There is something wrong with the way puppy linux way of handleling WiFi adapters

[joet12345]

I base this conclusion because I use multiple OS's e.g. windows 10, linux mint, other distros, and in puppy linux I keep having WIFI adapter issues... while in the other distros it is non existent.

Sorry but puppy wifi adapter support needs to improve... is not the adapter, it is the code...

Just one example, I have a few of these old WiFi adapters: http://www.amazon.com/D-Link-DWA-140-RangeBoos ... B0010T8X54

[joet12345]

And they work fine under Linux mint, ubuntu, others but puppy Linux struggles...

Same scenario with other adapters....one of the reasons it is stopping me to use puppy Linux daily... I remember Wi-Fi issues in the early stages of Linux so that reminds me of puppy Linux in the year 2021..

[ozsouth]

@joet12345 - the other OSes you mention are 5+ times the size of Puppy. Hence they contain many more drivers. Puppy is a fairly minimalist OS. From creation, that was always Puppy's mission. The Devs try to include the most common drivers, but Puppy will require added drivers for some folk. We try to help fulfill requests when made. Firmware is the same.

[dimkr]

@joet12345 What Puppy version have you tried? I'm working hard to improve hardware support, and future releases (Slacko 8.x, dpup based on Debian Bullseye) should be much, much better.

[joet12345]

@joet12345 - the other OSes you mention are 5+ times the size of Puppy. Hence they contain many more drivers. Puppy is a fairly minimalist OS. From creation, that was always Puppy's mission. The Devs try to include the most common drivers, but Puppy will require added drivers for some folk. We try to help fulfill requests when made. Firmware is the same.

Size doest seem to make sense to me because drivers should be really small in size... under one megabyte...

I am just sharing my experiences... puppy linux needs more fine tuning in the wifi adapters area...

Can you tell me how this works?

Ubuntu is using the same drivers for my WiFi adapter and puppy linux too...but puppy linux wifi sucks and ubuntu doesn't... same linux driver numbers... so why is ubuntu working fine with this old adapter and puppy is crap? You can't blame it on size wifi drivers are small....

[ozsouth]

@joet12345 - As an example, iwlwifi has over 130 firmware files. Many are 2mb. This is just one type of adaptor. Multiply that plus drivers for the many different types of adaptor and you get over 1gb quickly. That not to say we can't improve, as @dimkr says, just that the devs have a hard time trying to decide what to omit. Puppy has slowly got bigger over its 16 year public life, and projects like Fatdog etc are evolutions of it, currently quite different.

[rcrsn51]

Do the standard diagnostic test for wifi.

Run the command: dmesg | grep -i firmware

[joet12345]

@joet12345 - As an example, iwlwifi has over 130 firmware files. Many are 2mb. This is just one type of adaptor. Multiply that plus drivers for the many different types of adaptor and you get over 1gb quickly. That not to say we can't improve, as @dimkr says, just that the devs have a hard time trying to decide what to omit. Puppy has slowly got bigger over its 16 year public life, and projects like Fatdog etc are evolutions of it, currently quite different.

You need to work for Microsoft lol...with that mentality...sorry...

what?

[ozsouth]

@joet12345 - Haha - I'm no MS fan, but Google doesn't worry me unduly. It's just my long puppy experience, watching much brighter folk than me struggle to produce what we have, that makes me more accepting of limitations, especially my own. As a pensioner, this is as close to a day's work as I get.

[rcrsn51]

Do the standard diagnostic test for wifi.

Run the command: dmesg | grep -i firmware

No reply to this?

[rcrsn51]

Do the standard diagnostic test for wifi.

Run the command: dmesg | grep -i firmware

No reply to this?

I guess not. Complaining is easy. Problem solving is hard.

[joet12345]

Do the standard diagnostic test for wifi.

Run the command: dmesg | grep -i firmware

No reply to this?

I guess not. Complaining is easy. Problem solving is hard.

