Search results

Has this been confirmed? I just received a new Fire 7 and no amount of shorting could move past the "Serial protocol mismatch" error. Tried with various points labeled CLK, CMD or DAT0 on the back of the PCB.

If we trust the app, then also A/B updates should be enabled. Already did the build props test. Returns true. EU-version, shipped by Amazon.

According to the test app it does support Treble, but I'm not sure whether to trust it. Came with an update out of the box and I didn't check the version before applying it (which would have made that clearer).

Hmm, who might that be? ;) Yes, there's no setting allowing us to differentiate between flashing RGB and on/off notification lights. Yes, these settings don't know about notification lights that can only be on or off. I switched to Marshmallow and won't touch CM13 until it's at least a bit...

Oh, come on! Would you please stop spreading facts? Looking at the commit (superficially) I actually can't see a real reason why adding support for an encoding format should result in a "laggy" recents button.

Well, you're essentially spamming people. What did you expect?

I first noticed this a few days ago. It's really unbelievably ugly. Couldn't pinpoint the problem via git though and don't really have an idea which commit in which project might be responsible. Maybe next weekend when I find some more time.

No, actually you don't ;). I'm not really sure that it improves anything at all. It just makes an error go away that distracted me a little bit. Judging by the code that drives input devices, wrong configuration parameters are simply reported and then ignored.

Took me some time to catch up with this thread, but well, better late than never, right? I'm actually following the posts in here and am quite glad that there's someone providing a vanilla CM build so I can get sort of direct feedback. Don't hesitate to PM me "bug reports" – I can't promise...

I'm part of CM (standing on the shoulders of giants), so rather no ;)

People, stop panicking. :p There's no "hopefully, maybe, let's see". The Transformer Prime will get 4.1, and that's a fact, because its hardware is really close to the Nexus 7. I'm already running a completely stable version of Jellybean on my Prime, lacking a few features of course, but we...

I'm sorry that it broke your device, but I can't do more than pointing out that these files are invalid. So I point out that these files are not only useless (as most GPS fix instructions are) but also broken and invalid and what do I get? Some random guy jumping at me telling me how stupid I...

It is a config file parsed per line. This means that there shouldn't be any new line characters for lines that the system has to interpret as one. It's perfectly OK to use a local NTP-server, by the way. It just doesn't fix problems at all. The short differences in the time the packets take...

This is not needed for old stable releases and was fixed a few weeks ago (so recent RCs and nightly builds already contain that bugfix). All of the files I checked there are invalid.

Yepp, he is. We've got more people working on it though (like Rayman and myself).

Well, then you actually know more than I do. :) Maintaining the Gingerbread port doesn't mean I'm not looking into possibilities for ICS on the Legend – we're just not doing half-done crap-releases. CM is our baby, we treat it nicely. :D I'm still here, my Legend's still working and it (and a...

He's right, believe me. From the provider of the modified lib: He reverted a commit. He didn't add logic for reading new config values. The modified lib might very well improve GPS for a lot of you (I still need to test this on my device, maybe tonight). The modified gps.conf won't. Ever.

Please don't forget that even if there were binary drivers suitable to run msm7k on ICS we'd still need an updated kernel (kgsl, etc) to use them.

No, unless there's a miracle.

That nonsense is haunting this forum like nothing else. I'd say it's the king of an army of nonsense. Hail to the king. On a more serious note, let me quote myself: The short snippet you can see above was actually copy/pasted from the source code of our GPS module. This means the values are...

This is going off-topic. Sorry thread-starter, I know that. ;) Anyway, nope, I'm still maintaining and testing the CM7 branch for my Legend (it's my main device), but nothing apart from that. I ordered a tablet a few days ago though, let's see how much fun this is going to be. :)

The msm7k in our device was designed and manufactured to run at 600MHz. As far as my experience goes suggesting to overclock isn't really a good thing to do, because, as I said, whether your device is capable of running at a certain speed is dependent on the chipset itself. I got various...

Whether or not an overclocked device is stable doesn't depend on a random number (like 807) but rather on the hardware – in our case the build quality of the CPU – itself. I for example can't even overclock to 7xx without having my device crash after a few seconds (a minute at best). O rly...

As usual. I never said the Legend was going to be updated to CM9, even though I wished that was possible. Exactly. I'm aware of a lot of things (but thanks for the reminder on what I should do, I love it when people do that). I know users wanted to see a fully functional ICS running on their...

