I was among the naysayers. I first met Drew at pg's house just before Dropbox did YC. I listened as he explained Dropbox, and I immediately thought of a dozen reasons why it would be very difficult and probably fail (I'd recently worked on something very similar for a month or so just to figure out whether it was a direction I wanted to go with my own company, so had some familiarity with the scale of the problem...I also knew the allure of the simple parts of the problem).
I don't recall a whole lot about the conversation; I thought Drew was smart, and he seemed to have a pretty good understanding of all the problems he was going to have to solve. But, I still had my doubts, and walked away assuming Dropbox would not be one of the success stories out of that upcoming YC batch. We see who from that conversation is now a billionaire (or will be in the coming years)...so, it seems I was wrong. Or, at least, overly pessimistic about Drew's understanding of the problems and his ability to resolve them.
And I'm an example of how you can be right and still wrong because you let "other plans" get in the way. I'm on that thread with my two cents: "I'm impressed."
However, Drew took me out to lunch about 10 months later. They'd just signed their Series A with Sequoia (although it wasn't public knowledge), and were looking to hire their first couple employees. Aston had just been hired the day before as employee #1; Drew was wondering if they could interest me in being employee #2.
At the time, I had my own startup that I was trying to get off the ground with a friend from college. I could tell Drew was wicked smart, and he was way more prepared and ready to bring a product to market than we were. But my immediate response was "I can't leave my cofounder, and I want to see where my startup ends up going." He was very understanding, and said "Yeah, if I were in your position - when I was in your position [DropBox is Drew's second company] - I wouldn't give it up either." I couldn't help thinking, as I walked away, that maybe I'd just passed up the best opportunity I'd get in my life. They filled their positions with a couple MIT grads later that day.
A couple years later, after I'd moved out to the Bay Area and was working for Google, I saw Aston at Maker Faire, and mentioned that maybe I'd made a mistake in passing on DropBox. He said, "Naw, you made the decision that was right for you." And in hindsight, I think that's correct. I had 3 dreams coming out of college: one was founding a startup, another was inventing a programming language, and a third was learning how Google Search works. This way, I got to try my hand at all 3 of them. And I'm not sure DropBox would've succeeded the way they have if they'd hired me: they got some kickass early programmers instead, and I had some growing up to do before I was really ready to start something big.
I like your story, and your humility. I think that Aston is right. At the time, it was the best decision for you, based on what you knew, and wanted to do.
Nothing prevents you from inventing something spectacular at Google, now :)
Quite an interesting story. I guess I am a few years behind you, as in I just graduated a year back and get offers from startups which i accept/turn down. I often wonder how these startups would end up and where I would be relative to them. But I guess what makes the most sense here is the part that says "you made the decision that was right for you.". That kind of seems like the right line of thought.
FWIW, during that time period (07/08), I talked with about 10 startups about joining as a founder or early employee. DropBox is the only one still in business today. So I think that the Paul Buchheit post that Joe linked is pretty accurate: "How to be right 90% of the time: predict it'll fail." It's just that you only need one success.
I've had other reasons for staying at Google as long as I have, though, and I wouldn't necessarily consider it a failure.
From what I've seen, 'naysayer' is a term people most commonly use to ignore and suppress critical thinking. It's very convenient, because it sidesteps the question of why someone says 'no', and blames it on their personality instead, making all kinds of 'no' (substantiated or not) equivalent.
Besides, people who can't handle criticism are usually the same people who can't properly deal with new ideas either. Have you never seen large corporations going under because of abandoning their main products and chasing something new and shiny? That's a case where some intelligent no's would save the day.
it may not be of much comfort to you, but sometimes naysayers are important. Even if they are wrong about the overall trajectory, they may in the process of naysaying point out specific weaknesses that the visionaries might have missed, and, thus, ultimately allow for the success of the product.
Careful! A lot of would be Drew Houston's end up discouraged by naysayers. As a founder, it took me a while to know who to listen to. One of the toughest things about doing a startup is people telling you you're wrong all the time. One of the things YC (and pg in particular) gets right is that they're very encouraging of potential for success.
Perhaps this was your problem. Who is not important, what they are saying is.
You're being told you're wrong? Why? Do they have a point or does their argument break down to simply 'because' or 'because no one has done it that way before'?
If they have a point, at the very least, they probably just pointed out a problem you need to over come in the future and you can thank them for expanding your understanding of the situation.
> Do they have a point or does their argument break down to simply 'because' or 'because no one has done it that way before'?
You'll find that the vast majority of arguments that people give you have a point. Look no further than the arguments provided against Dropbox here. The real difficulty is knowing which points are more important than others.
For example, whether developing a desktop program has the potential to be the next big startup in part depends on whether or not you believe the desktop market will still be around in the next some odd years. There is no 'right' answer to that question unless you happen to be able to see into the future.
> A lot of would be Drew Houston's end up discouraged by naysayers.
I think this is one of the more troubling aspects of entrepreneurship. Most of today's really big, wildly successful tech companies sounded like really dumb ideas when they were first launched. I wonder how many Googles/Facebooks were never founded because their founders were discouraged by naysayers??
good point, not everyone has skin as tough as me (I'm in the sciences, where the most searing crucible is not your colleague's criticism, but the cold, silent, emotionless failure of an experiment).
I do get a kick out of all the complaints and snarky comments. Sometimes you just have to do one thing very right, and everything else takes care of itself.
I'm reminded of this Quora post on the popularity of Dropbox:
"Dropbox: Why is Dropbox more popular than other programs with similar functionality?
Well, let's take a step back and think about the sync problem and what the ideal solution for it would do:
There would be a folder.
