I could be extra corny today and shout "Mayday, Mayday!" - but I won't. Happy Worker's Day and for those of us slogging away regardless of whether it's a holiday or not, keep us in your thoughts - or send cake.

Let's dive in to this month's newsletter.

Tech Talk

Speed tests. I would love to tell you an ISP is required to perform one human sacrifice every time you run a speed test, but unfortunately that's untrue. Speed tests are the bane of every ISP simply because they don't necessarily accurately portray the provided speed of any given connection. So let's start with the basics of how they work.

First things first - MB vs Mb. "MB" refers to a megabyte, a unit of digital storage, while "Mb" refers to a megabit, a unit of data transmission speed. Essentially, MB measures the size of a file or storage capacity, while Mb measures the speed of data being transferred. One megabyte (MB) is equal to eight megabits (Mb).

We provide an "up to" speed in Mb, on the part of the internet we control which is our own network. This is like the speed limit on the highway between Somerset West and Cape Town. You can go up to that speed but not past it. Just like hitting rush hour traffic or a dirt road instead of smooth tar, your speed can be influenced by a whole bunch of factors. The road between Cape Town and Blouberg could have a completely different speed limit (aka another network), but you're on the Somerset West to Cape Town road with that specific speed limit and those road conditions.

Hey - remember these days?

Back to Mb and MB. Take the image above. The download screen shows how much of a specific file size is being transferred and at what rate - so MB are used. It's only squeezing 4KB of that 180MB file through every second.

Then there's your everyday garden variety speed test below:

This shows Mb - megabits - it's not showing the file size or data transmitted, it's showing the potential speed you can download or upload at with current network conditions. You'll notice the download is a lot less than the upload, and here's why.

I'm testing on a network that has a couple of hundred devices connected, all chattering away and talking to the internet. The speed test above shows as unbalanced (my devices download more information than they upload), meaning I have more "space" on my network to upload than I do to download because most of my capacity is being used by all those talkative devices downloading info but not uploading it. This doesn't show what the actual speed of my network is (I'm on a 1Gb connection), it just shows how much of that is currently available. Think of it as "someone has taken the rest of this pie, and here's what's left for me to eat".

On to how your network affects your speed test.

Cellphone to router - your speed test can be affected by the ability of your cellphone's tiny wireless antenna to deliver a specific speed, by its stable or unstable connection to your WiFi router, by distance and objects between them, by your router's WiFi broadcast signal speed, and by the router's connection to the internet. If there is more than one device connected to the router, your speed is divided up among all connected devices (like the pie analogy above). If one of them is trying to download the entire internet at max available speed, you're basically left with crumbs for the rest - or they completely lose connection to service.

Laptop on a network cable to router - your laptop's network port speed, background traffic (downloads, updates, syncing stuff etc) and router port speed all affect your speed test. There are specific scenarios where port speeds are set up on a router on client request for things like VoIP, or assigning a certain speed to a certain device - plugging into one of those will of course affect your test.

Router to internet - if you have a dodgy cable between your router and your wall, it won't transmit properly. This is especially true for those tiny fibre cables - one kink or sharp bend and light won't travel effectively.

Past your router, there could be an outage on a network between ours and the server that is hosting the speed test, meaning traffic is routed over a backup with a longer "road" to get there and more people trying to use it.

And finally the speed test server itself - it too has a specific connection speed to the internet, so will talk back at a certain rate. If everyone is hitting speed tests, that pipe clogs up with traffic and affects how your test runs. If you are testing to a server somewhere else in the world (or at a location that doesn't provide direct peering connections to most networks), your speed test traffic has to take a back road to get there - again affecting your test results.

And then there's DSTV and smart TVs... DSTV has a limited amount of streaming pie to hand out, and during rugby games there's a good chance you'll get none. Streaming services on smart TVs have similar issues, along with limitations on how your TV connects to your router. Please help it along by connecting it via a cable so it's not competing with your cellphone for WiFi and can focus on delivering your content. (Next month's newsletter will delve into streaming)

Speed tests are only one tiny part of troubleshooting a slow connection. They are not the be-all and end-all of tests. If our support team requests you to test your speed in a certain way (using a laptop, via a cable, disconnect all other devices using your router), there is a definite reason for this and will help them obtain a starting point to help you.

From the Archives

Once upon a time we had a tower on a tall residential block. Also residing on that block was a rather special box, tucked away out of sight from the general public. When we found out what it was for, we very quickly installed a camera...

Every year a pair of Peregrine Falcons used this box for nesting, and hung around the rest of the year scouting out their domain. Working on this building was interesting - we'd often get mock-charged on the tower by a bullet fast blur of bird if they thought we were intruding in their space. It was incredible to watch them soar way up, dive down, swoop in one massive power slide around the building.

Around September each year, mom would settle in and lay 3 to 4 eggs. It was an eager competition to see who could spot the first little head poking out from under her feathers a month later. Then started a relay-race of food supply. Mom and dad had a ready pantry of pigeons on the building (some of which apparently had rings, which we would find littering the ledge, clinging to leftover pigeon legs...oops). The pigeons generally didn't even know what hit them, these falcons moved so fast.

They were carefully monitored by a Cape Town group, ringed, weighed and photographed.

The more they ate, the more they grew. Before you knew it, they'd be struggling out the nest box, waddling up and down the ledge and peering into our cameras.

Flight lessons followed. We'd see them flap-flap-flapping, strengthening the wing muscles, racing back and forth, building up speed.

And one day they'd simply be gone.

Fair Use Policy Softcaps

I understand there has been some confusion regarding our implementation of our long-standing Fair Use Policy by means of softcaps on the three slowest speed fibre connections. I apologise if this was not clearly communicated earlier this year.

Softcaps were implemented along with our package changes on 1 March and published on our website, as a comprehensive package and product audit earlier this year revealed rampant abuse of our lowest speed packages beyond what they cost us to provide. We had a choice between dropping them entirely (as many FNOs have done with their smallest speeds), or introducing checks and balances to bring cost-to-deliver in line with package pricing and use.

With these in place, if you reach your fair use limits during the month, you are not disconnected but your speed drops according to use. During the last part of April we doubled the softcap limit to the maximum we can provide without blowing the budget on providing these. To clarify, here is how the softcap system works.

Fibre Lite Softcaps:

• 30Mb has a fair use softcap increased from 45GB to 100GB, then it falls back to
• 20Mb, which has a fair use softcap increased from 30GB to 60GB, which then falls back to
• 10Mb, which has a fair use softcap increase from 15GB to 30GB, and which then drops to 1Mb speed until the end of the month.

If you wish to upgrade to the 40Mb uncapped service at R599 per month, please email accounts, and we'll do so immediately.

May Competition

Want R100 off your June subs? We're giving away three! (valid on our own wireless/fibre network only).

Simply send your answer to the question below to hello@cape-connect.com before 15 May to be in the running. Random draw out a hat of correct answers nets you the prize.

"Where on our network can you apply for a 1Gb home fibre service?"

Shot of the month

We're often out on site at very odd hours and in very strange places - but this means we get some awesome views. Here's this month's picture perfect shot.

A marvelous May to you!

Michelle Bainbridge

Chief Internet Artisan

Cape Connect Internet (Pty) Ltd

Read Past Newsletters