Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ay
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
39 ms
·
361.
▲
by
ay
12y ago
Take a look at https://github.com/ayourtch/nat46 - I've coded it primarily for OpenWRT (so it requires forwarding), but the core portion is usable in userland as well - so I have this crude hack ( https:/
362.
▲
by
ay
12y ago
tunnelbroker.net - works pretty well and is easy to setup.
363.
▲
by
ay
12y ago
I am curious if this news affects how many HN readers which are involved into the product development are developing their products / websites with IPv6 support ? If maybe this helps folks managing news.ycombinator.com to click the but
364.
▲
by
ay
13y ago
A couple of links that may be of interest on this topic: http://www.ietf.org/proceedings/87/slides/slides-87-tsvarea-... http://www.ietf.org/mail-archive/web/tcmtf/current/
365.
▲
by
ay
13y ago
In the US, today, one in 15 clients accessing google.com / yahoo.com / facebook.com is doing it from an IPv6 address. And it's more than double compared to a year ago [1]. While indeed this still qualifies as "relatively
366.
▲
by
ay
13y ago
https://github.com/LuaDist/lua The repository size is exactly the same as the submission, but you get a fully-fledged language - which allows you to make parametrized configs, etc, etc, etc. And the syntax is quite con
367.
▲
by
ay
13y ago
People may have 1+ device per person and not share them.
368.
▲
by
ay
13y ago
You can collect the estimate of the units wanted + email address. This way you could both capture the HN's momentum and keep in touch later. I'd love to hear from you again when you have stock :-) (and I am in .be as well :-)
369.
▲
by
ay
13y ago
What about a slightly modified idea: (0) The server informs the client about the parameters of the bloom filter where it currently keeps all the contacts. (1) The client builds the bloom filter with the same parameters based on its address
370.
▲
by
ay
13y ago
+1 on all counts. I went to Wikipedia page and looks like it was written if not by him then by his fans - it does not contain a half decent description of the protocol. Just some propaganda-like fluff. I admit "squirrel" does soun
371.
▲
by
ay
13y ago
In github case since it's server side only and FOSS, they could probably go even further and go v6only and do stateless translation for ipv4 clients. But that's a bit too hardcore :-)
372.
▲
by
ay
13y ago
Depends on what you do. YouTube should work fine, netflix does as well. That more than 50% of traffic for some. The long tail is indeed long :-) There with no transition tech you get about 4.8% [1] I think there are two strategies: 1) keep
373.
▲
by
ay
13y ago
Speaking of latency: one of the big IPv6 news for me in 2013 is that now, on average, IPv6 when it works gives better latency than IPv4: https://ripe67.ripe.net/archives/video/43/
374.
▲
by
ay
13y ago
Transition tech should still provide decent IPv4. Would love to hear more details and see if there is something that has to be fixed. Sent you an email.
375.
▲
by
ay
13y ago
It's a pretty good excuse. But not a good reason anymore.
376.
▲
by
ay
13y ago
Try with nat64. Not entirely clean experiment, I know (IPv4 upstream of translator). But the result is quite usable. Ping me via mail if I can help with the setup, we did this setup several times with reasonable success.
377.
▲
by
ay
13y ago
Yeah maybe if we all ask pg nicely, he will do it ? :) As a backup plan - cloud-based NAT64 is a fair game for the day, search "go6 nat64".
378.
▲
by
ay
13y ago
Remember the World IPv6 day ? In 2014 a few of us (about a hundred of folks around the world at the moment) are going to do a "World IPv6 only day" (turn off IPv4) on the 06/06/2014, for ourselves. If we can't work
379.
▲
by
ay
13y ago
No amount of Moore's Law will change the speed of light...
380.
▲
by
ay
13y ago
http://www.epostmail.org/ ; - seemed like an interesting technology, at least by reading the papers. Unfortunately getting it to run involved sacrificing at least a dozen newborn animals - which is maybe the reason it nev
381.
▲
by
ay
13y ago
Of course. But it can't be turtles all the way down - at some level, the overlay traffic will be a TCP segment or UDP packet, sent from one node running overlay network to another node running overlay network. If both nodes in the over
382.
▲
by
ay
13y ago
You can connect to IRC because the IRC server is not behind a NAT. Try to run an IRC server on your host which runs behind the consumer NAT which runs behind the CGN. Yes, if the gear supports it, you can use uPNP for the first, and PCP for
383.
▲
by
ay
13y ago
FWIW, the C preprocessor can be used to calculate PI as well: http://chaos-pp.cvs.sourceforge.net/chaos-pp/order-pp/exampl...
384.
▲
by
ay
13y ago
I thought of a variant of this approach for distributed code storage/reuse. Store each function as AST with normalized variable names. Then, a cryptographic hash of this would uniquely identify algorithm. So, caching becomes quite easy and
385.
▲
by
ay
13y ago
Indeed. 25 is the maximum - mostly the shorter, the better. Though at least for me there is somewhere a limit (2-3 lines?) where the returns start to diminish - there's more noticeably more typing involved, and more mental context switching
386.
▲
by
ay
13y ago
Great advice IMHO. One trick that I found useful for myself (good chunk of my code is "hobby" code that I myself revisit some months later, so not really "production" problem but nonetheless almost the real-world maintenance example): - mak
387.
▲
by
ay
13y ago
A simple way to make this a bit harder: stty -echo; echo `read | sha1sum` LittleBitOfSalt | sha1sum | cut -c 1-40 | pbcopy; stty echo Then paste whatever ends up in the clipboard as your password. pbcopy is MacOS X specific I think, but IIR
388.
▲
by
ay
14y ago
A small (offtopic) comment on your blog entry about IPv6... /64 is not twice the IPv4 address space - it's twice the number of bits. This translates 2^32 times bigger than entire address space.
389.
▲
by
ay
14y ago
Unhelpful when you are putting in the keys - better strobed full-light than dim - the latter sounds cool but is impractical. The human vision is much more effective at flash-like lighting than at a constant near-zero one. At least, mine and
390.
▲
by
ay
14y ago
exactly. This is to say "soon it will be dark, press the reset if you are sure you need a lot more time and just hurry up if you do not". Most of the time it really is no-op though.
More ›