time before 2.2 is ported to G1.
Click to expand...
Click to collapse
How do you know it will be such a long time?
Depending on the code changes between 2.1 and 2.2, it may be little more than trivial to get 2.2 running.
Also, I suggest that the first step required in either backporting or building a whole system would be to actually wait for the availability of the source code. Then we judge what would be a more efficient use of time -- building a whole system or backporting chunks of it.
2.2 will support the ION, so in turn, it will support the Dream. We just have to WAIT for it, people.
Mod edit: not dev related, moved to q&a
Just a thought: during the google io presentation of google TV they make a presentation of this feature with the google box in 2.1. So I don't know if the sdk or the source code is published but can't we use that to make it working with eclair roms?
PS: here at 1:30
Time is relative here, by the time a new update is released from modders porting it to another much older phone version a new version is almost ready to be released.
This is not a bad thing and Im not talking down about it ether. But was simply asking the question if it would take too much effort to get just that .apk to work for even 2.1 that is now very stable on the G1.
2.2 was just released, and it will probably take a while to get it stable, to the point where 2.0 or 2.1 is now for the G1 anyways.
I just really like the chrome to phone and think its a super idea. Maybe I'm a bit impatient too.
The intent of this thread was to provide a consolidated answer to all of those waiting patiently (some, not so patient) for a complete Android ROM for the SE Xperia X1.
Many perceive Android as "a hacker's OS" because it is both open source, and supported on so many other HTC devices. Because of this, I suspect many of the "spectators" here (like me) incorrectly assumed that Android on Kovsky (or, any HTC smartphone) would be a straight forward affair, considering some of the amazing work done here with the Windows Mobile ROMs.
It seems, however, that creating a usable Android ROM for Xperia is far more difficult than creating WM6 ROMs? Android has been out for quite some time and there's still no widely available Android ROM for Xperia X1. It is for this reason that I've started this thread.
I'd like to hear from those that have/are working on Android about the challenges in putting together an Android ROM for Xperia X1. What has contributed to this being a difficult OS to port to Kovsky?
Is it driver support? Perhaps Kovsky has a unique selection of hardware that complicates the process? Or, has much of the effort been directed towards Haret vs. a native ROM? If so, why? Can the Haret work be merged into an effort to create an Android ROM or are these different approaches completely exclusive to one another?
Perhaps there are IP/DRM/license issues that are complicating the effort? I'd very much like to hear from ROM chef and devs working on this port.
While it may seem like a idealist perspective, I wonder if some dialog on the issues surrounding an Android port/ROM might result in greater community involvement, and ultimately a stellar Android ROM for one of HTC's most beautiful smartphones. I'm certain that interest in Android on Kovsky is very high. Are there areas where the community might be able to assist with direct hardware, financial or other contributions?
So please, sound off. I'm very interested to hear about chefs/devs experiences with Android and I'm certain there are many here that would like to hear more about this port.
I just didnt know how to tell all you told here... I really want to know because I love my x1 and I love Android.. but i can"t help
i have just check out the Kovsky project from Gitorious, but still don't know how to work on it. 'cause i don't have much knowledge about develop kernel and open source, why don't we have a tutorial thread about this?
You people asking when a Android rom is coming just about every day are so funny. First of all their are still major issues that need to be worked out. Android running from haret is not even every day usable yet so what is the freakin point of expecting a rom now or asking about it? Do you guys want a rom where you phone won't even last 2-3 hours, GPS dosen't work along with the camera etc......? We need All the issues fixed before we can even be thinking about a rom. When Android from Haret is everyday usable then lets start talking about a rom. So can we please give the Devs a break about this rom nonsense and let them work in peace, when their ready to release a Android rom they will release a Android rom.
wphoenix said:
i have just check out the Kovsky project from Gitorious, but still don't know how to work on it. 'cause i don't have much knowledge about develop kernel and open source, why don't we have a tutorial thread about this?
Click to expand...
