25 year old computer file?

[mcgriff]

As I was debating today, once again, on what reader to buy a thought struck me. My first computer was a C64 from the way back 80s. I don't have a shred of media from that machine and if I did I doubt I would be able to do anything usable with such ancient code. I do however have many books from the 80s that are perfectly readable today.

With this in mind, is stressing out about what device to get or what format to buy in to really that important? Fifteen years from now will it matter if one's ebooks are all epub or azw? The frustration of maintaining the usability of old computer files often outweighs the cost and convenience of simply diving into a new format. Yeah, txt, doc, jpg have been around for quite sometime, but should we expect the same from topaz or the like?

Seems to me ebooks are a 10-15 year investment after which time technology will simply no longer support hardware capable of manipulating current ebook formats. I am fairly certain a paper book in hand will last far longer than an electronic book. I'm quite certain this viewpoint has been discussed in the past. It simply gave me a new way to look at my future investment in an ebook reading device.

[Solicitous]

Originally Posted by mcgriff

Seems to me ebooks are a 10-15 year investment after which time technology will simply no longer support hardware capable of manipulating current ebook formats. I am fairly certain a paper book in hand will last far longer than an electronic book. I'm quite certain this viewpoint has been discussed in the past. It simply gave me a new way to look at my future investment in an ebook reading device.

I think the smart thing to do, and not just with ebook files but all computer files in general, is do not keep a file for 10-15 years then hope you can open it. Technology moves along, so if you have files that current software is starting to reduce support for, convert then and there. Think back to the Word Perfect days, the product didn't disappear when the product was essentially dropped.

[iandix]

The Digital Dark Age.

That is what historians may call the period we now live in. In a hundred or more years, will there may not be many photos, films, even printed books for them to gather the knowledge they need about how we lived.

I was reading NASA is having are hard time trying to save the data from the early space missions because they don't have hardware you need read it and it has deteriorated to a point that it's just about useless.

I use to know a guy with a massive VHS collection, I wonder what he does with them now :-)

[wallcraft]

Originally Posted by mcgriff

With this in mind, is stressing out about what device to get or what format to buy in to really that important? Fifteen years from now will it matter if one's ebooks are all epub or azw? The frustration of maintaining the usability of old computer files often outweighs the cost and convenience of simply diving into a new format. Yeah, txt, doc, jpg have been around for quite sometime, but should we expect the same from topaz or the like?

There are two related, but distinct, issues with very old files, a) the media, and b) the format. With not quite so old files we add c) DRM. The obsolete media issue is the easiest to avoid going forward - just make sure to copy to a new storage device every few years. DRM is the enemy of longevity, so never buy a format that isn't circumventable and immediately strip the DRM and keep multiple copies of the DRM-free file. The format is now a non-issue if you stick to formats that Calibre fully supports. The point being that we know how to handle these formats and this knowledge is in an Open Source program that can give you back whatever other format you want.

TOPAZ is an example of a format that Calibre can't handle (i.e. can't convert to another format). This was initially due to DRM and lack of documentation and its unique approach to scan conversion. Now, due to heroic reverse engineering, it is possible to convert TOPAZ to something else, but only with difficulty. I can't see any reason for TOPAZ to exist in 15 years time, although Amazon might still be a major player and might still include TOPAZ support in its readers. My advice is not to buy a TOPAZ ebook with any expectation of support on future reading devices.

[Greg Anos]

How much do you value your data? There are no standards that last. So you have to convert your files on a steady basis. If they are non-executable files. (i.e. e-books, music, pictures) You will just have to run them thru a conversion program every so often, to keep up with the new standards. But most people spend more time buying toilet paper than taking care of their data. That's why NASA has such troubles today, nobody valued the old data at all. not enough to even copy it to a new machine, years ago. Same thing happens with people; no backup, no copying, no nada. Nothing it ever supposed to wear out. But the buy cars every few years...

[sourcejedi]

From this point of view, EPUB is pretty good: they're a ZIP archive of HTML files. The subset of ZIP used in compliant EPUBs is a completely open standard (with complete open-source implementations).

So the worst case in 50 years time - you have to rename it to .ZIP, unzip it to a directory, and make sure your OS recognises the files as HTML (I guess some pathological cases might require you to rename them to add a .HTML extension).

Any PC EVAR will be able to read the resulting directory. I can even read it on my e-reader (OpenInkpot-based).

Just remember to strip any DRM *now* - there's no guarantee you'll be able to find the same DRM software in 5 years time (even the official reader software...). And remember to keep a PC. I'd be very surprised if you can do this unzipping and renaming on the current version of the iPad :-).

[HarryT]

Originally Posted by sourcejedi

I'd be very surprised if you can do this unzipping and renaming on the current version of the iPad :-).

There are a number of "unzip" and file management tools available for the iPad.

[Jellby]

Originally Posted by sourcejedi

So the worst case in 50 years time - you have to rename it to .ZIP, unzip it to a directory

If your "worst case" assumes that an ePUB no longer works... why do you assume that a zip would work?

[HarryT]

Originally Posted by Jellby

If your "worst case" assumes that an ePUB no longer works... why do you assume that a zip would work?