You need to ask the puppy linux team or guy/girl responsible for wifi stuff...not me...

I am just reporting my experiences as a user... I am not a coder nor looking to refine the code... heck the coder can borrow the code from ubuntu since it is open source... If I am not mistaken, isn't linux just code written on a text pad?

I checked and the ubuntu and puppy linux wifi adapter are the same number... then the open drivers are the same...

but

something in the code that is not working and gets the router mad that it does not like it... all I can understand is the hand shake for WPA is better in other distros and puppy has some bad code that confuses the routers... something in puppy linux needs fine tuning in a reliable WiFi connection...

[bigpup]

@joet12345 - the other OSes you mention are 5+ times the size of Puppy. Hence they contain many more drivers. Puppy is a fairly minimalist OS. From creation, that was always Puppy's mission. The Devs try to include the most common drivers, but Puppy will require added drivers for some folk. We try to help fulfill requests when made. Firmware is the same.

This is the simple reason.
This is not going to change.
We all know it, can live with it, and are always willing to help someone find, and install the missing drivers and firmware.

If Puppy Linux was 2 or 3 or 4 GB in size, like other OS's.
Forget being able to load all of Puppy into RAM.
Which is one of the big benefits of Puppy.
Puppy is still trying to work on a 1GB RAM computer, and still load all in RAM.

Plus, if you put every possible driver and firmware in the OS.
In about a month, there would be some new piece of WIFI hardware, that needs a driver and firmware not there.
Maybe not even supported by the Linux kernel in the specific Puppy version.

[dimkr]

I don't think it's Puppy's size that's solely responsible for missing support for some WiFi adapters.

It's super easy to misconfigure the kernel when you build it, and accidentally drop a WiFi driver or two, or get a faulty fdrv with missing firmware. We have bugs in kernel-kit (for example: at least at some point, it didn't stop with an error when firmware download failed), although it's true that the situation is getting better with time.

In addition, people in the Puppy community keep building more and more kernels, of different versions, and sometimes they don't go over the configuration copied over from an older kernel version, option by option. Some new drivers are left in the default setting that excludes them from the build. It's hard to pick a kernel version and build a configuration file that doesn't exclude anything useful.

The main reasoning behind the automated periodic kernel-kit runs in https://github.com/puppylinux-woof-CE/w ... el-kit.yml is to allow us to improve the kernel configuration files over time (i.e. add driver x when someone asks for it in the forums) and avoid regressions (like forgetting to add driver x when upgrading to a newer kernel branch).

Another solution to problematic WiFi support is https://github.com/puppylinux-woof-CE/woof-CE/pull/2457: a kernel very similar to that of another distro, which is almost guaranteed to work if WiFi works in that other distro. If you look at the numbers, they're not super bad: this "big" kernel that kinda goes against Puppy tradition, is only slightly bigger than the Fossapup64 9.5 kernel.

[8Geee]

Of course the OP might try a USB/wifi based on the Atheros 9271 chip... supported in most new Pups (Ath-9K).

[dimkr]

AR9271

Of course the OP might try a USB/wifi based on the Atheros 9271 chip... supported in most new Pups (Ath-9K).

AR9271 has some issues here and there, from my experience (https://github.com/qca/open-ath9k-htc-firmware/issues) - it uses the ath9k_htc driver, not ath9k (the good driver that has less issues and doesn't load buggy external firmware).

[April]

Do the standard diagnostic test for wifi.

Run the command: dmesg | grep -i firmware

No reply to this?

I guess not. Complaining is easy. Problem solving is hard.

I'm interested to know just what this is telling me ?
I ran it on mine and got the following . What does it mean thanks?

Code: Select all

dmesg | grep -i firmware
root# dmesg | grep -i firmware
ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
usb 2-5: ath9k_htc: Firmware ath9k_htc/htc_9271-1.4.0.fw requested
root# 

[BarryK]

I'm interested to know just what this is telling me ?
I ran it on mine and got the following . What does it mean thanks?

