True Aliases in Binder — a vital organisational tool!

De
December
Posts: 58
Joined: Fri Jul 01, 2016 7:02 am
Platform: Mac

Thu Feb 23, 2017 4:28 pm Post

I love this program — can’t imagine now how I ever lived without it. Seriously, what you guys have created is an amazing gift to writers. But I have increasingly been bumping my head — banging it, even — against a limitation that is driving me CRAZY. I have no idea if this is one of those seemingly-simple things that would actually be near-impossible to change given the underlying architecture of the software — but if you could fix it, that would EXPONENTIALLY increase the functionality of Scrivener for non-fiction writers, especially people who like to play with the logical structure of their work at a granular level (or who think in outline form!).

My problem is this: I’m in the process of taking an old (non-fiction) manuscript to pieces — two different manuscripts actually, on related topics — and fashioning something new out of the material. The sort of thing Scrivener is ideal for. BUT....I’m going nuts because there’s currently no effective way of introducing ‘aliases’ into an outline structure. I have a lot of original material which is usefully grouped together in topics & sub-topics. The last thing I want to do is to prematurely take that repository apart and scatter the bits across my draft! But I do urgently want to ‘borrow’ stuff, reassemble it into the outline I’m starting to construct for the new project. Several alternative outlines, in fact. In theory, this is exactly what Scrivener should be good at: letting me string together several chunks of thought and see how it looks.

BUT...if I use the Collections feature — that is, gather all the relevant docs together into a Collection — the outline structure of the material, its organization into folders, gets totally flattened out; the result is a unmanageable hodge-podge string of fragments with all trace of logical structure erased.

And if I rearrange it in the Binder, well then my original organisation is permanently lost. Equally important, there’s no way to look at — preserve — two different sequences at once. Yes, I could make copies of the relevant folders/docs and distribute those into my new outlines — but that way madness lies, because I continue to make emendations & additions to the material! I’ve been trying exactly this for a while now, and can report that keeping all the copies of crucial documents updated is completely maddening.

What I desperately want is the ability to drop an ALIAS into the appropriate place in the Binder, and have its contents show up in the Editor as if it were an original (the way Collection items are rendered)! Then one could properly sketch out alternative outlines, and start working with them. To be any use, this does need to be a true alias whose contents show up in Composite/Scrivenings view, not just a doc containing a link to the original — so you can actually see any subfolders and/or text!

Failing that, the ability to organise Collections material into subfolders (that is: new, temporary Collection-specific folders, not the ones the docs actually belong to in the Binder!) would help a lot — though that’s a second best (not least because it’s much easier to accidentally delete a Collection than to lose a chunk of the Binder).

I don’t know: maybe either of these changes would take too much processing power — but if it were possible, honestly, it would transform the usefulness of Scrivener. As it is, I’m beginning to think my only way forward is to print up all my material, get out the scissors and start pinning it to bulletin boards!

Which sorta vitiates the point of working in Scrivener... : )

Thanks for listening!


ETA
A few specific examples of why this feature would be so useful:

• Writing a long introduction. You might want to pull bundles of material/text out of the body of a ms and gather it in one place to cherry-pick bits for the intro. Much easier to see what you’re doing if you have the stuff in sequence, in a section of the Binder dedicated to the Intro, rather than having to keep looking up bits elsewhere as you go. But obviously you don’t want to actually remove material from the body of the ms!

• Topics that come up in two different places in your ms. All that material ‘belongs’ in both parts of your outline; you want to look at it as you flesh each section out. Sure, you could park a placeholder in one, but it’s annoying not to be able to open it out and view the subfolders when you look at that part of your outline. Or maybe the whole topic is relevant to one section of the outline, but other sections only need to refer to parts of it.... Aliases would make this process a doddle.

