How to Rescue an eBook Library Trapped in a Dead Reader Format
There is a particular kind of digital loss that creeps up on people. You find an old external drive, or a folder that has been copied from laptop to laptop for fifteen years, and inside it is a library - a few hundred books you bought and read on a device that no longer exists. The files are all still there. Nothing you own today will open a single one of them.
These are the formats of the first e-reader era, when every manufacturer shipped its own. This post covers what each of them is, how to get your books out into something current, and - the part most guides skip - the honest limits on what can be rescued.
The formats, and the devices that took them down
LIT - Microsoft Reader. Microsoft's reading app launched in 2000 with ClearType as its headline feature, ran on Windows and Pocket PC, and had a real catalogue behind it. Microsoft retired it in 2012. A .lit file is an LZX-compressed container of HTML, so the text inside is entirely recoverable - provided it is not one of the DRM-protected copies bought from a store that no longer exists.
LRF - Sony BBeB. "Broad Band eBook" was Sony's format for the Reader line, the e-ink device that arrived before the Kindle and, for a while, was clearly better. Sony eventually moved its own devices to EPUB and later exited the eBook business entirely, handing customers over to Kobo. Anything bought early in that era is .lrf, a binary format that only Sony's software ever fully understood.
PML - Palm Markup Language. The markup behind eReader, the Palm-era reading app that survived onto early smartphones. PML is a lightweight text markup, so these convert cleanly and are among the easiest to rescue.
PDB - Palm Database. Less a format than a container: .pdb is the generic Palm OS database wrapper, and several different eBook formats used it. That means two .pdb files can have essentially nothing in common beyond the extension, which is exactly why they are awkward to open today.
TCR - Psion. A simple compressed text format from the Psion handhelds. Text only, no formatting, and the smallest rescue job on this list.
HTMLZ. The odd one out - not from a dead device but from Calibre, which uses it as an export format. It is a ZIP containing HTML and its assets, and it turns up when someone exported a library years ago and forgot what the extension meant.
Start with DRM, because it decides everything
Before converting anything, it is worth understanding the one thing that determines whether a rescue is possible at all.
Books sold with DRM are encrypted. Opening one required your device to check in with the seller's activation server and receive a key. Those servers are gone - Microsoft's, Sony's, and most of the smaller stores of that period shut down years ago. The encryption did not expire along with them, which means a DRM-protected book from a dead store is, in a straightforward technical sense, no longer readable by anyone, including the person who paid for it.
No converter changes this. Anything claiming otherwise is either stripping DRM - which is illegal in many jurisdictions and not something we do - or quietly handing you an empty file. Our converters take the third path: a protected or unreadable book fails with an explicit message saying so, rather than producing a valid-looking EPUB with nothing inside it.
The good news is that plenty of these libraries convert fine. Free titles, public-domain texts, books from independent publishers, technical documentation, and anything from stores that sold unprotected files are all normally DRM-free. The fastest way to find out is to convert one book and see what happens.
Pick your output format once
Decide before you start, because converting a hundred books twice is tedious.
- EPUB is the right answer for almost everyone. It is the open standard, it is what Apple Books, Kobo, Google Play Books and most reading apps use natively, modern Kindles accept it, and the text reflows to fit whatever screen you read on. If you are not sure, choose this.
- PDF makes sense only when you want fixed pages - for printing, or for a document whose layout matters. On a phone screen a PDF of a novel is genuinely unpleasant, because the text cannot reflow.
- MOBI / AZW3 are the Kindle formats. Worth choosing if you have an older Kindle; newer ones handle EPUB directly, so EPUB is usually the better bet.
- DOCX if you intend to edit the text rather than read it.
- TXT if you only care about the words and want the most durable possible container. No formatting survives, and nothing will ever fail to open it.
Converting your books
Each converter takes one file and returns the result in a few seconds - no account, no software to install, and no need for the original device or reading app:
Additional targets - Kindle's MOBI and AZW3, plus DOCX and TXT - are available for the LIT and LRF formats through the API, which is the sensible route if you are working through a large library programmatically rather than a book at a time. The eBook tools page lists everything available, including conversions between the formats still in active use.
What survives the conversion, and what does not
Being realistic about this saves disappointment later.
Reliably preserved: the complete text, chapter structure and table of contents, paragraph and heading formatting, italics and bold, and embedded images.
Often lost or changed: the exact page layout, since a reflowable format has no fixed pages to preserve. Page numbers therefore will not match the original, which matters if you have citations pointing at them. Unusual typography, drop caps and decorative styling are commonly simplified. Bookmarks, highlights and reading position lived in the reader application rather than in the file, so they do not travel.
Where it gets rough: heavily designed books - cookbooks, textbooks, technical manuals with complex tables - were pushing the limits of what those old formats could express, and the layout tends to come across imperfectly. Plain prose converts close to perfectly. A novel will look right; a chemistry textbook may not.
Working through a whole library
A sensible order of operations if you are facing hundreds of files:
- Copy the originals somewhere safe first. Everything else in this list assumes you can start over.
- Test three or four books across the range - a plain novel, an illustrated one, something with tables. This tells you within minutes whether the library is DRM-free and what quality to expect.
- Sort by extension. Libraries from this era are usually mixed, and knowing you have 200 LIT and 40 LRF files makes the job concrete.
- Convert the important books first, in case you run out of patience halfway through. You almost certainly will.
- Keep the originals afterwards. They are the masters. Converters improve, and if a better one appears you will want the source files.
For a genuinely large collection, scripting the API against a directory is far less painful than clicking through a browser several hundred times.
Why these files are handled carefully
Obsolete binary formats deserve some caution. These are compressed containers designed decades ago, parsed by code that is now maintained mostly out of goodwill, and a deliberately malformed file is a reasonable way to attack a parser. That is a general risk with any converter, not a reason to avoid converting.
On our side, every one of these conversions runs inside a sandbox with no network access whatsoever and a hard memory cap per job, entirely in server memory rather than on a disk. The memory cap is what stops a decompression bomb, and the missing network is what makes a book with embedded remote references unable to fetch anything. Your file is discarded the moment the converted copy is sent back - there is no storage and nothing to delete later.
Which is a fitting way to treat a personal library. The books were yours to begin with; getting them into a format that will still open in twenty years should not mean handing a copy to anybody.