Any way to get WLU6331 wifi dongle support on OSMC + Pi2?

Hi all, I’ve been trying to get a wifi dongle with better signal for my pi 2 and i saw this one on the official raspberry pi shop:

http://swag.raspberrypi.org/collections/frontpage/products/official-raspberry-pi-wifi-dongle

Silly me assumed ‘works with raspbian out of the box’ also meant ‘debian drivers’. It seems OSMC does not see this dongle - if I plug an older tplink model in I can see it when I run connmanctl, but with the ‘official’ dongle nothing shows up. I take it I need to compile/install the driver myself? Would anybody know how I can do this or if there is support coming into OSMC at some point please?

Thanks,

Edwin.

Please post the output from dmesg so we can see what vid/pid combination this adaptor is - I don’t think any of the OSMC devs have one of these adaptors. (I know I don’t)

I have this dongle, it is Broadcom 4300 based. It should already be supported, because the Foundation gave me one in January. I’ll double check.

Drivers are not tied to Debian, they are closely coupled to the kernel that we ship.

Sam

Thanks for the help everyone. Here is the output of my dmesg command:

Hi everyone, is there any news on this? Thanks :slight_smile:

It’s not forgotten

Sam

Any progress?

Not yet – I am away at the moment so my ability to work on things is slightly limited.

Ok. Thank you for the quick reply.

Can you tell me whether it is a matter of getting it to work, or whether it can be made to work at all?

I don’t quite follow your question. OSMC (if it doesn’t already) will support the dongle.

Sam

1 Like

are there any updates on this? i bought the same wifi dongle expecting it to work aswell before i found this post.

Thanks

Still enjoying holiday.

It is on my list.

Sam

Answer to other commenters

Install Official Raspberry Pi Wifi Adapter (WLU6331 / BCM43143) on OSMC

Note: I have actually uploaded an auto-installer script at this point which is available on the above-linked blog post if you don’t fancy compiling the whole thing manually. The manual compile takes a good couple of hours so I figured I could save people some time. I have left the manual instructions up too incase anyone prefers that method.

Gripe at Sam

Seriously… this is such an easy fix and “I’m on holiday” is a bullshit excuse. I’m assuming it’s not just you on the team there. I’m a team lead of 18 engineers. When I go on holiday everything at my work does not grind to a halt.

1 Like

Thanks for demonstrating your ignorance and the huge sense of self entitlement you seem to have, especially when you only registered on the forum to have your little rant. :wink:

In case it escaped your attention, OSMC is an open source project that is developed by a small group of volunteers in their spare time. This is not software which was developed in the course of a paid day job that you have purchased and have an expectation of support for. The software is provided as is.

A quick look at the commit history of OSMC and you would see that the team of core developers is much smaller than your “18 engineers”, and that the core devs specialise on on different parts of the project. As a team leader I’m sure you also understand the concept of time management and the necessity of prioritisation of bugs/features in a project.

So while this particular issue might be a high priority for you that does not mean that it is for everyone else who uses OSMC. Everyone has their pet bug fix/feature request that is super important to them, but that doesn’t mean that that issue should have priority over all other issues that are competing for developers time.

Given a limited amount of development time, would you rather have a wifi driver for one particular wifi adaptor (used by a small percentage of users, especially given that it has only just been released) first, or would you rather have an Isengard release on time soon after the release of Isengard, something that ALL users will take advantage of ?

Just one example of weighing up the benefits of where development time is spent.

Instead of griping and frankly just being rude and ignorant, feel free to submit a pull request at http://github.com/osmc/osmc that adds this driver to our kernel build process and we will give it proper consideration for inclusion. You say it’s “such an easy fix”, so go to it.

So no pull request then ? Don’t let the door slam on the way out…

So your usual modus operandi on the discussion forum of open source projects is to wade in with an obnoxious rant as your first post ? Bet that makes you a few friends.

What exactly so offensive about Sam saying that it is on his todo list when he is back from holiday ? Is he not allowed to go on holiday ?

I’ve also replied to questions about this particular wifi adaptor at least 2-3 times saying that it’s not currently supported but that it’s on our todo list to support it - what else do you expect exactly ?

With all due respect, Sam is more than capable of adding support for this driver on his own. The issue is not that we don’t know how to do it, the issue is purely a matter of juggling the time with other issues that need to be worked on (such as Isengard) many of which have a higher priority. There are only so many hours in the day…

Hello @chrisaw,

Indeed – this involves adding support for the USB interface for Broadcom FullMAC drivers.

I am not the only team member, but I am the only team member with the device. I had received this device some time ago, but did not have time to test it. Firmware is included, as I am sure you can tell. A simple dmesg would’ve shown me there was no kernel module, but I had not received one to date. So naturally, being on holiday, without the equipment needed, I was unable to resolve this.

Firstly, I do not see how that adds weight to your argument. Secondly, things did not grind to a halt. Our team achieved the OSMC final release while I was on holiday.

Although you see the addition of the kernel module as a simple task, there are other considerations:

  • Potential regressions (VID/PID overlap). For example, RTL8192DU will overlap a PID with RTL8192SE if in-tree.
  • Power management: RTLx shows significant decrease in performance with this enabled.
  • Compile options. We preempt, so does -O3 treat this driver OK? What does upstream say?

I was aware of the issue, and needed to test it myself. I said I was on holiday because I did not want someone to waste their time submitting a PR when it was already being handled.

Maybe – but that money goes in to supporting the flagship device well as a priority. Hardware enablement is an extension of that, but only to an extent.

That’s not how the build system works. rbp-012* support package includes a defconfig, but that is not used.

Only if you build the whole kernel. The individual module which can be built in tree takes only 5 minutes.

My changes are branched and are pending for the July update.

You invoked the hostility here. Even a bump would have been more appreciated or to ask how progress was. You could have also posted your solution, to which I would have raised my concerns as mentioned above. I have afforded you some respect here, please be as courteous as to understand that OSMC’s development roadmap is of significant scope. It is favourable to us and users to support as much hardware as possible, and we do some in as timely a manner as possible. Even if I had resolved this issue some days ago on my holiday, this release would be deferred until the July update as there are significant kernel changes in the pipeline as well.

Should you have suggestions for how we can improve our development process, I would love to hear them. You can send me an email at sam at osmc dot tv. If you’d prefer a direct line, send me a PM and we can have a chat.

Cheers

Sam

1 Like

Wow! You pissed off Sam AND dbmandrake! That takes some effort.

What’s the turnover on your 18 person team like? Or did you mean that you run a team of 3 and have had to find replacements 15 times?

I mean I can understand how that could happen. Especially when you misunderstand what your team members mean when they ask you for a “PULL REQUEST”.

You started by disparaging the team here. What exactly did you expect?

When it became clear that you offended people, you had more than enough opportunities to realize your error and apologize. But you don’t seem to think you did anything wrong at all.

And now you say that the reaction was because we are annoyed that you provided a workaround, rather than your awful social skills?

What the hell is wrong with you?

Not if your first introduction is to insult the “team”. A team which you obviously don’t know the first thing about, but expect to be given licence to instruct on its activities.

You don’t deal well with bullshit, because you’re full of it already.

DBMandrake said not to let the door hit you on the way out. But I say, don’t worry about that. Given you touched it on the way in, we’ll end up removing it, cleaning it, painting it, and then burning it anyway.