Okay. Your first assignment is to solve the metadata problem:
A Scrivener project is not just a collection of files. The most important piece – what makes Scrivener unique – is the metadata that explains how those files relate to each other. To be Scrivener, your program must allow users to arbitrarily create new files at any point in the outline, to drag existing files into an arbitrary order, and to keep any simultaneous editing sessions in sync with these changes.
As I understand it, this is the technical issue standing in the way of a cloud-based version of Scrivener. Any suggestions?
For background reading, this thread explains how Google solved the simultaneous editing problem, but they still haven’t done metadata:
Throw out the idea of simultaneous editing sessions; I’m only ever in one place at one time. (I’ve no interest in Scrivener for group editing projects.)
Once you get rid of that, the cloud is just another place to put files. You do need to manage when to make local copies, deal with caches, and such but the files are the files are the files. None of that changes in the cloud.
As long as you can manage the commits, you have done most of what needs to be done.
Relational databases do this all the time, after all; I used to work in Pick (yeah, I’m old), which was a file system as OS, so that’s yet another example. Version management tools could help as well, if you want to go with the check out/check in model.
Any Scrivener features that don’t work with these models, of course, can be cheerfully tossed from the online version. As someone who only uses Scrivener to write and compile, you’d have at least one happy customer that way.
Keith, you had that early reaction to an iOS version as well, and now we are awaiting a release date. I understand your response, but would urge you to start thinking and planning soon, for a simple online version would integrate nicely with Google’s fast-evolving Chrome environment.
Since October I have used the Mac Chrome browser and have come to love its speed and ease of working with cloud apps. Everything works in a single interface: e-mail, contacts, calendar, tasks, phone calls, just a tab away. Everything saves to the 100 gb Google Drive. Change a file on one machine, and your other Chrome machines see the same file. That makes a snap of sharing files or collaborating.
I have not yet tried to export a Scrivener project and import it into Google Docs. If that works, then maybe I have the desired integration I’m seeking. But if it were that easy, someone would have done it already. Maybe out in the Scriv forum is an expert who knows how to make it happen.
Anyway, Chrome is now available on phones, tablets, laptops, and desktops. It would be great to see Scrivener there as well.
PS: if you have access to Chrome, see an app called Scriptito. It strongly resembles Scrivener, with Binder-like files and folders, and the import/export features are detailed here:
I’m not aware of any way in which a cloud file would be different from a local file, or how the relationships between files would be handled any differently in the cloud than locally.
I did not intend to suggest no editing of the outline. I’m only saying I don’t need support for group revisions; just me, on one machine at a time.
Any database would work with cloud storage, the cloud is just a disk somewhere else. All functionality is local; the cloud is just storage. (Unless you want web-editable, in which case, that’s just typical web code - and DB support is really, really wide, with a zillion options. The ‘program’ sits on a server somewhere, but to the cloud-stored document package, it’s just another chunk of code operating on cloud-stored data.)
So my argument, summed up, is that putting something in the cloud is not too hard if only one person is using the data at any one time. Not trivial, but not onerous, either. You should add things like remote closure (“I’m no longer on machine A; I’m on machine B”), local caching (with guaranteed commit) and some security for access, of course. Multi-user is always what creates by far the worst headaches (as you pointed out). Eliminate multi-user, and the job is much less demanding. I’m not aware of anything in Scrivener that wouldn’t translate to cloud storage. (In no small part because I use a modified form of cloud storage [with local copy]: dropbox.)
Come to think of it, if Scrivener only added remote closure (of a file open on another machine), Dropbox support would be vastly improved. It doesn’t even have to be real time; set a bit in the metadata that means something like “no more changes permitted on this machine until you close and reopen the file, and the file will not be saved on close” (this works because autosave will have already saved any prior changes).
I’m afraid it doesn’t work. Allowing programs other than Scrivener to directly edit Scrivener projects will corrupt the project. Please don’t try it unless you enjoy mutually frustrating conversations with the support team.
