[JSWolf]

Originally Posted by Greymage

Connect took about 600M for me at the 1400 book level... since the laptop I usually use Connect on only has 768M memory, that really slows things down.

One thing to try is to sort out your books into directories (folders) and then you won't have so many book for Connect to see at one time.

I've had Connect crash frequently when importing RTFs... but usually only on RTFs generated by one of the Amber converters, I haven't had issues with it on RTFs produced by Word. Usually I can load the offending file into Word, save it back out, and then it will import fine. So perhaps something in the RTF is not quite right, but (a) Word can handle it (b) Connect shouldn't crash even if it can't, and (c) a Connect crash shouldn't corrupt the database like it does, or at the least should be able to automatically recover.

The problem here is the Amber converter is crap. When my wife got her reader, I used it to convert something from PDF to RTF and the resulting RTF locked up her reader. The way to solve that was to load the RTF into Word and resave. So what is happening is Connect is crashing on a poor RTF made by a poor program.

I can see the advantages of LRF for some types of books, but my library is almost purely entertainment fiction, so lacking TOCs and images isn't really an issue for me, and up to now I've been batch converting my existing library, so I didn't want to spend a lot of time manually processing each book. After I catch up I might try it.

There are more advantages then you know. I have a book in RTF and the same book in LRF. The LRF is 626K and the RTF is 4,917K. At that size difference, you an fit about 7-8 times as many books on the readers internal memory. Also converting with Book Designer or the libprs500 suite is very easy overall. I can take a text file, toss it in word, clean it up in minutes and then let Book Designer have at it and it'll be done in no time.

I used a USB 2.0 card reader with my 500 since that has a USB 1.1 connection, but I still used Connect to move the content (one of the tips I picked up from this forum ). I tried loading it with the card in the 505 when I got it, but that still seemed slower than the reader for transfers. I didn't notice any difference in the long startup time.

What you didn't get is the fact that while it may have been faster to transfer the content to the SD card using your USB 2.0 card reader, it's overall slower. What happens is that Connect sorts out the books as it is transferring it so it handles all the formatting and pagination and writes out that info along with the book. And it takes longer for the reader to sort all that out then it does for Connect to do it when it transfers. That's why it took me less then 2 minutes for my reader to sort out about 200 books because I used Connect to send the content to the SD card in the reader.

I don't think blaming the software for crashing on an import and corrupting the database is unfair. Also I think the memory utilization is excessive - 600M for 1600 books when reading off the SD card? Considering it only takes up 40M when it first comes up, when it has the identical library in its database, I think it has issues.

It crashed due to a poorly written RTF from one of the ABC Amber converters. The RTF output and PDF output are both really poor. My guess is that it's taking up so much memory as it's reading the info from the book and storing it in memory since it doesn't have it elsewhere. 14090 books worth of info can take a LOT of memory. But if you do as I suggested, I think it'll work a lot smoother for you.

At 2 minutes for 313 books, that comes out 10 minutes for 1600, which is about the number I have loaded, but I get the impression the actual time is closer to 20-30 minutes. I wonder if lrf books load faster? If so that would be a good reason for me to make the effort to switch everything over. I'll try that someday.

I don't want to give the impression I hate the Sony Reader, I love it, warts and all (and very few warts on the hardware itself). I love the 505 even more than the 500. Part of it is that I'm a programmer, so I'm offended by shoddy workmanship in a shipped product... you can't get rid of every bug, but I just get the impression the Connect software was done by a bunch of junior programmers.[/QUOTE]
Connect is not perfect, but if you know what you are doing and some of the gotchas, it works quite well. I cannot speak yet for the reliability of Ebook Library yet as I've only used it twice so far. It wasn't 2 minutes to read 313 books, it was actually less time. I just didn't get the actual exact time. That why I used less then 2 minutes.