Comment by justsomehnguy | Hacker News Reader

Comment by justsomehnguy | Hacker News Reader

> I’ve brought my text files with me since 1990, from Mac to Windows to Linux to BSD

> But plain text? Always. Everywhere.

Quite obviously you have your files in ASCII. Anyone who needs more than 26 letters (or god forbid non-latin ones) are fucked. Sometimes quite literally.

> Your writing should outlive you. Depending on companies is not an option.

ASCII? Sure. EBCDIC? Ughh. UTF... 8? 16? 32?

> If you rely on Word, Evernote or Notion, for example, then you can’t work unless you have Word, Evernote, or Notion. You are helpless without them

And you are absolutely helpless if you can't access your ASCII/EBCDIC/UTF8whatever plain-text files.

> Plain text can be converted into anything else.

Date: 2022.02.03

Is this March 3rd or February 3rd?

> NEED HIERARCHY

Sure. But in the year 2022 I want not only a hierarchy but crosslinks too.

> CONCLUSION

> Reliable

If you know the encoding. IF.

> flexible

See #date

> portable

See #date

> independent

Sure, but what about dates and encoding?

> and long-lasting

If on the suitable media. Good luck accessing your precious files on Google Drive when the mighty AI decided you aren't worthy to access them. Or your CD-ROM isn't readable because the last machine with working [0] CD-ROM compatible drive were thrown out 5 years ago because it can't run any modern OS[1]

> Plain text files will be readable by future generations, hundreds of years from now.

[x] Doubt

Real conclusion:

Sure, plain-text files are way better than binary blobs with unknown structure, but they are not a silver bullet. You still need to think about how, where and in what format to store them.

5/10, would not recommend.

[0] The last machine in my house with CD-ROM drive is X301 which is .. somewhere. My desktop's NEC-45xx was dead last time (6 years ago) I tried.

[1] CentOS8 performance with Gnome on ThinkPad X301 is abysmal to say the least. Even Windows 10 (!) runs better on it.

PS: to the kind person who bothered to downvote this comment: do you prefer your UTF encoded files with BOM or not? And your preferred editing software?

I didn't downvote you but your tone is overly dramatic for some simple heuristics. Specifically you're just repeating two things over and over. Here's some simple answers:

1) you're right that ASCII was English only, but now utf-8 is the defacto standard, backwards compatible with ASCII, and good for every language. Use that, no BOM needed and the BOM is discouraged by the standard anyway.

2) dates can be YYYY-MM-DD as per the ISO standard. Also they're sortable as a string too as a nice side effect.

All good?

Why can a vapid blog post from a famous person make the front page, but a serious, if slightly aggressive, response to it gets downvoted? Did everyone just want to circle jerk about how they too are just like famous person?

I think "vapid" is generous. The author seems to have no grasp of concepts like encoding, representation, and medium. And the other responses to the critical comment (not troll) quite miss the point.

I have a large amount of digitized data, including programs I first hand wrote on paper and then transferred to punch cards, poems written on paper and later typed in, photographs I've scanned, photographs I've taken digital photos of, scanned books and articles that contain text in various languages along with images, digitized music and videos, etc. Some of these consist of "plain" (whatever that means) text and some don't. I've been accumulating this data since the 1960s. When I think about my collection and the ways I manage it, and my knowledge of encodings, representations, and media, and then read this article, my most polite response is ... smh.

That said, I think there is a valid point in all that confusion, and it is stated by several of the comments on the article: use Markdown or something similar (but MD is usually the best choice) when you can rather than Word or other proprietary binary formats. But MD is not "plain" text.

> Anyone who needs more than 26 letters (or god forbid non-latin ones) are fucked.

Anyone ... are fucked.

Hmm...

> If on the suitable media.

Hmm.

(Oh, and still February 3rd, obv. March 3rd would be 2022.03.03.)

Even if you don't want to use utf8, you can use any other text encoding, it's your files, nobody cares how you use them, do whatever works best for you.

I'll give you a 8/10 troll approval rank.

> ASCII? Sure. EBCDIC? Ughh. UTF... 8? 16? 32?

8.

> Date: 2022.02.03

> Is this March 3rd or February 3rd?

February 3rd, obviously.

But the periods are weird; the standard is hyphens.

> Sure, but what about dates and encoding?

ISO and UTF-8.

Reply on news.ycombinator.com