Firmware 4.33.19608

[Deobulakenyo]

Originally Posted by windirt

yes, I'm using the 0.11.3, but I saw the 0.11.4 just released.

0.11.3 is fine on my Libra 2. that version solved the problemes i had before.

Here is my config file, just edit out the parts you want removed.

[Quoth]

I've reverted to 4.32.x on the Sage as I can't see any advantage to 4.33.x and Advanced Notebook was affected by 4.33.x: handwriting would vanish while writing and reappear later. Notebook exporting needed two attempts. On 1st it would be ages if you didn't tap cancel and 2nd attempt instant.

[PeterT]

4.33 is meant to have made changes to battery management on the Sage and Libra 2

Sent from my Pixel 4a using Tapatalk

[Quoth]

Originally Posted by PeterT

4.33 is meant to have made changes to battery management on the Sage and Libra 2

Actual management or just the display of the status? In any case poor usability of the Advanced Notebook means 4.33.x isn't viable on the Sage N778. I didn't notice much difference to the battery usage anyway.

[NiLuJe]

I'm five minutes into running 4.33 on the Sage and I've already experienced a whole bunch of wonky-as-hell screen refreshes (reminiscent of how broken the display driver got on the Elipsa when you enabled the display driver's verbose debugging flags via debugfs).

Wheee.

[Quoth]

Originally Posted by NiLuJe

I'm five minutes into running 4.33 on the Sage and I've already experienced a whole bunch of wonky-as-hell screen refreshes (reminiscent of how broken the display driver got on the Elipsa when you enabled the display driver's verbose debugging flags via debugfs).

Wheee.

Yes, 4.32.x seems better for Sage & Elipsa. I only had a problem with the Notebook, but I have reading refresh set to every page.

I've not tested the Libra (original) much on 4.33.x, but it seems OK.
I was going to recall the Nia and update it from 4.32.x to 4.33.x, but as it's working and the user is happy I won't. I told them not to set up WiFi.

[NiLuJe]

No kernel update on the Elipsa (nor was there in 4.32), it's still a February build.

On the other hand, the Sage got an updated kernel both times.

[davidfor]

Originally Posted by NiLuJe

I'm five minutes into running 4.33 on the Sage and I've already experienced a whole bunch of wonky-as-hell screen refreshes (reminiscent of how broken the display driver got on the Elipsa when you enabled the display driver's verbose debugging flags via debugfs).

I noticed some weird updating when opening a new sideloaded kepub on the Sage over the weekend. I meant to have a look at the image used for the first page to see if there was anything unusual about it. I also had the ghosting get really bad when reading at some point. After starting the next chapter it was back to normal. But, I don't know what I had done just before this.

[NiLuJe]

That was on the home screen on my end, and was fairly easily reproduced by repeatedly showing/dismissing the same menu popup (be that the wifi or battery one, or NickelMenu's).

That sometimes lead to delayed and tiled flashing in roughly four quadrants (of varying heights, IIRC, and possibly refreshed in sequence in counter-clockwise order, I'd say?).

(That it seems to think that it requires an update, and worse, a flash, in largely unrelated parts of the screen at all is already a bug in the collision handling, but, oh, well ^^).

[windirt]

Originally Posted by Deobulakenyo

0.11.3 is fine on my Libra 2. that version solved the problemes i had before.

Here is my config file, just edit out the parts you want removed.

I do compare both of our ini files. the only difference are two options:

backgroundless=false
overlay=false

I set both options to true to get better looking contrast and borderless effects.

At first I was more suspicious of the autofresh option, but you also set it to true, so those are the only two options left that could cause the crash.

I'll try to install 0.11.04 and try the new config again.

[davidfor]

Originally Posted by NiLuJe

That was on the home screen on my end, and was fairly easily reproduced by repeatedly showing/dismissing the same menu popup (be that the wifi or battery one, or NickelMenu's).

That sometimes lead to delayed and tiled flashing in roughly four quadrants (of varying heights, IIRC, and possibly refreshed in sequence in counter-clockwise order, I'd say?).

