3,344 karma · joined August 2, 2011
I'm still a bit leery of this idea of fundsourcing, because it sounds like a way for retail investors to fund the most dangerous and riskiest round, and then once a startup has traction, VCs can come in with a lot of money and dilute the initial investors with impunity.
Is this a possibility, or are there ways for these initial investors to protect themselves?
Physically though, my back tells me otherwise... and usually wins that argument :)
I google just about any HTML issue I have, so I don't spend too much time trying to learn the intricacies of it. If anything, trying to understand why CSS was behaving as it was, was what got me closest to smashing my keyboard more than a handful of times. Stupid Wordpress.
But I don't think it's necessarily a waste of time for absolute beginners to learn and memorize HTML as their first exposure to "programming". Having a fundamental understanding of how making changes to a script or file affects the output to a web browser is something that I imagine most non-programmers aren't familiar with, so having a firm grasp of this concept will probably help them digest more advanced concepts later on. Being able to whip up a web page is probably one of the first things I would teach beginners if I ever taught a course on programming.
Personally I think C/C++ is still relevant, and it's my language of choice, but I'm not close-minded about it. For example, in developing my OCR functionality, I discovered how wonderful Python was, because it was a sane programming language, and allowed for much quicker development and progress than C++. I'm now a Python convert, although I still have many more miles to travel before I can become a competent Python programmer. In addition, Javascript is a lot more powerful that I imagined, and it looks like a lot of people are turning Javascript into something a lot more powerful, ex. Node.js. If I had to put my money behind the languages that might really blow up in the near future, it's Javascript and Python.
As to the career opportunities of C++, I can't really speak for that because trends change quickly. The key is to be flexible and open-minded, and not get stuck in a particular pigeonhole otherwise you'll find yourself like COBOL programmers in the late 90s. You will definitely learn a lot by going the C++ route, and being "closer to the metal" as some people say, you can and will learn a lot of different concepts that aren't available to many other languages, like memory-management, etc. It might make it easier to take up other languages in the future, as well, so it's not something I would discourage anyone from learning.
But I did come across similar things that the writer talked about. I first went about it by choosing Python/WebPy as my backend, and then switching to Wordpress as a front end for my website because I got tired of having to do every single little thing, and subsequently learning PHP. Then I started focusing on Javascript, jQuery and AJAX. My website is hosted on EC2, so I needed to understand how that all worked (not nearly as difficult). Then I tried learning how to make a Facebook app, so that ended up taking a few weeks, since the documentation is amongst the worst I've ever encountered in my career. I needed to figure out how nginx worked in order to get certain things working, and I also needed to learn the basics of Photoshop and Illustrator and basic graphics techniques. Luckily I have several friends in design that helped me get over this hump.
These last 8 months haven't felt like an exercise in learning how to program more so than 8 months of constant problem solving. Sure, I learned a bunch of new technologies, but my main goal was to get up some web-based projects, and every step of the way I encountered problem after problem, and my job was to solve these problems by any means necessary. Solving these problems entailed me having to learn Javascript, jQuery, Photoshop, Illustrator, etc. I didn't relish the idea of having to keep learning new stuff, but I did it because I knew it was the fastest way to solving my problems. I can't say I'm an expert in any of the technologies above (I constantly get Javascript and Python syntax mixed up in my head, and I google 80% of what I need), but I can solve just about any problem that I've come across.
I think this is the key to bootstrap yourself and doing anything related to technology, be it programming, sysdmining, etc. It's not so much the love of constant learning, but rather not giving up in the face of constant roadblocks, and constantly attack problems over and over and over and over again until you solve them. Don't get me wrong, I've very much enjoyed learning these technologies, especially Javascript which is a wonderful language (despite its flaws), and so much more powerful that I ever thought. But I know I've only started the scratch the surface.
As well, another big factor is keeping a high level of emotional energy in order to keep pursuing your goals. One of the things I was working on was using Tesseract to do OCR on uploaded image files. After 3 weeks of trying to get it to work, I achieved about a 97% accuracy, but because of some underlying bugs that I didn't understand, it wasn't good enough. I was frustrated and disappointed and finally gave up on it, and moved onto my next project, which I finished in about 2-3 weeks. But after that, with renewed energy, I decided to give it one more go at my problem, and spent another week at it. This time I used OpenCV in conjunction with Tesseract, and achieved almost a 100% accuracy, but more importantly, I figured out a way to get around the bug that had been dogging me on my first attempt.
I'm at the point now where I'm not a "great" web programmer, but I definitely can take a design from one of my friends, and implement it, pretty much to spec. I'm not fast, and I'm not creative, but I'm confident I can solve just about any problem I need to in order to finish my task.
