On the server side I know it transfers about
Never mind Dolphin uses Bytes (Byte - Wikipedia)
The way I learnt it is:
a small letter b stands for bits
a capital letter B stands for bytes
Unfortunately not everyone follows these rules.
This frustrates me, I studied electronics/radio/radar back in the '80s, so I would use MB/s meaning megabytes, or 1024 kb… and it’s easy to understand binary, powers of 2.
But sure, now we use IEC and this is where the ‘i’ comes in… and it gets a really silly name ‘Mebibyte’ which is 1,048,576 bytes.
So that’s basically just what we learned many moons ago… but now the term MB is Decimal.
KDE in general uses IEC units, which are are powers of two rather than powers of ten. So you get “mebibytes” (1048576 bits, or 1024 kibibytes, where a kibibyte is 1024 bytes) instead of metric “megabytes” (1000000 bytes).
Yes, this is awkward and doesn’t match the words and prefixes most people are accustomed to.
You can configure this in System Settings > Region & Language > Data and Storage Sizes.
We might want to look into changing the default setting.
Just be a little careful - because some applications can report MB when they mean MiB… back in the early 1980s, ‘MB’ was what ‘MiB’ is now.
I studied electronics, Radio and Radar, back in the early '80s…
So we learned binary, powers of 2 - 8 bit chips gave us 256 numbers, 16 bits went to 65,535…
For our ‘digital rotation’ system, we worked with 4096… and when you have learned this it really came as a shock that it was actually ‘wrong’.
Mebibytes MiB came up in 1998 to clarify (there were issues with storage contexts and RAM contexts).
ISO (like standard SI units) defines decimal prefixes…
- 1kg = 1×10³g
- 1MJ = 1×10⁶ Joules and so on…
1MB SHOULD mean 1,000,000 bytes… but (in digital radar systems back in 1980) it actually was understood to be binary, so for a ‘Megabyte’ in 1980:
1024²=1,048,576
So there’s an interesting question about your server stats, just stating 1M should imply 1×10⁶ - but it could be short for Mi.
I’d agree with Nate - the default settings should be adjusted.
I think computing would be a lot simpler and fewer mistakes made if the industry used units in decimal bits and no units defined as multiples of 8 or 1024.
Sadly this is not possible, because a BIT must be either 0 or 1… so by definition it must be binary…
Perhaps your comment is confused - the issue is simply that operating systems and marketing people still conflate them - and MANY people are still confused whether kB means 1024 or 1000… but NOBODY is confused by KiB.
Let’s see - stop using binary - multiples of 8, or 1024???
- UTF-8 would also be dead, also every CPU…
Basically, it’s only simpler for humans.
Computers prefer binary (powers of 2) and it’s VASTLY more efficient.
A 4KiB page (Binary, meaning 4096 bytes) is perfectly aligned with line sizes and disk sectors…
A decimal 4KB is a nightmare.
So this is why we have middle ground.
Sadly we can’t manage this with worldwide industries - especially as the largest market (USA) has it’s own preference for using completely non-standard units which are barely understood by anyone else.
Did you know, a car doign 26.6mpg in the USA will do 32mpg using petrol bought in the UK? Of course, we use Pounds in the uk - not Dollars - nobody uses pounds when there are kg 
I am not confused, rather it seems you misunderstood what I meant. What I am saying is that things would be a whole lot simpler if the unit Megabit was prominently used and defined as 1000000 bits, and units which are multiples of 8 (bytes) of defined as factors of 1024 were not used prominently.
Well, everybody who doesn’t know what “KiB” means is confused by it! And I would wager that amounts to most people, seeing as these power-of-two units are only used in technical contexts and KDE software. The rest of the industry either uses Metric/SI (Apple, Google, GNOME+other GTK desktops) or JEDEC (Windows).
I deleted my message because I was wrong.
The image shows MiB and I miss-read it as MB.
That’s true - but it certainly isn’t ambiguous, it only has one meaning.
- KiB is definitely ‘kibibyte’ or 1024 bytes - very distinct from SI units like kg or km.
- KB is ambiguous - officially 1000 bytes, but sometimes 1024 bytes for RAM or older applications… so it’s this ‘ambiguity’ that’s the issue.
- Gbps is bits per second - fine if only used for networking, confusing when transferring a file (how long does it take to transfer a 1GiB file at 10 Gbps
- 1Gbps = 125 MB/s or ≈119.2MiB/s (that’s just nasty).
- 1GiB = 8.589934592 Gb
The problem is a social one - we exploit ignorance…
How I exploited ignorance...
well, I certainly do. I once did carboot sales in the UK - I bought batches of 10 cheap gas lighters from an Indian gentleman opposite me.
