08-26-2013, 08:15 PM | #16 |
....
Posts: 1,547
Karma: 18068960
Join Date: May 2012
Device: ....
|
Yes you should have . So can see your point.
But the truncation is not something that would have affected me as I am agnostic to the structure of the Calibre library folders and when loading books into Calibre (drag and drop) I always do a visual title check first within Calibre rather than relying on the file name, etc. being the same and Calibre warning me, so I don't get duplicates (well not often, and if so it is my own fault ). |
08-26-2013, 10:42 PM | #17 |
creator of calibre
Posts: 43,749
Karma: 22446736
Join Date: Oct 2006
Location: Mumbai, India
Device: Various
|
Filenames have always been truncated. The truncation was made more strict in the new backend to workaround problems that could occur in some corner cases with the old truncation code, the full path length could occassionally go over the windows path length limit.
|
Advert | |
|
08-26-2013, 11:30 PM | #18 | |
Guru
Posts: 766
Karma: 6528026
Join Date: Sep 2012
Device: Kobo Elipsa
|
Quote:
And I care about how the book is named inside the OS's file structure because duplicate filenames cause problems when I try to move those files to the iPad. No OS (as far as I know) can differentiate between two files with the same name, so when adding files to the iPad, the first file is overwritten when the second file is added, unless I change one of the file names. Last edited by icallaci; 08-26-2013 at 11:43 PM. |
|
08-27-2013, 12:29 AM | #19 | |
....
Posts: 1,547
Karma: 18068960
Join Date: May 2012
Device: ....
|
Quote:
My approach is that I have a Library called Temporary and copy the book to be worked on into that (using the Calibre "Copy to Library" command) then open it (with the "Open With" plugin) with the editing tool one favours (in my case normally Sigil). I also use that library for conversions or for tidying new books up, if required, before adding them to my "permanent" libraries. When editing is finished and the epub, in my case, is verified I just delete the old copy in the library it originally came from and copy the edited one back using Calibre's "Copy to Library, delete after copy" command. That keeps everything inside Calibre, so the fact that the filename may get truncated differently when returned to the library is of no consequence (nor in my case, of any interest). On the odd occasion that I may edit a book file taken directly from the library folders themselves I return it direct to the folder myself with the filename unchanged, not by returning it via Calibre. To copy a file from the library folders oneself, and then return it to the library with Calibre is something that I would always suspect to have the potential for problems unless you delete the original copy from the library before adding back the new. In the end, even though I am quite at home working with files directly, it is much quicker for me to do it all within Calibre and doing so presents little opportunity for error. Last edited by AnotherCat; 08-27-2013 at 12:31 AM. |
|
08-27-2013, 01:37 AM | #20 | ||
US Navy, Retired
Posts: 9,864
Karma: 13806776
Join Date: Feb 2009
Location: North Carolina
Device: Icarus Illumina XL HD, Nexus 7
|
Quote:
Quote:
Last edited by DoctorOhh; 08-27-2013 at 01:43 AM. |
||
Advert | |
|
08-27-2013, 09:39 AM | #21 |
Guru
Posts: 766
Karma: 6528026
Join Date: Sep 2012
Device: Kobo Elipsa
|
We have multiple computers and multiple users in our house, so I can't always use the same computer to fix ebook files. And I never copy a fixed ebook file directly back into the Calibre book directory because often the file size has changed (and Calibre doesn't register the change unless I re-import it using the Add Books button). None of this should matter, though, because this method has always worked in the past, and Calibre does have a setting to "overwrite existing book formats," which it is definitely not doing.
Edit: Oops, sorry, I am mixing threads here. Over in Calibre's Library Maintenance section, there is a thread about Duplicate Formats, which is tangentially related to this one. (https://www.mobileread.com/forums/sho...d.php?t=220948) The problem of orphaned ebooks in Calibre 1.0 is related to the new way of truncating. That problem has now been solved (thanks, Kovid!), which has also solved most of my major headaches with truncation of file names. Last edited by icallaci; 08-27-2013 at 10:39 AM. |
08-27-2013, 10:49 AM | #22 | |
Well trained by Cats
Posts: 29,662
Karma: 54369090
Join Date: Aug 2009
Location: The Central Coast of California
Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A
|
Quote:
Editing the file CONTENTS and DIRECT replacement, should not be an issue as long as the existing file is not open. |
|
08-30-2013, 12:46 PM | #23 | |
Zealot
Posts: 105
Karma: 414068
Join Date: Feb 2013
Device: iPad Pro, Kobo Aura One
|
Quote:
I think there's something to be said about the simplicity of using Finder/ (Windows Explorer if it's still called that?) to do basic file management. The biggest missing feature of calibre, to me anyway, is full-text search. Searching metadata is mediocre at best when it comes to my workflow (I'm a grad student) and I use Spotlight (OSX's built-in full text search) to find books about 99% of the time. And Spotlight can find text in PDF's, epubs, mobis etc. And when I do find a book, I can just use QuickLook to preview it and copy text out of it or read a section. It's so fast and makes me productive and calibre just doesn't support a workflow like that (it can't even do QuickLook right now… you have to open PDF files). Anyway, there are options and you don't have to use calibre for everything. Calibre is a superb piece of software but its cross-platform nature and non-native widgets and weird user interface idiosyncrasies that sometimes make little or no sense are not its strengths. Metadata fetching and conversion are must-haves and irreplaceable features right now. I understand why Kovid is truncating filenames but that doesn't mean that the file structure of calibre is not important. It clearly is because if it was that critical to have bullet-proof file names, Kovid would have used hashes as filenames which would make it more robust on any FS (always 256bits long and filename's characters would be compatible with any FS). But that would raise even more ire since people would complain even more. |
|
10-09-2013, 04:55 AM | #24 |
Member
Posts: 15
Karma: 10
Join Date: Oct 2011
Location: London England
Device: Kindle 4 non touch,paperwhite 10
|
Joining this late but I can confirm that something indeed has changed in the way that Calibre (Excellent product btw) works in the way it renames files.
