I've also recently self published my book http://www.buildyourownlisp.com/ via createspace, having first converted it to an ebook and then to a print book, so I was really interested to see this article to read what was the same and what was different. I did speak to a couple of keen publishers, and I guess ended up self publishing for similar reasons. I wanted more control over the process and the final layout, and some publishers were not happy with having a free version provided online.
Creating the ebook was fairly easy. It ended up being a case of writing a bunch of scripts to fix and tweak various things about the HTML website. Constructing the print book was a lot harder work.
I explored rendering a PDF from the ebook but the font rendering was ulitmately too poor and the configuration of the ebook tools was getting really difficult. Additionally the text layout was not great with lots of code blocks getting split over pages and images in wrong places.
I knew I needed to use a program more suitable for publishing so I tried doing a manual conversion to Latex, but after the first couple of chapters I realized it was going to take too long. Instead I begun on some scripts converting my HTML into Latex automatically. Most of this work was done by `pandoc` and a list of regexes that redefines "unholy". God only knows how it all worked in the end. I found various Latex solutions for the pullouts, and syntax highlighting, that I liked ang which gave me as much control as I required, and in the end it was actually looking pretty nice.
Then came the copy editing, and like in this article it was long, tedious, difficult, and I'm still not completely 100% happy with the final result. It took the form of a bunch of scripts inserting pagebreaks before or after paragraphs and moving text, images and pullouts to balance the pages. Essentially it is very difficult to copy-edit a book with so many code blocks, even with the help of Latex. This is the only part of the process where I felt down on my skills, and that someone who does it for a living would not face the same problems (probably they wouldn't be using Latex in the first place). But in the end I was happy with the interiour - which still looked awesome printed in color - and I think I by far did the best I could.
I designed and made the cover in Photoshop to the vision of it I had always had. I think I did a fairly decently job but certainly it would have been better looked at by a professional. Unfortunately createspace doesn't seem to do a good job of printing the cover, so it didn't look precisely like the supplied image I made. Perhaps this is different when ordering non-proof copies.
The nice thing about my approach was that when people send pull requests and corrections to the website I can automatically integrate them across the ebook and print versions, and if I ever do want to make updates in future the process is relatively painless. I think I'd be happy to self publish again. Createspace was great and I felt good not having to rely on others for the process. The main thing I didn't like about it was the marketing - that is something that does not come naturally to me. I think it is unlikely I will write a programming book again. It was fun, but I felt like finishing it more of an obligation once the ball was rolling. Ultimately I think I made a "good" thing which lots of people read, enjoyed and appreciated, and I know it wasn't just an additional item of waste in the universe due to all the people who have reached out to me personally (thank you so much). But at the end of the day unfortunately it hasn't derived me satisfaction worth the effect it took to produce - which was a lot - and I think that is just due to the nature of the thing, not it's success or position in the world.
Oh, I just saw your post about this a couple of days ago. Your book looks beautiful! I was so tempted to do full color, but didn't think the finances made sense for my book. I'm totally jealous you went that route. :)
How did it end up working out? I assume it's process, not spot color. Is the colored text still sharp enough to be read easily?
> Instead I begun on some scripts converting my HTML into Latex automatically. Most of this work was done by `pandoc` and a list of regexes that redefines "unholy". God only knows how it all worked in the end. I found various Latex solutions for the pullouts, and syntax highlighting.
Smart! I considered this route a little bit, but I spent some of my formative years doing "desktop publishing" back when Aldus PageMaker was still a thing, so I looked forward to doing it a more hands-on, designery way. It worked out well for me -- the time I spent in InDesign was really enjoyable -- but I can definitely see how it wouldn't be the right choice for others.
> Essentially it is very difficult to copy-edit a book with so many code blocks, even with the help of Latex. This is the only part of the process where I felt down on my skills, and that someone who does it for a living would not face the same problems (probably they wouldn't be using Latex in the first place).
It's a relief to hear someone else say this. Trying to keep the code blocks from being split was really really hard. Halfway through the process, I started flipping through my copy of SICP to see how they did it. As far as I can tell, though guys are just absolute wizards. Every page is completely full and yet code blocks are almost never split across pages. Granted, Scheme snippets tend to be fewer lines of code, but they still did an amazing job.
> Unfortunately createspace doesn't seem to do a good job of printing the cover, so it didn't look precisely like the supplied image I made.
I've done some print work before, so I was fully expecting some variance here. Unless you have a carefully calibrated display and keep track of color profiles through the whole pipeline (and they do too), there's always some difference.
In my case, it came out surprisingly close to the image I sent, but I tried to keep the design pretty simple and not use too much detail or color.
> The nice thing about my approach was that when people send pull requests and corrections to the website I can automatically integrate them across the ebook and print versions, and if I ever do want to make updates in future the process is relatively painless.
