Quote:
Originally Posted by gweminence
Tell you what. Give me your longest, worst conversion file. If it takes my computer as long as you say, I'll publicly eat crow. If not...not.
|
I only keep the final epub output in my library. The other formats are sent to the bit bucket. But I'm sure someone must have an example lying around.
I recall that the original epub to epub conversion of Eldest by Christopher Paolini was one that took hours, but now that it has been cleaned up a epub to epub conversion takes only 2 minutes.
Quote:
Originally Posted by gweminence
Because here's a simple fact: conversions are a function of processing power, at their most basic level, it's a matter of how many cpu cycles it takes to make said conversion.
|
If this was so simple the epub to epub conversion that took hours the first time wouldn't take just 2 minutes the second time.
Quote:
Originally Posted by gweminence
It isn't a matter of what kind of file it is. It's a matter of what kind of muscle you bring to bear on it, and the algorithms used by calibre.
|
This is mostly true. It isn't a matter of what kind of file it is but a matter of what kind of crap is in the file. I have seen epubs that had started as Word docs. These epubs had 100+ separate html files within and each html file had 20-30 thousand lines of style crap that calibre had to sort through, flatten out and reassemble.
I may be wrong but unless some recent optimization has been done in this area each conversion is handled by one process and utilizes essentially one core at a time.
You are right file type is not a determinate factor but experience shows what is in the file is critical to how long it takes.