Jump to content

ScottishMartialArts

Members
  • Posts

    89
  • Joined

  • Last visited

About ScottishMartialArts

  • Birthday 05/05/1985

Personal Information

  • Flight Simulators
    IL-2 1946, IL-2 Sturmovik: Cliffs of Dover, Mig Alley, DCS: A-10C Warthog
  • Location
    San Francisco

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Out of curiosity can you give a ballpark figure? My back of the envelope math was based on a first day client load of 2500 or so people trying to download. That's a number I pulled out of my ass admittedly, but conversely I can't imagine we'd be talking about more than 10,000 or so clients -- the DCS community isn't THAT big -- which would end up putting us in the 300TB range, so definitely well short of the PB levels needed to get the good CDN deals. At any rate, thank you for acknowledging the frustration and trying to get stuff working. I'm feeling a lot happier now that I finally have the terrain file haha.
  2. Thank you for understanding the frustration, and working to fix it. Hope it goes better next time.
  3. I don't think we'd be talking about Petabytes of data transfer, but on the order of maybe 50-100TB: DCS is a niche product with probably only a few thousand people trying to download the 30GB alpha and terrain module. Out of curiosity, I'm looking at the AWS Cloud Front pricing calculator and you're correct that relative to the amount of revenue that ED likely received for selling NTTR, it would not be economically viable to scale up to fully handle the load. It would however have been viable to do some degree of scale up to support a smoother customer experience. I understand why they didn't do it, and I'm glad that they continue to make sims despite the razor thing profit margins for this market, but on other hand its hard not to have a sour taste in your mouth when you've been waiting for something for five years, have paid for it, but then can't have it because of infrastructure capacity issues.
  4. From the perspective of the CCNPs in the NOC, yeah network capacity has hard finite limits. From the perspective of the software engineers trying to get their application deployed through the infrastructure as a service that they have a service level agreement for, then the hard limits are an abstract detail obscured in the cloud part of the network diagram. The amount of data ED needs to move to its customers over the next 24-36 hours is a drop in the buck compared to the hard finite limits that the network engineers and DevOps people are dealing with at Amazon Web Services, Bitnami, Microsoft Azure, Cloudera, etc. etc. etc. If ED wanted to avoid this download problem they could have, and for the time period and data quantity we're talking about, it wouldn't even have been that expensive.
  5. Exactly. I appreciate Revelation's offer, but I doubt anything will happen with it. After 13 hours of downloading with no end in sight, I'm annoyed, frustrated, and disappointed that something that I've been looking forward to for nearly 5 years now still isn't actually here yet, at least not for those of us stuck in download hell.
  6. Seriously. While Revelation makes a good point that the desire for network capacity is infinite and budgets are finite, his comment about planning for average use + buffer is only true of fixed network infrastructure. When you're dealing with Infrastructure as a Service however, it's simply a matter of determining how much your lack of capacity is costing you in lost revenue versus how much it would cost to scale up the infrastructure for a few hours. Evidently, ED didn't feel like getting NTTR into the hands of its customers at a reasonable speed was worth the cost of scaling up for a few capacity hours. That was their business decision to make, but that doesn't mean I'm happy with it, and it doesn't mean I'm not going to think twice about purchasing one of their future products at launch.
  7. Yeah I suppose since they already received our credit card payments two months back, it doesn't make much difference to them whether their customers can actually access their purchase. And thus catastrophe was averted! /eyeroll
  8. Distributing files over the internet, whether it be through HTTP, FTP, or Torrent, is NOT Alpha technology. These are established protocols, with established server architecture designs, with established best practices, and with established professions whose job it is to ensure that when customers want to get files off a server or peer, and on to their individual computer, that they are able to do so easily, reliably, and quickly. Nothing about that is new, experimental, or still being solved. ED chose not to scale their file distribution infrastructure in preparation for what was arguably their biggest product launch -- whether an Alpha or not -- since the original Black Shark module. And as others have pointed out, this isn't exactly Steam on a Call of Duty launch day; at most we're dealing with a few thousand clients trying to get the DCS 2.0 files, and yet it's still just about impossible to get the files. My 105Mb/s connection is currently pulling 29.2KB/s down to get the NTTR file. Last I checked, it wasn't 1995 anymore, but it's real hard to tell tonight, given these download speeds and given how hard and experimental and alpha-level technology you make file distribution sound.
  9. I don't even care that much about the promised release date to be honest. It's the 30KB/s download speeds of a 27GB file that have me upset. It's a simple problem with a simple solution, but ED doesn't want to pony up the cash to scale up their infrastructure for the next 48 hours. If they aren't willing to pay what it takes to get their product to their customers, then it makes me less willing to pay what it takes to get one of their products.
