Copy link to clipboard
Copied
HI All,
Having quiet a bit of issues with FM12: Created a Book, with crash, it created Recovery Files as shown in the image below:
Now I am starting with new Book.
When I add these files, It does not add to the book (no message, nothing).
When I rename the files and then add to the book, I get message Not Recognized Format.
I would appreciate any response. Thanks.
re: I am trying to understand what is the purpose if these files, if they may be corrupt?
Often they are not corrupt, or the corruption may be limited to the last edit you were making, or could be isolated to just one spot of document structure elsewhere (and FM detecting inconsistent structure is probably grounds for aborting itself). The .recover, plus your .backup, usually get you back on track.
.recover is FM's best effort to save your work when it detects an error that is going to result in i
...Copy link to clipboard
Copied
.recover.fm may be a reserved name in FM, and it's wise to avoid using such an extension for production files.
If later FMs still work the same way as legacy FMs, a .recover file is actually a MIF file, and not a .fm binary, despite the .fm extension.
.recover files are only assured to be syntactically correct, and can have corrupt content.
Open each .recover individually. Verify contents and structure. Re-save without the .recover. Then bring into book.
Copy link to clipboard
Copied
Hello Bob Niland,
Appreciate your response. Now I understand that .recover files are infact MIF file and not FM file (despite their .fm extention).
I am trying to understand what is the purpose if these files, if they may be corrupt?
One thing I can think of that it would give your all your data in file, but even then it would be quite time consuming to bring in new FM file and then format it all again...
If anyone has any insight, I would appreciate it.
Thank You.
Copy link to clipboard
Copied
re: I am trying to understand what is the purpose if these files, if they may be corrupt?
Often they are not corrupt, or the corruption may be limited to the last edit you were making, or could be isolated to just one spot of document structure elsewhere (and FM detecting inconsistent structure is probably grounds for aborting itself). The .recover, plus your .backup, usually get you back on track.
.recover is FM's best effort to save your work when it detects an error that is going to result in it crashing. It presumably uses MIF so that if the crash crashes partway through, you at least have a plainttext markup file from which to recover some work. More usually, you can recover all of your content, and most if not all of your formatting. A partially-written .fm binary file might be entirely un-openable, even with Heroic Open.
If the writes complete, you get a files that FM can open as well. If the crash crashes (or the Window, VM, or whole OS does), you might get nothing at all, as was the case for the Error 7103 crashes I used to get in a former assignment - FM would just wink out without a trace. Save early and often. Backup the entire job daily at least.
Always keep in mind that the first thing FM usually does on opening file xyzzy.fm is to over-write file xyzzy.backup.fm, so when recovering from a crash, first make copies of everything.
Copy link to clipboard
Copied
Hello Bob Niland,
Thanks you so much for taking time explaining the purpose of .recovery and .backup.
We had inherited this project from other group and it was done quiet a while back in FM7. With not so good track-record, as a result, somewhere along the way (obviously) we have corrupted some files. For that reason, we are sorting through it a the moment.
In any case, as your response has helped me, I hope it will also help other users.
Thank you.
Copy link to clipboard
Copied
re: We had inherited this project from other group and it was done quiet a while back in FM7.
When working with a legacy project, the very first thing to do is preserve copies of all the files as-is, before even attempting to open in FM.
In the case of FM7 or earlier, bringing a project forward is going to result in automatic partial conversion of some of the text to Unicode. Complete conversion is likely algorithmically impossible, so you'll need to inspect carefully.