GigaSpaces Blog

GigaSpaces XAP (http://www.gigaspaces.com) gives you a big fraction of that. You can create an in-memory distributed grid that acts as both an object store and processing platform, that is both resilient (self-healing) and scalable. The CTO of Elasticsearch used to work for GigaSpaces. Disclaimer: I work for GigaSpaces.

And for a slightly more modern take on the tuple space, check out Java Spaces [1] or Gigaspaces [2]. There's still plenty of active research on the topic too [3] (disclaimer: I did my PhD thesis on distributed tuple spaces). I've long contended that a tuple space was basically a generalised key-value store, so it's nice to see projects like this one crop up. [1] http://java.net/projects/jini/ [2] http://www.gigaspaces.com/ [3] http://eprints.utas.edu.au/9996/

I don't think Azure needs to run Java to succeed, but it certainly wouldn't hurt. And they've been pretty out-front with supporting Java for a while: * Open source SDK for Java (and Eclipse): http://www.windowsazure4j.org/ * Gigaspace running their products on Azure: http://www.gigaspaces.com * More info (admittedly from a UK-based Azure consultant): http://www.theregister.co.uk/2011/09/09/java_on_windows_azure_cloud/ To me, this sounds like a pragmatic continuation of their 'big data' plans on Azure -- gotta go after what the current user base is used to using -- and probably compliments their own Dryad project (http://research.microsoft.com/en-us/projects/dryad/)

If you want a framework with scalability "baked in", so to speak, you should take a look at GigaSpaces XAP (http://www.gigaspaces.com/). Its not the same model as the others mentioned on this thread but will prevent that "big rewrite" in the event that you do indeed become successful enough to need it.

> I've never seen a company suffer financially because of inefficient code Well, how many companies that went bankrupt because of inefficient code have you looked at? ;) It's hard to see what no longer exists. See Survivorship Bias [1]. Also fun: "More than 10 years ago, Amazon found that every 100ms of latency cost them 1% in sales. In 2006, Google found an extra .5 seconds in search page generation time dropped traffic by 20%. A broker could lose $4 million in revenues per millisecond if their electronic trading platform is 5 milliseconds behind the competition." [2] [1]: https://en.wikipedia.org/wiki/Survivorship_bias [2]: https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co…

> extra 10 milliseconds you saved Strawmen arguments are no fun so let's look at an actual example: https://www.rippling.com/blog/the-garbage-collector-fights-back P99 of 3 SECONDS. App stalls for 2-4 SECONDS. All due to Python. Their improved p99 is 1.5 seconds. Tons of effort and still could only get 1.5 seconds. https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co… > Amazon Found Every 100ms of Latency Cost them 1% in Sales I've seen e-commerce companies with 1 second p50 latencies due to language choices. Not good for sales.

The data seems to indicate otherwise: https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co… > A 2017 Akamai study shows that every 100-millisecond delay in website load time can hurt conversion rates by 7% > 10 years ago, Amazon found that every 100ms of latency cost them 1% in sales

> going from 400ms to 100ms gets you...nothing. According to Amazon [0] that'd be a 3% gain in sales (assuming the inverse holds true as getting slower, anyway). [0] https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co…

Do you mean "people still use VSCode even though many interactions have unnecessarily high latency" or "people's experience using VSCode would not be improved by the elimination of these latencies?" Can you post the proof you are referring to? Here's some article about this topic: https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co…

If shorter load times there translate to higher conversions in the same way this connection is observed by Amazon [0] this guy just made Rockstar a little fortune. Hiring the blogger outright and giving them a VP of Something Something Role would be more adequate. [0]: https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co…

5 second page load is not fine. Having the page load in 5 seconds is an annoying experience that causes you to lose users -- Amazon lost 1% revenue for every 100ms in page load time: https://www.gigaspaces.com/blog/amazon-found-every-100ms-of-latency-co…

TLDR at the bottom. First off, thank you for taking the time for a thorough reply. Clearly, I am a fan of you and your team's work on workers. There are some issues that are beyond the scope of this reply. Overall though, a well thought out and executed product. On the specific point of the waiting room, however, I have to respectfully disagree. Let's just begin very generally by the mission of Cloudflare - a better internet. Does giving a waiting room for the internet sound like a better internet? Everything Cloudflare does is predicated on reducing waiting times. This is what has informed the expansion of PoPs, the edge cdn, human captcha, and even your own work with workers and v8 isolates to reduce cold start times. At the very least, the waiting room is inconsistent with Cloudflare's mission. Now, I understand Cloudflare has onboarded a bunch of new folks lately. But it remains important that new members are clear and consistent about the mission. Attention to this detail can fall by the wayside when you just need to increase the body count quickly. Next point is to qualify the problem at the root of the waiting room. How many websites can afford to have a waiting room? How many users of those websites will wait patiently in a waiting room for their turn? I am sure you are well aware of the impact of even minuscule delays on user retention. But, let's just point them out for others just in case. Amazon found that 0.1 seconds results in 1 percent less sales. Google…