Testing the so called Smashwords curse

[Turtle91]

Originally Posted by cromag

I've read quite a few pieces from Smashwords -- from short stories to longer novels, in both EPUB and MOBI formats. They've all looked just fine on my technologically naive Jetbook Lite. Certainly nothing I would call a "glaring" error.

Congratulations!

I would submit again, as gmw did above and ePub/eBook creators would all agree, that you may not "see" the errors if it is a "text only" type view. But those errors are still there. Whether or not those errors are noticeable on a users device/app are dependent on several factors....make/model/version of device/app to name a few...

Would you happen to have any non-DRM smashbooks that you are willing and able to legally share? I would like to open it up and see what it's like "under the hood". That is where you can determine the quality (and survivability) of your product.

Please PM me if you have one.

[cromag]

All smashwords ebooks are DRM-free. There are plenty available free, and you don't have to be registered user to download any of the free books.

[Turtle91]

Cool! I'll go take a look and see if I can find some. Thanks!

[gmw]

Originally Posted by avantman42

I'm aware of what Sigil is. I wouldn't call cleaning up a converted ePub, or creating an ePub in Sigil, "hand crafting" an ePub. To me, the term "hand crafting" implies writing the raw code in a text editor. In the same way, if I created a web page in a WYSIWYG HTML editor, I wouldn't say I'd hand crafted it.[...]

You're a hard man to impress.

Of course "hand crafted" doesn't really apply very well to computer created stuff at all. When I write software these days I use one of several different high-level programming languages. To truly "hand-craft" I would need to go back to punching cards - that would qualify, wouldn't it?

[JSWolf]

If you want to truly hand craft, all you'll have to work with are two digits. a 0 and a 1.

[Turtle91]

As cromag recommended I went to smashwords.com and downloaded a few books in the ePub format (Author's and Titles withheld to protect their sensibilities...but I will tell you in a PM if you really wish to know). I'd like to say it was random, but it was only random in the "fantasy" genre. It seems my fingers automatically go there without any conscious volition...
Here are the results of my quick...and completely unscientific...smashwords test:

Book one:
Opened the ePub in Sigil.
First thing I noticed, the cover looked good - centered on the screen and taking up the whole screen.
I then checked the code view of the cover - (this is a basic page, there shouldn't be much on this page at all) - and noticed that there were style commands in the header. That in itself is not a big deal, but it is non-standard and could cause confusion if you are trying to change the way a book looks. I checked all the other pages, and yup, they all had styles defined on the page. The bad part of this was that the actual styles used in the cover were all redundant - giving the same commands 3 times. Unnecessary/redundant commands in a document are what we call "bloat". In extreme cases I've seen bloat more than double the size of a file.
OK...some errors...but looked OK on the screen.

Next I looked through the actual text pages in "Book View".
First thing I noticed was the paragraph indent was WAY too large. "That's just style - the author wanted it that way" I thought. So I switched to code view....

....

...

I thought smashwords used the "meatgrinder" to convert their documents? Then why am I seeing all kinds of "Calibre" classes in the tags??!! For those who don't know, Calibre is a free open eBook management/conversion software. Now I am certainly not a lawyer, nor am I saying that smashwords doesn't have a right to use/license the Calibre conversion plugins...but it seems odd...

I also saw that every paragraph is defined with a particular class. This is not abnormal, but I consider it bloat. If it is a normal paragraph than it should just fall under the normal paragraph style...you don't need to tell it that it is normal.
I also saw NINE metadata tags in the header - NONE of which are required or even desired - And of course the previously mentioned style section.

OK...again...not TOO bad....fairly clean...I've seen much worse. The one thing I really want to check on is the paragraph indent.

I opened the stylesheet.

OMG!!!!

I didn't care anymore that they had put a fixed width paragraph indent of almost 1/2 an inch....1/2 an inch!! That pushes the indent to about halfway across my screen depending on what my margins are set at.

What I cared about was MUCH worse. To check what I thought would happen I threw the ePub into my calibre library (it was like the prodigal son returning home...) so I could use my OPDS server. I then opened the epub in Stanza. It is an older app but is still considered one of the premier readers...people are still waiting for a "Stanza replacement".
So the book looks fine when I open it up - as I expected. But with a flick of my finger I switched over to night mode...and guess what happened to the book??

It disappeared!

I started flicking through the pages and the ONLY thing I could see was the occasional hyperlink floating in blue in random locations on the screen.

Smashbooks had declared (or allowed the book to pass) with the fonts all colored "black". That looks fine in day mode with a light background, but when you go to a black background with black letters......not so good.

I would call book 1 a failure for that reason alone.

I did however want to be reasonably thorough so I checked it with Adobe Digital Editions and Sony reader for PC - it looked fine (except for the indent), but the PC versions do not let you change color settings. I tried to load it into Kindle Previewer but it failed import. I understand that smashwords will give you an already converted mobi format but well made ePubs should NOT fail import. However, that may be attributable to a cover image being too small or something equally ridiculous.

