Category Archives: Technology

audio, Computers, Technology

Home Baked

A question on a Facebook group triggered this: someone asked about the Brennan B3. Is it hifi? Is it worth the nearly £800 price tag?

I’ll leave the latter to your judgement, but the Brennan devices have interested me for a while.

They’re a combined CD ripper and hard-disk music player, nicely packaged and with an easy to use interface. The advertising seems to be aimed at the older generation (well, the kids all stream now…), and at replacing your CD collection and old player with one box and adding a bit of convenience.

This is something I’ve been doing for a while with a Rasperry Pi and Moode, RuneAudio, Raudio or Volumio software: all similar concepts of a modified Linux distribution for playing media.

Can you see where this is going yet?

The Brennan box is unashamedly and openly the same thing, tidied up as a consumer appliance. The latest B3 unit is the most complete: it takes a Raspberry Pi, a hard drive, a CD drive, and a DAC and class D amplifier, puts them in a nice case with a nice display, and glues it all together with some software.

That software is freely downloadable if you’re so inclined, and if you write the image to a SD card you can mount it and poke about.

It looks to be based on Ubuntu 32 bit, and uses lighttpd to serve pages, and omxplayer to play media.

So: is it hifi?

I suppose, for certain definitions of hifi. It’s not high-end, but then that isn’t the target market.

Is it worth nearly £800?

Not for me. I’ll stick with the homebrew solution. But some people love them, and at the end of the day it’s a supported consumer product and the manufacturer is upfront and honest about it. You decide.

I’ve got a certain admiration for it: it’s a clever, well-executed idea, and it’s not like using a Raspberry Pi as the basis of a product is unusual.

Cars & Driving, Life, Technology

Updated

At work, we’ve been busy.

We’ve been bringing some infrastructure up to date: an elderly VMWare install was migrated to Nutanix– because since Broadcom bought VMWare, pricing just got stupid, a legacy application with its roots in the 1990s, running on a very old Server 2003 box finally got retired, and at this point, we could start to address issues like a Windows Active Directory also running on Server 2003 that literally couldn’t be altered until the legacy application was retired.

The legacy application also meant an old desktop OS: it literally wouldn’t work on Windows 10 or 11, and was unsupported (but did work) on Windows 8.1. This meant old hardware for the users- in some cases purchased secondhand.

So, first of all, a Windows 2016 domain controller was introduced, this being the newest that was still compatible with the 2003 functional level and servers. This had been tried before, but the legacy server promptly shat itself: the mistake was patching the new server. Backing it out, then rebuilding it unpatched worked just fine.

Services were migrated- DNS, DHCP, RADIUS. The 2003 servers were removed, more services migrated from the 2008 Domain controller. A second 2016 DC came in, the 2008 server was removed, and finally, we get to raise functional level and remove the need for SMBv1 (an easily exploitable protocol). We might still have unsupported OSs, but at least now we can use supported ones on the network!

Now we were in a decent place to deliver new hardware to users, users that have been justifiably clamouring for workstations from this decade, which is why, last week, I found myself waiting for my non-driving colleague near a flyover in South Birmingham at 05:40 in the morning (and why I’d found myself at another colleague’s house at 06:00 just before Christmas) with a boot full of Thinkpads and a long commute ahead of us.

It’s your job to keep me awake!

This time, the target was an office south of the M25 in Surrey, so it was on to the M42 and M40, a short trip round the M25, and then the A3.

It never ceases to amaze me that in places, once you leave the M25, things can look almost semi-rural, and this was the case here: we drove through Ripley and it felt almost villagey, though admittedly a village with a Bentley Dealer.

We found ourselves onsite by 08:20, and after a quick drink got underway: all done by 14:20, which was a result, and then it’s just a return journey. I drop off my colleague, and drive home with a bootfull of antique computers, encountering the only really bad traffic of the entire journey at the top of the M5 as per normal.

