BionicDog (updated: 2018-06-04)

[Mike Walsh]

@ Fred:-

The silly thing is, it's not the desktop that's freezing-up; that still responds!

It's just the individual applications that are seizing solid. Figure that one out, if you can.....


Mike.

Weird..., let me understand exactly "it's not the desktop that's freezing-up".
You mean mouse and keyboard are still working, but that's all.. or... ?

Fred

Let me try to be a bit more descriptive. Mouse and keyboard still function, yes. I can click on things, and they respond, yes. I can open new applications. But after a minute or two of anything being opened, it ceases to respond to any controls.....including the window manager.

I've lost count of the number of things I've had to kill, because they refuse to close. I've never had anything like this happen before, which is why I'm so puzzled.....


Mike.

[Mike Walsh]

Mike: From the Xorg.0.log, it looks like the Radeon driver was loaded. I'm guessing that the RS480 card doesn't need any additional firmware, so X was able to start on its own.

Since you've got the right video driver, I don't know what else to suggest.

LateAdopter: Thanks. I didn't understand that either.

Hi, Bill.

Mm. I'm stumped, too. I do hope this isn't a sign of things to come...

I know I've heard of situations where a certain piece of hardware works perfectly for dozens, if not 100's of kernel 'point' releases.....then all of a sudden, it quits (or starts behaving in a totally different way.) Lots of people complain; within a couple of releases, someone at kernelHQ has got his wrists slapped for changing things that didn't need changing, and.....hey presto! things are back to how they were before.

Perhaps a kernel change would help. How do you actually do that with these 'Dogs'.....or can you perform the same 'manual' swap we do in Puppies?


Mike.

[fredx181]

Perhaps a kernel change would help. How do you actually do that with these 'Dogs'.....or can you perform the same 'manual' swap we do in Puppies?

Yes, worth trying, if you have time can you try this 4.4 kernel ? (it's from Xenial) EDIT: 32 bit
https://dl.dropboxusercontent.com/s/b18 ... ar.gz?dl=1
Contents of this archive:
- 00-kernel-4.4.0-124.squashfs
- initrd1.xz
- vmlinuz
Just put these in the "casper" folder.
Replace original initrd1.xz and vmlinuz with the ones from the archive.
(better make backup of original files first by renaming them).

I'd appreciate if you test this, then we possibly know if the problem is because of the 4.15 kernel or not .

EDIT: Ooops, I realize now that I don't know if you used 32 or 64 bit.
Anyway this is for 32 bit.

Fred

[Mike Walsh]

Hi, Fred.

Yeah, no worries, mate; be happy to. Won't be till the morning, though; it's 2.30 am here in the UK, and I'm more than ready for some shut-eye...

Bear with me, and I'll let you know what happens sometime tomorrow. We'll see if it makes any difference.

(BTW; I am using 32-bit, yes; it's my usual preference if I have a choice.)


Mike.

[zagreb999]

REGARDS

WHO KNOWS TO CREATE
OKULAR GTK IN BIONIC?

THANKS!

[Mike Walsh]

Morning, Fred.

Well, I don't know if this is what you were expecting, but this is what happened:-

Swapped over initrd1.xz and vmlinuz. Booted up. The boot process went OK, right up until the end.....then there was a load of stuff about the X server, amongst which was

Code: Select all

...Fatal server error. No screens found.....

.....and which ended up with

Code: Select all

...terminated with error (1). Closing log file.
xinit: giving up
xinit: unable to connect to X server. Connection refused.
xinit: server error
root@live: ~#

Attached is the current contents of /var/log/Xorg.0.log. Remove the fake .gz, of course.

I'll leave it with ya.


Mike.

[fredx181]

Hi Mike,

Looks like you did something wrong, I tested this and it should work booting with the 4.4. kernel.

Code: Select all

root@live:~# uname -a
Linux live 4.4.0-124-generic #148-Ubuntu SMP Wed May 2 13:01:51 UTC 2018 i686 i686 i686 GNU/Linux

In your Xorg.log I see you are still booting the 4.15.0-15 kernel

Code: Select all

[    35.627] Current Operating System: Linux live 4.15.0-15-generic #16-Ubuntu SMP Wed Apr 4 13:58:15 UTC 2018 i686

Ok, so you replaced vmlinuz and initrd1.xz with the ones from the archive, but did you also add "00-kernel-4.4.0-124.squashfs" in the casper folder ?

If you didn't, can you try again ? No need to hurry

EDIT: Oh, I see just now that your attached xorg.log is from yesterday... confused...

