amirmc

39 ms·

331.
by amirmc 12y ago
> "I was blown away by how good facebook's recommendations for friends were..." Search online for 'Facebook shadow profiles' to get an idea of why this was so good. It's more than just your phone number. E
332.
by amirmc 12y ago
I think OCaml is doing pretty well for itself. For example see the graph of package-growth at [1] and the recent news from people at Facebook about Hack and Flow (both in OCaml). Not all repos are on GitHub and it's not really fair to
333.
by amirmc 12y ago
> "For practical or business purposes, this is a nice bit of incomplete information to help make a decision. I want to take a serious, time-invested dive into a new statically compiled language, but which one should I pick? An old
334.
by amirmc 12y ago
> "So basically, if someone says "full stack", they likely mean that you can develop and run the entire product, not just a part of it." On the face of it, this sounds reasonable but it glosses over the fact that be
335.
by amirmc 12y ago
There's no fixed definition and I've found that people use it when they're asking for 'broad' skills -- where broad is defined by their own experience (cf. most of the other comments here). For example, most web dev
336.
by amirmc 12y ago
Indeed. This is a nice example of data porn, in that we can pore over graphs and trends and make lots of speculations (which will conveniently confirm our pre-existing biases). It's visualisation, not analysis so we should take it w
337.
by amirmc 12y ago
I should also have mentioned that Mirage is a core component of the stack for distributed personal clouds. We need to have resilient, scalable infrastructure if we want to own a piece of the cloud for ourselves. http://nymote.org
338.
by amirmc 12y ago
We're actually putting a paper together that uses Jitsu so you might find that answers most of your questions (the team is busy with eval at the moment). For the time-being, Jitsu [1] 'just' spawns a unikernel which serves re
339.
by amirmc 12y ago
Exactly, so if there is a breach, then we've learned about something we need to secure better (win). If there's no breach — and provided people did try to break in — then we're only incrementally more confident in the sta
340.
by amirmc 12y ago
DDoS edit: now back
341.
by amirmc 12y ago
I'm part of the Mirage team but when I say 'authors' I'm specifically referring to the people who wrote the Piñata code.
342.
by amirmc 12y ago
Yup, this is work in progress. The first part of this is Jitsu, which is a DNS server that can spin up Unikernels on an incoming request. https://github.com/MagnusS/jitsu
343.
by amirmc 12y ago
I bet the authors never considered that as part of their threat models.
344.
by amirmc 12y ago
I appreciate what you're saying but even then, the data you infer this way (apart from the breach itself ) is highly suspect. I don't think you really get a view for how much it costs to break and time is a poor proxy. It's
345.
by amirmc 12y ago
Most of the discussion is on the other submission: https://news.ycombinator.com/item?id=9027743
346.
by amirmc 12y ago
> "Break in and take it." But please tell us if you do. We'd like to learn from this exercise and improve the stack.
347.
by amirmc 12y ago
> "and want to find out/demonstrate how secure it is" As the page (and my post) emphatically state, this contest cannot prove anything about how secure it is. That distinction is really important.
348.
by amirmc 12y ago
There's some additional context at http://amirchaudhry.com/bitcoin-pinata/
349.
The Bitcoin Piñata(ownme.ipredator.se)
482 points by amirmc 12y ago|103 comments
350.
by amirmc 12y ago
The graph on slide 11 is pretty clear that the trend is for more time being spent in mobile apps than on the web. Unless you want to take issue with the data, there isn't much nuance there. To whoever downvoted my earlier comment (if
351.
by amirmc 12y ago
Isn't that just a technical challenge? In this world view, you are your own central server and everyone else is their own. Let's take your example and run with it (bear in mind I'm ignoring some rough edges here). You share y
352.
by amirmc 12y ago
There are plenty of ways of making money, including asking your users to pay. Take a look at the number of 'personal cloud' crowdwfunding campaigns to see how much latent demand there is for things like this. Most of those product
353.
by amirmc 12y ago
That's not the way the current evidence is pointing. People do like apps and so far the pendulum doesn't seem to be swinging back to the web. See slides 11 & 30 at http://a16z.com/2014/10/28/mob
354.
by amirmc 12y ago
He's asking for innovation in the wrong place. The OP conveniently ignores all the other mobile OSs out there such as Windows, Firefox, Ubuntu and also people like Fairphone. It's pretty lazy to just claim we need more, when what
355.
by amirmc 12y ago
If we want to do this in a way where everyone can run their own infrastructure then we've got to tackle some of the distributed systems problems. Specifically, life-cycle management of software/apps, how things sync/backup, a
356.
by amirmc 12y ago
And can go back if they wish. I'm not sure how many students eschewed college entirely (high-schoolers or true dropouts), but even then you'd be in a fantastic position to apply if you'd done the fellowship first.
357.
by amirmc 12y ago
I'm happy for the team but somewhat sad for the users. I wonder why they sold so quick? Google is the antithesis of using edge devices to their maximal advantage, and this was clearly not a product acquisition.
358.
by amirmc 12y ago
Yet Flickr did ok for itself. As did Twittr, and del.icio.us (sp?). I agree that names matter but the scenario you've painted isn't fatal by any means. In the modern connected age, 'word-of-mouth' is just as much about
359.
by amirmc 12y ago
Not alone and I suspect there are more and more people becoming uneasy with 'the cloud' for myriad different reasons. Obvious reasons include privacy and security but there's also the ephemeral nature of such services (they
360.
by amirmc 12y ago
I've actually been quite pleased with the events concept. I found albums annoying and manually tedious but that pics automatically get grouped has made it easy for me to go back through old photos and find the ones I'm looking for

More ›