It's quite simple. Paper is a near-perfect technology. What replaces it has to be similarly perfect. When people look at what replaces paper, they think screens. Screens are pretty-near perfect too. But screens are not what replace paper. Files are. Screens merely display information, files store it.
And file storage and manipulation is still in the Stone Age. It's about as far from perfect as any technology can be. To kill paper you have to make it perfectly easy for people to design applications without it. Because it's already perfectly easy to design applications and systems that use paper.
We have to fix software development before we can truly kill paper.
Office IT is almost uniformly terrible. I see a lot of companies trying to achieve a paperless office, but very few spending the money it takes to make it a reality.
Paperless workers need two large high-res monitors, not one crappy 21" 1080p monitor or a laptop with a 1366x768 panel. They need at least one tablet with a decent pen interface. They need a good document management system. They need the training and support to make use of it all.
And, to add to what you're saying, paper is infinitely easier for people to interact with. It's a physical, tangible object which they can see, touch/feel, and manipulate just like every other object in the world. The paper we use today is the product of hundreds of years of iterative design - conscious, and otherwise. Plus, of course, paper never runs out of batteries.
I do though believe we eventually will create computers which are as natural for us to use as paper is today.
While paper used to be the near-perfect technology for storing and distributing non-ephemeral information nowadays this has been turned on its head to the extent that using paper for these purposes has become downright detrimental:
Information stored on paper in some filing cabinet in some office for all practical purposes simply doesn't exist. It can't be easily searched, retrieved or accessed by someone who's not in the same room and doesn't exactly know what he's looking for. In other words: That information is lost.
Old habits unfortunately die hard, which is why people still seem to be thinking that storing non-ephemeral information or knowledge on paper is a good idea.
Information locked away on some local user's desktop also doesn't exist. If information isn't stored on a dedicated application-specific storage system staffed with people skilled in it's use and repair, then it's the exact same thing.
We can't just take information and dump it on a block storage device. It needs proper archival. This is the same with paper, only if you lose the key to your file cabinet or the drawers stop sliding so easily, you can still get at the information.
If your RAID backplane dies and you don't know how to fix it, you're in for a bad day.
There is a skillset you can learn to work with paper more efficiently. But most people can muddle through without those skills. Not so with digital.
I was recently researching some of things archivists have to do, and the general consensus is that their skills are needed even more in the digital age, not less, because it introduces a whole host of new problems: files that fork, forgeries, losing the ability to read old file formats, etc.
There was also a project by MoMA (I can't find the original article), where they had to get a whole team together to create formats that will ensure survival in the future by doing things like including the instructions to view the format in a readable way.
Reminds me of a quote from the early days of the internet. It went something like "The internet is like a library with every book in the world. But all the books are on the floor."
I work in this very field. My company sells enterprise content management (ECM) software that helps organizations go paperless.
In my experience, it's not just about archival, search and retrieval. The system also needs to have robust automation capabilities, and be able to do things like automatically rename the files in a consistent way based on storage location, file content (text, voice, etc.), and identify when people store things in the incorrect location and correct those mistakes or at least flag the files for manual correction.
The other thing to realize is that information doesn't get generated in a vacuum. It's generated via various business processes. So if you can automate those using software, and the automation routines are tied to the content management system such that things are filed in the correct location based on business process data (e.g. which department initiated the process, who approved what, etc.) then you're basically 80% of the way to paperlessness.
>Information locked away on some local user's desktop also doesn't exist. If information isn't stored on a dedicated application-specific storage system staffed with people skilled in it's use and repair, then it's the exact same thing
Office365, GSuite. Small and medium businesses have options.
Small and medium businesses can keep their documents like this because the expected lifetime of those businesses is less than time until major breaking change on the cloud office suite platform (or its shutting down).
That’s not true at all. You use one of any number of filing systems.
I worked on a pretty huge paper to electronic record conversion project... millions of pages. We made it much easier to access, although about 25x more expensive, but 95% of searches align with the indexing systems in place for the paper process.
Even in my life, I use a traditional 43 folders system for some things. Very accessible.
Any system probably is better than no system in this case.
Still, there's the fundamental problem with paper-based system that when you design an indexing system you have to correctly anticipate how people will use the system for the foreseeable future (i.e. until the next reorganization).
It's much easier to find scientific books and papers online than by using IR systems in university libraries designed for exactly that purpose by librarians, whose entire profession is built on organising information. The reason for this isn't that those librarians don't know how to do their job. It's that information stored on paper alone (i.e. without a digital representation) can't be easily searched for.