Yes I know that the physical file name shouldn't be an issue with the metadata etc however like others here I use the file name for other purposes (generating webpages, database links etc ) and the file-name IS important. having looked into this Ive too tried a few things. My Calibre library is set to write to C:\A\... so one of the shortest possible root there can be and its still truncating the filename - the longest one Ive found so far is 124 characters, well within the 256 (win 7 64 bit and XP versions) limits. I can also now seethat having kept a few older versions that this definitely changed with the release of 1.0 as 0.9.35 does not exhibit this behavior yet everybody (including Im afraid Kovid) seems to deny its happening.... Has anybody found a solution yet please (bar going back to 0.9.35) Thanks. |
10-09-2013, 07:18 AM | #25 |
creator of calibre
Posts: 43,749
Karma: 22446736
Join Date: Oct 2006
Location: Mumbai, India
Device: Various
|
Sigh, how many times do I need to say the same things.
Yes, the file name truncation logic has changed in 1.x. File names are now truncated a little more strictly than before, this was because of a bug in the old logic that could, in some corner cases, cause file paths to become longer than 265 characters. The length at which filenames are truncated has nothing to do with the location of the library. If it did, it would be a recipe for disaster. Suppose you create a library at c:\library and tomorrow decide to move it to c:\users\me\my documents\my school library. Boom, your files would become inaccessible. calibre allows for ~70 characters for the path to the library and truncates the file/folder names so that any library path upto 70 characters will work. And no those numbers are not exact, if you want the exact calculation, use the source, Luke. And yes, I know that you can use UNC style paths to access paths longer than 265 characters on windows, however, while calibre itself could do that, lots of other tools would choke and be unable to even see the files, including Windows Explorer itself. |
10-11-2013, 02:49 PM | #26 |
Member
Posts: 15
Karma: 10
Join Date: Oct 2011
Location: London England
Device: Kindle 4 non touch,paperwhite 10
|
Ok I understand
Anyway, can I suggest an update then to provide something to maybe highlight a too long file name, show the shortcut/edited/truncated name so you have an idea that it may change the name before you convert it..?? as I like I'm sure many others use Calibre to rename files like dgfdjjje.epub to something more meaningful/book title. It keeps folders etc nice and neat and makes finding stuff outside Calibre nice and easy... Also keep up the good work. it truly is an excellent product |
10-11-2013, 04:18 PM | #27 |
null operator
Posts: 20,421
Karma: 25894272
Join Date: Mar 2012
Location: Sydney Australia
Device: none
|
Suggestion for Kovid - if mouse hover on format in Book details show full path as a tooltip/hint.
BR |
10-11-2013, 04:36 PM | #28 |
Well trained by Cats
Posts: 29,662
Karma: 54369090
Join Date: Aug 2009
Location: The Central Coast of California
Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A
|
|
10-11-2013, 05:02 PM | #29 |
null operator
Posts: 20,421
Karma: 25894272
Join Date: Mar 2012
Location: Sydney Australia
Device: none
|
No need to guess, I know what it shows - the path to the book folder - not the path to the book (format) file. It was that very feature which led me to hover on a format.
The truncated author folder and book folder names do not map precisely to the truncation of author and title as they appear in the format file name - for starters there's no book ids in format file names and for dessert they are reversed i.e. title - author BR Last edited by BetterRed; 10-11-2013 at 05:21 PM. |
10-11-2013, 06:16 PM | #30 | |
Well trained by Cats
Posts: 29,662
Karma: 54369090
Join Date: Aug 2009
Location: The Central Coast of California
Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A
|
Quote:
In the metadata editor, you can see the timestamp of the format (this is useful if you wish to drop an updated format version) |
|
Thread Tools | Search this Thread |
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
File names in File explorer | vvilmos | Onyx Boox | 5 | 10-05-2012 07:40 AM |
toc.ncx truncated at 100 characters | garygrosso | Conversion | 2 | 03-08-2011 01:08 PM |
Truncated file | crutledge | Sigil | 11 | 09-13-2010 03:37 PM |
HTML file added to library is truncated unexpectedly | seth123abc | Calibre | 3 | 10-31-2009 03:57 PM |
File Names | Vulcan | Sony Reader | 4 | 01-02-2009 04:03 PM |