Hacker Newsnew | past | comments | ask | show | jobs | submit | more lambda's commentslogin

Ouch. Fairly nicely written piece about punctuation, and generally nicely formatted with justification, properly indented paragraphs, use of em dashes, and so on; and then it uses straight quotes everywhere. While I'm generally used to straight quotes everywhere on the web, the additional standard set by the rest of the typesetting and the topic meant that their omission here was that much more glaring.


Seriously? I can think of nothing about this article that bothers me less than that.

In my world, quotes are quotes and the distinction between ways of drawing them is infinitely irrelevant. That this is not true to someone else is actually surprising me.


"use of em dashes"

It's worth pointing out that this is not a universal style choice.

http://en.wikipedia.org/wiki/Dash#En_dash_versus_em_dash


The piece also asserted that "word" and "whirred" are "pronounced exactly the same".

They aren't. The first consonant is different; IPA [w] vs. IPA [hw].


In a good number of English dialects [w] and [hw] have collapsed together.


Yep, to the point of /hw/ being considered comical by many speakers.

http://en.wikipedia.org/wiki/Phonological_history_of_wh#Wine...

Edit: I just want to point out that despite what the map in the linked article says, the merger is complete in most of the urban areas in the Southeast as well. You'll usually only hear /hw/ in rural areas.


For me, it's the same consonant, but the vowel sound is very slightly longer in "whirred".


OTOH, that "wh" is pronounced /hw/ only reinforces his main argument.


What are you talking about? Daala is as open as you get.


As I read it, they plan to patent new stuff. See page 9, the third bullet for example.


Read the fourth bullet: "Use a patent license that encourages adoption and discourages defection."

It sounds like they are planning on a copyleft-style defensive patent licensing system.

Right now, one of the big reasons that open codecs like VP8 and VP9 don't see wider adoption is that there's a lot of FUD thrown around by people saying "they might infringe H.264/HEVC patents, it's too risky to use them without a patent licensing pool."

By filing patents of their own, and offering them under a free (libre) license that has a clause causing it to be revoked if you engage in some other related patent litigation, this adds pressure against anyone trying to use their patents against someone using Daala.

Not sure how well this would work out in practice; I know that some people have gotten patents under these kinds of defensive open patent licenses before, but I don't know if they've ever been used defensively in practice before.


Note that this is simply following the successful model of Opus: http://www.opus-codec.org/license/



So you're saying in practice copyleft and hence open source doesn't really work, you have to patent and then license for free to protect your users. How does this impact innovation based on the ideas you have patented?


We've talked about making a GPLv3 release to at least open up the patents we control to all copyleft software. But there are important details to work through, and like most FLOSS projects we have more things to do than people to do them, so it hasn't happened yet. If someone needed it to happen it should be a pretty easy conversation to have.

Keep in mind also, our goal is generally to stay out of court. See above about having too much to do.


I wish we could easily get anti-patent patents.


Why "they"? Why not you? There have been other Git frontends in the past; why not write a new one that behaves as you'd like?


A number of reasons: Writing your own means dealing with changing internal behaviour. It means following changing output. It means not being able to reuse parts that already exist but are not exposed. It means deploying separate software where you don't necessarily have root/install rights.

Even if the interface you see as a user doesn't change, the internals do - for example submodules handling has been changed from a standard git dir to git file with directory link. This broke some of my scripts that made git interaction easier.

I don't want to slap something on top of git and hope that there are no edge cases in state/output parsing. I want an actual first-class interface that has at least a bit of consistency between commands.

As you can see, I went that way and it's not usable long-term.

Edit: changed symlink to git file, bad memory


I don't buy this. Give me some concrete examples of how internals need to change. I have lot of git scripts which work off git plumbing and they didn't require any of what you mention. You just sound like an armchair critic otherwise.


I gave one example already. The implementation of submodule .git directory changed (git-file instead of a directory). The change meant I had to check the git version and have two different ways of doing the same thing or the interface I wrote failed.

For an example of difference in output, see http://stackoverflow.com/questions/3921409/how-to-know-if-th...

"""As of git version 1.7.3.1, git status doesn't say anything about the rebase status.""" - so you're left with "does some file under .git exist" approach. It's unlikely to change, but I don't believe it's part of any external interface - when it changes in the future, it changes.