Code: Select all

dmesg | grep -i firmware
root# dmesg | grep -i firmware
ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
usb 2-5: ath9k_htc: Firmware ath9k_htc/htc_9271-1.4.0.fw requested
root# 

It would be good to know what the next line from dmesg is. I suggest log all of dmesg to a file, and open the file in a text editor:

Code: Select all

# dmesg > dmesg.log

[April]

ACPI: Added _OSI(Module Device)
ACPI: Added _OSI(Processor Device)
ACPI: Added _OSI(3.0 _SCP Extensions)
ACPI: Added _OSI(Processor Aggregator Device)
ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
ACPI: Interpreter enabled
ACPI: (supports S0 S1 S3 S5)
ACPI: Using IOAPIC for interrupt routing
PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
acpi LNXIOBAY:00: ACPI dock station (docks/bays count: 1)
ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
acpi PNP0A03:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM

.......
usb 2-5: ath9k_htc: Firmware ath9k_htc/htc_9271-1.4.0.fw requested
usbcore: registered new interface driver ath9k_htc
floppy0: no floppy controllers found
usb 2-5: ath9k_htc: Transferred FW: ath9k_htc/htc_9271-1.4.0.fw, size: 51008
ath9k_htc 2-5:1.0: ath9k_htc: HTC initialized with 33 credits
ath9k_htc 2-5:1.0: ath9k_htc: FW Version: 1.4
ath9k_htc 2-5:1.0: FW RMW support: On
ath: EEPROM regdomain: 0x809c
ath: EEPROM indicates we should expect a country code
ath: doing EEPROM country->regdmn map search
ath: country maps to regdmn code: 0x52
ath: Country alpha2 being used: CN
ath: Regpair used: 0x52

[bigpup]

@joet12345 What Puppy version have you tried? I'm working hard to improve hardware support, and future releases (Slacko 8.x, dpup based on Debian Bullseye) should be much, much better.

Thanks for working on this hardware support improvement!!

As you say, the specific Puppy version is important in this.

Much improvements have been made with each new version of Puppy.

There is work right now on improving the network connection programs.
viewtopic.php?t=445
rerwin is working to make them all work better.
Give him some help testing.

These programs could be part of this persons reported issues.

Not so much now, but in the past.
The reason to have several different network connection programs in a Puppy version.
If one did not work.
Maybe one of the other ones would.
In the past, this usually was true.

[dimkr]

If one did not work.
Maybe one of the other ones would.

In Vanilla Dpup, I replaced them with ConnMan, connman-gtk and connman-ui:

https://github.com/puppylinux-woof-CE/woof-CE/pull/2514
https://github.com/puppylinux-woof-CE/woof-CE/pull/2724

connman-gtk and connman-ui complement each other: one is a "wizard" and the other is a tray icon. Both talk to the ConnMan daemon, so the configuration is always in sync between the two.

This setup remembers the password for more than one SSID, connections are restored after resuming from suspend, and it manages wired and wireless connections in one place.

[April]

How do I get it installed to give it a try?

Why is it that even packages in the repository fail to load when chosen?

Connman.png (369.77 KiB) Viewed 489 times

Also...Cant get a 303 Kib png image to load for some reason ? See below.reducing image dimensions to 800x 600 seemed to fix that? but by the time I got to this all detail had been lost . Ridiculous!

Connman-errors.png (6.99 KiB) Viewed 491 times

[bigpup]

If you use MtPaint to resize the image.
It should show in the forum clear to read.
.jpg images seem to work best.

I use MtPaint->image->Scale Canvus..... to change it.
I only change the first new number and select OK.
Let it auto select, what the other new number, needs to be.

Try taking a image shot of each program window.
Not all of them open on the desktop and a full image of the desktop.
Lowering that full image down is probably your problem.