There remains much to be done: Exchange has to be brought up to date (kind of) and there’s new Citrix to build. Some of my colleagues still have bootfulls of Thinkpads to dole out elsewhere in the UK, but at least we have a significant chunk done.

Technology

Your nearest exit may be behind you.

I’ve wittered on here before about consumer VPNs and why if you’re using one for security. you’re deluded.

There’s only two legitimate use cases for a consumer VPN:

* Disguising your geographic location for reasons such as watching TV services in a different country

* Hiding your traffic from your ISP or government

The important point about hiding your traffic, perhaps because you distrust your ISP, is that you then have to trust the VPN provider. At some point the traffic that is encrypted at your PC had to be decrypted, and that point is your VPN provider’s exit point. Do you know they’re not snooping?

Come to that, are you sure they’re not using your home connection, flogging it to all and sundry?

Here’s the thing: TLS (what a https:// website uses) is already encrypted and is strong enough for anything you’re doing anyway. Don’t fall for the bullshit.

Computers, Technology

Thunderbird is 20

I don’t often do open-source software stuff here these days, but it’s worth noting that the best email client (and one that works nicely cross-platform on Android, Linux, and Windows) has reached 20 years old: I think I’ve been using it most of that time., and as well as being as being good to use day-to-day, its flexibility makes it a valuable tool in my line of work.

Politics, Rants, Technology, The stupid it hurts

Cash is King

2 Comments

As the rise of cashless payments continues, the memes against it with certain types continue.

They’re all pretty much of a muchness, and the only valid point any of them make is “what do you do if the payment networks fail”.

Cash is king, they say, which I can’t help but agree with. There’s much in common with insisting on cash and having a monarch:

* It annoys some people
* It causes inconvenience
* It costs money, and the costs are largely hidden from you
* It’s an example of tax avoidance
* It’s an anachronism

I’m not suggesting we eliminate cash (yet?), but I would like to pick apart some of the myths that surround this cash is king idea.

One image doing the rounds on Facebook recently is a blackboard that says:

PLEASE USE CASH IF YOU CAN LAST WEEK OUR CARD MACHINE CHARGES WERE EQUAL TO A DAY’S WAGES FOR A MEMBER OF OUR TEAM.

So, let’s assume they only pay minimum wage. UK minimum wage, currently, costs the employer £102 per day, including all legally required employer costs, assuming a 40 hour week.

A relatively expensive pay-as-you-go card transaction deal- with someone like SumUp- takes 1.6% fees, with no other charges. A really good merchant account will have a rental charge for a machine, but much lower percentage fees.

My arithmetic isn’t that good, so you’re welcome to correct me, but if I have this right, to run up £102 in card payment charges you’d have to take around £6000 in card payments alone. Every day, which seems like a pretty good turnover.

The biggest error, of course, is the assumption that cash is free to handle. This is an assumption many make, because the charges for card fees are visible, but the costs of handling cash are largely hidden.

we pay commission fees on every card sale,lowering profits. No commission on cash

as one poster said.

If you take cash, you have to maintain a float (with change). You have to count and bag the cash, take it to a bank, and pay it in- paying a fee to do so, risking getting mugged, and spending time, and probably fuel/parking to do so.

The exception, of course, is if all your outgoings are cash, which I’m told is very convenient for minimising your tax burden, as they say, but obviously no-one would wish to avoid declaring income, huh?

There’s also the ones that say

The government and banks can track what you’re spending your money on

and while there’s truth here, seeing as how this government are just so incompetent, the chance of them competently tracking my purchases and finding anything worthwhile are pretty fucking minimal, even if they have access to the point-of-sale data from my local convenience store (because if they don’t, all they know is how much I spent, at what time).

Oh look- this guy has bought 4 cans of lager, a bottle of wine, and an Indian takeaway. Best we keep an eye on him.