The fact they got 23M shares of FB is simply delicious gravy on top. This means they get to participate in any upside on FB for free. The only thing that would suck is if they were somehow taxed on the value of FB shares when the deal went down, but I'm not even sure that would occur. I'm sure there's a way to structure the deal so that they wouldn't need to pay taxes until they sell the shares.
If, after 2 years since iPhone 4, they can't come up with a feature set that will push smartphones once again to the next level, then it likely means that the smartphone is for the most part commoditized, and I won't "miss out" on some great feature by exploring other manufacturers.
If they reach the point where it becomes a "faster-cpu-nicer-screen" type event, it means that smartphones have become fully commoditized, and it means the rest of the smartphones will catch up shortly, if they haven't already.
They've had 2 years since iPhone 4, so the expectations are pretty high, and most are expecting something jawdropping. This is the double-edged sword with being an innovator at the level that Apple has been over the last 10 years - people's expectations get bigger and bigger, possibly to unrealistic levels.
To be honest, I can't imagine what else Apple can do that would blow my mind... fingerprint scanner? Breathalyzer? If it's just another "faster processor, better screen, more flash memory", then people like me are going to be sorely disappointed. I'm hoping for the best though.
It seems like in the diagram, the West Coast Client, instead of making a direct connection to the APP servers on the right, is instead making a connection to the ELB on the left, which then forwards the traffic to the nginx server, which forwards it to another ELB, which forwards it to the App servers.
If the client connected directly to the ELB in front of the App Servers, they would incur the SSL handshake latency, but would avoid the four extra hops (two per send and two per receive) on the ELB and nginx.
Over the lifetime of the connection, is it possible that this latency could be longer than 200 ms?
Are you touching the code because you need to fix a bug in it? If so, then yes.
If you need to maintain this piece of code then you might as well fix it to the best of your ability and knowledge, but please be careful.
Are you touching the code because you came across it, and you don't like how it looks? If so, then no.
It's a shame that unreadable code exists, but you can't save the world. My mom used to tell me "Don't sit in someone else's pile of shit." Meaning, if someone else made a mess, and if you're the last person to touch it, you're going to look like the one who did it. There are bigger fish to fry in any software project than shitty-code-that-hasn't-been-touched-in-a-while-that-works.
I have two 2TB drives that I switch between every couple of months that I backup my ReadyNAS to. Because this data is basically my entire digital life (my photos, documents going back to about 1993, emails, etc), I don't trust a straight file copy, so I wrote my own simple bit-by-bit file comparison program. An engineer at NetApp told me all the horror stories that he's been involved with in terms of hard drive reliability, including things like the hard drive reporting to the OS that it wrote everything correctly, when it fact it didn't, so I've become pretty paranoid. The last thing I want is to make a backup, and then find that what was written on my backup hard drive was corrupt.
I have a couple of 500GB hard drives that contain some older archived data from 3+ years ago, so my older photos and documents have about 4 different copies. The only thing I don't implement, which is probably my biggest weakness, is that I don't store any copies offsite, so if there's a fire, I'm screwed.
Except for database and photos, my data growth has slowed down considerably. I would guess that my database data grows about 100 GB every year (I collect stock quotes every day), and my photos and videos are, on average, 10-20GB. This year was an anomaly because I went on a road trip where I used 2 GoPros to take time lapse photos and videos of the entire road trip, and that data itself is about 100 GB, and 99% of it is useless (and embarrassing).
All the Silicon Valley companies I've worked at give you automatic 2 x salary as a death benefit for free. This is without taking any sort of physical, etc. If you want additional benefits, you need to pay extra per month, submit to a physical, etc.
It seems like instead of the regular 2 x salary, Google is offering 5 x salary for free, which is really great.
Saying it's some sort of scare tactic is preposterous. That's like saying paying people extra is a scare tactic, because they won't be able to make the same amount of money elsewhere.
Here's an example of my question that was closed on serverfault:
http://serverfault.com/questions/342191/performance-of-sql-s...
I specifically asked why my throughout for my database was more than doubling when I decreased the number of vCPUs from 4 to 1. And it was closed as an "exact dupe" of a question asking about how to optimize a VM's performance. There's nothing to optimize in my situation, I had more physical cores than vCPUs, so there was no over-allocation. I get that obviously there must be some contention between the vCPUs, but to double the performance by dropping to a single vCPU? I wouldn't see that type of performance hit on a physical system otherwise multi-process databases wouldn't exist.
I had another question that was arbitrarily closed as a dupe, but I made a comment disputing it, and by some small miracle, the mod removed the dupe status. I can't remember which other question it was at this point since it was modified to remove my comment and the closing. The point is that there is no way to flag the question as being under dispute, and if the moderator decided not to do anything about it, then I couldn't do anything (except post to meta which I just learned today).
EDIT: I posted the wrong link, and just corrected it now.