The “Sync with External Folder” function would be the appropriate method. However, Google Docs specifically uses its own internal format, not RTF, and so the edited files would need to be reconverted to bring them back into Scrivener.
Sure. We already have this: put your project in Dropbox, and use Scrivener on your local machine to edit it.
But it didn’t sound like that was what you were looking for in your original post. It sounded like you wanted a web-based application that duplicates some substantial subset of Scrivener’s functionality, with the local machine no more capable than a generic browser.
No, that’s still not what I’m talking about. Sorry if I’m not clear. Dropbox does not equal cloud; it does use the cloud, but it has potential fatal problems (most importantly, if I leave a file open at home, I can’t work on it remotely).
I’m talking about a situation where the file/package resides in the cloud. You could then edit it at your desk (with the various facilities for caching I mentioned earlier, which would be typical of any cloud storage solution), or at Starbucks (adding remote close if you leave the project open at home), or online (via an online editing engine that doesn’t exist right now).
So: data in the cloud, editing done with various client apps (existing Scrivener, iPad, online, or anything else that might get conjured up). This would require various services I’ve mentioned, and probably a few that I haven’t, but none of them are difficult services to build compared to complete new client implementations (e.g., online).
It strongly resembles Scrivener because the developer was “inspired” by Scrivener, as he says himself. I’d be grateful, therefore, if we could avoid promoting it here. There are no plans for our own online version at this time, and you should of course use the software that best suits your workflow rather than hoping our business will change to accommodate it. If you’ve moved from your iPad to online software then I’m afraid that is a different direction to us - I personally don’t like online apps. Of course, I rule nothing out for the future, but it’s certainly not in the calendar for 2013.
The answer may soon come not from the cloud, but from Scrivener on Linux.
Boffins have figured out how to run Ubuntu on an Acer Chromebook.
That’s the one that sells for $199 in the USA.
Agreed. I think getting the hardware to work more like a computer is probably going to be easier than getting the web browser to work more like a computer, because it only takes a little effort to get an operating system installed onto the hardware, from a few people (and once the initial puzzle of porting to the hardware is solved, it only takes the effort of those who wish to do the installing). But getting your web browser to work like a computer requires every single developer of every piece of software you want, to make browser extensions/servers/clients/web sites/etc to all mimic the way software works (and no, sorry, “the cloud” is not the answer to that—in response to the post above. That’s just a buzzword for the manner in which servers and machines and software and all of the stuff we’ve all been using all along is presented to the user. It’s not a magic pot you can throw “stuff” into and suddenly get results out of. Developers still have to code all of the stuff that goes into making “the cloud” work).
I’d like to add my +1 to this topic not for myself but for the scores of younger writers all over the world that are growing up with web-based being their primary mechanism to do just about EVERYTHING.
We have computers and think of software as software. They have texting phones and use whatever computer they have transient access to. They also think of pricing differently, so a scaled by feature approach might be better received than a single purchase type.
A web version wouldn’t have to do everything that the desktop version does and certainly not right away—but it should do the core things that make Scrivener different and most useful.
Unfortunately, the wants of younger writers (or any other writers) don’t have much relationship to the capabilities of the various platforms. Scrivener for Web doesn’t just magically appear because some number of people want it, it requires a substantial investment in coding, infrastructure, and support. So far, it’s not at all clear that such an investment is justified by the demand.
Sure, lots of people are happy to play games, read email, and browse the web from whatever device they happen to be using. I use cloud services for both my grocery list and my main to do list. But I think you’ll find that the number of people doing serious work in the cloud is much much smaller, for all the reasons discussed here.
Software never just magically appears; however, expressing the desire for it is a way to communicate the demand for it. It doesn’t matter if 200 or 200,000 or more would buy it, you wouldn’t know if 0 people said something. Pointing out the advantages a web version would have is not expecting it to just appear nor is it expecting Lit+Lat to bend to our will.
