This is Royrapoport's Typepad Profile.
Join Typepad and start following Royrapoport's activity
Join Now!
Already a member? Sign In
Royrapoport
Recent Activity
@joshhyde, I think the message here is actually quite the opposite from "killing the DVD." Creating DVD-only plans (and creating a business unit committed to DVD) signifies (to me) a resurging interest in, and commitment to, the DVD shipping side of the customer experience. (Note that I'm not in any way addressing the effective price hike) -roy
Irrespective of any current implementation issues, this is a really neat idea. I hope they're very successful with it. -roy
@Rocketboy X, I'm curious -- what's your issue with mockbusters such as Battle of Los Angeles? I mean, other than the obvious "they're incredibly crappy movies" angle?
@vio, it's worth noting that in some cases, sharing Netflix accounts is pretty much a requirement. Until Netflix makes it easy to switch between profiles/accounts on NRDs, it makes no sense for an NRD shared between people to somehow switch between two different Netflix accounts (because it's too painful). That's why our Wii, XBox, PS3, and AppleTV all essentially require me to share my account with my spouse.
Speaking as someone who has a desk next to a printer with a giant whiteboard as one of his "cube" "walls," I approve of this effort :)
@Enfuza, This article is frustrating because it's creating an impression that this is actually an issue that Netflix has, or that Netflix has advocated to resolve by way of the legal system. To the best of my knowledge, this appears to be a situation where the recording/music labels lobbied for this law to be created. If you read the article, you see that the author is using the case of sharing Netflix passwords as an example of something that would be prohibited by law, but that doesn't actually mean that Netflix was the intended beneficiary of the law. I think it's also the case that Netflix password sharing is done commonly enough (and the Netflix brand is popular enough) that it's the easiest way to explain the law and what's covered under it. So you can talk about the impact on Netflix password sharing (and given that this is hackingnetflix, that's probably a reasonable scope), but this law isn't driven by password-sharing issues with Netflix, nor will it (and this is IMHO, of course) significant impact Netflix (other than in cases where people think Netflix drove this misguided law and end up retaliating against Netflix, rather than the labels). -roy
I really liked both the old Friends feature and the old Facebook integration feature. Alas, at the time neither apparently merited much by way of maintenance and ongoing support, so they got crufty. I know, for example, the facebook integration app was working intermittently and at some point it can, actually, make sense to slash and burn rather than rehabilitate. Like it or not (I have mixed feelings on this, not least of which because I've got a family member at Facebook), Facebook is becoming the Lingua Franca of social connections on the internet; I do like the idea of using that platform rather than trying to maintain multiple independent friend lists.
We've got a bunch of NRDs at home: PS3 Wii Apple TV XBox Samsung BD-P1600 (Sadly, my Chromebook seems unable to play Netflix at this time; same's true about my HTC Thunderbolt and my spouse's Droid X) By far my least favourite is the Samsung, because its interface is so limited and it will never again be updated; The Xbox is still boxed, so I can't comment on it; similarly, the Wii is not ordinarily used for streaming. Our primary two platforms for streaming are the PS3 (upstairs) and the AppleTV (downstairs). I really like the PS3 for its capabilities (5.1, CC, etc). Because Netflix can push a new client to the PS3 whenever it wants, it tends to (in my experience) be the fullest implementation of all the capabilities our NRDP can accomplish. I'm not crazy about the current interface, mind you (and tend to have to go through interface testing because my account is marked as an internal testing-enabled account). The PS3 is also my primary Hulu Plus platform, so having both Netflix and HP in the same platform is definitely handy. I love the AppleTV for the ability to find more movies with the actors in the movie I'm watching, and for its noticeably cleaner and less cluttered Netflix interface. I also love it because it is my primary platform for renting movies and for watching previews (I've been known to spend an hour or two just watching previews, including for movies I've watched before, including previews I've seen before). AND, of course, it allows me to stream music and video from my iTunes library. If I had to pick one, I'd go for the AppleTV.
Agree on the Avatar front. The day Avatar came out, we (the UNIX SA team) went to see it as a company-sponsored activity. Win! The problem was that the person responsible for ticketing had managed to get us into the 2D showing, and we didn't know that until we arrived at the theatre (the lack of glasses made it obvious). When I left the movie, I called a friend and told him, IIRC, "if you don't have the money to buy a ticket to see Avatar, borrow it, get a paper route, or borrow it from the mob. And then go and burn it, so you can at least save yourself the two hours of hell that is watching Avatar."