So, the first book is a fail.

I'll be back later to write-up the other two books...I need to go read to my kids for bedtime.

Cheers!

[JSWolf]

The meatgrinder is just some version of Calibre.

But as to night mode, that's not uncommon from Smaashwords.

[Turtle91]

Thanks Jon! I did not know that meatgrinder is just a version of Calibre. You and I both know that Calibre is updated pretty much on a weekly basis...it is updated so much that it sometimes gets annoying how often we have to install updates!! I wonder if meatgrinder keeps up with those updates, or if they are still in a previous - more buggy - version?

You also confirmed what I found out from checking these other two books. ALL THREE of the books had the same fatal flaw of mandating the font color value is black. That is simply unacceptable. The only saving grace is that there are some of the newer readers that override all of the publisher settings and display things there own way. Normally that would be a detriment, but in this case it saves these smashwords books.

I am familiar with Calibre's conversions - I've seen them many times - and had to clean all the ones I see - I don't think I need to go into too much depth about exactly what is wrong. In general, as an automatic process, it is not bad. There are many things that could/should be improved, but it is very difficult to get a program to think like a person. It just can't anticipate all of the situations it could be presented with - so it defaults to a generic value and forces everything to fit into that mold. I presume that is why smashwords is so strict on the author putting the book in a specific format before submitting it...

That is what I found with all 3 of these books...the styling was almost exactly the same...with only minor variations. For example Book 1 had an indent of almost 1/2 an inch, while Book 2 had 1/3 of an inch and Book 3 had 1/4 of an inch. Normally one would think that is just the style, but the problem is that it is a "fixed" value. You lose the benefits of reflowable ebooks when the user can't adjust paragraph spacing and indents to suite their OWN preferences. In addition, when the user changes the font size, the indent remains constant. As font size increases the apparent indent will decrease, and as font size decreases the apparent indent will become too large. This will look bad in some way to everyone.

Header Style
In Book 3 the headers/Chapter titles were almost indistinguishable from the rest of the document. There was no space around the title, it was the same size and font as the rest of the document - the only difference was that it was bold. On a darker background that would be almost impossible to see.

TOC
It was non-existant in all 3 books. There was the initial page, and then a pseudo "mid-point". Users expect to use the TOC functionality built into their reader...this is one of the benefits of eBooks/ePubs.

Massive Bloat
Honestly, I couldn't get the image of Pearl out of my mind when I looked at how much bloat was in these ePubs! (for anyone who is not part of the "in-crowd" Pearl is the super bloated vampire in the movie Blade... )
image {warning hollywood-fake-vampire-nudity} (If you can tell).


Summary:
Not as bad as I feared based on the source document being a word .doc and going through a process called the meatgrinder.
It looked OK. It was readable. But it certainly did not have that "polished" feeling that you would get when viewing a quality document.

But all 3 books had fatal flaws.
As a test, I had my 14 year old son who just started learning HTML 3 days ago take a look at the style sheets on these books. You can't get much "noobier" than that. I did not say anything to him, other than "what do you think?"

In less than 2 minutes he saw that the book would be unreadable with a dark background!

So, it certainly pays to have someone other than a computer check your books.

Oh, my son says he works for cheap! ( he is trying to pay off a $150 cellphone data charge... )

Cheers!

[VydorScope]

Just FYI: "Meatgrinder" is a collection of scripts that call other programs to do the work. They use Open Office in their tool chain too.

[Smashwords]

This meme of poor formatting is a myth. Smashwords has stricter formatting requirements than most of our retailers.

A few weeks ago we launched our new Smashwords Direct feature, which is now in beta. Ironically, many of the "hand-crafted" epubs we're receiving are failing epubcheck, or failing other validation required by Apple, which has stricter epub construction requirements than other retailers. Books produced by our Meatgrinder conversion system don't have those problems.

Also as a reminder, authors format books, not Meatgrinder. If an author follows our Style Guide, they'll get good, predictable results. Before we distribute a book to retailers, our vetting team opens each book and reviews the epub for proper formatting and function.

Yes, there are poorly formatted books on Smashwords, but they're likely not books that are earning distribution, or they're recently uploaded books where the author is still working through corrections. If a customer complains to us that a book they purchased is poorly formatted, we'll work with the author to correct the issue and if the author refuses or is unable to repair, we'll refund the purchase price.

One of the earlier commenters asked about description lengths. Two years ago (January 11, 2011), we increased the book description lengths to 4,000 characters. Authors have the option to create both short descriptions (used for Smashwords store merchandising) or long descriptions (distributed to our retail partners and also used at the Smashwords store). thanks.

