Smart quotes fail on Amazon look-inside

I’m glad you have things “solved” - sort of.

Regarding the HTML cover page option, do you mean you ticked that in Sigil? Because there is an option for that in Scrivener that should do the job.

At least now you know that IngramSpark seem to be using the same converter as Kindle Previewer, you can test the results. For future uploads, I recommend opening the Epub file in Kindle Previewer, exporting as a .mobi file from there, and then loading the file onto your Kindle (Into the “documents” folder of the device via the Finder). This should show you how the file is going to look once IngramSpark has uploaded it, and so should allow you to fix any issues (or come here and ask) before submitting.

Congratulations on getting hot list, by the way!

All the best,

Yep, sorry - the Scrivener option for HTML cover page … one day I’ll write proper England like what I was learned at schools.

Nearly good news. Oh, wait! It’s still screwed. But I’m hoping that’s just waiting for Amazon to deploy the new version.

So, Ingram has uploaded a new version and sent me a copy for my reference. For Keith’s benefit, it seems to have a few differences to the one you generated - their first line looks a little different that they use a <U+FEFF> tag at the beginning of the text (not the UTF encoding string you suggest and literally as the first few characters in the file), then replace all the quotes with their escaped equivalent - ‘ and so on … Happy to share the ebook format they provided (assuming it works) - but not via the forum.

I have no idea how long it takes for Amazon to propagate the change, so I’m deleting the book and reloading every few minutes on my reader. As I write, I’m still seeing the ASCII rendered sequences, but Ingram UK has assured me that they have checked the uploaded version and it’s working for them … Another evening, another lost sales opportunity on the social media campaign! Ugh!

If it’s still showing the escaped set tomorrow, I’ll risk upsetting everyone by simply uploading my Sigil version onto their portal. I have a copy of their file format, so I can always re-upload that if the Sigil tweak doesn’t fix things. Here’s hoping people are pre-ordering the paper copy and generally avoiding kindle … But I fear that’s not the case!

Also, as a side note - Ingram say they can’t trigger a refresh for the look-inside and keep asking me to contact Amazon. I dutifully contact Amazon, but as I don’t have a publishing account - nor have I published via them, I’m at a loss why Ingram insist on me asking - Amazon has, as expected, pushed me back to Ingram. I’ll break forum etiquette here and add another topic - any thoughts on this?

Concerning your last question (I can’t believe I’m saying this), I have to side with Amazon. If you published through Ingram Spark, and Ingram Spark is who Amazon has the relationship with (meaning the contract is between those two parties and any funds being exchanged for services rendered are between them), then it is ultimately Ingram Spark that is liable. They sold you a service that included making your work available on Amazon. Unless you paid Amazon directly, they don’t have to work with you. Would be a great gesture and good PR if they did, but there might also be legal issues they’re concerned with (they might not want to get between a writer and the publishing service the writer has a contract with).

Ingram Spark promised you a service. When you receive payment for an advertised service, it is legally assumed you know how to provide the service. If Amazon changed the format required, it would be up to Ingram Spark to learn how to submit it properly or advise consumers that they cannot aid in distribution through Amazon.

If the only reason they are asking you to contact Amazon is to put additional pressure on them, that’s one thing (the whole “squeaky wheel gets the grease” idea), but ultimately the responsibility legally falls on Ingram Spark. I have no clue what your revenue being generated is (and PLEASE DO NOT SAY, you just said you made it to #2 on the Hot List, so I’m guessing you have significant sales), but if this ends up seriously hurting your revenue due to refunds, then Ingram Spark could be held liable for that. It’s just like in a company I used to work for. We made faulty packaging, the food spoiled, customers got refunded, and the food company sought reimbursement from us for failing to provide the product we promised and it cost them money. It’s the same principal. In this case, you’re in the same situation as the food company, and Ingram Spark is the packaging company,

Likewise, if they have submitted everything to Amazon and Amazon is just refusing to refresh it, they in turn could seek damages against Amazon for what they had to pay out to you.

But all of that is assuming that there is enough money to get lawyers involved. Otherwise, the only thing you can do is hound both of them until they fix it and then basically raise awareness to others - if you make a big enough stink, maybe they’ll agree to give you a discount on your next book to minimize damage to their reputation.

You can email it to us for my attention at AT However, it sounds like they are going about things the wrong way - or at least a roundabout route. What they are doing will almost certainly work, but it’s not necessary for HTML5. Using ‘ instead of an opening single quote was necessary in HTML4, but it is not necessary in HTML5, which uses the UTF-8 encoding by default. So it seems that they are making unnecessary work for themselves there. They should either be converting to .mobi format using kindlegen (which avoids all of these errors) or forcing the UTF-8 encoding using the line mentioned up-thread.

All the best,

I haven’t ignored you folks, I just haven’t had an update … After giving a couple of days for Ingram’s HTML escaped version to propagate through the various systems, nothing changed. So I uploaded a version of the text per our Sigil/discussion above - that went through to their platform last Thursday. Look-inside now renders beautifully, but the main text hasn’t updated - I know this because I modified the capitalisation on the opening three words … They’re correctly displayed on the preview, per the original version on the file itself.

So …, it looks like Ingram’s file update process has broken somewhere. Ugh! Happy Monday!

That’s good news that the Sigil edit worked, at least, since that confirms that 3.0.3 should fix the issue on Scrivener’s side.

All the best,

Yep, just keep us posted. I’m sorry you have had to go through this, but I’m glad you started this thread. Even though Scrivener could export to .mobi and .epub, I thought about having the “professionals” do it anyway. But apparently they are sometimes flying blind, so I’ll just take the responsibility and save myself the money.

Finally! about 10 minutes ago, we got this issue resolved. So - the Sigil method to insert the UTF-8 tag in the header is definitely the key. With that in place, I managed to get files refreshed and Amazon previews and the main book all render correctly.

For anyone publishing through the likes of IngramSpark et al, it would appear that the end-to-end process to get a resolution is basically firing emails between (in my case) UK front-desk->US front-desk->2nd Tier Support US->2nd Tier Support UK->Amazon US->Amazon UK and then all the way back again. Turn-around per cycle, around 2 days.

Anyway, that aside - it transpires that once uploaded to Amazon, the original purchasers of my book could only pull back the copy they had bought. It requires the purchaser to contact Amazon support to push a new version of the file back out. This despite settings in the account saying ‘always sync to the newest version’ etc.

Bottom line here - to prevent content rendering incorrectly on Amazon - run it through Sigil (probably not needed after 3.0.3 comes out, by the sound of it) save it, and then upload. If you miss this step, you can refresh your content, but some [all?] who bought the original content will see the original malformed document. It’s only taken 4 weeks and cost me my launch campaign - but hey, it’s finally out there in a sensible format.

Thanks to all on this thread for helping troubleshoot. Being able to cite earlier entries proved invaluable.

Quick aside for Keith - I gave you a quick plug on my last blog post - thanks for a great product and for your help in getting to the bottom of this lot.


I’m glad you have it sorted at last! And thanks for confirming what fixed it, as that does mean that the fix 3.0.3 has in place should have this sorted, which is great - this topic has been very useful.

Thanks for the kind plug in your blog post!

All the best,

Congrats!! So glad everything is how it should be.