You'd put your stuff in it.
It would sync.
They built that.
Why didn't anyone else build that? I have no idea.
"But," you may ask, "so much more you could do! What about task management, calendaring, customized dashboards, virtual white boarding. More than just folders and files!"
No, shut up. People don't use that crap. They just want a folder. A folder that syncs."
To be fair, in retrospect everything seems obvious.
At the time I'm sure the Dropbox team had to make a bunch of hard decisions. For example, without knowing the success they would later have, it doesn't seem super obvious to have "one master sync folder" as opposed to being able to sync various folders around your computer (a la not-so-successful Live Mesh and perhaps other sync solutions)
Read the comments there. Now come back. This is why you don't ask engineers for business advice. I can't tell you how many times I've come up with an idea I think is great, go to work, and talk to my buddies in firmware. The first thing almost all the time out of their mouth is basically "why don't use solution x, in addition to y, which will basically give you the same thing" where x and y are great technologies, but kind of hack to accomplish what you're doing. It kind of always kills my energy.
Giving advice is cheap, and deceptively easy to make sound wise. I've found when you want to bounce an idea, you need someone smart, who will tell you what you're missing... but also be open to new ideas. A lot of people lean on either side of that line. Engineers for some reason tend to lean on the pessimistic side.
>> 1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
This type of response is so typical of an engineer and so out-of-touch. Not to pick on it, but it's just absurd, and shows zero ability to put oneself in the shoes of the layman. And yet these types of responses are all over the tech community. This disconnect is the exact reason someone like Steve Jobs is able to enter product categories very late but still blow away the competition.
Came here to post this. No, seriously, exactly this. That was my favorite quote out of the entire comments!
FTP with curlftpfs and SVN/CVS isn't "trivial" to 99.9% of the population!
It's the same reason that sites like SlashDot were filled with all sorts of reasons why the iPad would never work. Why would you use that?!? You could do the same thing with ______.
Dropbox is definitely a case where a single person's vision was required to create a revolutionary product. Judging by the comments, leaving it to HNers as a group would have just resulted in a faster usb drive.
"For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. "
Actually the thing that's infuriating about Netflix and many of its competitors is the weak or no support for deep linking to content. We don't need channels to curate content if bloggers, etc. can do it themselves.
I think the technology and its usability wasn't the problem.
The problem was, nobody thought people would be so crazy and store their stuff on the other side of the planet (performance) on the servers of complete strangers (security).
Dropbox doesn't really suffer from the first problem since you keep a local copy of everything and people have been trusting complete strangers with their email and bank account details since forever.
I almost choked on my lunch laughing after re-reading that. I remember reading the original thread and actually agreeing with him, without the slightest chuckle. Im amazed at how most people still find it hard to find pictures in their filesystems on their pcs. But think about it, how many people carry .jpg, .jpeg, .png etc file associations in their heads? Hilarious.
> Dropbox is definitely a case where a single person's vision was required to create a revolutionary product.
In the case of Dropbox, I think the main factor was (and this is the exception rather than the regular case) the perfect technical execution, not the vision.
I think a far more insidious fact is that Dropbox fills in a gap left wide open by incompetent OS developers. Its the 21st Century - the OS should be providing such features inherently, without requiring a third party involvement. Maybe HN'ers, at that time, were somehow conscious of this fact, and thus declined to 'see the revolution' that was headed their way. Either way, since there was no other option, Dropbox was "revolutionary" enough to have gained a following.
Dropbox support several OSes though, which is still better than any one OS providing a similar capability natively. In general, the OS should just do what it NEEDS to do, and let applications fill out the rest.
The fact that there are multiple OS's out there means there is a battle going on. Dropbox serves a need that should be being defined by co-existing OS participants. But we see this everywhere in the Application space, so your point is valid, there is just an OS border, and an App border. I think what I'm trying to say is that border moves around a bit, and I'm cranky about that.
Apple has started to do that with iCloud [0] and to a lesser extent, AirDrop [1]. But aside from Apple hardware, iCloud only supports Windows laptops and desktops — no official Linux, Android, or Blackberry support.
Though they have been very successful, it's a pity that in my experience they have turned into something of a big co in the way they deal with customers and quite sneakily hide important technical limitations from hackers who might want to use them.
After a year or two of happy premium-paying use, I noticed dropbox was using 100% of my CPU. Some googling suggested this was due to having too many files. Ok, fair enough, perhaps there are technical limitations meaning indexing >300k files is tough (very easy to get to that count if you're keeping open source codebases on DB), so I move files out of dropbox and clear its cache. After a week of constant 24/7 100% CPU usage and dropbox failing to update anything, I contact customer support and get sent copy + pasted boilerplate telling me to do what I've already done.
After more than one email to say 'I've done that, what next?' I get told it's due to symlinks in my dropbox folder. I have several in node_modules folders, and have never had a problem with them before, so I find this weird but remove all symlinks from my dropbox folder. No change after several days.
I try deleting files on the web interface - it refuses to do so for a folder with a large number of files in, and tells me to use the desktop interface (great...)
Also throughout this dropbox repeatedly overwrites work files while I'm working on them (thankfully with backups.)
At this point the customer support tells me how to delete my account if I'm not happy and they simply stop replying to my (polite) emails.
Googling around it appears this issue has existed for at least a year and a half, and yet there is very little mention of it (there's a bulletpoint hidden away on their website) nor does the interface warn you about it at any point. How hard would it be to at least add a notification like 'looks like you're adding a lot of files, please don't add too many more or I might stop working'.
