How To Implement Sign in DropBox

Yes it works via the Dropbox client with a local web server. It's probably very similar to the way you did it. Interestingly, it was also an intern project.

Go is quite a lot like objective-C, although it is even more C-like.  It is not dogmatic about object orientation, so the programmer is free to build a big shared dataset and have different threads work on it without copying it -- and it can do other such "dangerous" things -- without strange contortions.  Apps that work on massive datasets must be built with the assumption that memory is precious.  Go is pretty good for this, although the tradeoff is that the programmers have to know what they are doing.So, you combine the fact that Go doesn't have so many safeguards that eat memory and slow down the code, it is compiled to native machine code so it runs at least twice as fast as code running on the JVM, and that the average Go programmers is -- quite likely -- better at writing performance-oriented code than the average Java programmer is, and you have a bunch of good reasons to use Go.I also wonder if they are using GPUs to do some of the work.  The GP-GPU libraries like CUDA are callable in C.  Go can call C directly, without penalty.  Java cannot do this.  When using JNI, lots of memory must be copied between the JVM and the native machine interface, and this really screws the performance of memory intensive apps.Perhaps a better question is: why did they use Go instead of C?  C takes off the "training wheels" completely, and an ace systems programmer using C can usually write the fastest code of all, quite possibly with some inner loops done in raw assembly.  So the management must have decided that the last, marginal speed improvement of hiring an ace systems programmer wasn't worth the cost or difficulty of maintenance.

This video has been linked as an answer to another Quora, but answers your question.  In short, Dropbox handles their own traffic related to permissions and notifications.  AWS S3 handles the pure storage and user file traffic.

Yes!For a smartphone wiki with Dropbox sync, try Trunk Notes: http://www.appsonthemove.com/tru...I'm working on a system that turns plaintext files in a Dropbox folder into your own personal wiki: http://WikiPackIt.com. You don't access it via Dropbox directly, but from a web interface with your own customisable subdomain. Markdown + Dropbox = wiki