• Trying out radically different outlines for your project. This seems like the heart of what Scrivener aims to help writers do. Currently though, it can only do this for linear sequences w/out substructure (ie via Collections) — ok for fiction or narrative sequences, not so good for more complex logical argument. Pretty sure I’m not the only non-fiction writer who thinks ‘hey, what if I totally took this apart and rebuilt it this way?’’, but doesn’t want to shred their existing outline. Trouble is, the new arrangement won’t be logically legible unless it can be organised into a folder structure, which Collections doesn’t do.

an
anderswt
Posts: 15
Joined: Thu Oct 06, 2016 9:25 am
Platform: Mac

Thu Feb 23, 2017 4:33 pm Post

I support this request, highly usable.

User avatar
nontroppo
Posts: 1246
Joined: Mon Mar 05, 2007 5:22 pm
Platform: Mac
Location: Airstrip One

Fri Feb 24, 2017 3:58 am Post

I can't think of a reason why this should be difficult to implement[1], I think it is a matter of design. Problems could be for example what to do if you alias a Folder that itself contains an alias etc...

I agree with the fact that Collections "flattening" an outline isn't ideal for some workflows, and that document aliases would be a flexible alternative. Personally I'd rather be able to "outline" in a Collection, but I assume aliases would be technically simpler to implement?

----
[1] I assume Scrivener can't currently do this, but I regularly discover new tricks on these forums!!!

De
December
Posts: 58
Joined: Fri Jul 01, 2016 7:02 am
Platform: Mac

Fri Feb 24, 2017 4:46 am Post

nonttoppo wrote:I can't think of a reason why this should be difficult to implement[1], I think it is a matter of design. Problems could be for example what to do if you alias a Folder that itself contains an alias etc...

Oh wow I hadn't thought of that. But I think you've hit the nail on the head: because scrivening sessions actively 'call up' subdocuments, the program can't have aliases the way the finder does -- it would easily create an infinite loop! (ie the finder doesn't care if you put an alias for doc A inside one of doc A's sub folders, but Scrivener would have to render the alias as part of Doc A: infinite regress!!). Damn, damn, damn.

Unless there's some clever way of telling the software to detect that kind of embedded self-reference, and simply ignore calls to display the alias' contents in Scrivenings mode...? (please please please... : )

User avatar
kewms
Posts: 6440
Joined: Fri Feb 02, 2007 5:22 pm
Platform: Mac

Fri Feb 24, 2017 6:04 am Post

Make more projects.

If you drag material from one project to another, it makes a copy. That is, the source project remains intact, and you can throw paint around in the new project to your heart's content.

For example, I use one project as a "notetaker" for conferences and interviews. Each document relates to exactly one interview or talk, neatly arranged with all notes from a single event together.

Then, in the actual writing project, I'll split these notes into thematic chunks, shuffle them around, delete the ones I don't want, write more words to glue them together. I use keywords to keep track of sources: if you split a document, both halves will get all of the keywords assigned to the original.

I'd recommend treating one project as the "canonical" original, so that you don't end up with research materials all over the place. But for the text itself, make as many different "piles" as you need.

Just be sure that each project has a unique enough name to make it easy to tell which is which.

Katherine
Scrivener Support Team

User avatar
nontroppo
Posts: 1246
Joined: Mon Mar 05, 2007 5:22 pm
Platform: Mac
Location: Airstrip One

Fri Feb 24, 2017 11:38 am Post

December: I'm sure there is a workaround for the recursion to infinity and beyond :D

kewms: I agree that a Project is a better solution in this instance than a collection. But it still requires that, as multiple documents are edited, a user must keep careful track of each one and continually copy the changes over to the "slave" projects. An alias would be a much more elegant solution.

This sounds like the sort of feature request that has to have been asked before, and there's probably a long thread debating the finer points with a good reason by Keith not to allow aliases, but until a digital archaeologist unearths that thread I will support this motion!!! :P

User avatar
kewms
Posts: 6440
Joined: Fri Feb 02, 2007 5:22 pm
Platform: Mac

Fri Feb 24, 2017 5:35 pm Post

nontroppo wrote:kewms: I agree that a Project is a better solution in this instance than a collection. But it still requires that, as multiple documents are edited, a user must keep careful track of each one and continually copy the changes over to the "slave" projects. An alias would be a much more elegant solution.


