[DEV][UPDATE: 2-10-11][Optimus One multitouch fixes]

Sippolo

Senior Member
Feb 3, 2011
92
19
0
Any eta on a new version with performance improvements or anything like that or there's nothing else possible to do?
Currently the only improvement (which I'm not sure actually) that could be reached, is about a smoother transiction when workaround acts and switch the axis, but I dont think it's worth the job.
That's because any more in depth coords calculation would, in my opinion, break what Der_Caveman has done so far, which is a really good performance rework of the kernel procedures, discarding a lot of useless calls and variables handling.

I don't think any more performance, in terms of fps and CPU using, can be ever reached with this touchscreen kernel, furthermore it's actually quite better than before right now.


My 2 cents though ^^
 

franciscofranco

Recognized Developer
Dec 9, 2010
24,725
136,397
0
Carcavelos
Seems reasonable. Indeed the work Der_Caveman is pretty good, I'd rather not have a kernel with his work "broken" like you stated.

I do believe, like you've also said, that the amount of work needed to fix the axis inversion bug, totally, isn't worth the shot and the time. I also think that we have one, if not THE best android phone out there for the price/performance ratio, and all that thanks to this fabulous comunity, I couldn't be more happier with my android than I am now :)
 

bluviper

Senior Member
Oct 6, 2010
918
124
73
MKTICBD
See yanuk's link. These are just workarounds to compensate for the inversion(Sippolo and waleeq's kernel). Der_Caveman's work is for the optimization of the driver :)
so youre saying that Der_Caveman's V3 is the most we get for an optimisation.
im currently using Sippolo's V1.1?

how does that par with Der_Cavemans V3?
will you prefer V3 over V1.1?


thanks in advance mate!
 

Sippolo

Senior Member
Feb 3, 2011
92
19
0
@bluviper - Sippolo's work is based on Der_Caveman's kernel. If you really need the axis inversion workaround, go with Sippolo's. The only difference between the two is that the former has on extra function call (correct me guys if I'm wrong) that inverts the inversion :)
It's not even a function call, it's just a single block within the multi touch sampling code that acts the workaround if needed :)
As for the comparison with Caveman v3, don't forget that combined kernel v1.1 contains also the pressure sensitivity support from mik_os v1.1 kernel (optimized).

The kernel base it's Caveman v3, so in terms of performance, it should be exactly the same :)
 

Zerobarat1

Senior Member
Feb 1, 2011
1,003
59
0
Lol , the phone is stable and fast as it is :) . What more do you want ?
Either way since the source code has been released , we might see some other improvements in other departments :D.
I wonder on another note , when is LG gonna announce the 2.3 official update. Considering they're one step ahead of everyone else ( Samsung's Ace phone which kinda resembles the P500 bearly came out in Janurary ) and LG optimus 2x which is the first phone to sport the new super features dual core and Tegra 2 .
I hope LG will bring LG Chic and O1 the 2.3 official by the 2nd half of March . ... I hope
 

lupohirp

Inactive Recognized Developer
Jul 27, 2010
977
7,399
0
Avellino
Lol , the phone is stable and fast as it is :) . What more do you want ?
Either way since the source code has been released , we might see some other improvements in other departments :D.
I wonder on another note , when is LG gonna announce the 2.3 official update. Considering they're one step ahead of everyone else ( Samsung's Ace phone which kinda resembles the P500 bearly came out in Janurary ) and LG optimus 2x which is the first phone to sport the new super features dual core and Tegra 2 .
I hope LG will bring LG Chic and O1 the 2.3 official by the 2nd half of March . ... I hope
mmm....i don't think so....lg hasn't released updates with sd unmount problems....and we are at the end of february....
 

M..N

Senior Member
Jul 14, 2009
1,212
99
0
Hey devs do you have an issue with 2 point locking. Basically when 2 point get close together they instantly lock. Could this be looked at?

Sent from my HTC HD2 using XDA App
 

Der_Caveman

Senior Member
Dec 24, 2010
169
99
0
Hey devs do you have an issue with 2 point locking. Basically when 2 point get close together they instantly lock. Could this be looked at?
The answer is: melfas.
Either their controller is not capable of "mutual capacitance scanning" or it is not programmed to do so. There is no possibility to work around this bug by software.

I contacted melfas and asked them if their chips can do "mutual capacitance scanning" but I got no answer. Seems to be a "sales droid only" company :mad:
Maybe I ask them again and claim to buy 1000000 chips :(
The world would better do without such companies...
 
  • Like
Reactions: kpbotbot

M..N

Senior Member
Jul 14, 2009
1,212
99
0
Well the touchscreen i use is one for the hd2, (the device im working on) apparently this point lock/line detection was implemented by htc to support some apps. However im sure this can be disabled but i am not sure as where to look for this specific line in the ts driver source code, any direction would be helpful.
 

Der_Caveman

Senior Member
Dec 24, 2010
169
99
0
Well the touchscreen i use is one for the hd2, (the device im working on) apparently this point lock/line detection was implemented by htc to support some apps. However im sure this can be disabled but i am not sure as where to look for this specific line in the ts driver source code, any direction would be helpful.
Just add the touchscreen driver source as attachment and I'll look over it.
 

Der_Caveman

Senior Member
Dec 24, 2010
169
99
0

M..N

Senior Member
Jul 14, 2009
1,212
99
0
That does not seem to be the same touchscreen controller chip as in the optimus one. I see no position calculations in the source- the coordinates are directly sent to userspace.
Yes I know it is different, but is the any indication of where to modify line detection?

Sent from my HTC HD2 using XDA App