I used to hold up dropbox as a great example of a YC company that was technically innovative and something of a hacker's company, but this experience has left me quite massively disappointed.
I think the success of the Dropbox depends on the common people wanting to simply share/keep their files and not necessarily if someone can keep 300k+ files in there.
For me as a hacker, I have at my disposal ( due also to higher income ) more powerful tools or hardware ( local/private NAS, higher bw, Amazon AWS, etc) real tools ( rsync, versioning, can customize/choose the filesystem, etc ) to achieve the same thing.
Ok, that's fine, so what is there to stop dropbox from popping up a warning when this limit is exceeded? Or for there to be some means of deleting things from the server side without having to manually click through to delete N files at once (where N << total number of files)? Or customer support that doesn't drop the conversation completely out of the blue?
Having the system break irrevocably at any stage is totally unacceptable when you're dealing with people's data.
Presumably you'd have a problem if gmail stopped working irrevocably because you happened to have too many emails? Would that argument stand up for you then, or would you feel aggrieved you hadn't been warned and had no means of fixing the situation?
In some ways, it reminds me of Apple's attitude towards their Pro Users - these users make up less than 1% of their customer base, and their revenue stream (and effort/attention) is focussed on the other 99%. So, because 99% of the users will never run into this problem, they're going to put minimal/no attention to resolving it, while focussing all their attention towards the 99%.
80/20 rule. You have to do the least work that would bring the most value to the busienss. It doesn't make sense that they allocate that time to fix an issue for the 1% instead of 99%. I think this is just a logical business decision.
Good service/support is about recognising that sometimes your customers are not typical and accommodating them if at all possible rather than just serving the lowest common denominator.
And good business means providing good support and service and infrastructure to the 99.92% of your customers who fall within your statistical norm while still functioning for the other 0.08%, even if you don't bend over backwards to 'accommodate' that 0.08%'s wishlists.
Ok, so just pop up a warning in this situation. It's been a year and a half at least they've acknowledged this problem, so prevent somebody from having a non-functioning product - wouldn't take much effort.
The problem is they hide this problem away and in no way warn you, I fully accept that they might not want to prioritise fixing this, I just ask for common courtesy - I had a month with dropbox basically not working at all. Popping up a warning isn't bending over backwards.
I wonder what would happen if this occurred on gmail - you have too many emails so suddenly it stops working without warning, completely, irrecoverably. Would you still make the 99.92% vs. 0.08% argument? Or would you find it a bit off? At least warn + give a way out...
My friend, I don't believe you realize just how outside of the norm you are. 300k files is quite a large number of files for even a local file store to choke on.
But there are actually two things that put you outside the norm. First, it is likely only technical users who are going to acquire 300,000 files in their DropBox. Second, technical users are already typically aware of the downsides of storing tens of thousands of small files on a file system and how common it is for even dedicated local servers to choke on them. We all have copied 300,000 small files and have it take f-o-r-e-v-e-r and then copy one file of the same aggregate size of the 300,000 small files and it flies through the copy. Why would DropBox be any different?
I wonder what would happen if this occurred on gmail - you have too many emails so suddenly it stops working without warning
Here's an experiment for you to try. Write a program that sends you 60k byte emails as fast as it can. When you get up to 300,000 emails see how well your Gmail account works.
I'm not sure how your response is actually addressing the suggestion to pop-up a warning. If the company is aware of the issue and there is no technical reason not to implement a warning, even if it would help out only .1% of the users, then why not do it? I would understand if it was some large feature and they didn't want to waste resources, but it's unlikely the case here (undoubtedly they keep track of how many files you have synced and they already have the ability to do desktop notifications).
As I said, very easy to achieve as a developer (I hadn't done anything crazy), and the symlinks were working up to that point.
Even as a non-typical dropbox user (a developer, then), I think there should have been some kind of a warning a, some ability to recover from the situation b (I was left for nearly a month with a completely non-functional dropbox), and not suddenly getting dropped without warning by support c.
If I'd known up front that a deployment of not that unusual amounts of code (especially if you use git) would have led dropbox to irrecoverably fail for me, I wouldn't have used them in the first instance.
I don't think even as a 'non-typical' user the very poor, typically big co, experience I had was justified, esp. given that I am a member of the group of people who first started using dropbox.
Not to dismiss the problems you have encountered, but I too use dropbox heavily with 100k+ files. Not as much as you do but still quite a bit more than the average user.
I had tried out other common services (Google Drive, Skydrive) and they all choked with that large of a file collection. Skydrive would just crash and exit whereas Google drive had issues with file uploads. And both of these had terrible CPU usage (it was basically at 100%) which is not good for a laptop. This was a couple of months ago, so may be their syncing client have improved since then to deal with large file collections.
Dropbox on the other hand was the only one able to handle my humongous file collection with no sweat. But as you point out it too has its limits. It's good to know that once one gets past/at 300k+ files, there are likely to be issues. I agree with you it will be good for Dropbox to publicly acknowledge the true limits they have in their service.
> indexing >300k files is tough (very easy to get to that count if you're keeping open source codebases on DB … Also throughout this dropbox repeatedly overwrites work files while I'm working on them (thankfully with backups.)
To their credit, it sounds like you should have been using source control all along.
> To their credit, it sounds like you should have been using source control all along.
I reply to this only so people don't think I am negligent enough to not use source control!
Yes I use source control, but there are these periods where things aren't committed, during which you can lose work if some process overwrites files. The same goes for disk failure, unwise use of the rm command, etc.
To their credit that a file you are currently editing is overwritten with an old version by their software, but hey at least that software made a copy? Are you joking?