Nice idea!

Some of your icons aren't showing up for me on Firefox, as FontAwesome is being loaded from a different domain without CORS enabled:

  Cross-Origin Request Blocked: The Same Origin Policy disallows reading the remote 
  resource at
  http://support.hirerefugees.org/plugins/font-awesome/font/fontawesome-webfont.woff?v=3.2.1. 
  This can be fixed by moving the resource to the same domain or enabling CORS. 
  downloadable font: download failed (font-family: "FontAwesome" style:normal 
  weight:normal stretch:normal src index:1): bad URI or cross-site access not 
  allowed
  source: http://support.hirerefugees.org/plugins/font-awesome/font/fontawesome-webfont.woff?v=3.2.1 font-awesome.css
There's a ticket for this on GitHub that describes how to fix the problem: https://github.com/cdnjs/cdnjs/issues/755


I'll fix that. Thanks!


You could never recover 10/8. Simply impossible. Due to the fact that IP address space is running out, 10/8 what everyone uses for NAT; how are you going to solve the address exhaustion problem by removing the main tool people are using to fight it?

Well before that, you could recover 18/8, MIT's prefix; how many addresses does a single university actually need, even MIT? Stanford already returned their /8 prefix. Or you could potentially recover a variety of DoD /8s, or start assigning some of the ones that are "reserved for future use" like 240/4.

Of course, an even better use of 240/4 would have been to use it as a backwards-compatible way to move to a larger address space. IPv6 should have been specified with IPv4 compatibility via NAT from the beginning, so that it could be a gradual migration. I mean, we're moving that way anyhow, but with a much more painful period in the middle in which NAT is necessary but IPv6 isn't ready yet for people who want to end the pain of NAT.


> start assigning some of the ones that are "reserved for future use" like 240/4.

The 240 block is problematic because a huge amount of legacy routing equipment is hard-coded to blackhole it. The best thing for it is probably to reserve it for unspecified local use, which would allow you to use it to NAT IPv6 addresses into locally, or use as additional RFC1918-style address space, or as ORCHID-style address space for IPv4-only applications using the likes of cjdns, etc.


Hmm. I would argue that we're still in step -1, though the clean shaven folks have finally started listening to the neckbeards in a few circumstances and are slowly moving around towards step 0.

Most people still can't actually route IPv6 over their tubes without going through a tunnel broker. At least for home, small office, and companies that aren't tech heavy, it's still firmly in the neckbeard camp.


Well, you definitely can route IPv6 over the public backbone. If you get an IPv6 address in SF and I get one in NYC, we can already talk to each other. That is how we got past step -1. Dual stack is now a pretty standard setup on a non-trivial, though still small number of networks. For example, Linode, Comcast, T-Mobile all support dual stack on at least some of their networks.

What we have not quite reached yet is the point where IPv4 addresses are really expensive. You still get one with every VPS you sign up for and with every home connection. Once that can no longer be done, IPv6 will become standard pretty much across the board, starting in Step 3 and 4.


As another example, when I am using LTE on Verizon Wireless, I am connecting over IPv6 to sites running IPv6 (ex. Google, Facebook, Wikipedia, etc.). I can go to sites like http://www.test-ipv6.com/ and see that I have full IPv6 connectivity.


Hey, so I had posted the above primarily because I've been trying to get a routable IPv6 address for a while, and haven't been successful; I've repeatedly asked Comcast when my CMTS would be IPv6 enabled, but their techs have just said "try enabling IPv6 on your computer and see if it works", they haven't been able to give me a schedule or even tell me if it was supposed to work yet.

However, just checked last night and discovered that they'd added a new feature that would tell you when your CMTS was IPv6 ready, and it turns out that they've finally upgraded me. I could finally switch off my tunnel and use a real, routable address. Woo!


Major (consumer) ISPs in Australia offer native dual-stack.


Which ones? I'm with iiNet and was wondering if they offered it yet but couldn't see anything about it online.


The sad part about IPv6 is that there's only one loopback address; ::1.

In IPv4, you have an entire /8 at your disposal, just for talking to yourself!