No one ever said something about dropping older devices. We will still support them and do CM7 releases.

I myself couldn't find a statement about that. What you call "judgment" are assumptions based on your little knowledge about Android and developers. That's unfair at best, but I'd rather call it inappropriate. Moreover, because of that little knowledge you start to insult people - or at least...

If you look at the list of devices that are already supported by CyanogenMod 9, you will see that this statement is false. Sorry, in my world and judging by my standards a firmware is clearly not complete in any way – not even "almost" – if it doesn't support the camera at all. Or accelerated...

In the thread you referenced above the developers clearly state that camera isn't working. This is exactly what I tried to explain. As you can see the Legend is not the only device affected. By the way, I didn't understand half of your post. Care to rephrase it a bit?

Oh, you got me there! I don't know ****. Of course, in contrary to what I keep stating, it's not impossible to port ICS for older devices and their ancient binary drivers. You can try to rip out all of the shiny new stuff that is fundamentally incompatible and "replace" it – as if it was that...

We never got any source code supporting the Legend's GPS version. So we are using a chimera built of sources supporting older and newer hardware. This is a bit problematic, it's only guessing how the hardware might work. So if there are hardware quirks we'd have to address we can't because we...

I stopped telling everyone they are wasting their time with the so-called fix. At some point I realised that people obviously want something they can believe in and no matter what I tell them they will still rather believe than think (for some the fix that doesn't change anything even magically...

People stupid enough to copy/paste these commands would rather deserve /boot, /recovery, /system and /data wiped if you ask me.

Yes, there will be support for msm7x27 - mostly because we have such awesome maniacs like arcee :). Unfortunately msm7k devices differ a lot and what works on one device might very well be broken on another one. Those are "only" the basics. I won't comment on translation layers until I had the...

Currently I'm not working on an ICS port for the Legend. There are a few requirements (code wise) that make any ICS port for the Legend almost unusable. For a few drivers there's source code available, those are not that big of a problem. On the other hand, camera support was rewritten (again)...

I still don't get why people always think it's sufficient to wave some (most of the times ridiculously low amounts of) money in the air to make everyone dance to their music. :rolleyes: Fact is, you won't be able to afford paying a developer full time. You won't find a good guy that is not...

Said who? I didn't, for sure.

No big deal guys. Just put a few hundred hours of work into the ICS sources and it will probably work on the Legend. Or rather not.

Oh, cool, you are already hacking on it to make it work on the Legend?

They turn your device into dust. Radioactive dust, so quite unstable.

FYI: Apart from the server URL all settings in the so called "GPS fix" update are bogus. Believe me, I know the code of our GPS library quite well and those settings either don't exist or don't do anything useful at all.

Please, guys, don't get too excited about Ice Cream Sandwich. Chances are really high - actually a lot higher than for every other release so far - that we will never see a 100% working release for various reasons. Yes, from what I learned so far hardware-wise the Legend could very well be...

Even my very first releases of CyanogenMod for the Legend (going back as far as Éclair) did not use the LED to signal full charge. Solid green (or blinking, if you enable it in the settings) is used for notifications. Orange for charging or low battery. The blinking green LED is sometimes hard...

Actually they already developed that and called it "Dalvik" :D. It's what makes running the same app on completely different devices possible. Still, for a hyper-visor to work you need a hardware abstraction layer and that brings us back to the problem I mentioned above.

Just a small comment regarding kernel 2.6.35: In my tests 2.6.35 was not significantly (or at all) faster than .32, neither could I observe improved battery life. What actually held us back implementing .35 were the crappy WLAN and missing access point drivers. The sources HTC released are...

No, certainly not. So you mean that Google will release ICS supporting all those (roughly estimated) 400 completely different platforms and a huge load of even more devices? I absolutely doubt that.

First and most important rule: Never touch the hboot partition. Second and almost as important rule: If you really think you have to mess around with the hboot partition (but, believe me, you don't) make sure you know exactly what you are doing. Now, in this case knowing exactly doesn't refer...

If your tree was up to date before kernel.org repos went down (which is likely, Google didn't push an update for quite some time) you can sync using "repo sync -f". This will effectively only sync the CM repos on github which fortunately isn't down at all.

adb pull /proc/config.gz gunzip config.gz

Bet your ass I will. Umm, I'm pretty occupied already...

Please test the attached update package after installing CyanogenMod.