4 ms·
As a user of Paperless I highly recommend setting it up, it is so incredibly useful as soon as you have to deal with any kind of paperwork-intensive activity (b
by anaisbetts 5y ago
As a user of Paperless I highly recommend setting it up, it is so incredibly useful as soon as you have to deal with any kind of paperwork-intensive activity (buying a house, getting a mortgage, paying taxes). Take literally everything you get and throw it into Paperless, then when someone asks for all of these documents it will take you minutes instead of hours to track them all down - when you need it, it's magic.
- einpoklum 5y agoCan you explain how it is better than carefully scanning and arrangement the scanned files yourself? (Not being facetious, I'm really asking.)
- vertis 5y agoNot the parent, but I suspect it's not better in the outcome, just in the amount of effort you have to go into creating and arranging vs just throwing documents at it and searching later.
- anthropodie 5y agoYou can categorise those documents, tag them, OCR them, search them. Access them on phone or any other device with browser.
- nojito 5y agoHow is that any different from Onedrive/iCloud/etc?
- h0l0cube 5y agoI'm looking at this ecosystem and wish I'd seen it sooner. I've been using the Google Drive Android app to scan documents. It does auto-cropping and perspective correction, but it doesn't OCR, and I manually rename the files so that I can order them by date and cross reference with my banking transactions. With OCR I might at least be able to extract some useful data and avoid some tedious data entry.
- einpoklum 5y ago1. If I need a browser to access them, that's very similar to "can't access them". Do you expect my programs to start browser sessions to access files? Will my file manager have to connect to that silly web server? 2. I can't effectively OCR them for many non-English language (at least - unless it's made some advances in OCR for some languages for which it's lacking). Also, if I do OCR anything, I then need to correct the mistakes, which takes a h-u-g-e amount of time even when the OCR works relatively well.
- ishi 5y agoYou seem to come at this useful tool with a very negative mindset. - The scanned documents are stored in the filesystem in PDF format, so you can certainly access them without going through the web application. - OCR doesn't have to be perfect, because it's just being used to locate a specific document when you need it. So if you have 1788 documents and you want to find your apartment lease, you search for "lease" and there's a very good chance that word will be found in the correct document. That's the whole point - storing documents with a low amount of manual effort, in a way that makes it very easy to find them when you need them.
- KennyBlanken 5y agoYeah, that guy is being a tool. "WHY IS THIS THING I REFUSE TO SPEND EVEN A FEW MINUTES BOTHERING TO RESEARCH, SO BAD?!"
- anaisbetts 5y agoIt's better because it automatically OCRs everything and puts it into a search index. It means that finding documents now consists of typing a few keywords and immediately getting what you're looking for vs having to open a ton of PDFs
- einpoklum 5y ago"Typing"? Typing where? What's the integration with file managers like? And what's the API for file lookup via that "index"?
- KennyBlanken 5y agoWhy don't you expend some effort answering your own questions instead of expecting others to answer a combative person on the internet? https://www.youtube.com/results?search_query=paperless-ng https://www.youtube.com/results?search_query=paperless-ng
- matthewmcg 5y agoAs others have noted, the tagging, auto flagging, etc. are great, and it provides a decent, cross-platform, web-based interface. Also, you can flexibly configure how it stores the actual documents. So at the end of the day if you don't like the program, it gracefully degrades to a bunch of PDFs organized into a nice folder hierarchy that's easy to back up.
- addingnumbers 5y agoDoes it still require you to press a button to tell it every single time it's pointed at a new sheet of paper? I'm waiting for one of these apps to monitor a video stream from a camera and automatically determine when a new sheet is ready to scan. I already have a mobile app that can find the edges of the paper, crop it, and OCR all its contents. But it still needs me there as a human to tell it "the page you're pointed at now is one you haven't seen before" when that determination should be far simpler than doing OCR. I'm waiting for the day I can aim my phone or webcam and just flip through a bunch of pages and have it scan each one as it appears. It doesn't have to be super fast like Data from TNG or Johnny Five from short circuit, I just want to go from "flip, press, flip, press, flip, press" to "flip, flip, flip"
- KennyBlanken 5y agoIf you have so much paper volume that pressing a button is tedious, you should be using a scanner with an ADF. Fujistu makes a bunch and they've been on the market so long that finding a used one for less money is pretty easy. If you don't have that much volume...just improve the physical UI. Whatever floats your boat: a foot pedal, capacitive switch, Clapper, etc.
- addingnumbers 5y agoIt's not half important enough to me to justify buying more equipment. Especially when there is no technological barrier that keeps us from having software do it automatically. I'm not going to buy gear to do something I fully expect the software to be capable of in the next year or two. Until then, searching paper with my eyeballs once in a great while is less work than procuring and maintaining a document feeder or pressing a button three thousand times, no matter how convenient that button is.