Hmm. Good point, and one that doesn't come up in my particular application.

One way to deal with it, though, would be to treat secondary projects as disposable. That is, use them only for experimenting with different structures, and throw them away once a structure has been chosen. The "master" project can be duplicated again at any time, allowing structure development to be an iterative process.

Another alternative would be to do the structure development in a secondary tool, such as Scapple. In my experience, that works very well at the idea/topic sentence level of granularity, but it would probably be pretty unwieldy if you tried to put full paragraphs or larger into Scapple notes.

Katherine
Scrivener Support Team

PJ
PJS
Posts: 1185
Joined: Sun Jul 22, 2007 5:05 pm
Platform: Mac + Windows
Location: Upstate New York

Fri Feb 24, 2017 6:03 pm Post

My needs these days are considerably more simple — more direct, one might say — than they were years ago when there were clients and editors and such folk as those to amuse and appease. So it's likely that my workflow ignores many of your concerns.

Still, I wonder if the plain old "Duplicate" command might help solve some part of this problem.

ps
You can't conquer stupid — or cure it — with more stupid.

De
December
Posts: 58
Joined: Fri Jul 01, 2016 7:02 am
Platform: Mac

Fri Feb 24, 2017 11:17 pm Post

hmmm. I thought I had posted this in the morning, but it seems not to have 'taken'. Reposting now, w/apologies for some redundancy in light of subsequent posts and an addendum I hope addresses them.

kewms wrote:Make more projects.

If you drag material from one project to another, it makes a copy. That is, the source project remains intact, and you can throw paint around in the new project to your heart's content.

This is basically what I’ve been doing, though I’m keeping it all inside a single project to make sure Scrivener-internal links don’t go pear-shaped. That is: I keep the original stuff intact in one folder “Fragments”, and make copies I can play around with. My situation is maybe a little more complicated than yours because the ‘source’ material I’m dealing with is bits of my own work — essentially those thematic chunks which are the endpoint of your reworking — that I’m trying to repurpose. What I want to preserve is their grouping: each ‘source’ is a collection of ‘material related to topic X’.

I did originally try using keywords for this, but it turns out there’s no substitute for having stuff that belongs together BE together: visibly, tangibly, in a place where I can not only find it, but visualise it to myself. A bit of personal mental wiring, I guess: I think spatially, the sort of person who prefers icon view to list view, because it helps me remember things! Keywords are fine for the occasional search, but the groupings they create are evanescent, you can’t scroll easily between two different groups, or see at a glance if you’ve ‘left out’ a document from the group, the way you can w/ primitive ‘physical’ grouping within the Binder.

I guess I belong to the pre-virtual world : )

nontroppo wrote:kewms: I agree that a Project is a better solution in this instance than a collection. But it still requires that, as multiple documents are edited, a user must keep careful track of each one and continually copy the changes over to the "slave" projects. An alias would be a much more elegant solution.

I’m afraid this is exactly right. Even if it’s all in one project. Proliferating copies of docs and folders, so you have the stuff where you want it, is a major pain. Thank you for seeing my issue so clearly!

FWIW a couple more examples of why this would be SUCH a useful feature:

• organizing source materials as they come in: it’s not just the writing process; even gathering together materials, you might want to ‘stash’ a document under two topic headings so it’s there, visible, when you need it! Glancing over a sheaf of notes is a very different process from laboriously calling them up one at a time — which is what you have to do if you use ‘pseudo-aliases’ (ie documents containing a Scrivener link to the desired original), which is what I currently resort to. Or of course you can proliferate copies, but...

