10 years since Google said to “hang tight” about Linux support for Google Drive
abevoelker.github.io
abevoelker.github.io
However, there is a post on this site that has become infamous in which a commenter questions the need for Dropbox when stuff like SFTP already exists.
https://docs.microsoft.com/en-us/azure/storage/files/files-s...
I don't understand how SMB could work unless there was geographical replication to nodes closer to you, with some kind of load balancing to change the endpoint you're using to be closer to you.
https://docs.microsoft.com/en-us/windows-server/storage/file...
Tragically it was still trash.
https://www.digitalocean.com/community/tutorials/object-stor...
Object storage isn’t random access (the ability to read and write any part of a file) the way a disk drive is. This causes complications for software that expects to be able to randomly access a file like they would on a disk.
Partial reads are supported everywhere out of box just by using the Content-Range HTTP header.
Partial writes aren't standard, but some implementations e.g. SabreDAV can do those using PATCH with X-Update-Range header (and e.g. rclone supports this).
WebDAV has its warts and limitations but it's a very underappreciated protocol.
> for which there exist multiple utilities to mount the thing as a local file system
Which all have their own sets of intricacies. The most obvious one being listing a directory is either impossible, costly, or as slow as the heat death of the universe. Other (far riskier) ones (from s3fs-fuse, which I have the most experience with[0] ) are the lack of guarantees about atomicity and concurrent access.
The last MacOS update broke Spotlight indexing for my Google Drive folders. Really annoying not being able to use Spotlight (and by extension Alfred), so the only logical thing to do was move to iCloud.
In the process of moving off of Google Drive, I also decided to move my email to Fastmail and just ditch GSuite (or whatever they are calling it today) forever. I’ll sleep better at night knowing that Fastmail knows their place and (hopefully) won’t try and force a chat function into my web email client as a means to build engagement for a product I don’t care about.
I just tested with Slack, VSCode, and GIMP, and all three allowed me to browse and open from there.
I ended up buying InSync because it was 50% off in easter, and its Google Drive/Dropbox paths works perfectly even in the terminal. Unfortunately it lacks Google Drive streaming.
I disapprove of the "new" way (that is not so new, but it is the current trend) because it favors lock-in and general loss of control. But there are also some advantages such as being able to do versioning, syncing, collaborative work, searching and general database-like operations when your filesystem doesn't support it. A lot of popular open source "not evil" software use that model, for example, IIRC, Firefox has always stored bookmarks in an opaque sqlite database while Internet Explorer stored them as files in a folder, and I don't think I need to tell you which one is considered the most evil.
$ grep FIREFOX ~/.bashrc
export FIREFOX_DIR=$HOME/.mozilla/firefox/q3m93qpe.default-release/
$ cat ~/.local/bin/ff-tabs
#!/bin/bash
lz4jsoncat $FIREFOX_DIR/sessionstore-backups/recovery.jsonlz4 | jq -r '.windows[].tabs[].entries[-1] | .title, .url, ""'Some OS builders have been trying to push this model on the desktop but have not been very successful due to legacy. It only seems to be done with mobile apps running on the desktop (Windows with Android apps and macOS with iOS)
I also lament this model, I think we should always have full access to our devices. But there's a strong commercial push for lock-in and subscription models.
Non I've seen do a proper job of mounting as a drive, NFS or SMB would be nice.
WebDAV is kinda that. I subscribe to a (super cheap) hosted version from Hetzner and I'm happy with it even though the server is on a different continent. As mentioned somewhere in another comment, I also use rclone mount to use it, but there are other options as well (the client built into GNOME is reasonably good, the client built into Windows 10 is terrible).
While this fosters new ideas and opportunities, that conversion into long-term direction and execution can be pretty rough.
Consider Google+ also. They had plenty of reference points for that. It was just not very good.
Google tends to abandon development on new projects way too quickly, before they're ready for mass appeal. They probably expect it to explode like Gmail did. But I think they forget they lost a lot of promotors like us when they abandoned the don't be evil thing.
At the time I recommended Gmail to everyone, now I try to get people away from Google :)
"We couldn't figure out how to market it" was from an interview straight from the people who made it: The companies they pitched to responded with things like "why would we need this, we have email".
The first part of it is not true (but the second part is true, in that you can access chats from Gmail as well).