> 1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
Trivially use three different systems to simulate some type of not quite automatic syncing.
Why do Linux users often claim something is "trivial" and then go on to list obscure commands and software packages that have to be tied together in just the right way? To me that's "possible to do", not "trivial to do".
I suppose you have to come at it from the mindset of those actually using Linux. That is, if it already exists and can be glued together using open tools (as opposed to having to code it yourself), it is "trivial", since it has already been done before.
Intentionally or not, your comment seems to stereotype "those actually
using Linux" as unable to appreciate the value of the time and effort saved by a prepackaged solution. While that certainly happens (and is more common in the Linux community than in many others) it is not universal. The popularity of Linux distributions like RHEL and Ubuntu among developers and IT people speaks to that.
> and then go on to list obscure commands and software packages that have to be tied together in just the right way?
Which one of these is actually obscure? curlftpfs is just a remote mount tool, SVN/CVS/FTP - I don't think I have to say anything about them given we are on hacker news. And seriously, setting up a repository on a remotely mounted folder is a challenge?
If I would want to answer your post with as much polemic as you put into yours I'd ask why Windows users always need a specialized malware and adware infused tool just to solve simple tasks such as batch renaming a few hundred images.
Also, you have to remember the context of that comment. Dropbox was just some obscure third party client back then that you couldn't even get from your package manager.
Don't get me wrong, I do see the use case of Dropbox, especially for those without a technical background. But I don't understand the perceived reverence Dropbox gets in this thread.
Well, I have noticed that Linux users tend to approach problems selfishly, i.e. "How can I solve this problem for myself?" and the solution is trivial in such case.
However, the question is actually "How can this problem be solved for all people, including my mom?", and when posed as such, the solution needs to be much simpler.
If you're already familiar with those tools, the incremental effort to put them together is trivial.
It's no different from saying that driving to the store is trivial: People who don't already know how to drive would disagree, but it doesn't make the statement invalid.
This is probably the disconnect. My experience with Linux is generally limited to servers (headless). I only have one Linux box at home and that's for XBMC. So for someone like me that has a workable knowledge of Linux but not every package that's around, these types of solutions are non-obvious and non-trivial. A matter of perspective I guess.
data's stored on s3, and encrypted before storage -- there'll be another
option to enter in an additional passphrase (or private key) when installing
in order to encrypt your data before it leaves your computer (kind of like
what mozy does.)
It is sad to me that this never came to pass. I guess the desire to offer a web interface overrode the idea of encrypting the files before they left your computer. Not that you can't use your own encryption, but having it built in would have been great.
How exactly could you "transfer" files that way? You could just upload something to Dropbox and share your logon credentials and any form of dedup or DMCA won't help.
See the StakOverflow question for more details but the basics of it is that in order to optimize uploads the DB client calculates the hash of a file before uploading it and if the hash along with other attributes about the file like size etc... match then they assume it is the same. So you used to be able to trick the service into letting you download a file you never actually had if you knew the hash and other attributes.
It's also amusing that the top comment on this thread is a big rant about one specific customer service incident. Totally unrelated to the linked story, and as usual, not seeing the forest for the trees.
I strongly dislike the strain of negativity that seems to be typical of hn these days, but...
I don't think it's totally unrelated, as I always very much admired dropbox as a company and technically, and had the sense of them being something of a hacker's company (an hn thread in which the founder is replying to technical issues definitely gives one that impression), and I was very disappointed to have that experience. I rant here because they refuse to listen to me, I have tried to contact them a lot privately but have been ignored.
I'm not the only one to have a similar experience - just a tree in the forest of developers who are likely to experience similar if they try to put open source projects in their dropbox.
I don't want to belabour my points, so I'm deciding not to continue too much further in these discussions (I've come to find internet arguing is somewhat pointless regardless of the rights + wrongs of the issue at debate), however as I have repeatedly pointed out, it's very easy to hit a number like that if you store code in dropbox, esp. open source projects.
And like I've also pointed out repeatedly, fine don't consider my use-case important, but add a warning or at least don't drop supporting me because I'm too much effort, or if you are, say that's what you're doing.
If I had known dropbox wasn't suitable for developers (it having originated from yc with a technical discussion on that original thread suggests otherwise) - I wouldn't have used it in the first place.
I had used it as a developer for a long long time before hitting issues, then it become unrecoverable even after a month of patiently trying to fix it, and new work files were getting overwritten by old ones throughout. That's not an acceptable failure mode, nor is then being ignored without warning by support (at least say 'sorry you're too much effort'.)
I think that if pg can't even tell for sure what's going to be successful and what's not, really how surprised can you be that the common HN commented couldn't tell what's going to be big.
If you're a startup and you're pretty sure there's a market for your product, people telling you their gut feeling really doesn't matter imo.
I wasn't on HN at the time but I do remember being an early user of Dropbox, and I remember being totally blown away on first use.
So I looked into my email to see when I signed up, it was 14 months after this post. I also found a gtalk chat log with the friend that recommended it, and it looks like my memory is quite wrong, I was just as skeptical as much of the linked HN thread:
>the thing about 2gb dropbox
>is i carry 6gb on my keychain
>and 8gb on my phone
>and i don't exactly trust them with important data
>also my iphone has shared folders that look just like any other computer on my network
>the keychain is kind of a hassle though and i mostly don't use it, i should probably throw it away
It is actually took place, later on when there was an announcement regarding Dropbox raising from Sequoia .
This was the first time I heard of Dropbox.
Those days building a product which did similar to what dropbox were doing, except that mine used any distributed version control it could find on a computer (I had it supporting git, mercurial and bazzar) and push to servers with SSH.