Code: Select all

[    35.641] (==) Log file: "/var/log/Xorg.0.log", Time: Sun May 13 11:38:45 2018

Fred

[tlchost]

27818 root 0:00 grep blue

That indicates that the bluetoothd daemon didn't start. In my setups, it always starts as soon as you plug in an adapter.

If it doesn't start first, the bluealsa daemon probably won't start either.

Type: /etc/init.d/bluetooth start
What happens?

Which Bionic Dog version are you running? 32 or 64bit?

Didn't start....Maybe it's because I'm running on a usb drive and it's the 18.5 pae version?

Sorry to have troubled you....I don't think I'll chase after bluetooth any more...taking up too much of my time and yours.

[rcrsn51]

I released this project back in January. Since then, there have been many anonymous downloads and zero feedback.

So I can't say if your situation is the exception or the rule.

18.5 pae version

I don't know what that is.

[tlchost]

I released this project back in January. Since then, there have been many anonymous downloads and zero feedback.

So I can't say if your situation is the exception or the rule.

18.5 pae version

I don't know what that is.

UPup Bionic Beaver (upupbb) 18.05 (May 2018)

Not a problem...Glad it works for others...thankfully I have decent speakers that do not require bluetooth.

[rcrsn51]

UPup Bionic Beaver (upupbb) 18.05 (May 2018)

??????
That's a Puppy. BionicDog is not. BT4Stretch is designed to work in a full Debian/Ubuntu environment. Puppy doesn't have that.

Glad it works for others.

?????
If no one reports, I have no idea how it works for them.

[rcrsn51]

@majka: Any update on this situation?

[Mike Walsh]

Hi again, Fred.

Yah; my fault, I think. I only put the initrd1.xz and vmlinux into /casper. I'll be honest, I couldn't see what the 3rd item (the squashfs one) was supposed to do!

It all boots up OK now. However, I still can't import bookmarks into Quantum..... Can't convert my Google-Chrome 48.0.2564.115.pet into a .deb, either; pet2deb says the file doesn't exist, it's hit an unrecoverable error.....now closing. Mind you, that could be a permissions thing, because of the file's location when I was trying to convert it. I'll experiment with that a bit more.

I can't quite figure it out. The FF-Quantum portable you put together works brilliantly in all my Pups.....yet the version you've built into BionicDog seems to have a bug in it. At least, the copy I've got hold of seems to....???

(The 'freezing-up'/unresponsive thing appears to be confined to the Firefox 'Library' window during my repeated attempts to import my bookmarks from an HTML file. It's not affecting anything else.....YET.)

Still, we've made progress. If I can install gDebi from the repos, I might see if I can install a newish version of Iron, instead. It's my favourite Chrome 'clone'. I'm guessing I'll need to log-in as 'puppy' for that, though, since they definitely won't run as /root.....


Mike.

[fredx181]

Hi Mike, looks like the older kernel solved your problem.
That's good news ! (but also bad news, because it seems that the newer kernel doesn't support some hardware anymore).
I'll make similar package with the 4.4 kernel for 64-bit later (as fallback, there may be others with the same problem)

Can't convert my Google-Chrome 48.0.2564.115.pet into a .deb

Just my opinion: I wouldn't recommend using pet packages.
Btw, you can ignore all messages like "Cannot stat ...." when converting pet to deb.
(it needs to be fixed, by suppressing these messages, on TODO list...)

Here's a deb for google-chrome 48 32-bit.
After install there should be menu entry "Google Chrome (as puppy)" (starting script: /usr/local/bin/chrome-puppy.sh)
https://fredx181.github.io/bionicdog/No ... 1_i386.deb
First do (if not recently done):

Code: Select all

apt-get update # update the package lists

Then can be installed by right-click from pcmanfm > Install deb (no need for Gdebi)
Or from terminal:

Code: Select all

dpkg -i google-chrome-stable_48.0.2564.109-1_i386.deb
apt-get -f install # force installing missing dependencies

EDIT: Strangely, the first time running Google Chrome from menu it doesn't start, but when I tried again it does start (and everytime after)

The FF-Quantum portable you put together works brilliantly in all my Pups.....yet the version you've built into BionicDog seems to have a bug in it.

Mmm.. strange, because I've put it together including the same binaries.

Fred

[rcrsn51]

@mikeslr: On May 3, you started a discussion about PeasyWiFi and WEP encryption.

You never returned, so I set up another router with WEP. It worked for me.