I just checked, and Google Chat exists as a separate app on iOS, Android, and web (chat.google.com). Even though the web version redirects to a gmail url (mail.google.com/chat), it is still a separate web version, not embedded into gmail. The only thing it shares with gmail is its subdomain, the page itself has zero mention of gmail or any gmail-related functionality.
I used to have this app installed, then one day it started telling me I had to use the GMail app for chat instead.
I wouldn't be surprised to hear that they were rolling the out the migration to accounts incrementally, or they've reversed course. But I deleted the google chat app because it wouldn't let me use google chat.
This is true today, but 3-4 years ago Spaces was actually a completely separate product that was so bad that most Googlers didn't even know it existed.
I have never met a single person, including on the Spaces team, that used Spaces for anything. Ever.
A decade and a half of instability: The history of Google messaging apps
https://arstechnica.com/gadgets/2021/08/a-decade-and-a-half-...
How do they keep failing over and over, in the same way, when these failures are a well-known embarrassment?
You'd think the CEO would either just ban development of new messaging apps or pick a winner and give it the same kind of care and feeding as GMail.
Killing a chat product that has always been a net loss for the company might not appear to be a failure.
Doing that once, yeah, doing that over and over, no so much.
At this point the smartest thing to do it just kill them all and definitively say: "No more, this is the end of Google chat apps."
What do you mean? It's always good move to kill a product that loses money for the company.
Why do they start so many new products that are losers? That's a different question. Possibly the incentives for starting new products are poorly calibrated, which is the common narrative. But possibly they aren't and they are quite happy to make a mountain of shit, including doing almost the same thing over and over, for the small chance of a diamond. That's entirely consistent with the rest of the start-up industry.
> What do you mean? It's always good move to kill a product that loses money for the company.
Look at the Ars Technica article I was responding to. Google has failed at chat apps repeatedly for a long time, to the point where it's an embarrassment AND they've created a self-reinforcing vicious cycle of failure. The smart decision for users is to avoid any Google chat app like the plague, because it will inevitably be killed, which means those apps will also fail for lack of users.
A business doesn't have infinite tries to get something right. Eventually they burn up all their credibility. They should kill them all, and stop developing new ones, completely exiting the space.
The question is what their strategy is. Obviously they know this too. The idea that nerds on HN know what is best for the company's brand or reputation is dubious at best. Google is (allegedly) one of the most valuable brands in the world, and more trustworthy than Microsoft or Apple or Amazon (https://www.zdnet.com/article/facebook-tiktok-least-trusted-...).
Sure us supernerds will guffaw and chortle into our neckbeards to one another and deride Google for killing lots of things, but how much does that actually hurt their bottom line or brand value?
Google.
* create stealth startup to compete
* wait???
* pounce on userbase when Google crashes and burns
* profit
Angular 2010-2021? Seriously? Angular was simply superseeded after 11 years by better technology, and same goes for a lot of other projects and services on that site.
I’m not a Google fanboy and I know about the few instances where they screwed over users by turning stuff off but cmon now.
These days I greatly prefer to own my data more directly, which I accomplish using a NAS. I don’t use Dropbox OR Google Drive, I use Syncthing, which can do stuff that neither of those could ever dream of in terms of syncing between machines.
Not only do I not want Google Drive, I don’t even want anything like it.
That said, I totally get why it’s still important. I do use Google Drive at work and for that type of use case I would not argue in favor of self-hosting. Syncthing is great for a single person or even a fairly large group of people, but not for a big organization. Not to mention you probably use Google Docs or Office 365 anyways, both of which are integrated with their respective company’s storage offerings. Syncthing won’t give you directory sync, docs integration, or granular ACLs.
But that’s actually perfectly fine, because I don’t need any of those things. Hell, I flat out don’t really want them. You could probably get something similar with ownCloud which I did try running, and I’m not even particularly enticed. I’m sure Synology has a full suite as well, but it’s not what I want out of my NAS. Part of being happier with my technology was just as much realizing what I didn’t want as it was realizing what I wanted.
As far as permissions go, I guess I am mostly ignoring granular permissions. I’m generally only concerned with permissions on a per-share basis. This way, file permissions mostly don’t matter and can be a one-and-done affair. Then, I can split files up into separate shares.