Aside from ::/128 (undefined address) and ::1/128 (local loopback), the entirety of ::/96 was reserved for ipv4-compatible addresses and had been deprecated for a while [1]

It is safe to assume that they won't be allocated in the future. You can easily assign them to your loopback interface. e.g. on linux:

  $ ip addr add ::2/128 dev lo
You will then be able to use them over loopback just like ::1, and can populate your /etc/hosts file accordingly.

  $ ping6 -c 1 ::2
  PING ::2(::2) 56 data bytes
  64 bytes from ::2: icmp_seq=1 ttl=64 time=0.078 ms
  1 packets transmitted, 1 received, 0% packet loss, time 0ms
  rtt min/avg/max/mdev = 0.078/0.078/0.078/0.000 ms
[1] http://tools.ietf.org/html/rfc4291#section-4

EDIT: formatting.


Likewise for Java.


Correct me if I'm wrong, but I think if Dalvik were released under a license compatible with the Oracle runtimes, this whole thing between Oracle and Google goes away? I agree allowing copyrighting of APIs is a bit daft, but the point is that if you're trying to draw a parallel between MySQL/MariaDB and Java/Dalvik, it doesn't apply. (Assuming the MariaDB fork has a compatible license, which I believe it does.)


Sure, there are going to be a few performance sensitive workloads that you can't use it on, but for the vast majority of issues I've used it for, it's been far more useful than perf or dtruss (on Mac OS X, I've never used dtruss on Solaris so I don't know if it's in better shape).

The big advantage is that it already knows how to decode most syscalls, decompose flags into symbolic arguments, decode things like stat return values, and so on. dtruss on Mac OS X can decode some syscalls, but doesn't generally know how to decode structures or pull apart flags, so there's some information you just can't see, and some which you have to laboriously manually decode. perf, as you note, decodes even less than dtruss. When I'm trying to debug an issue, the more information I can gather and see at once, the better.

Very few of the occasions I've had to use these tools have been particularly performance sensitive. Your example of the huge difference is because dd is copying fairly small blocks at a time, and thus making a fairly high number of system calls relative to the amount of data being transferred. Usually, I'm just looking for why an operation fails, or why a process is stuck; if it takes a little longer to get to that failure, that's OK, and if the process is stuck then stracing it generally won't affect performance.

sysdig looks pretty interesting, thanks for the reference! It's too bad they can't use the existing perf infrastructure in the kernel, and need to add their own kernel module; that makes it that much more cumbersome to deploy and use.


Except, you don't necessarily know that it's the database that's the problem. It might be trying to do a DNS query, and that's timing out. Or spinning in a loop trying to read from a socket that has no data available. Or blocking on a filesystem operation on a networked filesystem that has become disconnected. Or any number of other things.

The nice thing about strace is that it pretty quickly tells you exactly what the program in question is doing with the outside world, which many times can provide you the quick clue you need about what's going on.

strace is a pretty simple tool, at its most basic usage it's just "strace <pid>", and the nice thing is that it's pretty universal; it comes in handy many, many times, for solving many different kinds of problems. Have some third party code that seems to exhibit different behavior on different machines, but don't know what is different about its environment that's making it behave differently? strace it and see what files it opens for configuration data. Your program not making any progress, and it doesn't seem to be CPU bound as it's only using a few percent of CPU? strace it and see what system calls its blocking on.

Note that for this kind of use case, where the process is hanging, you generally don't get way more output than you can handle; the fact that it's hanging on this particular call means that that's the call that you will see at the end of the output for several seconds until it times out.

It's certainly not the only tool in your arsenal; there will be other cases where more specific tools are more appropriate. But it's a really useful general purpose tool.

My general purpose debugging arsenal, when I need to debug something that I haven't written, generally consists of tcpdump/wireshark to capture and analyze any network traffic, strace to see how the program is interacting with the world outside of it, turning up logging on any relevant systems to their maximum value and looking through that output, and then finally if all else fails attaching to it with GDB and stepping through the code.

Sure, it's always better to have more specific tools, to have turned on appropriate monitoring, and so on. But you don't always have that luxury. Sometimes more specific tools haven't been written. Sometimes your monitoring may find that your database is up, as it is accepting connections, but its actually hanging when you try to do anything, or it's succeeding for reads but hanging on writes, or the like. Monitoring code is never going to be as thorough as you like, there will always be something that you miss, and so you need general-purpose tools that can be used in any situation to get a handle on what is going on quickly so that you can narrow down on the problem.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: