I have a large technical reference library, in paper format. Some of it is up to date, some I use a lot every day, some is dated, some is obsolete. But I hang on to it, just in case.
If I get the opportunity, I get electronic versions of the books, just in case. From many sources.
O'Reilly Books emailed me with a special deal. Two for one of their ebooks, and you get all four of their formats, with free updates. They have the ebooks in PDF, EPUB, Mobi, and APK for Android. This was a good deal, especially as I've been considering buying a Kindle or a Nook. It's interesting to see where the ebook market is going.
I read books in each of them. Not technical books though, they wouldn't work very well on such a tiny screen as the iPod Touch. Kindle, B&N, Classics, eReader, iBooks, - I buy them in whatever format they are in, they get downloaded and I read them. I look for cheap deals, free books, and sometimes just for the hell of it, I'll buy ebooks at regular price and read them.
But I have a lot of books that I have collected in various formats over the years. Mostly in text format, some HTML, and reluctantly some in PDF or RTF. Most of these, I can convert to one format or another and load into Stanza and read them there. I've been doing this for a while.
But not the technical books. I can't read them on the small devices. So I get them in PDF and read them on the computer using Acrobat Reader (and watch it spin into oblivion on Linux and consume 100% of the CPU), or I get them in CHM format and use kchmviewer on Linux and get a good reading experience. It works because I can view them in large format. I don't know if reading technical books on an ereader device would work. I will find out one day.
CHM is a difficult format to work with. It's Microsoft's Compiled HTML that they use for their help system, and a lot of publications have produced books in CHM format. I can read them with kchmviewer, but can't do much else with them.
I did experiment with conversion. I tried some Windows conversion tools that would convert CHM to PDf, but they all wanted money, and I was reluctant to spend money when the trial versions did such a poor job of creating PDFs. On the Linux side, I found chm2pdf, which is a set of Python scripts with some heavy dependencies to convert CHM to PDF. This worked really nicely on some CHM files, but didn't work at all on others. I had to do a lot of reading and found that you need to do two passes, one to decompile and extract the HTML, then you edit the files and fix incompatibilities, and then the second pass builds the PDF. THis might work for one or two CHM files, but I have hudnreds. I didn't go further down that path.
I did keep looking for an ereader. It's between a Kindle and a Nook. I'll make a decision soon.