NFS also seemed appealing, but honestly it does seem to bring a lot of complexity that SMB/CIFS does not, and everything supports the latter well enough for my use cases.
Remote access is another issue, since you should probably not do SMB over the internet, but I guess that can be solved with Tailscale or ZeroTier.
server min protocol = SMB3 smb encrypt = required
to the [global] section of your smb.conf.
That said: for the moment, I still have some SMB1 clients inside my network. Yes, this is terrible, and frankly it is surprising Samba still supports it (I won’t be too surprised or too angry if it gets removed now that it’s separated out.) I should probably segregate SMB1 clients onto a different fileserver on a different VLAN or something. Until then, it would be annoying to force SMB3 on all clients.
(As for why Anybody would use SMB1, the answer is retro computing. I also suspect that Open PS2 Loader is an SMB1 client, though I have not checked.)
--without-smb1-server
and the SMB1 code is no longer compiled into the binary :-).
The problem with seafile I had was if I tried to sync down files and the sync client stopped, it would have no problem uploading the incomplete downloads as changes to the server, as well as pushing missing as deletes to the server.
I didn't lose any data over it, thank god, but I moved away from that.
The moral of the story is sync is a hard problem. And corner cases need to be tested before you rely upon any sync service for your data.
this surely does not seem like a 'corner case' and would be absolutely terrifying if true.
could you be more specific in regards to the circumstances?
Why are you running Ubuntu in Proxmox, and not directly on the metal?
Thank you! I might set up a NAS in the very near future.
just use raid1 if you are not in a position to write an academic paper on the differences.
I'm sure the team would like to support Linux, if they had a bigger budget, if they could get out from all of their tech debt, if they had one or two specialized engineers added to their team. But the giant corporate behemoth isn't even remotely aware of those problems, nor will it care. There would have to be a clear case that adding a Linux client would create real business value. But considering how small Linux users probably are, and without an additional revenue stream to offset the cost of development, it's a non-starter.
- The specific actions of a company will _always_ come from a specific subset of people who are interacting with and sometimes constrained by the broader organization. Should this mean that we can never talk coherently about the actions or statements of a company?
- Isn't the company overall responsible for creating the organizational constraints and obstacles? Why is that team that size? Why is that their remit, and how has it changed since _someone_ decided to say "hang tight"? Why do they have the staff they have?
You're right about the motivations: they must not view it as creating real business value, since that's how these decisions get made.
My only point here is that it IS a choice someone in an office is making: it's not worth it to support Drive on Linux. They do know about it, because it was on their radar 10 years ago. Somebody took it off their radar.
Demonstrably not.
To propose otherwise implies that all similar actions/statements made in a similar way are not made by the company, which is both ludicrous and also causes Google to not exist.
How long since Google said a Google Drive Linux client is coming? - https://news.ycombinator.com/item?id=24183399 - Aug 2020 (109 comments)
How long since Google said a Google Drive Linux client is coming? - https://news.ycombinator.com/item?id=9434643 - April 2015 (59 comments)
Insync works well, and it's 50% off for a couple more days: https://www.insynchq.com/ Not affiliated, but $15 is not a lot to pay, as opposed to waiting for something that probably won't happen.
Rclone has support for Google Drive, and it's open source: https://rclone.org/
There's a command line client that uses a push/pull workflow: https://github.com/odeke-em/drive It was written by a member of the Google Drive team.
Gnome supports Google Drive, or at least used to, directly in Nautilus. I don't use Gnome, so I can't comment.
There may be other options I've missed, but the point is that there is already good support in multiple forms. I'd be interested to know what support Google could provide that's not already available.
I'm using it right now. It is not particularly responsive (perform an operation and it looks like no operation is happening until the remote end responds) but works well enough for copying files in and out via nautilus if you don't expect instant feedback.
I don't see how 1st party could do any better than the handful of 3rd party options.
This bit sounds interesting, could you explain it a little more, how you managed it?
Is there something that's, say, 1% as complicated as nextcloud that can sync multiple devices without giving a cloud provider access to cleartext (while also keeping an encrypted backup in the cloud)?
There's got to be a reason that Google Drive isn't interested in implementing linux mounts -- I'd be surprised if it is market share alone. The reason may be interesting/unexpected.
The other interesting question is: what are Linux users using instead? Because that's Google's competition for this software, which implies how much they could make competing.
- Data shows not a lot of customers use Linux
- Which leads to poor Linux support
- Which leads to not a lot of customers using Linux
That's not specific to this either. I've seen companies discount continents of people due to lack of customers on those continents, which was often due to inanely easy-to-solve problems. But if you only have 0.1% of your market in China, why bother with servers which are inside the Great Firewall? If 0.1% of your population is Spanish-speaking, why bother with i18n?
Data driven companies have an inherent bias towards /current/ customers, and away from /potential/ customers....
Linux could change it if every single copyright holder consented, but that's even less likely than the FSF adding a field of endeavor restriction to the GPL. Google owns partial copyright to Linux. Other copyright holders are gone, and nobody knows who the heir is.
Google is a pretty big Linux contributor.
Now, it's a "our project is already dead", automated banning hellscape, where the only assistance is from other yet-to-be-AI-killed fellow user attempt to barely help. Or gods forbid, find someone here for a "social media escalation".
I used to care. I gave up caring WRT google ages ago. They are not worth the trouble.
If you already have an AppEngine App you can always keep it and create a CloudRun app to handle the WebSocket part and they communicate well.
[1]: https://cloud.google.com/run/docs/triggering/websockets
And happy that there is software from Filipino developers and software out there!
Ripcord doesn't have the best UI and I can't open Teams calls or even links that are too long (Slack broke their API at some point) but at least it's not wasting my ram and it's responds quickly. The rare times I have to open Slack on the web I sigh and wait 2 minutes for the app to load.
I automated Prituln VPN via bash just not to have to deal with another Electron client for a damn popup on my tray bar.
I just like, upload and download files as needed to Drive? I don't really have a huge "need to sync files to the cloud" quadrant in my personal or professional life.
There's a Google-wide shared file system that is mounted on all gLinux workstations, but it has nothing to do with Google Drive.
Alphabet do want data from the masses, not much from a small cohort of smart users. Invest for them is not interesting.
There is missing context.
I’m sure that it’s not native, but what features would I be missing? I was very pleasantly surprised when it just kinda worked.
https://wiki.gnome.org/Projects/GnomeOnlineAccounts/Provider...
Last time I tried it, that only supported the virtual drive thing, so no actual files locally.
Look at Nest. Barely any useful new features at all in years. They could easily add more useful features like car detection, etc, but it hasn't changed in several years now. Google has become a graveyard where ideas go to die because no one cares there.
Source: https://help.dropbox.com/installs-integrations/desktop/syste...
But I understand the frustration and that some people do miss it...
Wouldn't I be paying for gdrive and thus expect to get working software? The free tier isn't enough for me and I would have paid if they had a linux client.
Its okay if they don't think linux is worth it, but your argument that i should pay for the software again seems really weird to me.
Google writes, maintains, distributes, and supports a working Drive client for Linux, called DriveFS.
https://github.com/thejinx0r/DriveFS
It is the top hit for "google drivefs source code" on google. "google drivefs" just returns hits about clearing a cache on macos or something.
The duck duck go results are arguably more relevant, though one is a link to DriveFS.exe, which I don't plan to download.
At least they should sponsor someone motivated to write a client for it.
Perhaps Google sees no benefit in building a service in a space where users can self serve making it.
Chrome and all of Google's services make being in Linux-land way easier...
Targeting a locked down OS like Windows or Mac isn't too difficult, because those compromises can be carefully implemented in a way that avoids accidental data corruption but doesn't negatively impact user workflow too much. On a Linux system, there's hundreds if not thousands of configurations that folk would expect a filesystem to work in, and so it's a lot more difficult to strike that balance.
Years ago, I remember a coworker dragged a directory around on his MacBook, and it completely flattened the company's entire drive directory structure.
For dev environments, you can workaround it using Windows + Drive filestream + WSL. https://github.com/microsoft/WSL/issues/2999#issuecomment-91...
I'm not saying this is the right move, but I'm explaining why this project hasn't been funded.
At my current workplace I use a FUSE driver to get Google Drive in my file browser and I wouldn’t have it any other way.
There is FileStream or whatever it’s called, but it’s never coming to Linux so who cares.