I put them on my table - with 3 at the front and a sign saying 30p each or 3 for £1 and I then kept a count.
From the opposite stall (selling 10 for £1) he sold 80 lighters for £8. (and 50 of those he sold to me)
From mine, I sold 90% of them in batches of 3. One ‘smart alec’ said he wanted to buy three lighters, separately, and pay only 90p…
End of the day, I bought 50 for 5, sold them for about three times what I paid because people don’t see what’s under their noses, and they don’t think… the Indian guy was laughing a lot because I made a lot more than he did (from lighters anyway) even though I bought them from him.
Perhaps one reason folks find it offensive is that they are susceptible - and rather than take more time to learn to think, they just carry on regardless…
TL;DR
Nah, I’d prefer to scrap using decimal bits for networking, because it’s insane that you can’t transmit a 1GB file at 1GBs on a 1Gbps connection - networking looks ten times faster than it should.
There’s always going to be some GB confusion, GiB spelled differently should fix this, but you cannot fix wilful ignorance.
All the more reason to use Metric/SI units IMO.
Networking uses powers of 10 for speeds right not powers of 2?
Worse than that, Networking uses bits, not bytes… so it’s base 10 but also ×8.
That’s why Networking at 1Gbps is equivalent to 125MB/s (decimal) or 110.2 MiB/s… and it’s sold like that because nobody wants to pay the same for a 100M connection when they can go for Gigabit 
However, I set up my conky to show how many Kbps and the conversion also to KiB/s (which is more meaningful when you download a file)… so a proper mix of Decimal bits and binary bytes
A fair bit of arithematic in the form of a lua script does the lifting… normally conky shows Decimal bytes (MB not MiB) which is also fiddly:
speed.lua
-- ~/Dropbox/Admin/conky/speed.lua
function format_speed(kib)
local bps = kib * 1024 -- bytes/s
local bits = bps * 8 -- bits/s
-- Format bits (decimal Kbps/Mbps)
local bit_val = bits
local bit_suffix = "bps"
if bit_val >= 1e6 then
bit_val = bit_val / 1e6
bit_suffix = "Mbps"
elseif bit_val >= 1e3 then
bit_val = bit_val / 1e3
bit_suffix = "Kbps"
end
local bit_str
if bit_val >= 100 then
bit_str = string.format("%.0f", bit_val)
else
bit_str = string.format("%.1f", bit_val)
end
bit_str = bit_str .. bit_suffix
-- Format bytes (binary KiB/MiB)
local byte_val = bps
local byte_suffix = "B"
if byte_val >= 1024*1024 then
byte_val = byte_val / (1024*1024)
byte_suffix = "MiB"
elseif byte_val >= 1024 then
byte_val = byte_val / 1024
byte_suffix = "KiB"
end
local byte_str = string.format("%.0f", byte_val) .. byte_suffix
return bit_str .. " (" .. byte_str .. ")"
end
function conky_download_speed()
local kib = tonumber(conky_parse('${downspeedf enp3s0}')) or 0
return format_speed(kib)
end
function conky_upload_speed()
local kib = tonumber(conky_parse('${upspeedf enp3s0}')) or 0
return format_speed(kib)
end
You can also consider adding a fudge factor to account for tcp/ip overhead as well.
Woah, just as I put a Caramel candy in my mouth… 
I’ve been doing this computer thing since 1993, I come from DOS, and I had no problem with kilobytes and megabytes, but they were always expressed as KB and MB. Then I arrive at linux and I see KiB and MiB and some unexpected number, and became confused. It’s just not a notation that was ever used in the mundane world and as such, for average users it’s devoid of meaning. So yes, this is something that should default to the expected, and if you need that particularity, then set it that way.
Now I’m confused - what are you saying is ‘the expected’?
Also, you’re talking about a computer user… a technical field that would possibly expect bytes to be in binary format by default and never decimal at that time.
Are we to assume that normal folks didn’t go to school to learn that Kilo is 10³ and Mega is 10⁶ ? This might be feasible in many 3rd world countries (like the USA perhaps - schools aren’t always up to scratch) but for Europeans, there’s absolutely no confusion.
In short, the way ‘we’ did it in 1981 (for you 1993) was just wrong - we ‘kinda’ had an understanding… but that ‘understanding’ isn’t what is taught anywhere.
What’s expected: 1k = 1000 and 1ki = 1024… so the reason for our confusion was simply not being familiar in the early 90s with the new Ki format.
But that’s 30 years ago - so your voice is lost now in the mists of time.