Click to collapse
would like that too, i've only been reading for some good time about the advances, and would like very much to take part in the development of the kernel, but don't really know where to start from
Also, what Viper says its very true... the work is probably a long way from done, depending on how much help the dev's get, which is, from what i've seen, not much.
about Toe_Cutter's comment "I wonder if some dialog on the issues surrounding an Android port/ROM might result in greater community involvement, and ultimately a stellar Android ROM for one of HTC's most beautiful smartphones" ... well, yeah, i think it could... at least i hope so.
xD
cool down all your heads and lets just help in every way we can
...as (ilgreco112) does xD... cheering up isn always bad hahaha
ps: first post!! ... Hi!
Viper89 said:
You people asking when a Android rom is coming just about every day are so funny. First of all their are still major issues that need to be worked out. Android running from haret is not even every day usable yet so what is the freakin point of expecting a rom now or asking about it? Do you guys want a rom where you phone won't even last 2-3 hours, GPS dosen't work along with the camera etc......? We need All the issues fixed before we can even be thinking about a rom. When Android from Haret is everyday usable then lets start talking about a rom. So can we please give the Devs a break about this rom nonsense and let them work in peace, when their ready to release a Android rom they will release a Android rom.
Click to expand...
Click to collapse
As a dev, I bet it IS tiring hearing all the spectators asking about an Android ROM.
But what do you expect when spectators ask about the issues, and receive nothing in response?
In my original post, I chose my words very, very carefully, because if you go back and read it again, I never asked when, I only asked the why and how.
Why is it that no one can provide a response to this question? I searched the forum, and read thru dozens of posts trying to understand what this port might require, and I got nada (except several polls asking about an Android ROM, and a few more posts telling the spectators to quit asking about an Android ROM).
Anyone care to break the cycle and help the community understand the issues?
Well, the problem is that you can never guess what's required and how to achieve it. First of all, developers really know nothing about the hardware before they begin hacking it. Until you finish the job, you can't estimate the time it takes to complete. The major problem is that most of us devote just a tiny bit of our time to the porting, because there are lots of interesting things to learn and to do except it, but mostly because everyone's either studying or working (or even both).
But if you really want to know what's needed.. well
1. Fix the LCD panel (add the initialisation code)
2. Fix the power saving (that's the greatest PITA right now) and 3D (the latter should be easy and is not really critical)
3. Implement Bluetooth, FM Radio, Camera (Should not be really hard on the driver side, but rather on the RPC protocol side)
4. Fix USB (should be easy.. actually, everything's done on the device driver side, the problem seems to be with the msm usb driver itself.. just needs real debugging)
5.Make a version of rom that boots from NAND (well, once you get the LCD to work correctly it's not a problem. It actually boots very well off nand, but without LCD and USB working it's not much fun, to tell the truth)
6.Clean up the code, drop the ****ty-droid and head forwards to SHR/Maemo MER/Meego or some other real linux distro with X11 and other things
7.Write up misc stuff (like optical joystick, illumination, new keypad driver and others)
But really, it you want to help, just begin researching everything yourself and exploring wince drivers. You know, Theo de Raadt once said: "shut up and hack", and that's the only way to get things done.
full ack
@sp3dev
full ack your post. the main problems ar not at android 2.x this source we have. but the kernel an its's modules (for kovsky) are the problem.
regards
Toe_Cutter said:
As a dev, I bet it IS tiring hearing all the spectators asking about an Android ROM.
But what do you expect when spectators ask about the issues, and receive nothing in response?
In my original post, I chose my words very, very carefully, because if you go back and read it again, I never asked when, I only asked the why and how.
Why is it that no one can provide a response to this question? I searched the forum, and read thru dozens of posts trying to understand what this port might require, and I got nada (except several polls asking about an Android ROM, and a few more posts telling the spectators to quit asking about an Android ROM).
Anyone care to break the cycle and help the community understand the issues?
Click to expand...