Yeah, this part will be rough for me. The print version is basically a fork at this point. Any changes I make, I have to manually make in both the markdown (eBook + web) and InDesign (print).
> But at the end of the day unfortunately it hasn't derived me satisfaction worth the effect it took to produce - which was a lot - and I think that is just due to the nature of the thing, not it's success or position in the world.
For what it's worth, I'm really glad you made it, and I'll be buying a copy.
Thanks for the support, and congratuations to yourself on publishing. It looks like you've done a really great job and totally nailed the process and the print version!
> How did it end up working out? I assume it's process, not spot color. Is the colored text still sharp enough to be read easily?
Yeah it isn't spot color and the text looks really great and easy to read. The print quality of the interiour seems very high quality and it really makes it seem alive with the images and colored sections. I think it keeps it fun to read, which was always one of my intentions. I was totally over the moon with the interiour when the first proof copy arrived.
But as you mention financially it isn't going to make much sense. I don't make much more off the print book than the ebook sale, and yet it costs the consumer ten times as much! So in some sense it was kind of vanity project - but I still enjoyed it a lot, and wanted to prove I could do it additionally.
> It worked out well for me -- the time I spent in InDesign was really enjoyable -- but I can definitely see how it wouldn't be the right choice for others.
Actually if someone had mentioned InDesign to me I might have given it a try. I mainly picked Latex because I knew the font rendering and formatting would be awesome and it was something I knew already so I could get started. As you say, InDesign is fun, while working on a complex project in Latex isn't exactly a joy. But I think the end product was still sufficiently professional :)
> I've done some print work before, so I was fully expecting some variance here. Unless you have a carefully calibrated display and keep track of color profiles through the whole pipeline (and they do too), there's always some difference.
Yeah I probably should have been prepared for this. And additionally there was some error in how they were cutting the bleed which looked poor with my cover design. If I did it again I would definintely try to find a professional to consult before tacking the cover for any friendly advice he/she had to offer.
> For what it's worth, I'm really glad you made it, and I'll be buying a copy.
Thanks! I feel bad ending on a downer because I really have gotten a lot of pleasure out of it, so I don't want to paint the process in a negative light. As a learning experience it was really unique and great. There are just too many fun, great and unique areas of computer science to play around with as well!
Creating the ebook was fairly easy. It ended up being a case of writing a bunch of scripts to fix and tweak various things about the HTML website. Constructing the print book was a lot harder work.
I explored rendering a PDF from the ebook but the font rendering was ulitmately too poor and the configuration of the ebook tools was getting really difficult. Additionally the text layout was not great with lots of code blocks getting split over pages and images in wrong places.
I knew I needed to use a program more suitable for publishing so I tried doing a manual conversion to Latex, but after the first couple of chapters I realized it was going to take too long. Instead I begun on some scripts converting my HTML into Latex automatically. Most of this work was done by `pandoc` and a list of regexes that redefines "unholy". God only knows how it all worked in the end. I found various Latex solutions for the pullouts, and syntax highlighting, that I liked ang which gave me as much control as I required, and in the end it was actually looking pretty nice.
Then came the copy editing, and like in this article it was long, tedious, difficult, and I'm still not completely 100% happy with the final result. It took the form of a bunch of scripts inserting pagebreaks before or after paragraphs and moving text, images and pullouts to balance the pages. Essentially it is very difficult to copy-edit a book with so many code blocks, even with the help of Latex. This is the only part of the process where I felt down on my skills, and that someone who does it for a living would not face the same problems (probably they wouldn't be using Latex in the first place). But in the end I was happy with the interiour - which still looked awesome printed in color - and I think I by far did the best I could.
I designed and made the cover in Photoshop to the vision of it I had always had. I think I did a fairly decently job but certainly it would have been better looked at by a professional. Unfortunately createspace doesn't seem to do a good job of printing the cover, so it didn't look precisely like the supplied image I made. Perhaps this is different when ordering non-proof copies.
The nice thing about my approach was that when people send pull requests and corrections to the website I can automatically integrate them across the ebook and print versions, and if I ever do want to make updates in future the process is relatively painless. I think I'd be happy to self publish again. Createspace was great and I felt good not having to rely on others for the process. The main thing I didn't like about it was the marketing - that is something that does not come naturally to me. I think it is unlikely I will write a programming book again. It was fun, but I felt like finishing it more of an obligation once the ball was rolling. Ultimately I think I made a "good" thing which lots of people read, enjoyed and appreciated, and I know it wasn't just an additional item of waste in the universe due to all the people who have reached out to me personally (thank you so much). But at the end of the day unfortunately it hasn't derived me satisfaction worth the effect it took to produce - which was a lot - and I think that is just due to the nature of the thing, not it's success or position in the world.