Hacker News Reader

Comment by mmjaa | Hacker News Reader

Dropbox has been, for me, a most disappointing outcome of the whole Internet'ization of computers.

I truly believe that Dropbox is a lousy solution to a problem created by ignorance on the part of the OS vendors. There is absolutely no good reason that the computers of today cannot share media seamlessly between themselves, without requiring the use or involvement of an external third party - the only reason its necessary to have Dropbox now, is because OS vendors fell asleep at the wheel and forgot that they're producing Operating Systems.

I think this will change, though - OS vendors are going to catch up to this market, and Dropbox will become less and less important as the OS starts to do its job again, and lead on these kinds of services.

View on HN

I think you're being a little naive about what really drives features from an OS vendor here. That $vendor_A could enable seamless sync with $vendor_B is absolutely true, but the realities of competition and the market mean that technical feasibility isn't the driving factor.

I can't imagine Apple or MSFT ever offering something that would truly replace it - anything either does is likely to privilege their own OS over any other, and Dropbox has no reason to do that.

The other reason that services like Dropbox shine, at least for me, is that by providing a cloud-mediated sync solution, you don't have to coordinate a synchronous conversation between the devices in question. That's valuable, especially since it enables easy sync with > 2 devices (which also opens the door to selective sharing).

That always-there aspect is also how Dropbox became the de facto mobile file system for iOS (and Android, maybe? I'm guessing -- I'm an iOS user). It's also the default mechanism for sharing files too big or numerous for email (e.g., Aunt Sally's holiday pictures).

My tl;dr is that while Dropbox's initial value prop might not be earthshaking, it's done what it set out to do very, very well for me since I started using it 8 years ago, and I'm very glad to pay them for their service.

This is a head scratcher for me. First: doesn’t Microsoft offer OneDrive which is functionally equivalent to Dropbox? You seem to be saying MS doesn’t have incentives to do something they’ve already done.

Second: I realize everyone’s experience is different, but the three times I’ve had the misfortune to interact with Dropbox shared files (once in a corporate setting, twice with files shared by acquaintances outside work), it just didn’t work. So the meme that Dropbox is a reliable mechanism for sharing to others confuses me.

Yes, I agree. There is what I would call a little more hubris than necessary in this side of the argument.

After all, it was once totally conceivable that a multi-million dollar company would be founded on the basis of, basically, a device driver.

We're still doing that sort of thing, its not a new phenomenon that some crack in the design/methods/ethics of a technology vendor is exploited by some un-controllable third-party. Gone are the days when Microsoft, by way of example, would have waged war and just built-in local- and remote- peer discovery, rsync++, etc. in response to this daring challenger to the hegemony.

I guess, I really do sort of miss that. Dropbox really sucks. I only use it because my friends do.

Just for clarity: do you mean it sucks politically, e.g. for privacy reasons? I'm familiar with those objections for sure, but if you're saying it sucks at some other level I'd love to hear why.

It sucks from the perspective of the usurpation of agency in pursuit of luxury, which in my mind has led to decadence.

Um.

In what way is Dropbox more guilty of this than any other tool/service provided by a third party?

Sure, they have a work-alike, but the original comment seemed to suggest that this sort of thing should just be built in at an OS level without a "product" or add-in being required to support it. I think, but can't quite articulate why, that building a competitor to Dropbox (which both Apple and MSFT have done) isn't quite the same thing.

Obviously your experience is your own, but I have yet to have a Dropbox sharing event fail for reasons other than user error (typically by unsophisticated relatives or coworkers, so I'm not suggesting at all that this was the root of your problem).

We should probably both be cautious of overgeneralizing from our own experiences with Dropbox, but the general presence of "it's easy and it works" even among the nontechnical suggests that your experience is atypical for some reason. Did you ever figure out the cause?

>There is absolutely no good reason that the computers of today cannot share media seamlessly between themselves, without requiring the use or involvement of an external third party

If you're talking about some sort of decentralized peer-2-peer, the average non-sysadmin non-techie user isn't going to want to leave their laptop up 24/7 as a node to share files.