  10. By the strict definition of the term, releasing late in the day on November 30th is "late November". But when your infrastructure can't scale to meet demand and your customers can't actually download the files in question, the fact that you barely met the deadline isn't saying much. Look, I don't know what kind of server infrastructure ED has, but infrastructure as a service is very easy to rent and get put into operation. If they already get their infrastructure through a cloud provider, it would literally be as simple as purchasing more capacity hours. If they don't, it would be a several hour job at most to scale up capacity to meet demand through some virtual server rentals.
  11. I never once said I expected a bug free product. I expected access to a playable, but definitely not bug-free, product in the time frame which was advertised when Eagle Dynamics sold me a license key. The alpha-ness of the product is an entirely separate issue from their digital distribution platform and its inability to scale. NTTR is in alpha; after years of selling stuff online, their ability to deliver the bits that make up NTTR shouldn't be in alpha. So who exactly is not understanding things here?
  12. And access to the alpha was one of their primary selling points. Reread the press release from September. In the final paragraph, they were trying to encourage us to preorder immediately, because A) we'd get access to bonus campaigns, B) we'd get a discount, and C) we'd get access to the alpha in November. Those are called selling points. They were trying to make a sale on the basis of those points. The fact that it's an alpha that they were SELLING access to does not excuse an inability to distribute said alpha, otherwise they never should have listed it as a selling point.
  13. We may be guinea pigs, but we're guinea pigs who paid for the privilege. Remember: you can't do anything with the DCS 2.0 Alpha unless you've purchased the Nevada map. They offered the Nevada map for sale on the premise that they would have a playable version available in November. We paid money for an, at the time, undelivered product on the basis of loyalty to the company, and out of trust that they would in fact do what they said they were going to do. While it is hardly the end of the world to have to wait a long time for something to download, there are a lot of things that could have been done to make the launch smoother. It's very easy to scale up server capacity from a cloud provider for a launch, and then to scale back down once the demand decreases. That would have cost them money, but then they are charging us money for their product; to not have scaled up server capacity reeks of cutting corners. Look, I get that flight sim development is a tough business with a small customer base. I understand that pennies have to be pinched, and revenue found where it can. But don't sell something on the basis of promises you can't deliver on. If you aren't sure that you can make a November release, then don't sell something with a November release date. If you expect your customers to promptly pay you, then take the steps necessary to ensure that your customers can access their purchase when you said they'd be able to access it. "The servers are getting hammered" may have been an acceptable excuse in the days when server capacity depended upon physically ordering more machines and integrating them into your infrastructure and then being stuck with said machines after demand died down. It is not an acceptable excuse in 2015 when all it would take would mean logging into your Amazon Web Services, or Microsoft Azure, or any of the million other cloud infrastructure providers, account and placing an order for some more virtual servers to be spun up.
  14. FC3 is essentially legacy content, which has merely been ported to the modern DCS engine with a few substantial enhancements/added features. In other words, you're buying a 12 year old sim and it's priced accordingly. The DCS modules are much newer projects, projects which, when released to beta, have yet to pay for themselves through sales. Likewise, the depth of simulation requires a substantially more expensive development process for the developers, all of which has to be recovered through sales, and yet the deeper the simulation, generally the smaller the audience of people whom you can sell to. Something has to give, and the least worst option is price. All that said, some of the beta modules really aren't worth the money. The A-10C, MiG-21, and UH-1H betas were far enough along in development and shipped with enough content that they very much were worth the price of entry. In contrast, despite being my all time favorite military aircraft, the F-86 module, while not without merit, does not provide a comparable amount of value for its $50 price tag. With that in mind, I've been taking a break from DCS for most of this year, just waiting for the engine and environment to catch up with the module releases so that the modules can become more worthwhile purchases.
  15. I was never able to successfully do it until I got the Warthog HOTAS. Then, I nailed it on like my second or third try. That said, one key is constant micro-adjustments of throttle, kinda like how flying the Huey demands constant micro-adjustments of the cyclic. Likewise, you will want to find a reference point for the tanker empennage in your canopy. In otherwords, find a visual reference point in your canopy, and keep the tail section of the tanker fixed in that reference point. Ignore the boom, just keep the tail section in your reference point and you should get a good connection. Once connected, focus on the reference point but out of the corner of your eye watch the scale on the boom, and microadjust throttle to keep the scale in the green.
×
×
  • Create New...