There are good reasons for doing so, not related to the current mediocre tags implementation: better indexing, search, better import and encrypted sync over the provider of choice - which can be as well iClould. Just accept that file systems are tools which were never intended for an advanced AI-supported knowledge management. Thus, some people (like me) are pretty happy to rid of them completely at least for some tasks.
P.S.: Sync is seamless, and so is export. Actually, I could access all files even without export, it is just a package with files + index metadata + backups. Sorry, but this is FUD or a complete lack of knowledge.
I don’t mean to hate on DevonThink (or anyone). They’ve come up with brilliant solutions where Apple dropped the ball. My main point was that this all came about because Apple didn’t provide excellent indexing, search, and sync themselves, but I don’t see why they couldn’t. Also, to nuance the point about “headaches”: what’s been frustrating has been trying to see stuff inside packages alongside stuff outside packages as results in the same search in Finder. I’ve had less experience with DevonThink, but my experience with bibliographic databases like Sente (RIP) and Bookends, is that you can’t see many of the keywords without opening up the application.
Honestly, I don’t think it was ever their intention. Both are very specific applications for a very specific kind of problems. It is not something one would put in a file system. Even if all mentioned critic on current implementation of tags would be solved, we would still get only a “devonThink light” as an output. Tags solve only one problem - file systems are strictly hierarchical and file-based, but we mostly work with documents and don’t care so much about file system inodes. In a similar fashion, Spotlight does its work for a fast search and access. But some people need auto-classification, complex search queries and so on. As a launcher, it fails short of LaunchBar, but would you let your mother use LB? I would certainly not.
I think, what most people does not get right, when they talk about macOS, is why it is so great. Not because it is jack of all trades (it is somehow), not because it has so many utilities (it has), but because it has done its basics right - there is a lot of hooks and interfaces, so third-party software can use it. This is the reason for all the Hazel, KM, LB, Alfred, DT and so on around us.
So, do we need apple to build all this? No. But we need it to do the basics right. Sadly, the tendency in last years is just to throw half-backed solutions, which make automation between apps only more difficult(Hello, Mojave, and your stupid and non-transparent permissions). I have a feeling, that they don’t care that much about powerusers anymore. JXA was released some years ago and completely forgotten after the release - still full of bugs with no proper documentation, so I prefer to use AS, as ugly and strange as it is. Spaces lack some basic polish and so are tags. We have Marzipan on our radar now trying to get us on the level of iOS and I fear that should it work macOS automation will be dumbed down on the same level as iOS. Just because it is not a priority.
I completely grant your points about what’s so amazing about macOS. I get 80% of my work done on a 5K iMac, and if that number were 100%, I wouldn’t have written the post. It’s just that I’d really like to go iPad-only when I’m away from home. Instead, I find myself bringing my 2012 MacBookPro with me (in addition to my iPad), because the file management on iOS is so half-baked that I regularly have to turn to the MBP to find mission-critical files.
The fear that implementing this across OS’s will require dumbing down aspects of macOS that power users rely on… I get that. But I’m not seeing how such dumbing down would be required by any of the points I mentioned under the heading “What iOS needs to inherit from macOS”. Basically, I’m just looking for (1) easy tagging interface on iOS and (2) and complex search queries on iOS. None of this requires a UI that overwhelms anyone grandparent, either. Apple just needs to finish baking what they have in the oven.
I agree. That would be a nice feature improvement. I’m not much of a tagger, but when I need it, I need it to work universally. I think another hurdle, other than Apple, is getting developers to incorporate tag recognition, display, and editing into their interfaces in a consistent manner on both macOS and iOS platforms.
You should be always aware that Dropbox can and will read your stored files. So I would never push important files unencrypted over cloud services - including iClould.
@ptarh Thanks. I will have to read their TOS closer because their security pages talk about everything being encrypted at every stage. That article was from 4 years ago - is it still current?