It was all automatic, built with python, and monitored FS for changes. Supported any number of directories, etc.
So I felt I have this great prototype which I considered starting working on this full-time, till that morning when I read the TC article and I realized it simply been done, and by people who now have $6M in their pocket to make it even more awesome.
Given the effort and dreams I built upon my own version, I remembered how I could not use dropbox for quite some time.
Honestly, I still think the idea of dropbox is ludicrous. There are many ways to share files, and sending them to a third party to host for you is the worst one of them all. Aside from the few people that really need to multiply their bandwith by many orders of magnitude, a simple file sharing server on their own PC or a server they own would do the job just fine. Besides, dropbox was just yet another iteration of online file hosting (I'm pretty sure rapidshare and megaupload predated dropbox by years), so if file sharing was going to blow up, it would have already, right?
And therein lies the true genius of dropbox. The technology itself had already been done to death; the key was to convince a critical mass of people that this was the solution to their problems. Or even better, convince them of a problem they didn't realize they had. Yet again we see that many times success comes down to the better marketer than truly game-changing technology.
(to be completely fair, their syncing mechanism was the best up until then, plus their add-free freemium model was likely the missing key to success in this space)
"a simple file sharing server on their own PC or a server they own..."
I think that's the part you're missing. What fraction of the general population can set up a file-sharing server? What fraction of the population even owns a server? What fraction of the population even knows what a server is?
DropBox faced this objection over and over again from the investors they pitched, and Drew's answer was variations on "But how many of them have made it easy enough for your grandma to use?" That was the killer innovation - making filesharing so easy to use that you didn't need to set up anything, just install DropBox and drag your files to a folder. And that's been the killer innovation for many startups - Apple, Google, Amazon, Intuit, etc. all let you do things that many other companies have let you do before, but they made it so easy that even the average Joe can do it.
That's the thing though, how many "grandmas and grandpas" really need a file sharing service for blobs of data? There already exists "better" solutions to share common file types: flickr/picasa for images, youtube for video, email for word documents, etc. The vast majority of people simply don't need an arbitrary file hosting service. Somehow they've managed to convince people they do.
On the other hand, if you give something away for free that seems to have value, people will flock to it regardless of need. I'm curious if dropbox is actually profitable. I know of many people who use dropbox that I wouldn't have thought would even have heard of it. But not a single one of them pays for it. Perhaps what we're attributing to success is simply a company being sustained by the VC bubble. Even in the face of the "success" of dropbox I don't see the service in its current incarnation being self-sustainable.
Grandmas and grandpas is a bit extreme. Just about anyone I know uses Dropbox nowadays because it's just simpler than the alternative. I even use it to share files w/ my wife when we're in the same house. I have a SMB share that I use for some things but when I have to send my wife some pictures I just put them in our shared Dropbox. It's easier and she can access it later on her phone too even. Convenience is a very important factor.
My grandfather interviews holocaust survivors and helps make small books of their stories. I set up a dropbox account for him to share transcripts of the interviews, and now he uses it along with the other volunteers. A half dozen retirees, with minimal computer skills, barely able to read English, all happily using dropbox. That really impressed me.
You say a bunch of stuff about how it's the same old same old, and ma and pa can set up a file sharing server and figure out syncing between all the computers they use.
The comments on TFA and yours all miss the point. You touch on it in parentheses at the end. It's easy. Normal people don't have an always on server they can administer and have reliably available. It is literally sign up, install tiny program, put files in a folder. Done. This isn't a problem people didn't have, it was a very real and common problem, hence the success.
Your "to be completely fair" was, for me, the reason Dropbox was a valuable tool.
I know that I can set up FTP to a remove host and store files. I'm sure I could write Automation scripts that looked for updated files, opened connections and uploaded files when I saved them. Then more to regularly open a connection, poll for updates, and initiate downloads. On all my computers, at least. I'm sure that'd work.
With dropbox, I don't have to invest the time to do that, or to keep it maintained. I don't have to think about it. And I have seamless access to these files on any machine I go to, as well as my mobile phones. It's simple, and as I haven't exceeded my 20GB of free space, it costs me nothing.
Yes someone else has my files. But if I bought a virtual server from a hosting company someone else would still have them, and if I used a single machine at my home, it's susceptible to the whims of my Internet provider, a modem/router failure, and the added complexity of getting and maintaining a static IP address to allow the connections (as well as spending more time to harden and protect that machine as it's exposed to the Internet).
Quite a few friends and family have put Dropbox on their personal machines at my suggestion. Because of a whole range of factors (trustworthiness, cost, business model, reliability, usability, design, ease of installation, features) I felt comfortable recommending it, and they felt comfortable installing it. The same could not have been said for Rapidshare, Megaupload, a custom FTP server, etc.
I have a NAS in my house and I still put a ton of important files on Dropbox. Before I had Dropbox I would suffer from pretty significant anxiety every time I went on a trip. What if someone breaks into my house? What if there's a fire? I would literally hide my NAS in my closet when I went away for more than a week. I don't treat Dropbox as a single source of backups but I do treat it as redundant storage to my NAS. It's simple and it just works without having to mess with a lot of settings.
Disclaimer: I work for Dropbox and I trust the people I work with with my files.
"a simple file sharing server on their own PC or a server they own would do the job just fine"
For your own sake, I hope you get more real-world experience before making investment decisions or building technology to sell. To you, somehow the existence of a massively profitable, multi-billion dollar company must be evidence of people's irrationality and not an unmet need.