The point about zip is that it is a well-known algorithm. Even if old zip programs don't work, you'd still be able to use the algorithm to create a new tool for whatever systems are around at the time.

[AprilHare]

I have an Amiga 500 here (not used very much obviously!) and I remember using the Textcraft Plus text editor once, a lot.
The technical and practical problems involved with moving those documents (assuming they haven't been lost to disk corruption! ) to a modern computer and then to a liseuse are not trivial.
I may track down some (or generate some) and tell of the tale, just for illustration purposes.

Originally Posted by mcgriff

As I was debating today, once again, on what reader to buy a thought struck me. My first computer was a C64 from the way back 80s. I don't have a shred of media from that machine and if I did I doubt I would be able to do anything usable with such ancient code. I do however have many books from the 80s that are perfectly readable today.

With this in mind, is stressing out about what device to get or what format to buy in to really that important? Fifteen years from now will it matter if one's ebooks are all epub or azw? The frustration of maintaining the usability of old computer files often outweighs the cost and convenience of simply diving into a new format. Yeah, txt, doc, jpg have been around for quite sometime, but should we expect the same from topaz or the like?

Seems to me ebooks are a 10-15 year investment after which time technology will simply no longer support hardware capable of manipulating current ebook formats. I am fairly certain a paper book in hand will last far longer than an electronic book. I'm quite certain this viewpoint has been discussed in the past. It simply gave me a new way to look at my future investment in an ebook reading device.

[sourcejedi]

I agree with Ralph a little bit as well. The first challenge is to actually keep the files.

I buy ebooks on my PC, organize them in a folder on my hard-drive, and use a file synchronizer to keep my e-reader in sync. So without any extra effort, I have two copies of everything. When my e-reader or PC spontaneouslly combusts, I'll still have one copy of everything. (And I'm fortunate enough to be able to run backups as well).


I'm a bit leery of the "keep converting your files" idea though. It kind of relies on every conversion being perfect. But e.g. Calibre's conversion definitely isn't perfect. IMO it takes more of a pragmatic approach that lets you read the 10s of existing formats on your existing e-reader. That's why it has different output profiles for different devices (even for the same format e.g. EPUB readers).

At the very least, I think you'd need to keep the original, and a "master" version for each format you convert into, so that you can always back-track if necessary. IMO that's not convenient enough that people will do it as a matter of course. Using Calibre would help, but I don't know that it makes it easy enough to convert to both <my XYZ e-reader> and <generic format with lots of future-proofing>.

For example, I've snarfed a few Baen-CD ebooks. I've taken the Mobipocket/Palm files because they work best in FBReader (there's also one-webpage-per-chapter HTML, but that's not as quite as convenient for reading, and a pain to download and organize). But in the long run these are dead file formats, and you'd have to consult the computational archaelogists.

So I've also downloaded the RTF versions. RTF is still a proprietary format - MS keep on adding new features that no-one else can read - but it's more a sort of proprietary HTML. You can actually open these RTF files in a text editor. I could write a five minute script to strip the formatting commands and just dump out paragraphs of plain text. More realistically, RTF is an important file format and there are lots of readers, editors and converters for it. FBReader can also read it, so it's not a complete dead weight.

I wouldn't rule out buying mobipocket or rtf in future, it's just that I probably won't have to make that decision. I can't read DRM'd files, and anything thats not DRM'd is going to be available as EPUB. (Other than PDF, but in practice my e-reader can't read that either).

[sourcejedi]

Originally Posted by HarryT

There are a number of "unzip" and file management tools available for the iPad.

My impression was that individual apps and their files were completely isolated from each other, with a few limited exceptions for built-in tools. Thanks for the correction .

[sourcejedi]

Originally Posted by HarryT

The point about zip is that it is a well-known algorithm. Even if old zip programs don't work, you'd still be able to use the algorithm to create a new tool for whatever systems are around at the time.

Or more technically, "recompile the Linux version of the unzip command".

It's not proof against digital apocalypse, but then neither are the ebook files themselves. If my ebook files survive, then so will the C programming language. If you're worried about it, keep the unzip source code alongside your ebooks. It's probably the last thing worth worrying about though.

[Jellby]

Originally Posted by HarryT

The point about zip is that it is a well-known algorithm. Even if old zip programs don't work, you'd still be able to use the algorithm to create a new tool for whatever systems are around at the time.

Yes, that's true, and that was my point. In a "worst-case scenario", the zip algorithm could be obsolete at the time and you'd have to work a bit harder to get a working unzip, may dig into some library code, deal with old programming languages, try to make it work with current compilers... Saying that in 50 years you'll be able to just rename and unzip the file looks like saying that in 50 years you can just open the .epub with ADE (keeping in mind that we are in a "worst-case").

[luqmaninbmore]

Originally Posted by Jellby

Yes, that's true, and that was my point. In a "worst-case scenario", the zip algorithm could be obsolete at the time and you'd have to work a bit harder to get a working unzip, may dig into some library code, deal with old programming languages, try to make it work with current compilers... Saying that in 50 years you'll be able to just rename and unzip the file looks like saying that in 50 years you can just open the .epub with ADE (keeping in mind that we are in a "worst-case").

There are some who would consider opening ePubs in ADE a "worst-case" scenario period.