(That it seems to think that it requires an update, and worse, a flash, in largely unrelated parts of the screen at all is already a bug in the collision handling, but, oh, well ^^).

I can do that as well, but, only if I don't let the popup finish painting. If I wait for the popup to finish, it doesn't refresh the rest of the screen I don't see the full screen refresh. And even when I do it to quickly, it takes few taps to get the flash. And I think when it happens, I end up with multiple popups opened.

I don't find that a problem. Isn't this a case where the number of changes to parts of the screen trigger an escalation of the refresh to the full screen? And it is happening when you are doing an unusual action. But, the frequency of the updates might be what is triggering the escalation. It doesn't happen on my other devices, but, I don't have a Libra 2. That has the same type/model of screen. Does it happen there?

[Colonel Cathcart]

I used to get updates fairly quickly (just an impression, I didn't take note of the actual time since release) but still haven't gotten this one.
I wonder if Kobo made changes to how update propagation works.

[Lykke]

Had a freeze yesterday night, while reading Peter Singer's ramblings. A book from Overdrive, not sideloaded but downloaded by Kobo (so I assume it is KePub). Seemed random to me; was just going to turn the page when I noticed nothing responded. Can't see that the book is any heavy load on the device; 950+ pages but no images. Forced reboot fixed it.

[NiLuJe]

Originally Posted by davidfor

I can do that as well, but, only if I don't let the popup finish painting. If I wait for the popup to finish, it doesn't refresh the rest of the screen I don't see the full screen refresh. And even when I do it to quickly, it takes few taps to get the flash. And I think when it happens, I end up with multiple popups opened.

I don't find that a problem. Isn't this a case where the number of changes to parts of the screen trigger an escalation of the refresh to the full screen? And it is happening when you are doing an unusual action. But, the frequency of the updates might be what is triggering the escalation. It doesn't happen on my other devices, but, I don't have a Libra 2. That has the same type/model of screen. Does it happen there?

I don't *remember* the flashing thing happening on mxcfb hardware, FWIW, but the latest one I have is a Forma, not a Libra 2 .

FWIW, I wasn't really spamming taps, those were fairly spaced out taps, so the duplicate popups thing doesn't ring a bell to me .

And I don't *think* the Home screen is set-up to escalate to a flash (ever), either .

[NiLuJe]

Definitely not happening on a Forma (at least on 4.31, will update once the battery goes above 20% [EDIT: Same on 4.33]), and the refresh is perfectly fine and localized to the icon & popup:

[pid   193] 14:48:31.176864 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=0, left=1056, width=90, height=139}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=175, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000694>
[pid   193] 14:48:31.311992 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=0, left=1056, width=90, height=139}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=176, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000596>
[pid   193] 14:48:31.313827 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=144, left=676, width=721, height=415}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=177, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000342>
[pid  1256] 14:48:31.355594 [71b974ec] ioctl(46<socket:[9058]>, FIONREAD, [631]) = 0 <0.000738>
[pid   193] 14:48:34.170359 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=144, left=676, width=721, height=415}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=178, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000622>
[pid   193] 14:48:36.635961 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=0, left=1056, width=90, height=139}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=179, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000605>
[pid   193] 14:48:36.745592 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=0, left=1056, width=90, height=139}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=180, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000601>
[pid   193] 14:48:36.747433 [71b974ec] ioctl(3</dev/fb0>, MXCFB_SEND_UPDATE_V2, {update_region={top=144, left=676, width=721, height=415}, waveform_mode=WAVEFORM_MODE_AUTO => WAVEFORM_MODE_AUTO, update_mode=UPDATE_MODE_PARTIAL, update_marker=181, temp=TEMP_USE_AMBIENT, flags=0, dither_mode=EPDC_FLAG_USE_DITHERING_PASSTHROUGH, quant_bit=0, alt_buffer_data={phys_addr=NULL, width=0, height=0, alt_update_region={top=0, left=0, width=0, height=0}}}) = 0 <0.000170>

It goes icon highlight -> icon unhilight -> show popup -> close popup -> etc.

And nothing ever flashes, either in Home or Library views.