Are they "massively profitable" though? I haven't really heard anything regarding their financials. They're well funded, but that doesn't really mean anything these days.
There's certainly a need, don't get me wrong. It's just that for most people (paid) dropbox is the wrong solution. And at some point the right solution will present itself and suddenly dropbox will see its market disappear.
As a private company the financials are private of course, but take this as you will:
"Dropbox’s ascent has been just as stunning. The 50-million-user figure is up threefold from a year ago, and it has solved the “freemium” riddle, with revenue on track to hit $240 million in 2011 despite the fact that 96% of those users pay nothing. With only 70 staffers, mostly engineers, Dropbox grosses nearly three times more per employee than even the darling of business models, Google. Houston claims it’s already profitable but won’t reveal margins."
But has the percentage of premium users stayed constant as they grew? I wonder what percentage of users must be premium for dropbox to be/stay profitable. Or perhaps what percentage of traffic must be "premium traffic". Either way, there are a lot of unknowns, but I don't think this market is sustainable. Like Jobs said, dropbox is a feature, not a product. They are able to exploit a market inefficiency for the time being. I don't see that lasting for much longer.
Dropbox's long term future rests on them evolving into something that isn't easy to duplicate or pre-bundle. And even then, unless dropbox pivots massively, the market just won't be as big. File sharing simply is not a multi-billion dollar market.
(I can't help but wonder who I'm talking to seeing as you whipped out a throwaway for a dropbox discussion thread)
I assure you, Dropbox is laughing all the way to the bank!
Anyway, do you realize who uses Dropbox? Many are people in IT related fields and it is simpler and quicker to share files than asking their sysadmin's to setup an FTP or allow file-sharing or use VPN etc. It just works and you don't have to get into all this nonsense of helping your friends configuring FTP clients.
The problem as I see it is a that "show HN" is simply "here what do you think of this". If I - as a sort-of web-designer - look at the page and see a problem with alignments, lack of call to action, poor SEO optimisation or something then I want to offer that up; that's what I think.
Generally HN isn't the best forum to offer these more trivial thoughts - you really want to say "what do you think of this concept" or "what do you think about how I've implemented {concept X}".
Usually then if there's something that looks wrong but is not appropriate for discussion - like a spelling mistake - I'll take a glance to see if there's a simple contact form and use that instead. Emailing seems too much trouble unless I'm hoping you'll ask me to do some work for you(!).
Ugliness of the launch page seems quite important however and as it's subjective it's a point that could be discussed or on which the votes would help gauge the importance of the point.
If you feel the details are fine, or don't need work, then ignore the comments. Ditto for this comment.
When all you have is a launch page and no functionality demonstrating the features of the site, the points you mention are what many people are likely to focus on.
Not to be that guy, but since no one else has mentioned it: a file in Dropbox is a file shared with the NSA.
I was a happy paying Dropbox customer since 2008 but downgraded my account to the free tier a few months ago. I no longer consider Dropbox trustworthy for anything except (1) trivial files and (2) files encrypted client-side before they're put into Dropbox.
Even with the above, I had two specific use-cases that only recently did I resolve:
- 1Password Sync. Dropbox is no longer necessary here since 1Password natively supports iCloud sync across Mac and iOS.
- Arbitrary file-sharing between Mac and iOS. Dropbox is no longer necessary here ever since I've been running BTSync[0], which has worked flawlessly in my experience.
This is a perfect example of why you should ignore the naysayers and "experts" on HN when you first pitch your idea or product. Also, a great example of why the best startup ideas look initially like bad ideas (http://www.paulgraham.com/swan.html)
Let's be fair though, there's a survivor bias here. It's probably more dangerous to blunder on and simply tune out any criticism.
If you look at the top rated comment on that thread it is certainly a valid criticism, namely that corporate computers are often locked down and won't allow installation of third party software without getting sign off from IT. In the case of dropbox that managed to sidestep this by mainly targeting smaller businesses and individuals and also have the fall-back of letting users access their files over the web.
I remember being sceptical at the time (lots of people were). The consensus here on HN seemed to be that Dropbox was trying to solve a problem Microsoft had already tried to solve a million times (file-sync) and the fact that Microsoft had never been able to sell it and get a decent user-base for their service was proof that this wasn't something people actually wanted.
Funny how a good enough implementation and good marketing managed to turn that around.
I don't remember what was wrong with Microsoft's solution, but I remember not buying into it.
> I don't remember what was wrong with Microsoft's solution, but I remember not buying into it.
It might be the fact that it's Windows-only. I use Windows exclusively and wouldn't touch SkyDrive because it didn't have clients for other operating systems.
> I use Windows exclusively and wouldn't touch SkyDrive because it didn't have clients for other operating systems.
I don't understand your point. If you only use Windows what is holding you back from using SkyDrive? (Except that it is US hosted and you must assume all that data is public, but Dropbox is no different there.)
Because you might want to share files with somebody who isn't using Windows or you might want the ability to switch platforms yourself in the future without worrying about having to migrate all of your stuff.
I love this flashback, thanks for sharing. Between being a reminder that group thinking isn't the best thinking, no matter the quality of the group - and encouraging me to get out there - fail and be criticized. . . Great.
Group thinking is not bad, as long as we're defining group : (small) set of people with lot of common characteristics, goals (if applicable).
The opposite of the "mass" which usually is a large set of people whom highest common denominator is to be human.
The question is whether HN can be considered a group or a small mass =)
Plus you have to keep in mind the average pov will be biased in a different way than let's say marketing101.net.
There's people who try everything that is new and enjoy innovation, and there's people who tend to be skeptical at first. The first kind is busy trying out the software while the others are busy expressing their negativity on a website.