@BP, I'm flattered. Sadly, I fear your faith in me is misplaced. Let's get the disclaimer out of the way and note that I don't speak for my employer (now, or ever on hackingnetflix). I don't want to address too specifically what Netflix is doing with Cassandra, partially out of my own ignorance, partially because I try to not disclose things that we may prefer to not disclose, and partially because that's not particularly interesting to me. I do want to address some other points, however. Firstly, reference: Ted Dziuba's NoSQL rant Ted's rant is certainly entertaining, but I don't find it particularly compelling. Some specific points I take issue with are: It's about a year old, and Cassandra 0.7 addresses his concern about downtime for column family definition changes. 0.7 is stable, BTW; His condescending message to DBAs ("In the meantime, DBAs should not be worried, because any company that has the resources to hire a DBA likely has decision makers who understand business reality.") is entertaining, but misplaced. I've talked to our DBAs. They're excited to work with Cassandra (and, in fact, we've repurposed some DBAs away from Oracle and onto Cassandra. This was an entirely voluntary effort on the part of the DBAs involved); I want to make a different case, though, and it's in reference to two comments y'all made earlier: NoSQL is in its infancy; and Man, Netflix sure seemed a lot more stable back in 2007. I think a lot of what we're doing -- and that's a pretty global definition of "we," which includes not just Netflix, but also Twitter, Facebook, and others with explosive growth -- involves having to either manage technology that is in its infancy or build new technology and capabilities (and if you're building something new, by definition it is in its infancy). And I think that'll inevitably cause you problems. A slight detour: Back in 1997, I took a job working as a senior unix/network engineer for Macromedia. Macromedia at the time was getting more and more web traffic against its (static) website. We were hosting the shockwave and flash players, and they were becoming very popular. In 1998, the Macromedia website was hosted on five Sun Ultra Enterprise 2s (UE2s). These were moderately-priced (IIRC, $20-$40K) servers, and the site ran so well on them that when we had a load balancer failure that resulted in 4/5 systems being out of load, we found out because some people reported the site was running "kinda slow." The site used very simple technologies (mostly Apache and static HTML files), and was rock solid, bullet proof, and basically hard to kill. Around 1998, I got a new CIO (my boss^2, if I recall correctly). His name was Stephen Elop, and you may know him better as the current CEO of Nokia. Stephen went on a full-blown evangelizing campaign focusing on the assertion our static site sucked, and really what we needed was to have a fully dynamic, personalized website. 13 years later, this seems obvious, but at the time this was a considerably interesting and fresh perspective. Macromedia's new, personalized, website (engineered with the help of a big consultancy) was just choke-full of potential. Emphasis on "choke." Our personalization software (Broadvision) was buggy; the vendor recommended we reboot our servers "as often as the business will tolerate;" and the hardware requirements to run this thing were staggering. We went from five UE2s running the site to a lovely three-tier model with Sun Enterprise 4500s in the app and db roles. These were heinously expensive -- on the order of around $100K or so, not fully configured -- and when we inevitably had to pump them full of RAM (they were awesome because they could support up to 14GB of RAM (!!)), we paid through the nose. I seem to recall we were paying about $7000/GB to fill these machines up, and we were running them all full of RAM. And the site's reliability and uptime suffered horribly. Which, as one of the people responsible for the site being up, made me sad. You could question Stephen's timing, somewhat; you could certainly question his choice of personalization platform. But you cannot, even with the benefit of hindsight, question his advocacy for us having a dynamic site. Features and capabilities are complications. This page about Watch Complications is really interesting to me, and I really like its statement that "The more complications in a watch, the more difficult it is to design, create, assemble, and repair." That's true not just for watches. The truth is that the more complicated what you do is, the more you offer your customer, the harder it is for you to support what you've got. And the earlier you do this, the higher the pioneer tax you pay. One of the things I love about working here is that Netflix is so committed to increasing the the scope of what it offers its customers. I've been a customer since around 2001, with a break around 2002, and then solidly since 2003. In that time, I've gone from being able to get DVDs to DVDs, Blu-Ray, be able to watch movies on my PC, then Mac, and now today, at home, I've got a Wii, PS-3, XBox, AppleTV, Samsung BD player, and two Android phones which someday soon will be able to stream Netflix. We've gotten tons more content, and better at offering it to people. We've expanded the device UI from "here's your queue, and god help you if you want anything outside of it" to "you'll never have to log into the website again" (well, almost). We've been adding complications (in the horology sense) all over the place. At the same time, Netflix has a breathtakingly courageous approach to risk and to trailblazing (one of the reasons why we chose to go to the Cloud -- in some respects before it was really ready for us). "Risk-averse" isn't in our dictionary. We try to take smart risks, but it's better to fall on one's face than sit on one's butt. In short, I've never before worked at a company that was more brash than me :). This combination of tendencies -- expand capabilities, adopt early -- can (and does) hurt uptime. And I've got to tell you -- that sucks. I hate it when customers can't stream. I hate it when family members call me to report a problem with the site. I hate it when I sit down (as I did Sunday evening) to watch the episode of Luther I was in the middle of earlier ... and fail. It'll get better. I really do believe that. The 2007 website, in context of what we can do today, sucks for a full-featured online offering these days. Nobody wants it back. But as time goes on, and this stuff gets more established, I do believe our reliability (which, mind you, isn't necessarily terrible -- I suspect we may be doing better than Twitter, but don't have any numbers to compare) will improve. Part of my job these days is to actually help get this better by getting us aware of issues faster; helping us understand when an issue is ongoing where it's originating; and giving people vastly smarter than I am the tools they need to dig in and fix their stuff quickly, so you, and my sister, and my spouse at home see some moving pictures the next time you hit 'play' on your choice of NRD. I've gone and written a post of practically Jollyesque length. Sorry about that -- this is interesting stuff to me. If you've actually managed to read to the end of this, I hope it's not been a waste of time. If not ... that's OK, I'm used to TL;DR.