Actually since google docs is finally a full-featured office suit, there are many businesses that use it to do everything they used to use MS Office for. Some of these businesses give employees chromebooks instead of normal laptops because they can do all their work online. The number of people doing serious work in the cloud is growing, and as this thread grows older (and tech advances) I think a web version will become less impossible (from a development stand point). However, that doesn’t mean I think it’s reached the break even point. Doing everything from a web browser is still pretty new compared to doing everything locally (be it a PC or a tablet).
Though once the ipad version comes out, it might fill the need of most people who would be interested in an online version. At the moment I would like an online version, but that because I don’t have a normal laptop (I have a desktop). Once the ipad version comes out, I plan to buy an ipad so I can write where ever I want (riding in cars, coffee shops, the library, my living room, etc.), and so I can have other ipad apps like Photoshop. So rather a web version would get the audience to be worth the investment probably depends on the market success of the ipad version and the ipad itself.
But you seem to have missed the point that Keith has publicly stated that the “audience” is irrelevant.
Not many people seem to read the thread at the top of the page entitled “BEFORE POSTING …” which has been there since 2006. Keith states his ideas there very clearly, and says:
“Scrivener is … based on my personal vision of what I want for my dream writing software”
It is that vision (rather than what the market has demanded) that has made Scrivener what it is, and it would be as well for us all to remember that. Anyway, I would recommend reading that thread, as understanding the philosophy of the developer might save a lot of needless debate. And as Keith also writes in the thread
“Sure, sometimes I say no and then change my mind, but once I’ve heard your suggestion, provided I haven’t misunderstood it, I am aware of it and can change my mind or not at my own leisure”
Google itself only claims that around 5 million businesses use Google Apps, most of them small. That sounds like a lot, but it isn’t relative to the number of Microsoft Office users. And there’s no way to count the businesses – my own included – who sometimes use Google Docs for convenience but mostly turn to MS Office for power.
Similarly, most estimates of Chromebook sales are well below a million total units (all models). That’s rounding error next to the total PC market of more than 25 million per month.
It’s accurate to say that some businesses are experimenting with Google Docs and/or Chromebooks, but neither has yet established itself as a serious alternative.
I would like to voice my user vote against on online version. As a user who loves the current software and is looking forward to the iOS versions, I’d like to see development bandwidth go towards those projects instead. I like the smoother performance of native apps. If Apple would relax its rules and improve it’s developer tools/info for iCloud, having the options to store data there would be great, though.
I agree. I really don’t see a huge need for this once the mobile platforms are addressed. Sure, there are some people stuck on web browser based machines, but I don’t think that fad will last. The Internet just isn’t ready to become a distributed operating system, yet. So right now the Chromebook offers only a very small subset of what a normal laptop can accomplish, we can’t bend to whatever gadget is currently interesting; that’s a very expensive way to make mistakes. For portable stuff, the other reason a web interface is often desirable, covering iOS and eventually Android should be enough for most people.
There was some talk upthread about thumb drive versions, and this is something I’d really like to see made available. A user mentioned that the windows version of Scrivener will run from a thumb drive, but I know it leaves information in the registry. What I’d like to see is a leave-no-trace version, one that doesn’t touch the registry or anything on the host PC but operates fully within the thumb drive or external hard drive. Is that doable?
Also, does anyone know if the mac version leaves information behind on the host machine when run from a thumb drive? I’ve only had a mac for a month so I’ve no idea if macs have a registry or anything comparable.
As far as an online version is concerned - I know the iOS version is coming soon, so for Apple folks, I can’t imagine there’s a huge need for online. I’d love to see a version for Android for all of us Android users (yes I know that’s a whole separate ball of wax) - with that and stealth thumb-drive versions I can see very, very few instances where an online version would add any benefit at all.
There are no plans for our own online version at this time, and you should of course use the software that best suits your workflow rather than hoping our business will change to accommodate it. If you’ve moved from your iPad to online software then I’m afraid that is a different direction to us - I personally don’t like online apps. Of course, I rule nothing out for the future, but it’s certainly not in the calendar for 2013.