Click to collapse
The ROM question has been answered time and again throughout various threads on xda-dev.
There is no real issue "porting" Android to kovsky. Android is a virtualized environment running on a linux
kernel and it is that kernel that has to be ported. "Android" runs 100% on kovsky with minimal modification.
So. There is only one issue with the linux kernel - that issue is when the linux kernel for kovsky
is good enough for someone to make the effort to pack up a ROM with that kernel,
a bootloader and the android system.
Now, the consensus is the kernel isn't as yet good enough for daily use -
if someone with the skills decides it is and makes a ROM, then they will
obviously do so.
The list of kernel stuff to be fixed is broadcast all over this board and asking kernel devs
to repeat themselves ad nauseum won't get it done any faster
To convince someone that FLASHING the ROM on their GalaxyS or Desire is better than STOCK ROM, most people in general will not do this kind of thing lack of knowledge on the subject plus the thought of breaking it
So in order to make someone feel the need to upgrade from FROYO to GINGERBREAD with S-OFF bootloader etc.......how could you simplify it in a way they will understand
I don't convince them. Its their phone and their prerogative. I do link them up to XDA though, and go through basic benefits like smoother operation, better optics, overclocking, some great root only apps etc. But, in the end, I leave it up to them if they want to flash a Custom ROM or not.
Besides, another reason is because I wouldn't want to play tech support every time when each of them run into potential problems flashing ROMs, or experience ROM bugs. Another thing is that most people I know with Android phones get intimated and give up reading the first few intro threads for their device, no matter how simple.
efcgenius said:
To convince someone that FLASHING the ROM on their GalaxyS or Desire is better than STOCK ROM, most people in general will not do this kind of thing lack of knowledge on the subject plus the thought of breaking it
So in order to make someone feel the need to upgrade from FROYO to GINGERBREAD with S-OFF bootloader etc.......how could you simplify it in a way they will understand
Click to expand...
Click to collapse
Some of them simply don't want the hassle.
Gingerbread has security fixes and added features. They like music? What about all those equalizer apps only available to 2.3? I'd also mention the option (for low internal memory devices) that rooting allows (in CM at least) apps to be moved to the SDCard that stock does not. I'm sorry if that wasn't what you were looking for, but those are some reasons I gave Sparky (and He's clueless about this stuff)
Sent from either my N1, NS, or Sensation4G....
Babydoll25 said:
Gingerbread has security fixes and added features. They like music? What about all those equalizer apps only available to 2.3? I'd also mention the option (for low internal memory devices) that rooting allows (in CM at least) apps to be moved to the SDCard that stock does not. I'm sorry if that wasn't what you were looking for, but those are some reasons I gave Sparky (and He's clueless about this stuff)
Sent from either my N1, NS, or Sensation4G....
Click to expand...
Click to collapse
A lot of cool reasons, but most of them just wants a phone that works...
I've been there, and I have managed to get a few (just a few) to move over to Android. Another girl I know is planning to move up to Android as a legitimate smartphone instead of BB. And almost everyone in my church will be switching over to Android (including my pastor). So... I guess tech-evangelism works?
if his your son then sure he will listen to you , but to convince people to do that its quit harder reason because they are afraid maybe or something
usually to convince someone to buy something you have to show him\her what this thing can do , so probably you need to show him some vanilla love or just let him\her suffering of Froyo
3xeno said:
i don't convince them. Its their phone and their prerogative. I do link them up to xda though, and go through basic benefits like smoother operation, better optics, overclocking, some great root only apps etc. But, in the end, i leave it up to them if they want to flash a custom rom or not.
Besides, another reason is because i wouldn't want to play tech support every time when each of them run into potential problems flashing roms, or experience rom bugs. another thing is that most people i know with android phones get intimated and give up reading the first few intro threads for their device, no matter how simple.
Click to expand...
Click to collapse
i def got tired of being tech support so now i just say check out xda you wont regret it