The lawyers with the legal MS Word docs, and accountants with their MS Excel spreadsheets want to shut down their computer, drive home, or get on an airplane, or go to a meeting that may not have wifi connection.

If we ask users to save it to some sort of store & forward storage so their files are accessible to others (especially with others who are outside their particular company's boundaries) ... well... you end up reinventing an entity like Dropbox.

>f you're talking about some sort of decentralized peer-2-peer, the average non-sysadmin non-techie user isn't going to want to leave their laptop up 24/7 as a node to share files.

The trouble is, precisely this argument can be made about every single layer, upon layer, upon layer, of cruft which requires a lot of things to be in place, just to make Dropbox work.

So, I'm sorry, but I don't think you've entirely thought this argument through. Dropbox should be a built-in.

Why isn't it?

Not OP but: I would love it to be built in. However, neither OneDrive nor iCloud currently work with the same ease of use and seamless function as Dropbox. And Dropbox (as with Google Drive) works across OS. So, my work computer is a Mac, and my Home Computer is a Windows PC. My Phone is an iPhone. I can access all of my files, anywhere, from any of these devices.

None of this would be un-feasible with a built-in service on any of those OS's.

Microsoft has been moving in that direction by bundling OneDrive with Windows 10 and making Office save files to OneDrive by default.

Unfortunately, OneDrive is little more than a half-baked clone of Dropbox. The first thing I do after setting up a Windows 10 machine is uninstalling OneDrive.

What about multi OS. You think a Windows pc, a MacBook pro, and an android phone will all sync files at the OS level? Really?

Yup. Technology to do this has existed since the birth of filesystems. This is a non-issue.

It doesn't matter if the technology exists (which it does) what matters is if the vendors will cooperate (which they won't).

If there was a standardized protocol/system for "cloud syncing" of files it's entirely likely they would all support it. Kind of like how they all support SMB/CIFS now.

It's not a protocol issue.

A universal protocol would only handle the "data in flight". The other issue is "data at rest" which is storage problem.

As analogy, SMTP is a universal protocol and yet email giants like GMail/Hotmail/Yahoo exist. They store emails. Average users don't want to point DNS MX records to their laptops running home mail servers 24/7.

- Dropbox-the-client-software simplifies "data in flight".

- Dropbox-the-cloud-storage simplifies "data at rest".

Since storage costs money (e.g. see Amazon Cloud Drive drops "unlimited" plan), it means there would be conflicting business interests from OS vendors at that infrastructure layer. This why a company that was _not_ Apple/Microsoft /Google/Redhat such as Dropbox was able to fill a gap.

There is nothing in the world wrong with me wanting to serve 4 terabytes of content off my own local machine, and know where it is going, and to whom, without the involvement of any third-party at the content layer. That is a feasible, realistic, and tangible goal in the current state of things. Dropbox has a lot of data. They're not a file-system vendor; they're a neo-google.

Yes, thats precisely my point. Dropbox is only relevant because the OS vendors are no longer servicing their end-users needs.

I'd like to see you substantiate that claim. I see no evidence that it's true.

https://en.wikipedia.org/wiki/Rsync

You can't seriously be claiming:

1. Rsync has existed since the birth of filesystems.

2. Rsync will seamlessly sync between devices in the way that Dropbox does without the use of a third party. It clearly does nothing to overcome NAT.

It seems like referencing RSync does nothing to substantiate your claim, which still appears false.

They can & already do p2p files & audio/video/data using WebRTC(even Apple got onboard yesterday).

The problem dropbox tries to solve is allowing access to files when either party is not connected to the internet or to grant publicly share a set of files.

These are totally operating system functions. They should not be vendor'ed by what, I think personally, is a "Pepsi"-Co.

I think that if businesses can build their organiations on top of the Dropbox platform then they should do well. For individual consumers I'd love easy family/share services. You can probably already do this but nothing is really compelling.

I believe it should, and ultimately will be, a built-in. The Internet is going to route around Dropbox, eventually.

Reply on news.ycombinator.com