Calibre Too CPU Intensive

[freddyzdead]

The trouble is that Calibre is written in Python, possibly the least-efficient "programming "language there is. On my Windows 7 Toshiba laptop, Calibre is using almost 500 MB of RAM. There are 4 Calibre-parallels running and 4 Conhosts as well. This is the biggest piece of bloatware I've ever seen. And all it is doing is loading a few thousand books. It has been going for over an hour now and is just halfway through. So how many days is it going to take it to convert these PDFs into ePubs? Python is for script-kiddies, not serious programming.

Moderator Notice
You posted to a thread over 2 years old. Moved new posts to new thread. Do not resurrect ancient threads.

[PeterT]

Feel free to rewrite it

[GoSharks]

Originally Posted by freddyzdead

The trouble is that Calibre is written in Python, possibly the least-efficient "programming "language there is. On my Windows 7 Toshiba laptop, Calibre is using almost 500 MB of RAM. There are 4 Calibre-parallels running and 4 Conhosts as well. This is the biggest piece of bloatware I've ever seen. And all it is doing is loading a few thousand books. It has been going for over an hour now and is just halfway through. So how many days is it going to take it to convert these PDFs into ePubs? Python is for script-kiddies, not serious programming.

Ummm....so don't use it. By the way, reviving a 2 1/2 year old thread just to troll doesn't exactly raise your credibility. Welcome to the ignore list.

[cybmole]

Originally Posted by PeterT

Feel free to rewrite it

or buy the epub versions

[theducks]

Originally Posted by freddyzdead

The trouble is that Calibre is written in Python, possibly the least-efficient "programming "language there is. On my Windows 7 Toshiba laptop, Calibre is using almost 500 MB of RAM. There are 4 Calibre-parallels running and 4 Conhosts as well. This is the biggest piece of bloatware I've ever seen. And all it is doing is loading a few thousand books. It has been going for over an hour now and is just halfway through. So how many days is it going to take it to convert these PDFs into ePubs? Python is for script-kiddies, not serious programming.

So! 500M? that is 1/6 of the installed RAM that my W7 Toshiba came with.
W7 itself, uses more than 500M

You can READ all those book in A DAY? (IMHO the only reason you would need to convert them all at once. in which case it would have been better to by a device that supported viewing PDF's.)

[HarryT]

Why are you using such an ancient version of Calibre?

[cybmole]

why do folks bother to complain, when the time needed to type in a complaint is greater than the time need to just wait for the conversion to finish.

the conversion will run just as quick, or slow, if you go away & do something useful - as in get-a-life. you don't HAVE to sit & watch it

PS ripping / converting a DVD takes FAR longer than converting a book but I don't see folks moaning about the time needed in their software forums

[DoctorOhh]

Originally Posted by freddyzdead

It has been going for over an hour now and is just halfway through. So how many days is it going to take it to convert these PDFs into ePubs?

Do the math, an hour... half way through.... hmm... it could take anywhere from two hours to ?? days. Wait until you see the quality of the PDF conversions, they will range from OK to horrible because PDF is a page oriented format and even Adobe has a heck of a time converting PDFs to a reflowable ebook format.

Originally Posted by freddyzdead

Python is for script-kiddies, not serious programming.

This statement just shows how your skills and/or imagination are truly limited. The script-kiddie behind this program has managed to make a living off of this program for a number of years and there are currently over 12 million users of calibre. Any way you slice it that is serious programming.

If you ever manage to get calibre working for you...

Good Reading!

[freddyzdead]

It's obvious to me that the trouble is that Calibre is written in Python, which is hard to believe, but there it is. Calibre has been converting a bunch of books for a week now. CPU has been stuck on 100% the whole time, it is currently using about 1.2 gigabytes of RAM and the machine is barely able to do web browsing never mind anything serious. Some idiot accused me of trolling, nothing of the sort, I just cannot understand how anybody could have used Python for this, it was never intended for this sort of thing. Apart from the resource-hogging and speed issues, Calibre has destroyed the original directory structure and the books are now sorted by title not author/title.

I would like to know if there is another converter for PDF to ePub that doesn't have these issues?

[DoctorOhh]

Originally Posted by freddyzdead

It's obvious to me that the trouble is that Calibre is written in Python

Moderator Notice
Moved post to existing thread. Stop necro-posting to threads that are 2 years old.

[JSWolf]

Originally Posted by freddyzdead

I would like to know if there is another converter for PDF to ePub that doesn't have these issues?