A confirmation from you would be nice.

[fredx181]

For info:
MBinnun made a new version of MBPUP (remastered BionicDog) v1.3, very enhanced, with different "flavors" to choose from :
https://github.com/mbinnun/MBPUP/blob/master/README.md

Fred

[Mike Walsh]

Mornin', Fred.

A query for ya. Can we, or can we not, add Ubuntu PPAs in BionicDog?

I'm only asking, because one of the necessary commands for adding any PPA is

Code: Select all

apt-add-repository

.....and bash insists the command isn't recognised. Or can't that be done with the Live version of Ubuntu?


Mike.

[rcrsn51]

I believe that it's add-apt-repository.

[Mike Walsh]

Hi, Bill.

D'you know, I'm inclined to agree with you. All the time I've been looking at this, I've been thinking it doesn't look right. I'll try it the other way round before the afternoon's out; see what happens.

Meanwhile, I'm glad I've got you on, mate. Do you like a good mystery, Bill? How about you, Fred? If the answer's 'Yes', then try this on for size....

The subject here is audio. Specifically, audio recording & playback. Note the last word. That's the key item, here.

-----------------------------------------

My Linux journey started a little over 4 years ago. In May, 2014, to be precise, when I downloaded and installed the then-new Ubuntu 14.04 'Trusty' LTS.

Since I started with Linux, I have used the same identical hardware, all the way through. Yes, peripherals have changed and been upgraded, but the motherboard has remained the same.....including the audio chip. A RealTek circuit, built-in to the SB400 southbridge on this MSI mobo.

I often recorded streaming audio during the first year or so I used Ubuntu, before I began getting graphical freeze-ups, and switched to Puppy....but that's not the issue here. The point is, I recorded audio, then played it back again later. And every time, it played back perfectly.

When I moved to Puppy (first dual-booting with Ubuntu, then all-Puppy), naturally, I wanted to do the same thing. And y'know what I discovered? Oh, Puppy would record audio quite happily.....but playback was another kettle of fish entirely. It always played back quite slowly, and sounded like the audio was being dragged through a tub of molasses. TBH, it sounded crap. And no amount of adjustments or re-routing the audio output seemed to make the slightest bit of difference.

I got to the stage where I'd just accepted that this was one aspect of Pup that would never, ever work properly.

---------------------------------------

Fast forward to this last few days, when I decided to try out the 'Dogs' for the first time. This BionicDog is, of course, based on Ubuntu again. Just this morning, I thought about this problem, and, for a laugh, I decided to see what would happen when I tried to record a short video clip with GuvcView.

I've recently treated myself to a top-of-the-range Logitech webcam; the c920 HD Pro, so much better quality than what I've been using for the last couple of years. The clip recorded fine; Bionic recognised the cam's video chip, and configured it all correctly.....even Pup does that, because I've always made sure I buy UVC-compliant cameras, which are automatically supported by the kernel's UVC driver module.

What flabbergasted me was that BionicDog also not only recognised the cam's audio chip, and configured it properly, but the audio played back perfectly when I ran the resulting .avi clip through the default mpv media player.

No 'slowdown'. No 'treacly' quality to the sound. Just perfectly crystal-clear playback from the cam's stereo microphones.....

Remember; this is the same identical audio hardware since day 1 of my Linux 'journey'. It can't have owt to do with the kernel, either, since I'm running Pups with newer kernels than this. Yet not one Pup can play back recorded audio correctly. (I have no problem with playing back in media players, or streaming through the browser; in those, audio works correctly.)

-------------------------------

So; my 'question', therefore, is this:-

What the HELL is Ubuntu doing with this audio sub-system that Puppy seems incapable of emulating? What's different? It's NOT the hardware, obviously.....it's got to be the software, or the way it's implemented. But what is being done differently?

Anybody got any ideas? Suggestions? Even a 'hint'or two might point me in the right direction, because this has been bugging me for a very long time.....

Please don't tell me it's something to do with PulseAudio. Even though that's about the only aspect of all this that seems as though it might be at the 'root' of the problem.....because it's always been a PITA to get working in Pup..!


Mike.

[rcrsn51]

The first thing I would check is the ALSA version.

Code: Select all

cat /proc/asound/version

Are new Puppies getting a new version or are they recycling the same old version out of woof?

Linux developers are building their apps on a standard mainstream platform. Puppy is not a standard platform.

Just because you pulled a bunch of packages out of Ubuntu doesn't guarantee that they will work like they do in Ubuntu.