Best workflow for lots of figures + Sente?

Scrivener Scrivener for macOS
1

The previous entry has made me question whether I really understand Scrivener, and I need to understand it before I commit a large project to the software. (And I am really worried that I won’t be able to update to 2.0!)

My intention is to write a book with 300–500 illustrations. I need to develop the illustrations as I work on it, but probably some or all will be passed to a real graphic artist later for fine tuning. So I work in Photoshop, Illustrator, or other to produce a document. I then save a copy of this image as a .jpg to import as a place-holder in Scrivener. This way I can produce a draft which can be reviewed, with all illustrations. When final production comes, years from now, they will be dealing with TIFFs of course, and redrawing some or all.

So I just experimented with “Compile Draft…” I find that

If I control-drag an image from the inspector to a document, the image is not incorporated in the Compiled Draft.
If I compile to an .rtf images are incorporated but are invisible if I open the .rtf with Bean, TextEdit or Pages '08. I can see the images if I open the .rtf with NeoOffice. (I can tell that the images are there due to the size of the rtf document).
If I compile to an .rtfd document, the images are visible in Bean, TextEdit or Pages, but NeoOffice can’t open it. Also, I can’t share an .rtfd with any colleagues who are not on Macintosh.
I want to run the resulting document through Sente after I am done. My previous experience with Sente was that it would change rtf documents while processing references in them, and I had better luck using Pages format.

So it seems that my workflow has to be

  1. Drag images as needed from the Inspector bar into the document. Update them as necessary by hand.
  2. Export as .rtfd
  3. Open the .rtfd in Pages and save as .pages
  4. Scan the .pages in Sente. This will replace references and produce the bibliography
  5. Open the post-Sente .pages document and save as something that a non-Mac colleague can open

Does anyone (Amber especially :wink:) see any problems with this workflow? And how will I be able to upgrade easily to Scrivener 2.0?

2

The only real problem I see with your checklist is procedural, and that is down to preference. Personally, if I were working on the project you describe with the formats you are using, I would just forget trying to keep images updated and use intentional placeholders—something close enough so the graphic designer and page layout person (or yourself if that be the case) can work with them. Trying to keep it all up to date all of the time seems to be adding a lot of micromanagement to something that will essentially have to be done anyway at the end once the final high-resolution illustrations are prepared. So unless you have the type of personality that really needs things to look as they will (roughly), I would revise step one to just dropping in placeholders and forgetting about it until later. Naturally, you can keep references in the pane to all of the actual, updated pictures so if you do need to reference an illustration while writing, you can still get at the media and ignore the placeholder.

I guess the key thing here is that Scrivener is a drafting tool, not a book production tool. In some simple cases (and one very advanced exception, MMD->LaTeX), the two can be combined, but when you start talking about images, footnotes, bibliographies, page layout and high-resolution finals—that’s way out of scope for what Scrivener is meant to be, a high-powered manuscript drafting tool.

Given the outcome of the other thread, I’d really avoid using aliases since it is a bug. Not only will compile fail, as you noted, but it will make your project get stuck to a particular version of the software. If this project will indeed take a few years, that’s a long time to be using what will by then, an ancient version of the program.

Someone might suggest MultiMarkdown to you because it does image referencing, but I’d hesitate to recommend that workflow for two reasons:

  1. It seems you definitely want images in the document, and with MMD you have coded references to those images in the text. They’ll always stay up to date with your drive, but visually it’s just a text marker to an image. Visually that might not be what you want.
  2. At the moment its RTF workflow is weak and will probably not produce what you want anyway. It’s really more suited to LaTeX and XHTML workflows.

As for the rest of the steps, I’d not worry about it too much until you get down to that point. Yes, definitely develop a workflow and make certain it works, but a year or two can change the landscape of what is possible by quite a bit. Right now Pages doesn’t support RTF very well, which is why you are having to use RTFD, neither does the OS X system in general (which is why you don’t see properly embedded images), that might change in two years. Maybe in two years Sente will work better with Nissus, or RTF in general—you get what I’m saying—get something that works, so you can be confident, but then don’t worry about the details too much. Things generally tend to improve with time rather than regress. Chances are if you have a working workflow now, it’ll either be identical or improved after a few years.

3

Having images updated in the text automatically is certainly problematic, and there is no real provision for that. It’s not really a bug so much as a limitation with the text system Scrivener uses; you would find the same issue in many programs, excepting those dedicated to page layout. And that’s probably the problem - Scrivener isn’t really a layout program, so its support for images is rather basic.

It should be if you exported to RTFD format - the RTFD format is the only one that supports image aliases.

This is a limitation with TextEdit, Bean and Pages (although I thought Bean had worked around this). TextEdit and Pages have only very basic RTF support, and do not support images in RTF - they strip them out. If you open the files in Nisus or Word, you will see the images.

Yes, RTFD is an Apple-only format, and only supported by programs that use the OS X text system. It’s not a good format for sharing.

(I can’t comment on Sente as I still need to do more research into bibliography programs; others haven’t reported any problems with Sente and RTF though.)

This should work fine, although bear in mind that you won’t be able to use footnotes if you are going this route - but then there is no route that gets footnotes across to Pages properly anyway.

As for 2.0, that would be difficult given that 2.0’s internal format is changing to RTF for cross-platform compatibility purposes. I’m not sure there’s a good solution at the moment, but I’ll think about it.