> It can't be easily searched, retrieved or accessed by someone who's not in the same room and doesn't exactly know what he's looking for
To a lot of people in a lot of scenarios, that is a feature. Modern cryptography has yet to match the ease of use, comfort and security of a single piece of paper locked away somewhere only you would think to look.
> Information stored on paper in some filing cabinet in some office for all practical purposes simply doesn't exist. It can't be easily searched, retrieved or accessed by someone who's not in the same room and doesn't exactly know what he's looking for. In other words: That information is lost.
You can't find anything if you don't know what you're looking for, be it through a search box or a pile of drawers filled with paper. A search engine is just a sophisticated index. Information, in any form, is only retrievable once a certain hierarchy is imposed upon it (be it scarce or not) and some sort of index is there (be it in your mind---the invoice is in New Folder (3)---or actually written down somewhere, or somehow generated and operated by a search tool). Yes you certainly can't grep paper (you can try though, there's OCR if one wants to go to those lengths, and that certainly can be useful for lots of tasks, even in the personal scale), but you can't grep your text files if you don't have a search term at hand.
I would say when files become easily storable and accessible. Right now untold amounts of human labor are destroyed whenever the computer systems those humans are using fail. When I think about how most people use computers, I get the distinct feeling like I'm watching an episode of Primitive Technologies.
Systems like Dropbox are a good first step, but the archival of files is still far too complicated. This stuff needs to be easy enough for your 80 year old grandma to master.
I'd say we're probably 20-50 years out. We're eventually going to figure out a basic file format that all applications will use and extend. Then it'll become necessary for all applications to support it. It'll have solved the encoding problem, and it'll have built-in error correction.
Sort of a Unicode for file formats.
Once that's in place, then cloud storage and backup solutions will finally become useful.
IMO one of the problems with cloud storage in general is how complicated it is. There are so many moving parts and so many things that could go wrong that there are inevitably many states a file could be in and asking them to troubleshoot these is impossible.
For example, my wife is syncing several GBs of data up to Google Photos as we speak. Periodically, the sync process will complain about something and stop. There is nothing we can do except ask it to exit and we try again. What went wrong? Who knows. We certainly don't. The app UI isn't telling us anything. So we cross our fingers, restart the sync app and hope it works. This isn't the future.
"A file" needs to be have the qualities we need as a paper-replacement as an intrinsic property of itself, not conceptual software design properties that are merely abstractions.
I conjecture that error detection (and recovery) are layers beneath the file format it's self. They belong in the filing system and storage layers within/under that.
Application code should never need to worry about error recovery.
We're kind of close to this for text files, if only Windows and related programs would join the rest of the sane world. Stop adding 'BOM Marks' and just assume everything is UTF-8 (at least for any txt file created after 2000 and all application stdin/out/errs).
However the biggest barrier here is a complete disconnect between interested parties. Adobe wants to make money and keeps 'pdf' locked away. It's also bad because paper size is enforced, instead of having an aspect ratio, and positioning for different elements. Even HTML rendering isn't that simple, and NO designer is working using those tools. Additionally that complexity is FAR above what the average user can handle and WAY above what they want to deal with even if they can.
Any format which /is/ user friendly will have to support things like 'drawings' / sketches which may or may not be auto OCRed for searching.
I'm of the opinion that this won't really be solved until we have //strong// AI to do transcription and formatting for the lazy humans.
> We're eventually going to figure out a basic file format that all applications will use and extend.
The general trend seems to skip that and is moving towards renting hosted software applications. Google docs is already pervasive in certain corners, and not only can I not tell you what version of Google docs I'm using, but I can't really tell you what its file formats are, or even where it's even actually stored other than "in the cloud".
Assuming that trend continues, what, then, even are file formats?
The hiding of file formats behind the application isn't new, but that doesn't make them go away. Should the app ever die, necessitating a migration, then the file formats will suddenly become painfully relevant again.
We need a universal file format similar to plaintext but has standards-based extensions for storing data in that all applications can extract to. That way software can finally achieve paper's universality.
It's the first step on the road to making information technology truly worthy of the name. The next would be easy-to-use programming platforms. Like Excel or Visual Basic but not shitty.
Not the OP, but 98% of the time that someone sends me a .key file for a Mac Keynote presentation, my Keynote application is out of date and cannot open the file. Additionally, it sometimes cannot update either, unless other components are updated, such as the entire OS itself.