I think it is current. You have to distinguish between encryption of your connection to Dropbox and encryption of data. Dropbox does the first one, but not the second one. Actually, it is even better - they do the second one, but they store all the keys and thus are able to decrypt your data every moment. The equivalent of burglar having the keys to your door. Here a small article on that: https://www.comparitech.com/blog/cloud-online-backup/make-dropbox-more-secure/
Ummm that article says DB checks the hashes of files you share against a list of known hashes for copyrighted files. First, hashing can be done without reading file content ( https://security.stackexchange.com/questions/168492/how-does-file-hashing-works ) . Second, the article says nothing about files you’re not sharing.
There are probably reasons to be concerned re DB but you should be clear on what the article is actually saying.
It is not important whether they are reading it on regular basis, the important things is that they are perfectly capable to read all your data - there is enough evidence for that. Your data is not private anymore if you upload it unencrypted to Dropbox. Whether they actually ever feel the need to browse through it is an entirely different question. Since I do not want to base my data security on beliefs and luck, I encrypt it.
I agree. While Apple provides basic tagging in Photos, we see it (and Google, and now Adobe) using AI to do the necessary analysis on images to make tagging superfluous for most occasions. For those without sophisticated (or neurotic) needs, being able to search ‘blue house’ or ‘puppy’ or ‘Aunt Agatha’ is much better for most people than making them have to manually tag. And I think Apple would not want to change its OS in any way that put more of a burden on the average user, so making tags more systematic and sophisticated is unlikely to be anywhere on a priorities list.
But nothing in that article says that - it says they may reveal your identity to law enforcement if you’re repeatedly sharing copyrighted material and presumably there’s a ECMA takedown order. Again what you say may be true about their encryption but it would be wrong to infer that from the article.
I’m late to this party. But the original post is fantastic – I agree wholeheartedly. I also agree with a lot of the comments. For me, using tags has been tremendously helpful and solved a major problem. I was on PC for a long time and used Evernote. Tagging on Evernote was wonderful because it allowed a note to live in multiple places at the same time. But I preferred using the file system for managing files, and I came to loathe Evernote.
When I moved over to macOs many years ago and discovered that the file system embraced tagging, it allowed me to transition out of Evernote, retain the one feature I really liked, and never look back. Now that I rely on tags and have access to them on iOS, I want the iOS implementation to be rock solid.
This is really great. The OS does not need to do everything itself. It needs to provide the features that third-party applications can tie into to provide the functionality. This is where iOS being so locked down creates a major hindrance. As it stands, it falls on iOS to be able to do everything we users want. I get the security issues and having a locked down system really does mitigate a lot of the security problems. The smart engineers at Apple need to figure out how to maintain the secure environment and still allow programmers and users the freedom they need to get their work done using the methods and tools they prefer.
I’m a late responder because I found this topic after responding to @JoelAnderson’s great post in a more recent topic and then looking at Joel’s other top posts.
This has been a much-needed feature in Finder. For now I do it in Leap.app, where it’s easy to do such multiple operations with tags.
This used to be possible in Finder. In Mavericks, you could do a Spotlight search from the Finder toolbar, with the search scope set to the whole computer, and the search results would include items in Mail (and I think most of the other Apple apps), which could then be tagged and would appear in subsequent searches for that tag with the search scope set to the whole computer (or in saved searches for the same). So this feature request is, in part, request for a “feature resurrection”!
AND and OR of tags in Spotlight can be done. It doesn’t seem to be 100% reliable because typing in the search field can be tricky! It seems to be more reliable if you make sure to use “tag:” and not have it substitute a tag token. Also is problematic if the tags are more than one word.
Thanks, @tomalmy: I knew about AND / OR / NOT operators in Spotlight, but I failed to mention it, so thanks for that follow-up note.
What is cool about Leap.app, as you may know, is that you can add tags to the search by clicking on them in the tag cloud, and the tag cloud shows only the tags in the current search scope, so since the search scope changes each time you add a tag, the tag cloud changes to show only the remaining available tags. You can navigate to a particular folder, easily exclude certain subfolders from the search scope, and see a tag cloud of only the tags in that particular folder but not the excluded subfolders. That kind of sophisticated tag-viewing/searching logic would be great to have in the Finder, as @JoelAnderson suggested. Fortunately we already have Leap.app.
