Format Wars: ASCII vs EBCDIC

The first episode of a new series I'm introducing, Format Wars.

This episode focuses on the character encoding standards ASCII and EBCDIC, a fight between backwards compatibility and common sense.

[thewiirocks]

Good overview! A few notes: 1. It's pronounced "ebb-sih-dick", 2. EBCDIC is also one bit off between capital and lower case. My boss used to tout that as a feature of EBCDIC. (Until my coworker pointed out that ASCII had the same convenience.) 3.Technically there are multiple Unicode standards. UCS-2, sometimes called UTF-16 was the best known one early on. However, 16 bits was actually very limiting. Several more Unicode standards were created with the world eventually settling on UTF-8. UTF-8 is awesome because it's looks like ASCII. But those values 128 and above actually tell decoders to use more bytes for the character. This means that UTF-8 characters are of a theoretically unlimited length! This was a huge improvement over even 32 bit Unicode which promised to be wasteful and STILL not be big enough in the long run.

[vera]

Just binged your entire channel. Awesome videos my dude, keep it up!

[Fopenplop]

man, I sure am glad that the unicode consortium solved the problem of universal character encoding standardization for all time with such a flawless and perfectly supported system.

[RaymondHng]

I've always heard EBCDIC pronounced as IB-suh-dik since I have been using computers in 1974.

[gerhardbiebl]

When i was developing Text conversion software in the late 80's / early 90's, there was a joke about EBCDIC:
On the classified internal EBCDIC Documentation of IBM they printed the words: "Burn before reading!" 😊

[AiOinc]

>Support for other languages
>Also emojis


There is no reason why I would want to use this. I'm switching back to ASCII.

[jeberle]

From what I read, Bob Bemer developed the early ASCII charset to unify the disparate charsets that were used inside of IBM. It's a sweet irony that the only player not able to take advantage of ASCII's success was IBM itself. Insane.

[okaro]

ASCII was pretty much established in the microcomputer world before IBM PC. IBM PC's great innovation was to support full ASCII with various letters needed in Europe. Earlier computers had replaced characters like []\{}| with various national characters.

[therealjammit]

I like standards. There are so many to choose from.

[lawrencedoliveiro]

2:19 “Preventing standards clashes” ... by *not* adopting the international standard A-series for the size of a sheet of paper, like most other countries did?

[FindecanorNotGmail]

Unicode is far from "perfection". I find it to be a real mess. Instead of making it convenient to use, it has the goal of encompassing the codes from every other character set - including EBCDIC. For instance line endings: CR, LF CR+LF ... and "Newline" from EBCDIC. Then because those can be used in different ways Unicode added their own "unambiguous" codes "Line separator" and "Paragraph separator" - and all are valid line endings. Then there are many pairs of letters that _look_ the same but are actually different characters: a property that has often been used to scam people on the Internet.
Characters are in _logical_ order, not the order they are displayed on screen, which means that laying out for instance Hebrew and Arabic (that go right to left) is a complex task.
Many Japanese and Chinese characters occupy the same code space, which meant that you have to encode what language it is on the side
Also, control codes are scattered all over the code space.
All this means that code that handles Unicode correctly and safely in a non-trivial way has to be quite complex.

[vwestlife]

It's pronounced "Ebb See Dick". And ASCII was already the universal standard for personal computers long before the IBM PC came long. So if anything, it was ASCII that helped the success of the IBM PC, not the other way around -- if the PC had used EBCDIC, it would've been seen as archaic and proprietary.

[stumpybear]

Eb-sid-dic is the correct pronunciation.

[FlameRat_YehLon]

Before Unicode become the de-facto standard for everything, in China things are basically a mess. Even nowadays, there is still quite a few software using either GBK or Big-5. This does cause quite some issue if say a Big-5 software is being used on a GBK computer. And even in cases like, accidentally stored some GBK data in a Unicode software and it's unreadable.

One factor that contributes to such a mess a lot is that Windows defaults to use GBK rather than UTF-8 or UTF-16 on the Simplified Chinese version, most probably for backward compatibility for DOS software.

But, well, things aren't completely perfect on Linux as well. If the system language is set to Chinese, you basically have to use a graphical SSH terminal as the native terminal can't display Chinese. (Have no idea what genius made it that Bash would output Chinese error message.) This is not a format issue though.

[altrogeruvah]

I very much enjoyed this, thank you for this video.

[DamianShaw]

We still use EBCDIC at work for IBM Mainframes, but it was always pronounced eb-sid-ick. Do you know the official pronunciation?

[rkpetry]

[00:48] what are you saying—almost like 'hasskey'—and we say, 'eb-cee-dic'...
[02:45] "charcter" (sic: 'character'), 𝒶𝓃𝒹 𝕨𝕙𝕖𝕣𝕖'𝓈 𝓎ℴ𝓊𝓇 5-𝒷𝒾𝓉 𝔙𝔞𝔲𝔡𝔬𝔱 code...
P.S. we use UTF-8 mostly these days but we really need UTF-X for plaintext...
(we, want, all, letters,  superscript-able, subscript-able, italics, faces,  colors...).

[TardisRepairService]

People pronounced it with an S sound for the first C. Also, ASCII had gotten popular way before IBM released the PC.

[RonJohn]

2:58 Think *octal.* Decimal 201 is octal 311. Decimal 209 is octal 321. IOW, the alphabet is chunked into groups of 8.

(ASCII is still better.)

[JustWastedHoursHere]

Anyone old enough to remember BBSes knows that some pretty cool graphics (even animated!) could be made with various extended ASCII character sets, such as on the Commodore 64. In fact, even some games made use of them.