@Cubic Z, this is the same topic, we just randomly shuffled this topic's place in the overall Hacking Netflix site to the top again, for your convenience. :)
I wish these announcements separated out the DVD and streaming offerings, given how terribly lame I am at returning DVDs (I just returned three DVDs in one envelope because it's been so long since I got them, that I lost all my envelopes and had to borrow an envelope from a friend).
Let's get this clear right now, because frankly the idea of a Netflix-like subscription service for underwear horrifies me like few other things can. Manpacks isn't like Netflix because you're actually paying per item you get, rather than a flat monthly fee. It's more the PPV model. More importantly, and this is really the key point here, is that what I first feared it was (get a new pair of underwear, wear it, return it in the provided envelope, get another one) is not in fact what it is. You don't return the underwear to be worn by the next person for whom they were at the top of their queue. And I think we can all agree to be happy about that.
I'd start with Princess Bride. And I'd end with Last Night (the 1998 version). Conveniently, Last Night (1998) is, in fact, currently available for streaming.
@Edward R. Murrow, How much Netflix pays its CDNs is one of those things a lot of people would like to know, and it's also one of those things I don't think Netflix has ever actually stated. This means that I'd guess you're not likely to see Netflix employees answering that particular question here.
Netflix didn't start with the iPhone. Subtitles have been available on the PS3 for a while now. I believe subtitles are platform-independent; the only issue is that the client needs to be updated to know what to do with them. Netflix can more easily push new client versions to the PS3 and the iPhone than other platforms, I'm guessing (other than Macs/PCs, which have also had subititles for a while now). -roy
As of about a week ago, we cut the cord -- my wife's a big Hulu fan; I'm a fan of Netflix streaming (with exceptions for four shows on Hulu -- Community, The Office, 30 Rock, and Modern Family). Now that I've moved from AT&T DSL to Cable, I get all my Netflix stuff in HD (well, as HD as 720p gets), which is even better. Yum.
I work in Los Gatos; I'm a relatively new Netflix employee -- only about 223 days. For me, Netflix is pretty much the best place I've ever worked. It's high-performance, with high expectations. I get paid very well to do interesting things and solve interesting problems. I'm fond of, or at least deeply respect, the vast majority of my coworkers. (This may be going a little off-script, but this is the sort of stuff I tell people I interview here w/o NDAs) Netflix has its share of significant technical challenges -- not everything we deal with today was designed, back in the day, to support the number of customers, the amount of activity, we've got to deal with today. In some respects, it's the most screwed up environment I've ever worked in -- and the environment that has given me the most freedom I've ever had to fix the problems I find. People get hung up on the vacations, and some get concerned that in an environment where it's not clear how much you're supposed to take, nobody will take any. That's not my experience. I've always left companies with fully maxed out vacation hours; at Netflix, because I know that's not going to happen, I tend to try harder to take vacation time. In the time I've been here, I've probably taken something close to 20 vacation days or thereabouts -- that's pretty high for me. Those are some of the things I like. What don't I like? Well, the socialist egalitarian in me isn't happy that there are really at least two tiers of employees at Netflix -- the full-time, exempt, Los Gatos crew, and the Hillsboro crowd. There's also the hubs, but I don't really know much about how those folks fare. I know that I've seen a lot of unhappy chatter from Hillsboro people, both current and former Netflix people, and I wonder if we could do better at applying the Netflix culture to the hourly folks than we have so far. Another thing I like variably is -- depending on your point of view -- either the lousy hours or the ability to get stuff done. My boss is fond of saying his most productive hours are 6PM-midnight, where he actually catches up on email. I know that there are times when I get far less done in a given day than I'd like to get (at least in the sense of getting far less of what I would like to get done done), and so I get to either adjust my expectations, or continue working. I was joking with people on Friday that I was really looking forward to the weekend because I could catch up on the work that I hadn't gotten around to during the week and, sadly, that wasn't entirely in jest. My wife's been known to refer to herself as a Netflix Widow, and I've not been as good at dealing with that as I'd like (every few weeks I shift my routine so I leave by 5:15PM and don't do any work in the evening barring emergencies; that works pretty well). In some ways, I could say Netflix is extremely libertarian -- you've got the freedom and responsibility to figure out how to be successful. It's not the kind of place where people will look out for your work/life balance -- you'll either manage it yourself (in the face of pressure), or ... you won't have one. It's not for everyone. I've got friends -- very good and capable friends -- who I wouldn't want working, and who wouldn't be happy, here. For me, it's a dream come true. (As a point of reference, I'm a technical lead in the Systems and Storage group in the IT/Ops organization. I've also got my full name attached to this post, so that should tell you something about how comfortable I feel being open about working here, and my management's (and HR's) strong commitment to openness and blunt honesty).
Royrapoport is now following The Typepad Team
Feb 7, 2010