Yes, there are other programs that will convert PDF to ePub. But NONE WILL DO IT WITHOUT ERRORS. So why are you bothering? Do do it right, you will then need to A/B compare the PDF to the ePub and it's not worth it. Are these legally obtained PDF?

[geormes]

Guess you should ask for a refund!

[DoctorOhh]

Originally Posted by freddyzdead

It's obvious to me that the trouble is that Calibre is written in Python, which is hard to believe, but there it is.

There are limits to everything, but Python works fine for this application.

Originally Posted by freddyzdead

Calibre has been converting a bunch of books for a week now.

Mass converting PDF files is a foolish waste of time. To have a chance at decent PDF conversions you need to review the converted file and adjust conversions settings as needed and do as JSWolf stated an A/B compare between the original and the conversion to make sure it survived as a readable document.

Bottom line PDF conversions are hit and miss because for so many years PDF was strictly a page layout format limited/confined by the size of the page. It was never previously meant to be able to be reflowable to fit different page sizes.

Originally Posted by freddyzdead

CPU has been stuck on 100% the whole time, it is currently using about 1.2 gigabytes of RAM and the machine is barely able to do web browsing never mind anything serious.

If you allow calibre full use of your machine it will use it. If you had decided to limit cores (Preferences - Miscellaneous) or job priority (Preferences - Behavior - Job priority -> Low) you could have been using your machine with ease this entire time.

Originally Posted by freddyzdead

Some idiot accused me of trolling, nothing of the sort

When you twice resurrect 2 year old threads for the purpose of whining before you even investigate the features built into the program, by definition you are trolling.

Originally Posted by freddyzdead

I just cannot understand how anybody could have used Python for this, it was never intended for this sort of thing.

Nothings perfect, but given calibre's success I would say that Python is well suited for this sort of application. The plugin architecture has allowed other developers to expand calibre's capability beyond its core function and the template programming language allows users to do all sort of customization of their library.

Originally Posted by freddyzdead

Apart from the resource-hogging and speed issues, Calibre has destroyed the original directory structure and the books are now sorted by title not author/title.

Trolling much? The original directory structure has NOT been altered in the slightest. When books are added to calibre the books are copied to calibre's database so your original files and folders are not touched.

Originally Posted by freddyzdead

I would like to know if there is another converter for PDF to ePub that doesn't have these issues?

No program converts PDF well. But you can use Adobe Acrobat to save books as html (even this is far from perfect) then use other programs to convert the html to ePub. Calibre is great at quickly converting html to ePub.

Good Reading.

[freddyzdead]

Well, thank you for your patient response and I apologize for bumbling into your forum. I'm using Calibre 0.9.5 btw. Not old. I can't help noticing that I'm not the only one having clogging issues. My Toshiba is not top-of-the-line or anything, but it is dual-core and has 2 GB RAM and I've never had any speed problems. Until Calibre. What I didn't realize before is that when I first noticed the trouble, all it was doing was LOADING the books, not even converting or anything. That's what took 3+ hours. The converting is what's taken days on end.

Up until now, I've regarded Microsoft Office as the pinnacle of inefficient programming. But the only thing that comes close to Calibre is Opera when I've got about 50 tabs open. I try not to do that as it uses huge amounts of memory.

But it's obvious that Python is the problem here; it has an enormous run-time library, that no one ever mentions. It has to execute masses of code just to do even the simplest of tasks. I am old enough to remember when real programs were written in Assembler and a program that actually did useful stuff was less than 64 kilobytes. And that was when programs had to provide their own drivers for video, audio, etc. The trouble started with Basic, which allowed almost anybody to write "programs" just by learning a scripting language. Things only got worse from then on. Now the landscape is littered with compilers and top-heavy runtime modules. Maybe it's Dennis Ritchie's fault, because most of them look a lot like "C". So now we need multi-core processors and gigabytes of RAM to get the same performance we used to get from an 8086 with 640KB. Well, almost.

What I mean is, we've just about lost contact with our roots; even Microsoft has stopped worrying overmuch about backwards compatibility. 64-bit Windows can no longer run 16-bit programs. So what? you might say, but it's important.

Maybe the shift toward mobile computing will slow down the rot.

Actually, Calibre just now finished the conversion and is now idling, for which it is using a mere 90 MB. Good grief.

[DoctorOhh]

Originally Posted by freddyzdead

Well, thank you for your patient response and I apologize for bumbling into your forum. I'm using Calibre 0.9.5 btw. Not old.

The comment by HarryT about using an old version was because you resurrected a thread that stated the OP was using calibre version 0.6.24. When this pops up new, folks ask why such an old version. Yet another reason for not resurrecting 2 year old threads. When in doubt start a new thread.