And there's of course those who find it a big shame, that Dropbox and other cloud services have become completely unusable thanks to the host country's government.
The top two comments are critical (though not mean).
The first comment is pretty interesting.
'My suggestion is to drop the "Throw away your USB drive" tag line and use something else... it will just muddy your vision.'
He's more or less correct. 'Like a USB' is a bad analogy. Dropbox only replaces some of a USB's use cases and does lot of things that a USB doesn't. OTOH, he's wrong because there is no other 3 word sentence that could have done a better job. 'Like a USB' is probably the best starting point even if it only gets across 25% of the message because 25% is better than nothing. 25% (assuming it's the right 25%) might get the user to install it. Then they might get to know the backup, file sharing/sending, versioning, or whatever subset of functions they use.
The comments here are a clear proof why founders should take feedback from different sub-set of users. For example, HN audience is build up of programmers, their feedback alone is not what founders should focus on. Dropbox is used by almost everyone I know because of the convenience it provides and the problem it solves. Most of the what commenters here are saying is not something layman users care about; they don't have any technical skills or time to learn to build their own custom sync-systems.
I wonder how many people that initially thought/said that they wouldn't need/want/use Dropbox are actually using it now. Would be particularly interesting to know what reasons convinced people who didn't believe it would work to end up using it themselves. Particularly, would be cool to know if it was due to lack of understanding of the product, lack of a clear enough pitch, or something else that established the wrong expectations.
When first shown it by a friend, I didn't want to use it because I was ideologically only using open source stuff, and servers I have control over. I carried on using Unison and some dodgy shell scripts.
Some years later, I got a free 60Gb Dropbox account with a new Samsung phone. It seemed a good way to sync my music and photos, and to let me access text files on the phone.
I'd meanwhile got fed up with my hand made scripts, and tried a few other commercial backup solutions, and got fed up with them all. So Dropbox seemed not a bad part to a backup strategy to use from my desktop as well.
I think this year's privacy news on the Internet has shown my original view was correct. And Drew has shown that a mature, usable, cross-platform syncing client is nevertheless something I use!
And therein lies a fascinating fact about Dropbox. Drew made the tradeoff between convenience and security/ownership one that convenience won out, particularly for a lot of people that from the outset were not expecting it. I think Dropbox's adoption in the smartphone world has been one of the most ingenious partnerships. I had pretty much the same experience in that I really only started looking at Dropbox as a serious solution to me when my Samsung phone shipped with a free account (+ I got a whole lotta free space through my university finishing high in the space race).
I'm not a daily linux user myself, so what particularly interests me is what Dropbox's adoption rate is like in the linux world. This seems to me to be the toughest nut to crack, since most linux users are power users in the sense that they're savvy enough to setup their own Unison/rsync etc. system.
My conclusion after thinking about Dropbox more is that it's a testament to the fact that to the majority of people convenience wins out over most other things (so long as the execution is bang on).
As a linux user (though I wouldn't consider myself a power user) I shopped around for a long time to find a cross-platform cloud backup solution.
After poking at a lot of the other options out there, Dropbox ended up being The One. The ease of use, lack of (unnecessary) IU (I'm looking at you, SpiderOak), and general un-buggyness after some time of trying out the free account sold me on it.
Since I started paying for Dropbox, I have become concerned about privacy and have encrypted my more sensitive data with Encfs. There have been some other companies started in the last year or so that I'm considering (Copy looks nice, and I want to see what appears based on http://camlistore.org/).
Roll-your-own backup systems are cool and all, but I've got shit to do that's not file management.
Maybe the most remarkable thing is how short the comment thread is -- but I suspect that is more an artefact of how much smaller the hn community was back then, than anything else.
Interesting how dropbox managed to succeed in an area with so many competitors.
Finally, I still think it's complete crap for me. But I also see how it's a great packaging and reselling of s3 -- and I'm certainly not surprised it took of (Not saying I necessarily would've bet on dropbox in 2007 -- but the sorry state of webdav in in windows left the market open for anything that offered user-friendly, secure cloud storage, and dropbox ticked (the most commercially important) two of those boxes.
edit: Ok, complete crap is too strong -- but it's a product I have extremely limited use for. While it is easy to migrate away from in the sense that it just stores files, it's not Free software (important for me for anything I use to store my files) and it has no privacy and questionable security (although dropbox+encfs patches up some of that). Still surprising that people didn't seem to see the commercial value -- I absolutely see that (much as I see how people would pay for google apps even if I never would).
Thank you for posting this. Having launched my own landing page yesterday (as Drew did in 2007), I thought that I should explain the positive feedback I left for Drew on Hacker News in 2007.
Launch posts were full of naysayers, even back in 2007 when Hacker News was called Startup News. I had done my own startup for 3.5 years at the time, so I wanted to give members of Hacker News useful, insightful, and positive feedback about their project.
The link was a landing page with a video about "throwing away your USB drive."
In my Hacker News comment, I proposed a list of ways to replace "throw away your USB drive" with a non-technical metaphor that people of all ages could understand (a personal secretary who automatically makes duplicates of anything you create for safekeeping, stored the way you want and accessible from anywhere.)
I've observed that newer Dropbox videos started to incorporate my feedback more and more. It's really been awesome to see Drew and the rest of the guys kick butt!
Sharp criticism from know-it-alls is often a sign that an idea is good. It's amazing to me how, even though this is common knowledge, good ideas continue to polarize. I'd hate for you guys to convince everyone to be so nice that we skew this long-trusted signal :)
So many people today are still harping on the security aspect (Not ~free~ software, you don't own the servers).