A loosely connected bit of bullshit is this one:

First and foremost, using cash is a great way to support local businesses. When you pay with a banknote, that money stays within the community and can be used by others for a wide variety of purposes. For example, imagine you have a £50 banknote in your pocket that you use to pay for dinner at a local restaurant. The owner of the restaurant can then use that same note to pay for laundry services, and so on. After multiple transactions, the banknote will still retain its original £50 value, benefiting everyone who has used it as payment.

Which is the biggest load of toss. Unless every single transaction remains as cash and none of the businesses use a bank or a credit account, or have a mortgage or bank loan, it isn’t true. Obviously, the note retains value- but at the point it re-enters the banking system, someone is paying for it to be deposited, and often far more than 1.6%. As to the local business angle, people will either use local ones or not, and the cash isn’t simply a factor.

The only valid reason to insist on cash is to hide transactions: i.e: to avoid tax. And you’ll be turning away customers who can’t be arsed going to an ATM.

By all means accept cash. Hell, don’t take card payments if you like, but don’t kid yourself it’s saving money, helping local commerce, or is a great statement of freedom.

Technology

TV Freedom

Following on from the Internet and phone move to A&A, we just lost Virgin’s TV service when the cancellation period expired. This was the one bit I wondered about: A&A’s Internet offering is definitely an improvement, with few drops (and if it does drop, they tell you, rather than you having to guess), and latency and speed way improved (this really showing up in work Zoom meetings), and a router placement that works better for us now.

The phone, too, is good. We don’t use the “landline” much these days, but having had it for over 30 years, i wasn’t ready to drop it. Having it in SIP adds flexibility- I can have it present on my mobile if I want, and the cost is trivial.

Back to the TV though. We now realise just how many of the channels Virgin offered were just crap- noise in the TV guide- almost too much choice, and terrestrial Freeview combined with the free streaming services actually fill our requirements pretty well. A bit of fiddling with my 14 year old Onkyo AV amp and ARC settings and we can get surround too, and one less remote. Our TV can stream the common services as standard, and the fact the router is closer means it’s never going to struggle for bandwidth.

The only thing we’re missing so far is the ability to pause or rewind live TV, but if we really, can’t live without it, we can buy a Humax box with the savings from just 1 month’s subscription.

Computers, Technology

Flat Out

2 Comments

Warning: DNS/Email nerdery below.

For my sins, one of the things I had to do recently was complete DNS domain authentication, DKIM, and DMARC for some email domains at work in Mailchimp. Mailchimp is the spamming engine mass email system of choice, and to be fair, they’re responsible and force compliance with good practice. They also try to increase the chances of your spam legitimate mass email being deliverable, which is where this comes in.

The big email providers such as Yahoo, Hotmail/Outlook, and Gmail now insist that if you send over a certain threshold, you:

* Send from a domain you own
* Configure DKIM, SPF, and DMARC
* Don’t send shitloads of actual spam
* include a one-click unsubscribe

This is all part of the spam arms race: electronic means to detect spam are less effective, so this is an attempt to stop the problem at source by making sure the email is coming from where it claims to: spoofing email is trivial, so this effectively adds a signature.

Well, we already had the domain, and SPF, we don’t actually spam people, and Mailchimp handles the unsubscribe, so that left DKIM and DMARC. This actually isn’t that hard, just involving publishing a few records in DNS that match up with the email servers. The DMARC is little more than a published policy of what people should do if the mail seems to be unathenticated- and “take no action” is acceptable. The actual authentication is done by DKIM, where you publish a public key in DNS, and the email is signed by your outgoing server with a corresponding private key. If the encryption key matches, the mail is deemed as being legitimate.

So, I went ahead and did all this for our own infrastructure, as it seems silly to set it all up only for the mass mailing. Pretty simple, too, in the end.