But I do agree about the criticism of the moderators. I also think that there is almost a "Stanford Prison Experiment" level of power-wielding amongst some of the mods. There are examples of someone/me asking a question, and then someone closing it as a dupe, when in fact the question isn't a dupe. It might be superficially similar, but the mods don't take enough time to actually read the question and see if it's a dupe or not. And the rest of us are powerless to do anything about it.
But this is the minority of cases for me. For the most part, SO is a completely invaluable tool, and I love it immensely.
I think removing "Thanks" is a bit too abrasive. Maybe 10+ years ago, the NYC subway system tried getting ticketing agents to stop saying "Thanks" because they thought it would make things more efficient, and all it did was piss off the passengers. It's trivial, but there has to be a base level of civility between participants. I get that upvoting is "Thanks", but we're still humans (at this point), and trying to make communication more "efficient" by removing natural human responses makes it too terse, and almost insulting sometimes. It's a small thing, and I think it's better to just leave it in, because it's nice to say "thanks".
1) The target audience is definitely those people who are too busy to get their own groceries. If they are too busy to get their own groceries, aren't they too busy to cook a meal? I would think that take-out or going out to eat would be their #1 competitor in this area. When both my wife and I were working full-time, at the end of the day we were so exhausted that the last thing we wanted to think about was what we were going to eat, and then actually cooking it. We ate out probably 5 days out of the week. Ordering take-out will be quicker, and less of a hassle.
2) If the delivery person selects items such as meat, eggs, milk, etc that are expired or very close to expiry, it will likely cause a complete lack of confidence in the service. It will only take 1 bad experience buying meat, or a couple of busted eggs in a dozen, and they could potentially lose a customer forever. I'm not sure what their return policy is, but if they have one, it will likely be pretty costly for the company. There also could be incentive for people to return their close-to-expired food to Instacart, claiming that the delivery person selected it. It may seem petty, but for example very high chargebacks rates are one of the biggest challenges that Square faces right now.
3) Free delivery is interesting, but the 3hr delivery window I think will turn out to be pretty inconvenient for customers. It means that they need to deliver their goods at home, because who wants to be stuck at work waiting for their groceries to be delivered? I'm not sure what the delivery times were during their beta testing, that would be interesting to know, but my best un-educated, armchair-analyst guess is that the highest amount of deliveries will be from 7 to 10pm, and on weekends. If Instacart takes off, it means a disproportionate about of work required between 7pm and 10pm, and a lot less at other times. This sounds like a lot of people sitting around waiting for deliveries.
4) Safeway has their own delivery service as well, with a 1hr time window. What exactly does this service provide that Safeway can't? Is it cheaper? Or just faster?
I compare my experience with Newegg, and although I've bought some stuff from Newegg, I will generally go with Amazon, even if it's a few dollars more expensive, because I trust Amazon so much more, and the entire experience is so much better. I'm basically their model customer, and they have me hook, line, and sinker, especially with Amazon Prime.
The fact that they pieced together all this information from multiple sources, including Amazon's ability to add credit cards over the phone, to getting the billing address through domain name registration, to hacking into Apple iCloud really makes me feel... I guess depressed is the word.
We really have no control over our own data security. I've been super paranoid about things like identity theft, and I got my identity stolen, which is something I've been dealing with over the past 2 years or so. Somehow, my birthdate, addresses, etc were all wrong, and I had to jump through hoops to get it changed. As well, I currently have an unpaid credit card linked to my account, and the credit agencies and the collection agency won't remove it. The collection agency required me to submit 3 copies of my signature, a police record, copies of my identification, etc, before they'll remove it, even though THEY were the ones who made the mistake. I went to the police station to file a report, but they needed documentation that I didn't have, since I had already changed most of the information through the credit agencies. At this point, I froze all my accounts through the credit agencies, and I've given up.
The safety of my email, etc, is something that I also take extremely seriously, and now I'm being told that there's a possibility of being hacked via clever hackers piecing together information from various sources, each of which have different security procedures. We literally have no data security except "security through obscurity", meaning that the likelihood of being randomly hacked is low, but if someone wants your account, they can and will get it, pretty easily it seems.
The industry NEEDS to standardize on very rigid set protocols on things like what information they give out, how accounts are reset, how things like credit cards are added to accounts, what information they leak, etc. This is ridiculous.
But what if your website is secured behind an Amazon EC2 or Linode CSR? Isn't Instagram and Netflix run at least in part on EC2? I have no clue what the security schemes are for either of those service providers, but if they allow CSRs to change passwords, then it's the same thing. If the CSRs can be paid off, or fooled over a phone call, then it might be cheaper to just do that if they want to inflict potentially millions of dollars worth of damage to a rival.
Having the security of your entire business behind a single CSR or a cell phone is the equivalent of millions of dollars worth of Cisco firewalls being outdone by a $20 wifi-router plugged into the internal network.