I disagree with the consensus that you just need practice to become a great programmer. That goes without saying. That's like asking "how do I become a world class cyclist" and being told, "take more bike rides." You need practical experience to become a good or competent programmer. It takes much more to become a great programmer. Here are some of my suggestions. The primary thing you need is mentorship. You simply can't see your own faults and bad habits. This can come in the form of formal classes, with a professor or TA who corrects you. It could be from a code review at work. It could come from a similarly skilled peer during a pair programming session. It could even come from reviewing other programmer's code and seeing their mistakes (which you probably also sometimes make). We all have bad habits. Everyone that is pointed out so you can correct it makes you a better programmer. You could find something to improve in anyone's code.You should take some time to read the classics. This article is a good place to start building your bookshelf: Programmers Don't Read Books -- But You ShouldAlso here are a few of my own favorites:The Mythical Man-MonthThe Design of Everyday ThingsThe Pragmatic Programmer: From Journeyman to MasterDesign Patterns: Elements of Reusable Object-Oriented SoftwareEventually you will need a specialization. Computer Science and even just Software Engineering are broad fields. No one can be an expert in it all. It would be like a medical doctor knowing as much about eyes as an ophthalmologist and at the same time as knowing as much about feet as an podiatrist. Whether a language, an industry (biotech, finance), a platform (windows, apple, mobile, web), or a sub-field (usability, security, localization, quality, data, networks, performance), or some combination -- you should have things that you are better at than other things.At the same time, diversification is also necessary. If you are a world class Java programmer, I would bet that you also know other languages. It gives you perspective. Helps you make implementation decisions. Generally makes you a better programmer. I would say any "great" software engineer should have a little experience in an OO language, functional language, compiled language, scripting language, low level language, high level language.I'm no security expert, but I still made an effort to learn the basics, go to the occasional conference (they're also just fun), and keep up with what's new in the field. I have a much stronger interest in usability. I'm an engineer, not a designer, but I still eat up design and usability publications with as much fervor as if I were one. It also takes a bit of passion. Software is a fast moving field and it takes effort to stay on top of it. From just the new and trendy to the new standard way of doing things, you're going to need to read off the clock.Know more than average about Programming Languages. PL is just another sub-field in CS, and some engineers will have more interest in it than others. But IMO, knowing a little more than average about programming languages is a requirement to being a great coder. A great coder picks the best languages for the job. And to do that she has to know what makes it the best language. And she has to know how to take advantage of key features in any languages that she's using. Similarly, if there is any official "Guide" to becoming a great programmer, it is probably a PhD in PL. To be a great anything is more of a journey than a destination. The more you know the more you will be aware that you don't know. I would be suspicious of anyone who called themselves a "great programmer". There are a lot of similar questions here on Quora that might give you some more ideas. Best of luck on your journey to greatness!--------------------------------------------------------------------Update May 26, 2015 -----------------------------------------I feel compelled to give a reply to Aideen NasiriShargh's mention of my answer. This is a question that deserves different point of views. As we both give reference to, greatness is an elusive thing. Aideen says, "I don't call myself completely qualified to answer this question". I say, that greatness is more of a journey than a destination and "I would be suspicious of anyone who called themselves a 'great programmer'" (meant to imply that I also don't call myself a great programmer).However, since he goes out of his way to reference and misquote me, I feel it merits a reply.Aideen and I are coming at this question from slightly different angles and experiences. Specifically our answers differ...(1) First, our take on Formal Education. I have great value for my Computer Science degree. I feel it complements my practical experience and allows me to write code at a level that I would never have signNowed with practical experience alone. Based on Aideen's third point, it seems he did not share my positive experience. (2) Second and more relevant, who is asking this question / reading this answer. Something certainly up for interpretation. While I value my CS degree, I'm not assuming the reader already has one or would benefit from starting one. If you have a degree in CS, then you are probably already heavy on the Theory side and in that case more practice is probably the best prescription for you (and you probably know that). On the other hand, from the sense I've gotten, the audience here is has a lot of people coming from the self-taught experience. In that case, you probably want to beef up your theory. Combined with my own theory heavy background, I felt more qualified on giving some tips on filling in that side. I think one needs both to signNow "greatness". I grant that my list is more theory centric than Aideen's and that this is certainly not what everyone needs. Again, this is a question that deserves multiple answers. Again in where we are coming from, since there have been so many comments comparing our answers. When I answered this question, there were a dozen answers that simply said "you need practice" and "you need passion".  When Aideen answered, my answer was the top answer by over 1,000.I took the position that practice is a requirement for "competency", but should go without saying when the conversation moves to "greatness". I put passion toward the end of my list for two reasons. First, while it is worth a mention, we hear a lot about following our passion and I didn't think I needed to emphasize it further. Second, "have passion" is pretty useless advice. For the second part of my update, Aideen NasiriShargh misquoting me...Aideen quotes me with:I don't want to look like a jerk, but the fact that thousands of people upvoted "PhD in PL is the best Guide" and "The primary thing is having a mentor" just blew my mind off.PhD in PL...My original answer mentions diversification and specialization in different sub fields of Computer Science. Then, as my very last point, I mention that if one is pursuing a career in Software Engineering, then the sub-field of Programming Languages is specifically worth learning a bit more about. I give some reasons why I feel it is useful in the real (working) world. I have the impression that Aideen only skimmed my answer, but it also seems he skimmed the question. The question details are, "Is there any guide to becoming a great programmer?". My entire mention of getting a PhD is this one-sentence paragraph as an afterthought to my shout-out to Programming Languages:Similarly, if there is any official "Guide" to becoming a great programmer, it is probably a PhD in PL. I will now explicitly write out what I thought was implied in that statement. There is no guide to becoming a great coder. The majority of people should not pursue a PhD, but that is really getting outside the scope of this question. Mentorship...I've had a few discussions in comments with people about a better word for what I call "mentorship". I give 4 examples of what I mean by "mentorship" and where one might find mentorship. Not a single one is "having a mentor", as Aideen says I say. Perhaps I should have said "outside influence" rather than "mentorship". I just like the word better, and it is my answer :p I stand by this being #1 (when you already assume practice, as I stated I do). The next most popular answer after ours (at the moment, anyway) simply lists "1. write code everyday" and "2. hang out with other people who code". So despite it "blowing your mind off", it's not so radical an idea. In Closing, Dear Aideen,So, Aideen, I hope I'm not sounding like a jerk now. I think your answer is a great addition to this question, with the obvious exception of the first paragraph that is simply attacking and misquoting mine. It's different from my point of view, and that's a good thing. Again, this is a general question that could have a 100 useful answers. I've enjoyed reading the others. I certainly don't think mine should be the only one and I was (pleasantly) shocked by the large and positive (except, of course, for your) response to it. Your answer lists topics including passion, persistence, bravery, and trust. To me, this comes across more as a motivational speech than an answer to a question. Yours is better than the average expression of the same sentiment and does gets into a few specifics. However, IMO this idea has been repeated as nauseum and is not extremely actionable advice. To me, it comes across more as "this is what you should have been born with" than "these are some things you might not have thought of adding to your arsenal to help push you to the next level". So ditto to you, the fact that you received 3k upvotes "blew my mind off" (though not really, I would have never cared if you didn't first misquote me).And in case you actually care, which I doubt you do, slightly misquoting me in a way that completely changes the meaning of my words before then insulting that new meaning, does make you a bit of a jerk. Cheers!

In addition to some user authorization and authentication (for example using devise and CanCan) you could use Paperclip combined with aws-s3.Starting from creating very basic upload form once you setup the paperclip after uploading your file it should be automatically stored under specific location following given paperclip pattern. In order to have some basic security and keeping your files private. You can do very basic checking for given user_id matching user_id assigned to a given file. You would probably also like to have some share/make public functionality. That would be also easy to implement making just another column “is_public” default to false where you would update to be true and check during security check written above. In order to use folders I would recommend using this gem: acts_as_tree.It allows you to use methods, like folder.children to access sub-folders, and folder.ancestors to access root folders.That should be it. We could go deeper and fancier using some html5 properties to allow drag&drop but this is just a frond-end and backend should stay the same.Keep in mind Rails is just a tool - as long as you know what you are trying to do you apply the same almost in every language.

I’m going to answer this one from a tech executive perspective, and compare it to another highly popular IPO, that of Snap Inc (Snapchat).The key numbers:Revenue of $1.1Bn in 2017, up 30% from 201611m paying users out of 500m totalCash-flow positive: generated $330m free cash flow in 2017, up from $252m in 2016Flat ARPU: $111 in 2017, slightly down from $113 in 2015Gross margin doubled from 2016 to 2017, now at 66%, because cost of revenue (COGS) fell from $391m to $369m 2016->2017 despite 30% growth in top-line.GAAP loss of $210m in 2016 and $111m in 2017On the face of it, this is a very healthy business that it throwing off meaningful cash and growing well at 30%, which is very good for a $1.1Bn business. Remember: cash is fact and profit is a matter of opinion. Losing money on a GAAP basis is just fine when your net margin is 33% and you’re rapidly growing cash reserves.Strategically, the company is also impressive in its technical decision-making. Crucially, it completed a move off Amazon’s Simple Storage Service (S3) in December 2016, choosing to lease colocation space (data centers) and build out its own storage infrastructure. For most companies, this would be a mistake — but for Dropbox with an exabyte (one quintillion bytes)[1] of data to store, it is the source of the company’s doubling of gross margins in the last 2 years and impressive resulting unit economics. If the company was still using S3, we would have expected its storage costs (cost of revenue) to have risen in line with the growth of its usage — but in fact, it was able to spend less on storage in 2017 than it did in 2016. More on that in a moment.This is also an impressive feat of execution, because it was able to migrate one exabyte of data off AWS without users noticing, and without losing a single bit. It essentially duplicated the data first, and then began deleting it from AWS S3. It was also able to hire a team that had the technical chops to implement and operate low-cost storage at a scale most companies never get to. This is a level of technical challenge only experienced by the Googles, the Amazons, the Microsofts, the Apples and Facebooks of the world.A second impressive strategic decision was to start pruning its inactive customer base, which enables it to recover storage space from long dormant accounts. It didn’t have to add a single bit of new data storage to accommodate its user growth in 2017. That’s terrific. It’s also not repeatable, most likely, so there shouldn’t be the same expectation for 2018. But the decision-making smarts that this shows will serve the company well in the future.Also on the technical front, it is smart enough to know what it is good at and to stick to that: it still uses AWS for compute services, because that’s still a better choice for running its applications than trying to do so itself. It must be tempting to start running its own compute service. I’ll bet there are engineers at Dropbox loudly proclaiming at they should build that, because… well, it’s a fun technical challenge, they already have data centers full of servers etc. But they’ve been smart enough to ignore the Siren Call[2] of doing that themselves. That’s great discipline and very valuable to the company.In terms of corporate governance, this is a good IPO. Dropbox is selling ordinary Class A shares with one vote per share, not some cynically made-up equity class with no voting rights (looking at you, Snap Inc). Dropbox spends the most on the areas most likely to improve future results: R&D (engineering the product) and sales and marketing. General and Administrative is a fraction of these amounts and reasonable for a billion-dollar company. Again, this is not like Snap Inc which spends more on G&A than on sales and marketing because apparently it’s more important to have fancy offices, throw lots of parties and fly the founders and execs around in private jets and helicopters.What should we be concerned about?Growth. Growth is slowing. Dropbox spends almost nothing on marketing externally because it looks to groom its 500m user base and entice them onto paid plans. The good news is that it experiments extensively with digital marketing to find out what works. The bad news is that there may well be limits to how effective this can be, especially for corporate accounts where it has to be successful to grow ARPU signNowly. It might want to think hard about whether it wants to start marketing to corporate buyers and not just end users. Maybe run some tests on that?Product pipeline: the core product (file sync and share) is incredibly good, drop-dead simple and this ease of use has shown itself to be a durable competitive advantage. But Paper, a new document creation app? I totally get the problem statement articulated in the S1 that people spend a lot of time on the “work of work”. But it’s a totally different proposition for Dropbox to go from simplifying something users already know and love (files and folders/directories) to convincing them to adopt new cloud-based productivity software — Paper.Real estate. this is a far lesser concern, but apparently, it signed itself onto a long-term contract (until 2027) for its current HQ in San Francisco that it now doesn’t need because it’s moving to a new, larger HQ in 2019 with a $827 million lease commitment. ** Edit ** Folks who know better tell me this is just a fact of life for SF commercial real estate where large spaces are hard to find *** It can likely sublease the space and may even be able to sell on the lease.Overall, this looks like a solid IPO, especially when contrasted with recent over-subscribed IPOs like that of Snap Inc. Dropbox has a real, cash-generating business and demonstrates good decision-making skills.Footnotes[1] Exabyte - Wikipedia[2] Definition of SIREN CALL