Jump to content


Photo

[WIP] PLi-DarkOS Skin


  • Please log in to reply
2041 replies to this topic

Re: [WIP] PLi-DarkOS Skin #501 littlesat

  • PLi® Core member
  • 57,122 posts

+698
Excellent

Posted 28 November 2023 - 14:27

I agree....


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: [WIP] PLi-DarkOS Skin #502 littlesat

  • PLi® Core member
  • 57,122 posts

+698
Excellent

Posted 28 November 2023 - 14:39

But I just also confirmed that going from 'blend' to 'on' has also strange effects.... So this is not a real solution.


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: [WIP] PLi-DarkOS Skin #503 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 14:40

What effects except crappy rendering of the images?


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #504 littlesat

  • PLi® Core member
  • 57,122 posts

+698
Excellent

Posted 28 November 2023 - 14:43

Look e.g. at the HBBTV icon... on a hisi box.


Edited by littlesat, 28 November 2023 - 14:43.

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: [WIP] PLi-DarkOS Skin #505 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 14:46

Yes that is what @Huevos told you. When you change alphatest to "on" images that have alpha transparency with antialiasing will have this halo around them and you will see under them....

Thats why i am not agreeing to have that changed in the skin. Simply because that will be the effect on all boxes.


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #506 WanWizard

  • PLi® Core member
  • 70,396 posts

+1,807
Excellent

Posted 28 November 2023 - 14:53

I can't really imagine a HiSilicon chip having such an issue with this, after all, it has an ARM Mali GPU, like many millions of other ARM based systems.

 

Neither Kodi nor Android exhibit the same issues when run on HiSilicon hardware, so perhaps the problem is more with the way E2 addresses the hardware (broadcom derived code instead Mali specific code)?

 

If you look in the GDI cpp code, it is BCM everywhere, and in several places there just isn't any code if not run on BCM, So instead of using hardware specific code, it is just generic painting. No wonder it is slow as shit...


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Ultimate (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: [WIP] PLi-DarkOS Skin #507 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 14:54

I think it work great on Vu+ because libgles is used for the blending. Seems HiSi boxes doesnt work well with that or dont work at all with it.


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #508 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 14:59

 

If you look in the GDI cpp code, it is BCM everywhere, and in several places there just isn't any code if not run on BCM, So instead of using hardware specific code, it is just generic painting. No wonder it is slow as shit...

 

Event some BCM based boxes have issue with alphablending....


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #509 littlesat

  • PLi® Core member
  • 57,122 posts

+698
Excellent

Posted 28 November 2023 - 15:03

Yes it all depends on 'drivers' and when you support so many boxes you get limitations.... brrrrr....


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: [WIP] PLi-DarkOS Skin #510 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:06

Some HiSi doesnt suffer from that issues or at least not so visible. Like qviart dual 4K as reported


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #511 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:13

So for info. I have pushed to the skin repo the latest changes. Screen paths are added in menu and all appropriate screens. As well as other screen changes ;)


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #512 littlesat

  • PLi® Core member
  • 57,122 posts

+698
Excellent

Posted 28 November 2023 - 15:14

Where can I find them so I can sync?


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: [WIP] PLi-DarkOS Skin #513 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:15

Repo is here

 

https://github.com/D.../E2-DarkOS-skin


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #514 WanWizard

  • PLi® Core member
  • 70,396 posts

+1,807
Excellent

Posted 28 November 2023 - 15:17

Event some BCM based boxes have issue with alphablending....

 

Yes, but "issues with" is different from "not supported at all".

 

The problem here, with it taking over a second to paint a simple unitform grey square png image, has imho nothing to do with alphablending issues, but with missing code to use the Mali GPU.

 

Even the FORCE_NO_BLENDING_ACCELERATION flag @Huevos mentioned earlier is a workaround used in BCM specific accelleration code, leading me to believe nobody bothered to add Mali support when the first HiSilicon SoC's were introduced, but rather hacked their way around issues with the existing BCM specific code.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Ultimate (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: [WIP] PLi-DarkOS Skin #515 WanWizard

  • PLi® Core member
  • 70,396 posts

+1,807
Excellent

Posted 28 November 2023 - 15:18

Some HiSi doesnt suffer from that issues or at least not so visible. Like qviart dual 4K as reported

 

What is "that issue"? As the movie I posted was exactly from a Qviart Dual 4K.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Ultimate (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: [WIP] PLi-DarkOS Skin #516 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:23

Interesting....I am speaking for the same issue with rendering of infobar....But it was tried on OBH with same box and it was not reported to have such a visible effect...

Maybe because of that sync/delaying code that you said there is on OE-A based images...


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #517 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:26

 

Event some BCM based boxes have issue with alphablending....

 

Yes, but "issues with" is different from "not supported at all".

 

The problem here, with it taking over a second to paint a simple unitform grey square png image, has imho nothing to do with alphablending issues, but with missing code to use the Mali GPU.

 

Even the FORCE_NO_BLENDING_ACCELERATION flag @Huevos mentioned earlier is a workaround used in BCM specific accelleration code, leading me to believe nobody bothered to add Mali support when the first HiSilicon SoC's were introduced, but rather hacked their way around issues with the existing BCM specific code.

 

 

Possible. HiSilicon is used in E2 STB from not long time i guess....


Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #518 WanWizard

  • PLi® Core member
  • 70,396 posts

+1,807
Excellent

Posted 28 November 2023 - 15:28

That what I wrote, indeed.

 

OE-A has hacked some workaround in so you don't see the buildup on screen, but instead nothing happens for over a second when you press the OK button before the infobar is shown.

 

Which imho is even worse, as it happens everywhere in their UI, and it makes the image feel very sluggish even though the actual response time is the same.

 

It is the same as seeing a website load versus displaying a spinner before showing the complete page. Human perception is that the first load quicker, even though a stopwatch would disagree.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Ultimate (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: [WIP] PLi-DarkOS Skin #519 DimitarCC

  • PLi® Contributor
  • 1,566 posts

+68
Good

Posted 28 November 2023 - 15:31

So i've been told that for svgs is forced alphablending even if it is not enabled....so thats why it could be sluggish on problematic boxes if you have many svgs

 

But even removing them will not fix anything on HiSilicon boxes...

 

Infobar doesnt have svgs at all but still doesnt render well.


Edited by DimitarCC, 28 November 2023 - 15:33.

Vu+DUO4KSE, DM920UHD, Vu+Uno4KSE, SF8008Mini, 2xPulse4K, Vu+Solo2, Dreambox DM500HD, Triax 78 (7E,9E,13E,19.2E,23.5E) & 2xTriax 78 (39E)


Re: [WIP] PLi-DarkOS Skin #520 WanWizard

  • PLi® Core member
  • 70,396 posts

+1,807
Excellent

Posted 28 November 2023 - 15:32

Possible. HiSilicon is used in E2 STB from not long time i guess....

 

Vendors switched to HiSilicon after Broadcom made a deal with content producers and refused to make chipsets for the OEM market anymore. As a manufacturer you can only get Broadcom SoCs now if you make closed source provider boxes.

 

And now the problem is that although HiSilicon designs the chipsets, they are manufacutered by TMSC in Taiwan. Which is no longer allowed to do so due to US sanctions against China ( Huawei, the parent company of HiSilicon ).

 

I know some are trying to develop new boxes using AMLogic chipsets, but these are so bad that until now not much luck.
 


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Ultimate (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.



7 user(s) are reading this topic

0 members, 7 guests, 0 anonymous users