conversion troubles

[N13L5]

since I just got calibre, I'm doing a lot of books all at once. Some of these I've had since the 90's and they've been through various conversions as my preferred readers changed.

So I've been setting up a load to convert before I go to sleep, and overall, it works admirably well, if it can't convert something cause I've mangled it too much, it thros up an error message but continues converting the rest.

however, there are some, where it just chokes, produces no error and will just sit the rest of the night without moving on.

this happens mostly on chm or pdf files. (I made a lot of the pdf files myself, and i'll be damned if I can remember what tool I used when for those...

my DRM'd files won't convert but Calibre just raises an error and moves on without a hitch.
So while I'm sick of wasting money on DRM'ed content I end up not really owning and can't read on my phone in the end, that's not whats stalling my conversion progress.

But there's some error condition Calibre fails to handle and it'll just sit there until I come to the computer in the morning and cancel that particular job.


Is there one of the many input settings I can change to avoid that? The defaults work satisfyingly well for conversions, so I didn't really want to mess with them, unless it fixes getting stuck on things..?

Or can I get a log of the stalled conversion somehow to try to see what I might have to fix in the input file?

[itimpi]

Originally Posted by N13L5

Is there one of the many input settings I can change to avoid that? The defaults work satisfyingly well for conversions, so I didn't really want to mess with them, unless it fixes getting stuck on things..?

Without knowing what the problem is you would just be taking a shot in the dark trying to get around any issues by changing settings.

Or can I get a log of the stalled conversion somehow to try to see what I might have to fix in the input file?

If you click on the Jobs icon on the bottom right of the Calibre window to bring up the dialog showing the list of running jobs, you can then double-click any job to bring up its log.

[DoctorOhh]

Originally Posted by N13L5

however, there are some, where it just chokes, produces no error and will just sit the rest of the night without moving on.

You really have to check out the jobs detail in the lower right and look at the on going log of the job to see if it has stalled out. I have never seen it stalled out. It is either still working or eventually it gives you an error. The progress bar can be on 1% until completion when it is working hard on a difficult book.

For the record I had one epub - epub conversion last week that had been running for 5 hours when I went to sleep. When I woke up it had completed successfully (I never did look at the jobs detail to see how long it ended up taking) and the book looked fine. This is the exception since most epub - epub conversions take less then 30 seconds.

[N13L5]

ahh... thank you again, Itimpi...

I was thinking how can I get the error log when there is no error...


Double clicking on the job is just too simple to figure out;
They need to change it to something more difficult that involves using several components of the systernals suite or something

[N13L5]

Originally Posted by dwanthny

You really have to check out the jobs detail in the lower right and look at the on going log of the job to see if it has stalled out. I have never seen it stalled out. It is either still working or eventually it gives you an error. The progress bar can be on 1% until completion when it is working hard on a difficult book.

For the record I had one epub - epub conversion last week that had been running for 5 hours when I went to sleep. When I woke up it had completed successfully (I never did look at the jobs detail to how long it ended up taking) and the book looked fine. This is the exception since most epub - epub conversions take less then 30 seconds.

thanks much for this info.... when I woke up and it was sitting on a job at 1% at 380 minutes with 75 jobs undone, I had pretty much assumed it was all over, and just killed it.

what would make a book "difficult" to take so much time? Images? Most text processing is near instant

Also I've never converted epub - epub, what is the purpose of doing that?
I just try to get out of old formats that I either have no reader for or that won't reflow for my phone, i.e, pdf and chm

umm, you think I could do a pdf to pdf conversion for a roll of the dice on maybe fixing some issue?

[DoctorOhh]

Originally Posted by N13L5

thanks much for this info.... when I woke up and it was sitting on a job at 1% at 380 minutes with 75 jobs undone, I had pretty much assumed it was all over, and just killed it.

what would make a book "difficult" to take so much time? Images? Most text processing is near instant

The book I had was 2000 pages and chock full of Microsoft html junk with 10s of thousands of lines in the CSS files.

Originally Posted by N13L5

Also I've never converted epub - epub, what is the purpose of doing that?

I'm 51 and I like my books with a larger font (to avoid reading glasses), spaces between paragraphs and an indent on each paragraph. So I sometimes do a epub-epub conversion to get the book the way I am comfortable reading it.

Originally Posted by N13L5

umm, you think I could do a pdf to pdf conversion for a roll of the dice on maybe fixing some issue?