Best,
Keith

P.S. I see AmberV beat me to it. :slight_smile:

4

If you play around with it, I think you’ll find that aliases are not exported. I have always found that “should” is not very helpful with conversing with a computer.

Thanks for all the other advice, you and Amber make a great team

5

Actually, if you play around with it some more, I think you’ll find that they are. :slight_smile: I just tested it - try opening such an RTFD file in TextEdit. The trouble is that Pages strips the alias. Which means that even if Scrivener could maintain and export file links, export would be a problem, unless it converts them to real images on export - but then you wouldn’t be able to change the images in the exported file automatically.

I’ve been looking into the RTF syntax’s support for linked images, but don’t seem to be able to find any examples that actually work.

Best,
Keith

6

I may be slow, but it seems to me that you want to use HTML (or some other ML variant) in a WSIWYG editor. Think IMG tags.

That said I don’t think you could pay me to try to write anything longer than a couple of paragraphs in an HTML editor. I think Amber’s idea with a little perl or MMD is your best alternative. Write your “stuff” and toss in the needed image tag or two then run a post export filter to build the real HTML for viewing.

Not sure you want to go that way, but that is how I would tackle it.

7

If you can get a working process with an HTML file then that is not a bad option, so long as you don’t need footnotes or endnotes at the pre-Sente stage. A well-formed HTML file to NeoOffice is actually a pretty good method. Headers will be actual style headers and so on. But, to get real headers and image links in Scrivener turned into a well-formed HTML file, you need MultiMarkdown. It’s easy to learn, but you should definitely test it a bit to make sure it will do what you need, and that you can live with bold instead bold, and image name instead of a graphic in the page. Going that route, your method would be: Scrivener+MMD -> MMD/XHTML -> NeoOffice (rtf/doc/whatever) -> Sente -> colleagues.

And yeah, that is how I would do it too, but it’s definitely the “geeky” option, though not as geeky as using Perl to update your RTFDs. :wink:

8

I have been trying for the better part of a decade to not have to learn perl, and I really don’t want to start now. I used to love sed and awk, but that was a long time ago…

Thanks, I’ll look at MMD.

9

If you, at any point in your life, ever loved sed or awk, you’ll be just fine with conceptualising MMD. Especially so with the HTML workflow which is practically “quirk” free and with Scrivener, point-and-click.

Note if you do go down this route, there is a handy tool in Path Finder that lets you copy the full URI of a file by right-clicking on it. If you use that file manager, that can save a lot of time in pasting in image names. But I would test to see if images even work in NeoOffice first. I’m pretty sure I remember testing that way back, but it’s been a while and I could have just imagined it. Does Pages support HTML import? It might even work there, though it would shock me since their RTF support is so bad.

10

I’ve been looking at this, and I think I may have a solution for 2.0. Currently Scrivener’s RTF-saving ignores symbolically-linked images (because I hadn’t realised you could import them at all :blush: ), so such links get lost. I’ve looked at RTF syntax and it doesn’t handle this easily, but that’s not a big deal - I can save some of my own custom tags as I do with things like annotations to save linked image destinations in the RTF, so I could still preserve symbolically-linked images for 2.0. I could also convert linked images to real images when the text gets exported, and offer an option for them to be flattened even for RTFD - in this way, you could have linked images in Scrivener that would update when edited, but they would be incorporated into the exported text upon Compile.

Best,
Keith

11

That would be a nice addition. For large projects like this with hundreds of images, keeping media (especially print-ready media) out of the Scrivener project will only keep things much tidier in general. The project file will be easier to backup, forward in e-mails, and so on.

12

What’s the best way to send a big sloppy kiss via email? (Not that, probably, it would incentivize you to implement this feature!)

Isn’t it 10:30 PM in Cornwall? Shouldn’t you be out at a pub?

13

I think I’ll accept the sentiment without the sloppy kiss if that’s okay. :slight_smile: It is indeed after 10.30pm here, but instead of being in a pub I’m watching the so-so reboot of V, sad soul that I am. I’ve done a little work on the linked images and it seems feasible with only a few hours’ work, so I hope to finish it off tomorrow. The hardest part is going to be figuring out how to allow the resizing of the image - currently it’s not possible to resize linked images, only embedded images, because resizing changes the underlying image scaling. The easy option would be to disallow the resizing of linked images, which wouldn’t be as bad as it would be in 1.x because 2.0 restricts the horizontal viewing size of images to the width of the text view, so enormous images wouldn’t look as bad in 2.0 as they do in 1.x. But obviously it would be better if you could set the size, so I need to look at that.
All the best,
Keith

14

That’s worse than watching Caprica, all the while knowing that Caprica is most certainly going to end very badly.

But I did it anyway. Sigh.

15

Ha, after the unforgivable travesty that was the BSG finale I stayed well clear of Caprica (I did catch one or two and am still baffled at how it got any good press at all). Oh well, at least Baltar is now in FlashForward… (I watched the original V again when I heard about the remake - the new series is certainly pointless by comparison.)

Anyway, that’s completely off-topic… :slight_smile:

16

Are you trying to gain admission into the ranks of the +3?

17

Why not. Make it a round number. Like +3.1415926535897932384626433832795028841971693993751058209749445923078164062862089986280348253421170679…