Which format is best for long-term storage?

[eschwartz]

Originally Posted by Nova5

This is more what I was intending to say. Don't convert to LIT from epub/lrf/etc. But if you have a LIT file, keep it around for conversion from as the master file for any new file type that may be created.

Um, no. It doesn't matter what the master format is, LIT or otherwise. You still keep it. Because it is the master file, and every conversion from anything --> anything will lose data.

For instance, calibre conversions flatten the CSS, so if you ever want to edit it, you'll want your master file.

[JSWolf]

Originally Posted by eschwartz

Um, no. It doesn't matter what the master format is, LIT or otherwise. You still keep it. Because it is the master file, and every conversion from anything --> anything will lose data.

For instance, calibre conversions flatten the CSS, so if you ever want to edit it, you'll want your master file.

Actually, I disagree. For example, Mobi is a pain to edit and it's actually easier to convert to ePub, make the edits, get it looking as you want and then convert back to Mobi.

[eschwartz]

Originally Posted by JSWolf

Actually, I disagree. For example, Mobi is a pain to edit and it's actually easier to convert to ePub, make the edits, get it looking as you want and then convert back to Mobi.

If you create your own master file through painstakingly tedious editing effort, obviously you don't need the original file.

If you use off-the-shelf conversions to put on your new device, always convert from the original.

[Timoleon]

BookDesigner makes beautiful customized lit files, which Calibre converts into beautiful epub and mobi files. I keep all my books saved as lit, epub, mobi, and htmlz files, and store them on an external drive as well as on OneDrive (and of course my C:drive). Keeping ebooks in various locations has saved my bacon numerous times, since 1986. From 5¼" to 3½" floppies to CDs to zip drives (remember them?) to DVDs to external drives to flash drives to the cloud --- my library has managed to survive numerous disasters and user stupidity over the years. Hell, in a fit of idiocy, I just installed the latest Madeira Linux to the "wrong" partition on my D:drive --- one filled with data. So my advice is obvious: save your ebooks in as many formats as you can, in as many places as you can. You may not think that something will happen --- but it will.

[JSWolf]

BookDesigner is one program I do feel might not be long for this world.

[GERGE]

I keep them in txt, ePub and KePub.

[HarryT]

Originally Posted by JSWolf

BookDesigner is one program I do feel might not be long for this world.

It's what I use to create all my books. Works absolutely fine with Windows 8.1. Why do you think it's "not long for this world"?

[RichL]

For the hardware I have no idea. For the formating I would say save it as .txt. Plain old text files have outlasted every other format I can think of and is showing no signs of disappearing anytime soon.

[HarryT]

Originally Posted by RichL

For the hardware I have no idea. For the formating I would say save it as .txt. Plain old text files have outlasted every other format I can think of and is showing no signs of disappearing anytime soon.

Plain text files have no formatting. You'd lose a great deal of information by storing a plain text file. That really wouldn't be a terribly good idea.

[ApK]

Originally Posted by HarryT

Plain text files have no formatting. You'd lose a great deal of information by storing a plain text file. That really wouldn't be a terribly good idea.

Right.
Which is why .txt isn't at the very top of Calibre's preferred conversion input format list.
If you take what Kovid said purely literally, clearly ALL other ebook formats are indeed a superset of text....

Our goal is protect our "books" as whole, not merely the knowledge in them.
We are guarding against technical obsolescence, not really just ensuring that cave men will be able to rebuild civilization after the collapse.

And truthfully, maybe if civilization did collapes like that, the cave men would be better off re-learning everything again for themselves, like they did the first time around.
Apparently anything we'd have to teach them will have led to the collapse of civilization. Better they start fresh and try again.

[HarryT]

Originally Posted by ApK

Right.
Which is why .txt isn't at the very top of Calibre's preferred conversion input format list.
If you take what Kovid said purely literally, clearly ALL other ebook formats are indeed a superset of text....

A good compromise would be plain text files with some sort of additional "markup" to provide additional information. We could call it something like "hypertext markup language" .

[jandrew]

Originally Posted by HarryT

Plain text files have no formatting. You'd lose a great deal of information by storing a plain text file. That really wouldn't be a terribly good idea.

Well, some people have different ideas of what makes a plain text file. My LaTeX files, html files, markdown files, xml files, css files, rdoc files ... are all plain text files.

[ApK]

Originally Posted by jandrew

Well, some people have different ideas of what makes a plain text file. ..

...to the point of making discussion meaningless, if you're going to play that fast and loose with terms that are widely understood perfectly well.

[eschwartz]

Originally Posted by jandrew

Well, some people have different ideas of what makes a plain text file. My LaTeX files, html files, markdown files, xml files, css files, rdoc files ... are all plain text files.

No, those are formatted text. You can tell because of the mimetype.

Anyone who feels the need to describe it as plain text vs. binary (text? ) is causing themselves and everyone around them unnecessary confusion without, as far as I am aware, having any meaningful point.

Just in case, I will explain it for you:

There is binary and then there is text. Text is readable because it is based on characters we understand. Binary is NOT readable because it uses computer codes that need to be interpreted by specific programs. With me so far?

Text comes in different types, like:

• text/plain -- Just Regular Words, like the kind you see on a sheet of paper. Commonly referred to as "plain text".
• text/html -- text with textual tags which can be read but obscure meaning, unless you happen to know which stuff is meta info, and even then requires additional effort to mentally figure things out. A type of formatted text.
• text/css -- textual info with no meaning outside that given it by some sort of (web) processing engine.
• text/x-markdown -- text with an agreed-upon style of contextual sheet-of-paper-like contextual info. (Mostly) readable as text/plain but only with a subset of the meaning. (Because of things like links.)

and others.

There is probably some sort of reason, silly though it may be, why people come to agreements about the meaning of specific terminology.

[jandrew]

Originally Posted by eschwartz

Just in case, I will explain it for you:

There is binary and then there is text. Text is readable because it is based on characters we understand. Binary is NOT readable because it uses computer codes that need to be interpreted by specific programs. With me so far?

Thank you. I have been a software developer for a couple decades.

Yes html (and sgml, LaTeX, etc) is considered formatted text or rich text, but still plain text. In the context of a discussion of preserving content for the future, the distinction is important. Some people in this thread seem confused to the point of thinking that a text file stripped of all html is more future proof than the html file itself. This is a misconception that only amounts to a loss of information.