A bare minimum requirement for me would be, "a computer can be reasonably expected to open the file". Yes, a computer without the correct application installed could not reasonably be expected to open the file. But here is a computer, perfectly capable of displaying the file, but choosing not to do so. Note that these are not files that are created with the new Keynote that I may not have installed. These are normal Keynote files created in the same version I have - but now my version isn't good enough to open files any more, it would seem.
Sharing files is unbelievably inconvenient. Sharing files to my mum in the other room involves copying files to a USB drive and moving them manually, after I reformat a USB to ntfs or fat32. It doesn't matter that both of our computers are on a LAN because by default we have no pre-existing convenient ways to share files. With her being on Windows and me being on GNU/Linux, I'm not even sure which program I would pick if I could install one. For anyone that isn't in USB drive proximity, I have to brainstorm a solution every single time I want to share a file for every person I want to give it too. This is all despite the fact that theoretically, there are a million different ways that we could share a file.
You can share directories between Windows and Linux. The Ubuntu machine I'm typing on has a folder shared with the Windows 7 machine next to it. You can set this up entirely from the GUI. It's called "Samba" on Linux.
For these cases, I use a trivial program I wrote myself that just serves a specified directory over HTTP, optionally with basic auth.
Most home routers have dynamic DNS updates enabled, so the only instruction for the other side is "put $hostname into your browser window", sometimes followed by a "prefix it with http://" if their browser insists on searching Google for that name.
It requires some effort and knowledge on my side but tends to work pretty well for the other person. Even better when you have some instant messaging channel available where you can just send a direct link.
Most alternatives are really inadequate, especially for large files.
When you are in the same room or even city, I should be able to just move large amounts of data from one computer to a physically adjacent computer without having to go through the internet.
For long distance transfers, I agree with various cloud storage. Upload some files and create a share link. But even this is an annoyance for a large file.
AirDrop on Mac solves this problem. But how do you propose getting Microsoft and Apple to implement a cross platform standard when they have financial incentive to not do so? Cloud storage is the best answer, even though it is an extra technically unnecessary step for local transfer. That said, Even my 60+ technically illiterate mother can do it reliably. I think this is an excellent example of letting “perfect be the enemy of good enough”.
AirDrop is brilliant. I shared 10 pictures with my mother with one click. She accepted with one click and the pictures were instantly available on her iPhone, iPad, mac and Apple TV.
the only problem I have with it is that it can take a few seconds before nearby devices shows up on the sharing screen.
But isn't it a bit sad to send files to some cloud-storage or email server farm, probably half-way around the world, just to share between machines which are on the very same LAN?
All of the options I listed are cross platform. Why must a standard exist in this case? FTP is standard and cross platform. This is nothing more than a UX problem.
When you read about Kay's Dynabook idea (and get past how the machine mockup looks) you see that he's essentially describing the "computational equivalent" to paper. With such a medium, one could create digital forms as easily as they draw printed forms on paper. And we do not have these types of machines or systems today.
Additionally, there is no better read experience than paper. And consider archiving: digital documents are much harder to preserve than paper documents, and this has enormous implications for public policy, history, and democracy.
Paper is far from perfect, but it does have some features that certain business processes have come to rely on which files lack. For example: the ability to physically bind an authorization (a signature) to the information that authorization is associated with by putting both on the same sheet of paper.
Paper also doesn't require batteries. It's much more robust in the face of certain kinds of physical abuse.
But paper takes a lot more space to store, and it's a lot harder to transfer from one place to another.
Even as display it sucks. Where's the e-paper that was promised decades ago?
Paper (and pen) have stood the test of time for thousands of years. Warped tech nuts believe they can be replaced by substandard junk electronics. I'm glad consumers aren't so gullible.
Paper isn't perfect. For one thing, it costs a hell of a lot more to store paper than to store digital files; we pay a huge amount to store files offsite. And paper used for ephemeral content is inherently wasteful.
Paper is slowly but surely going away. We just finished digitizing all our employee records; all records from now on will be digital. We used to print reams of reports and store them in filing cabinets, shelves, etc but we don't do that anymore. We aren't paperless but it's been a slow and steady march for decades now.
And file storage and manipulation is still in the Stone Age. It's about as far from perfect as any technology can be. To kill paper you have to make it perfectly easy for people to design applications without it. Because it's already perfectly easy to design applications and systems that use paper.
We have to fix software development before we can truly kill paper.