[PainMike]

This thread is iron good-to-know stuff, cheers. Also more frightening than anything I have written, if looking from the side of casual author who don't care nuts and bolts behind the typewriter. I'm simple in a way that I know the nuts and bolts and writing, but quite not simultaneously. It may be a delusion but techwise it compares to preparing medieval manuscripts, if skipping the slaughter of few hundred cattle for parchments.

[avantman42]

Originally Posted by gmw

You're a hard man to impress.

Of course "hand crafted" doesn't really apply very well to computer created stuff at all. When I write software these days I use one of several different high-level programming languages. To truly "hand-craft" I would need to go back to punching cards - that would qualify, wouldn't it?

As you say, "hand crafted" is difficult to define in terms of computer-created stuff, and I don't think it's a term I'd use for anything I'd made with a computer. This is one of the reasons why I asked Jon exactly what he meant.

I've written software in a variety of languages, and I've used SCADA packages to create interfaces. I've never written anything on punch cards or tape. I wouldn't call any of my software hand crafted, but I would say it was written by hand.

Originally Posted by Turtle91

I suppose you could say that only he can explain what he means...but what I described is kinda the common understanding...

You may be right, but I'd have assumed he meant not using a WYSIWYG editor.

Originally Posted by Turtle91

If your definition of hand crafting is only "writing the raw code in a text editor." Then you can "Hand craft" an html document in Sigil...it is a text editor - albeit specialized. You go into "Code view" and just start typing on a blank sheet...codes and all. And yes, I have done exactly that with Sigil.

I've not used Sigil much, but I don't really see a difference between using code view in Sigil and using a text editor. In either case, you're writing the HTML tags by hand. That's what I'd take "hand crafting" to mean in this context.

Originally Posted by Turtle91

Even with a text editor (my favorite up until recently has been notepad++) there are built in tools that make your job of writing the code easier...if I tell it that I am writing in xhtml it will turn all the appropriate selectors/properties/values pretty colors and highlight matching tags. The editor also has certain functions that are built into it...like being able to save the file into a particular folder in a given format - open multiple files up from the same folder - rename the files, etc. Using those tools/functions in no way lessens the "hand-craftedness" of my html work. Does the fact an artist uses a saw and sandpaper discredit their work?

The same goes for Sigil. It has an added set of ePub specific functions that I use as tools for me to create an ePub.

I use Geany on Linux, which has similar code highlighting features for various languages.

I think I should clarify something here. I am not making a value judgement. I'm not criticising Jon, or saying that he should write ePubs in Notepad. I simply wanted to clarify what he meant. Earlier in the thread (post #31) he said "Now, I might consider buying your eBook if you hand crafted it and put it up and Kobo and B&N." To me, hand crafting a book implies writing the raw HTML, CSS etc in a text editor. On the other hand, I've seen Jon talk about Sigil before, so I didn't know if he meant writing raw HTML etc, or if he meant using a dedicated ePub editor like Sigil, as opposed to writing in Word then converting.

Originally Posted by Turtle91

Did you know that there are many different ways to achieve the same result when it comes to formatting an html document?? Different people will do things different ways. The act of going through a document line by line tweaking tags and THINKING about what is wanted - rather than letting a computerized algorithm make assumptions about what is intended - is the difference between "hand-crafted" and "automatic conversion".

Yes, I'm very aware of that. I've been writing HTML since 1998

Originally Posted by JSWolf

I would say that Sigil is fine for hand crafting. It's your words, your code, your format. Sigil just helps you build the shell. What goes into the shell is all yours.

Also, I would agree that hand crafting is also taking a converted ePub and fixing it so the code is neat and easy to deal with.

I just finished fixing up an ePub that now no longer longer look (internally) like it used to. All the messy code has been replaced. I would call it hand crafted given how much I changed. None of the original CSS exists.

Thanks for the clarification, and I'm sorry if you (or anyone else) thought I was implying that you should edit the HTML in Notepad (or vi, or ed, or whatever even more extreme example you can think of).

[teh603]

Originally Posted by Smashwords

This meme of poor formatting is a myth. Smashwords has stricter formatting requirements than most of our retailers.

True dat. One of my other distribution channels has an incredibly short style guide- it'd probably be less than two pages in print.

[VydorScope]

Judging by my sales at Smashwords, and judging by the test I did (covered in the OP), I standby my opinion that most readers open the Smashwords book on their Nook, Kindle, etc and you know, read it. Shocking I know! These readers do not have any problems, which is evident by their continued purchase of more books from Smashwords.

It is only a select few that will ever look at the code, and of them a small precent of them will care that the code is not as clean as they would like. That tiny group will never be happy.

[frahse]

Independence Day - the movie.

Time - Phrase

01:20:08 ...signal. I don't know how, uh... helpful I can be.

01:20:12 Why don't you show him what you've discovered so far?

01:20:16 Figure out what they haven't thought of yet.

01:20:19 We'll see if you're as smart as we all hope you are.

01:20:34 What code?