• segregating current work from older stuff: before I switched to Scrivener, I used Finder aliases to help me keep recent work together, even if it was dispersed across a project. Maybe this is more of an issue if you are re-thinking an old project, as I am, and have a body of ancient work that is now being overlaid by new ideas; but anyway, in my previous Word/Finder workflow, I would have a folder of Current Work, where my latest brilliant ideas and thoughts were stored, but I’d also distribute aliases of the new docs into the folders of my old work, so the new thoughts went into the topics where they belonged. Yes, Scrivener has various meta-data tools that would let me mark new work as recent, even if it was dispersed among my old files, but that’s not the same as just keeping it all in one virtual ‘box’ where I can see it and flip through it!

nontroppo wrote:This sounds like the sort of feature request that has to have been asked before, and there's probably a long thread debating the finer points with a good reason by Keith not to allow aliases, but until a digital archaeologist unearths that thread I will support this motion!!!

My thought too. My primitive searching skills didn’t turn anything up, but I imagine the thread is out there. It’s such an obvious thing to want — intuitive, powerful and incredibly useful — that there must be a good reason it’s not available, alas.

ETA
kewms wrote:One way to deal with it, though, would be to treat secondary projects as disposable. That is, use them only for experimenting with different structures, and throw them away once a structure has been chosen. The "master" project can be duplicated again at any time, allowing structure development to be an iterative process.

A good solution for the problem of trying out alternative logical sequences, thank you! Though still leaves unresolved many of the other uses I'd like to put aliases to (described above), which basically involve dropping aliases of useful documents into my work wherever they might come in handy and working with them there — not really a temporary, 'give it a try' scenario. I guess I've just been spoilt by the Finder which makes this so easy!

PJS wrote: Still, I wonder if the plain old "Duplicate" command might help solve some part of this problem.

As you'll see from the above, I came to the same conclusion, and I'm doing exactly that! Great minds, etc... : ) Still the fact remains that proliferating multiple copies of documents gets complicated and confusing and wastes a lot of brain time I'd rather be spending thinking about content. Frustrating!

I think we're basically all agreed here that:

a) there are some just-about-viable run-rounds (though frankly after today's latest session wrestling with my insane confusion of duplicate docs, I am seriously considering returning to scissors and paste...). Many thanks to all of you for taking the time to share your suggestions here.

b) true aliases, if they could be incorporated into the software, would be an infinitely more elegant and useful solution. Basically everything else is an ugly kludge which confuses the brain and clogs up the workflow. If it can't be done, it can't be done, but I would be interested to hear back from the development team sometime about what the issues are....

oooh, that got a little long. Sorry.... (*grin*)

User avatar
nontroppo
Posts: 1246
Joined: Mon Mar 05, 2007 5:22 pm
Platform: Mac
Location: Airstrip One

Sat Feb 25, 2017 5:22 am Post

I got 7 pages of posts, several others are the same wish, e.g. http://literatureandlatte.com/forum/viewtopic.php?t=34294, so you are not the first to ask.

The one with some meat on the bone from L&L is from last year, http://literatureandlatte.com/forum/viewtopic.php?t=33688, where we get this answer from Ioa:

AMBERV: We did in fact entertain the idea of “multiple binders”, and thus multiple structural treatments—but whew that gets real messy real fast even just purely at the conceptual level. What to do with 32 files that in structure A are nested beneath one group that has been trashed from structure B? Do we now have two items, one for A and one in the trash for B? Do the 32 files just get tacked on somewhere having been orphaned from the tree? And then what happens when you empty the trash for structure A, having forgot about the container in structure B that got trashed? Well, that’s just one problem. :) We gave it a good stab though as I recall, before concluding a simple list of files stored separately from the primary tree structure did 95% of what people want, with far less interface and potential for confusion, and with several months less coding and design time at that.

Maybe some day, we’ve learned much from iOS synchronisation, even solving some of those problems since syncing conflicted projects is by definition the resolution of multiple structures—but I wouldn’t keep my fingers crossed if I were you, certainly not for anything in the next few years.