How does that matter at all when selling to a large consumer base? How many customers of dropbox know what those words mean?
Like so many commercial offerings you could built it from source and get some hacky scripts going on your own but 99% of the world isn't going to do that.
Regarding specifically the "free software" statement, It's a holy war. One where I'm on the side of the religious zealots, for once. The special extra fanatical wingeing about big software like DropBox not being free comes from seeing the founders as smart people who have a base ripe for promoting your (my) view (that free software is just plain better for the world), and if they just saw it our way, maybe things would start to change. So we complain. Well, I try not to outright complain, because I know that's a shitty way to get what you want and that my views are ultimately tangential to their cause, but you get the idea...
Thats fine and I understand the advantages to open source.
But if you're in the business of selling software (especially to end users and not b2b) like many people here claim to be, that isn't a valid concern or argument against a business model.
Another secret to success: have good parenting, DNA, education, friends, and an interest in creating things of value (as opposed to the other million things you can do with your time)
This is extremely subjective, taking the example of dropbox, most of the comments against it at 2007 were dismissive because "it didn't had enough value".
Dropbox doesn't replace USB drives for me. I still use them for sneakernet. Installing the newest version of Ubuntu or OS X. Copying a 2GB file to another computer because downloading it again, even over local wifi, would take too long.
What Dropbox replaced for me is the 'My Documents' folder. It allows me to access my files on my work PC, my laptop, my iPad, my iPhone, or anyone else's computer that has an internet connection.
There certainly are USB use cases which Dropbox flat-out doesn't account for.
At one point during the setup of my new gaming PC, I somehow hosed things to the point where I couldn't install Windows (motherboard BIOS update, maybe?). I was getting driver detection failures during the install process, my peripherals (keyboard + mouse) wouldn't work, no NIC functionality, my CD drive was unusable. I ran out to Walmart at 1AM to buy a 16GB USB drive/16GB SD card, tethered my phone (my only functional computer at the time) to my PC, downloaded the required ISO to my phone's new SD card, moved the ISO to a USB via a live instance of Ubuntu running off of another USB, and installed Windows via USB (and things magically worked from there).
Why would you want to keep your data "in the cloud", if the government uses the argument that as long as your data is on a 3rd party's servers, then it's not yours?
I'm still not using Dropbox. I absolutely detest the idea that I'm relying on a 3rd party for a feature that I consider should be built into every modern Operating System. At the point when OS vendors relinquished to the Web 2.0 Cult their responsibilities for such features as easy filesharing, the world lost something.
Absolutely, its great that people can share files this way. But absolutely, its terrible that it requires fragments of an OS feature to be distributed among multiple, external, unreliable entities.
I think the biggest OSes _have_ integrated something like this, but unfortunately they don't work well between operating systems. So if I'm on an Android phone, an iPad, and dual boot Windows and Ubuntu then it doesn't really help. Dropbox does allow me to backup and sync on all of those.
I would love a simple integrated file hosting solution built into OS X, where my computer is actually the host instead of iCloud.
I think, though, that the lack of high upload bandwidth for almost every Internet consumer makes this less useful today. However, if a big OS would actually develop and ship it, perhaps we'd have users demanding greater upload bandwidth and help push forward the idea of the "person as Internet producer", versus just the current "person as a consumer".
Umm .. because accessing such resources as files in a shared way is a duty of the OS, after all. Or, to put it another way, lets rephrase your question: "Why should OS vendors supply drivers for filesystem _X_?" "Why should OS vendors provide a way for users to share files, at all?" The answer: because it is the duty of the OS to provide such features.
Another thing: who says we need a Cloud model in order to deliver the same results as Dropbox? It could be just as easily possible to enable filesharing on a local workstation that is accessible to the public/chosen-users over the Internet, after all. The technology exists to implement these features entirely without relying on a Cloud at all: its called 'the computer in front of you'.
The "cloud" aspect of the feature is not actually necessary to deliver this kind of filesharing functionality - this is only a marketing gimmick that is used to justify 3rd-party, commercial access to ones information. Remember, if you're not paying for it you're not a customer: you're a product.
The features that Dropbox provide should already be in the OS. I create a Shared folder, I administer it with local OS features, I publish.
I believe that OS vendors were asleep at the wheel during this Web 2.0 cult phase, but I hope they wake up soon and relieve us of the security and privacy nightmare that is Web 2.0.
The thing that helped Dropbox take off (at least in my circles) was that it is on the cloud. I don't have to have a computer somewhere running all the time. In fact, I don't. Almost nobody I chat with regularly does.
I think the point is that your computer connected to the internet is your cloud storage. There's no reason why that functionality needs to be farmed off to a third party.
Besides vectorjohn's point, which is completely true, there's also a point about upstream speeds. Considering that most of us have asymmetric connections, sharing a big file with multiple people would be a pain if I had to upload it over and over again, while with a intermediary I can just upload it once and let it push to clients at greater speeds (and in parallel).
I don't recall a whole lot about the conversation; I thought Drew was smart, and he seemed to have a pretty good understanding of all the problems he was going to have to solve. But, I still had my doubts, and walked away assuming Dropbox would not be one of the success stories out of that upcoming YC batch. We see who from that conversation is now a billionaire (or will be in the coming years)...so, it seems I was wrong. Or, at least, overly pessimistic about Drew's understanding of the problems and his ability to resolve them.
I refer to this pretty frequently to try to remind myself not to be the naysayer in the room: http://paulbuchheit.blogspot.com/2007/03/how-to-be-right-90-...