You can roll the dice on anything, PDFs are great for what they are designed for, typically fixed sized presentations. Often things don't seem to go well when converting PDFs to a reflowable format. Personally I avoid PDF files as my source if at all possible. Now I have to go wash my hands.

[N13L5]

thanks for the explanations, I didn't know you could rejigger indents and paragraphs in an automated way like that! If I wanted to change a format bad enough, I usually took a detour to word... I shall put some more work in figuring out what all the input and output options do in Calibre!

And yes, it makes perfect sense that PDF's are not a good source for conversions to formats that reflow. but for some things thats all I have, and so far, Calibre has been doing a better than expected job at it.

Occasionally, I seem to get some index or ToC weirdness, like the same index twice in a row on top of the book, I figure I can repair that manually with an html editor of some sort...
But other than that, stuff remained very readable, with code samples still perfectly placed and readable.

[N13L5]

After looking at the log of a choked chm >>> epub conversion, I can see what's making it choke / take forever.... there are so many parsing errors, CSS errors, name space problems, it doesn't end. after 3 minutes, the conversion log is longer than the book you're trying to convert :P

Must have been crazy to write the conversion code to handle so many errors and wrangle so many details that can go wrong...

[kovidgoyal]

It was a challenge, yes

[Solitaire1]

dwanthny wrote as part of a post:

Originally Posted by dwanthny

You can roll the dice on anything, PDFs are great for what they are designed for, typically fixed sized presentations. Often things don't seem to go well when converting PDFs to a reflowable format. Personally I avoid PDF files as my source if at all possible. Now I have to go wash my hands.

I agree that PDF is great as an end format, as long as they are formatted for the ereader's specific screen size. I can count on them displaying on my ereader exactly as intended at my ereader's default size. But I don't use PDFs for any other purpose, or count on doing any type of reformatting on my ereader.

[Manichean]

Originally Posted by Solitaire1

I agree that PDF is great as an end format, as long as they are formatted for the ereader's specific screen size. I can count on them displaying on my ereader exactly as intended at my ereader's default size. But I don't use PDFs for any other purpose, or count on doing any type of reformatting on my ereader.

Well, PDF used that way takes away any of the more advanced features (most) ereaders provide, like being able to select a font size. AFAIK, the format is used as a pre-print stage and really works well for that, as you can depend on seeing the exact same thing on paper as you did on screen. But ebooks? No, I don't think so.

[DoctorOhh]

Originally Posted by Manichean

Well, PDF used that way takes away any of the more advanced features (most) ereaders provide, like being able to select a font size. AFAIK, the format is used as a pre-print stage and really works well for that, as you can depend on seeing the exact same thing on paper as you did on screen. But ebooks? No, I don't think so.

Creating PDFs fixed to the reader's screen size means you have already selected the font and font size you prefer. This is the one time where the PDF format works great for a ereader. I still avoid PDFs but if you are willing to format them to fit your reader you end up with an excellent book.

[Manichean]

Originally Posted by dwanthny

Creating PDFs fixed to the reader's screen size means you have already selected the font and font size you prefer. This is the one time where the PDF format works great for a ereader. I still avoid PDFs but if you are willing to format them to fit your reader you end up with an excellent book.

This assumes that you mostly read at a fixed size and font. While that's mostly true, I sometimes find myself changing either the font or its size. And yes, it's often enough to make PDF an inconvenience. I only use PDF when I have no other options, like in scientifical texts- equations don't work with epub, as far as I know, at least Calibres conversion either chokes on them or produces garbage.

[DoctorOhh]

Originally Posted by Manichean

This assumes that you mostly read at a fixed size and font. While that's mostly true, I sometimes find myself changing either the font or its size. And yes, it's often enough to make PDF an inconvenience.

This doesn't assume anything. My point was your not Solitaire1.

I agree with you about PDF files. I guess my pet peeve is when one user patronizes another user. Solitaire1 stated that the format gives him exactly what he is looking for and that he doesn't "use PDFs for any other purpose, or count on doing any type of reformatting on my ereader."

That is a plain and simple truth. It doesn't need any restating of the obvious that was already clearly implied in the original statement, nor does the fact that he prefers it on his reader need your blessing or your scorn.

But ebooks? No, I don't think so.

My intent isn't to attack you, I'm just trying to make my position clear.

[Manichean]

Originally Posted by dwanthny

My intent isn't to attack you, I'm just trying to make my position clear.

Chill. I was just using this to make some general comments not addressed to anyone in particular.