One thing I do know, though, is that like you said, security is likely going to get tightened across the board, and that means that it's going to get a lot more inconvenient for all of us. I guess that's a good thing, but it will definitely impact the usability of these services.
If it means that all vendors will tie their services to a two-factor authentication scheme linked to our phone, well that might just stop me from using the services altogether.
Depending on how valuable the data is to you, it might be easier to just pay off a CSR, and then fake a phone call where you pretend to convince that CSR that you are that person. The person will get fired, but probably won't go to jail unless they can prove collusion. And then they can either find a new job, or depending on which country the person is living in, they can live nicely off of the money for a while.
I'm not sure how to solve this problem, except by having highly paid and specially trained CSRs that do the account resetting, or by never allowing resetting ever, and if you forget your password and your security questions, you're SOL.
I have to admit this only makes me more leery of putting anything on cloud storage, although my own personal data is pretty useless to anyone, which is my only saving grace. Others who are more important might need to think twice about relying on these types of services.
The problem is that they haven't been able to follow this up with another home run. They have all these users, and unfortunately their best way of monetizing these users so far has been ads. And even then there is anecdotal evidence that these ads aren't very effective. And even worse, their users are moving onto mobile far quicker than they anticipated, and because that space is unmonetized, they are losing a large source of income. They are facing the perfect storm of poor monetization of their current users, a mobile platform that is completely unmonetized, and quicker-than-anticipated migration of their current users to this unmonetized platform.
The Facebook desktop experience is a lot less sticky than they believed, and now they need to scramble to fix this gap in monetization, and until they do (and if they can), their stock will be under pressure. The movement of users from Facebook desktop to Facebook mobile is something that probably could have been picked up a lot sooner through their analytics, and something they should have addressed a lot sooner, instead of publicly scrambling now for a solution.
I thought 2-3 years ago that the logical next step for Facebook was to invest heavily in their Facebook app ecosystem, so that people would log into Facebook and stay there the entire day. They could have email, IM chat, world news, tv shows, music, word processing, etc all from various Facebook apps and keep their users engaged all day long. It would follow the natural progression of Operation System to Browser to simply Facebook. And with the iPhone App store as a model, maybe they could charge 30% and make loads of money that way.
But this never seems to have taken off, which seems like a strategic misstep. I find the Facebook app space to be relatively inactive, except for games like Zynga, but even Zynga's results show that Facebook app usage appears to be dropping. Because the Facebook app experience is relatively non-existent, it means that the only really useful feature of Facebook is the newsfeed and photos, both of which can be reproduced very well on the mobile. If they had a healthy Facebook app ecosystem, it might slow down people's migration to strictly mobile, and allow for better monetization of users as well.
As a side note, I've been trying to develop a Facebook app just to learn, and I have to say that the developer documentation is amongst the worst I've come across. Things have changed so quickly that doing Google searches is relatively useless. And even the latest documentation doesn't appear to be proof-read, lacks good explanations or examples, is confusing in many places, and some links don't even work. Thank God for stackoverflow! There are a lot of weird loops that have to be jumped through as well, which feels hacky, but I imagine there are technical reasons for it. But still, it feels like the entire experience of developing an app should be cleaner and easier. Part of it is probably because you can't host your apps on Facebook servers, so all this indirection causes technical issues, so maybe the solution (if it even matters anymore since app development doesn't seem very popular) would be to offer their own hosting service.
People have already seen the demise of MSFT's Zune products, RIMM's Playbook, etc. If I'm going to choose between an iPad or a Surface, and they are the same price, will I really gamble on the Surface when I know it might not be around 1 year from now? I know the iPad will be around. I can't say the same about the Surface, and if it's gone quickly, then I will kick myself for not buying an iPad.
This is something that the Amazon Kindle Fire did right. They priced it low enough so that there isn't as much risk, and their Kindle brand makes me believe it will be around for a while.
Of course not everyone can become an Olympic athlete just by trying. And of course not everyone will be the next Steve Jobs or Zuckerberg no matter how hard they try, either. That's not the point.
The point is that most people don't even bother trying, and if you try, if you even just show up on a regular basis, you can probably reach above the 50th percentile in anything.
I would bet that almost anyone who puts a concerted effort into any endeavor could get to the top 25th percentile in just about anything. Hard work and hustle can bridge a lot of gap that natural ability and luck may create.
But if he is somehow just now coming to some epiphany that hard work can't trump luck and/or natural advantage (either thru ability or status), then my answer is: "Welcome to the real world." You need to temper any goals you have with maturity and realism. I'm a terrible basketball player, but I'm sure if I spent 1hr/day every day playing pickup for 1 year, by the end of a year, I'll be better than at least half of the regulars. But if my expectations are that I'll be able to dunk and beat Michael Jordan, I'd be delusional.