Now, obviously, if someone (Mailchimp) is sending mail for you, you need to publish their key for DKIM. The logical way to do this is for them to publish it, and then for you to alias an entry in your DNS (with what is called a CNAME record). That way, if they change the key, it keeps working without changes to your DNS.
Continue reading Flat Out →

music, Technology

Android Bluetooth “Feature”

I’ve been though a few pairs of bluetooth headphones since this, progressing to Cambridge Audio Melomania 1+ (which got washed and are no longer available…) and Lypertek Z7 true wireless ones (the Z7’s have a weakness, broke, and got refunded under warranty), and I ended up with What Hifi award winning (but budget-level) Sony WF-C500s, which work well and sounded pretty good, but while listening on the balcony on holiday, I noticed that they were using the inferior SBC codec paired with my Motorola G50 5G.

The previous headphones supported APT-X, so this hadn’t been an issue.

Poking about in Developer Options, HD Audio and the ability to select AAC was greyed out.

Android developer options screenshot

It seems that in their “wisdom” Motorola maintain a whitelist of devices that are able to use AAC, and if the device isn’t on that list, it reverts to the inferior SBC codec- the cited reason being stabilty issues. Deselecting the “Bluetooth AAC Whitelist” option and cycling bluetooth off then back on enables you to turn on HD Audio and use the AAC codec. Why this is buried in Developer Options is anyone’s guess.

I suppose the question is “did it make any difference”? I told myself I could tell the improvement with AAC, but that could be confirmation bias 😉 .

audio, Computers, music, Technology

rAudio

Following on from my forced evaluation of my Raspberry Pi audio player, as I’d got it in bits I went and looked at Runeaudio’s web site to see if there was a newer system image. There wasn’t, in fact there hasn’t been one since before this country lost its collective mind in 2016.

What I did find is that someone has forked and updated it, just like I mentioned here about all the other variants.

These players are all variations on a theme: a web interface, MPD, and a variety of other open-source things to provide connectivity and features like Apple Airplay or UPNP: the beauty is that if you have the skills and time, the components are all there for you to build it. If not, well, someone’s probably done a good enough match, which brings me to rAudio.

A proper open-source project this: no flashy website, just a Github and a concise set of instructions, but technically it’s great, with recent builds based on Arch Linux.

Booting it up, it just works: it reads the external disks just fine, recognises my Cambridge Audio DAC, it plays. There’s a few whistles and bells- you can do proper multi-room on a budget, and there’s Bluetooth and DAB if you have the hardware.

Best news is I’ve got rclone on to the machine, and a quick entry in crontab should see the music files automagically backed up to Google drive with no manual intervention.

audio, Computers, music, Technology

In the Clouds

I briefly mentioned that the mirrored disks for my Runeaudio music player broke. I then compounded the issue by not RTFMing about the RAID enclosure- which also seems to have had a fault- with the result that I lost my FLAC collection.

I tried the excellent testdisk (which didn’t find any partitions) and photorec (which recovered files but no folder structure and with possible corruption), and decided I needed to bite the bullet, copy what I had on my phone back from there, and re-rip the rest. I’ve found as I have 2 CD drives, I can rip 2 together, and still use my laptop.

2 terminal windows showing 2 abcde rips running together
Ripping 2 CDs at once: that’s as many drives as I have.

This post is looking like an exercise is open-source software promotion: the files were originally ripped to open source FLAC (and MP3 for the car) with open source abcde, recovery was attempted with open-source tools, the player itself is open-source, and there’s one more thing to add: I don’t want to have to do this *again*, so I need a backup solution, and preferably something that requires minimum intervention.

I already use Google Drive for other backup, and pay for a decent amount of space, so that seems ideal. The idea now is to use rclone on the Runeaudio device so that whenever I transfer new music to it, it ends up on someone else’s computer in the cloud as well, automagically. Initial tests on my laptop are encouraging:

rclone copying my music
A test run of rclone, copying 5GB of music up to Google drive

So now it’s a case of installing rclone onto Runeaudio and seeing if it works.