So that is another potential problem, how to handle deletion. So, a potential for circular-references and what to do with existential-conflict :D — both issues are quite easily solvable I think with some clear rules:

  • If a folder contains an alias of itself, the alias is not expandable. [simpler alternative: only single docs can be aliased, not folders]
  • A "file" is not actually 'moved' to trash until all aliases are. I think the Trash is anyway just another virtual binder right? [simpler alternative: aliases now shows "DELETED FILE in blinking text! :shock: ]

As Ioa mentions, with iOS they have now made a much more robust Project format that is built to deal with conflicts. The use of unique file numbers for docs and their references in the binder XML already exist; in effect currently a "document" in the Binder is already an alias, so I think going from one-to-one to many-to-one Binder map shouldn't be such a big leap. But of course Keith will know this better than armchair software architects like me!!!

-----
EDIT: first post at least 10 years ago: viewtopic.php?t=2930 — and in 2009 Keith said he like the idea but the backend didn't easily support in in Scrivener 1:

Re: Aliases: a Binder entry in two places at once?
Post by KB » Mon Jan 19, 2009 9:20 pm

Hi Dan,

This is something that has been suggested before and it is something I would be interested in adding one day, as I think it would be quite a nice addition. However, the way the document structure is set up internally at the moment means that to add something like this there would need to be quite a back-end overhaul, and although I like the idea I don't like it quite enough to rewrite as much as I would need to! So, it's an idea that is still on the backburner for the future - not for 2.0, but maybe for a later iteration of 2.x or 3.0...

viewtopic.php?t=5337 — I think that this would be feasible for a V3.x release now?

db
dbc183c7
Posts: 144
Joined: Sat Oct 25, 2014 12:59 am
Platform: Windows

Mon Feb 27, 2017 9:39 pm Post

[Only a small emendation from this' (!?) original iteration; in the last sentence, a relativistic change.]
I do appreciate Scrivener: It is a qualitative improvement over what I had been using (Info Select [infoselect]) -- regardless S.'s $100 cheaper price! I look forward to getting back into full swing once I've copied into S. everything I'd already developed in IS: ~ 13k RTF documents yet to go; already in, 10,216/996 RTF/TXT docs accounting for 250MB/331KB. (80 MB more of PDF and mostly pics.)
[Though as noted elsewhere, I still have to endure my S.'s long initial indexing ... another time.]

And one thing that I like much (and had asked for a decade ago in IS), the ability to have 'collections' of document shortcuts -- precisely so I did not have to update duplicate and many other copies of the same working source text whenever I changed it.

Putting in my 2-cents worth (even in these inflated times) ...
1st penny: The flatness of collections is a disappointment. (It hadn't occurred to me as an issue before S. cuz my separate 'collection' folders in IS were simply that, new folders, and copying documents into them was simply select/drag. (The S. document-duplication is less simple.)
2nd penny: I don't see making more projects helping my developing a single project -- Due to the lack of cross-project searches, i'd lose the serendipity that enhances the context and depth of THE project, particularly early on.
(And re 'early on', even were I to have substantially different possible layouts, none are 'secondary', but are rather alternatives, each is a prospective 'primary'.)

I've taken to doing two different workarounds:
For the likely candidate documents that are duplicated and more throughout a project, I use the yellow caution triangle item icon for them to remind me (not for the source). And I have a suitable status for finding them all.
(I'm also thinking about using another, non-default icon for the actual source working document rather than the label I use now. :roll: )

In a 'technical' set of folders, I have Miscellanea folders in which I have empty documents with 'titles' that are simply an appropriate sign for book divider classes (Parts, Chapters, Sections ...) and another 'class' that is a line.
(Yes, because of more 'developed' Collections still awaiting formalization into a 'draft', each of these divider class markers needs to be replicated since each single Collection can't have more than one link to a particular item (?) -- which 'restriction' I do otherwise appreciate, so no foul.)

[I have one Collection of Shortcuts to jump to the Misc. folders whence I click/drag dividers as I need them. (The Collection has substantially more to it than that, but that too would be for another time.)]

Two cents and some footnotes ... :wink:
(In my 'perfect' world -- or at least in that nearly perfect world already represented by Scrivener 8) -- outline-able Collections would be a reality.)