| 00:46.06 | *** join/#utah emcnabb_ (~emcnabb@cpe-065-190-176-153.nc.res.rr.com) |
| 00:59.12 | *** join/#utah mabes (~mabes@68.69.166.132) |
| 03:28.57 | *** join/#utah tiwula (~lane@63-248-200-215.static.sdyl0430.digis.net) |
| 03:56.23 | *** join/#utah jalbretsen (~jalbretse@c-24-10-129-200.hsd1.ut.comcast.net) |
| 04:02.12 | *** join/#utah har-od (~dshipp@c-71-195-216-67.hsd1.ut.comcast.net) |
| 04:42.40 | *** join/#utah ljenkins_ (~ljenkins@2001:1948:212::a) |
| 04:47.05 | *** join/#utah lahwran (~lahwran@unaffiliated/lahwran) |
| 04:48.35 | *** join/#utah jnbek (~jnbek@pdpc/supporter/active/jnbek) |
| 04:54.35 | *** join/#utah jnbek (~jnbek@pdpc/supporter/active/jnbek) |
| 07:28.59 | *** join/#utah ljenkins_ (~ljenkins@2001:1948:212::a) |
| 08:31.20 | *** join/#utah ljenkins_ (~ljenkins@2001:1948:212::a) |
| 09:55.40 | *** join/#utah CamargoBP (~CamargoBP@66.89.100.211.ptr.us.xo.net) |
| 12:02.12 | *** join/#utah redbeard2 (~jphall@techhat.dsl.xmission.com) |
| 12:38.05 | *** join/#utah redbeard21 (~jphall@techhat.dsl.xmission.com) |
| 13:12.55 | *** join/#utah DonaldStrachan (~mushroom@mail.wangthai.co.za) |
| 13:12.58 | DonaldStrachan | hello |
| 13:27.22 | *** part/#utah DonaldStrachan (~mushroom@mail.wangthai.co.za) |
| 13:27.26 | *** join/#utah emcnabb_ (emcnabb@nat/redhat/x-fnbmahbizujpzpxo) |
| 13:28.22 | *** join/#utah eggyknap (~jtolley@unaffiliated/eggyknap) |
| 13:44.52 | jima | i know some people who are bad at directions, but a utahn in .nz? that's just ridiculous! ;-) |
| 13:46.19 | goozbach | ~. |
| 13:46.19 | ibot | methinks ~. is not the escape sequence you're looking for. |
| 13:47.07 | jima | holds a memorial for goozbach's ssh session. |
| 13:54.32 | goozbach | jima: it's ok |
| 13:54.39 | goozbach | everything is running in screen |
| 13:54.48 | goozbach | well everything important anyway |
| 13:56.43 | jima | good man ;-) |
| 14:04.14 | jima | man, coming up on 17 years i've been using screen. |
| 14:10.19 | fozzmoo | Sounds about right. I was using Screen on VMS in 93-94. |
| 14:13.34 | jima | Initial release: 1987 (according to wikipedia) |
| 14:22.35 | *** join/#utah dmdavis (~dedavi@sas15254.nat.sas.com) |
| 14:25.05 | *** join/#utah har-od (~dshipp@fw.oremut02.us.wh.verio.net) |
| 14:26.30 | *** join/#utah macnewbold (~Adium@76.8.223.101) |
| 14:27.15 | *** join/#utah eggyknap (~jtolley@74.36.73.22) |
| 14:27.16 | *** join/#utah eggyknap (~jtolley@unaffiliated/eggyknap) |
| 14:29.11 | *** part/#utah dmdavis (~dedavi@sas15254.nat.sas.com) |
| 14:37.13 | *** join/#utah mabes (~mabes@68.69.166.132) |
| 15:00.08 | *** join/#utah fadein (~Erik@c-67-171-115-158.hsd1.ut.comcast.net) |
| 15:02.03 | *** join/#utah TimRiker (~TimRiker@bzflag/projectlead/TimRiker) |
| 15:02.04 | *** mode/#utah [+o TimRiker] by ChanServ |
| 15:06.12 | *** join/#utah redbeard2 (~jphall@BEYOND-OBLI.car1.SaltLakeCity1.Level3.net) |
| 15:08.03 | *** join/#utah jasona (~jasona@nat-out.fw01.slc.dechocorp.net) |
| 15:10.34 | *** join/#utah fungus (~olsonl@bromine.sosstaffing.com) |
| 15:17.05 | *** join/#utah tiwula (~lane@70.89.246.241-Busname-UT.hfc.comcastbusiness.net) |
| 15:17.29 | *** join/#utah natgeorg (~george@97.75.169.100) |
| 15:17.51 | *** join/#utah giffordrb (~giffordrb@216.49.181.128) |
| 15:22.22 | *** join/#utah mabes (~mabes@24-104-67-234-ip-static.hfc.comcastbusiness.net) |
| 15:22.26 | *** join/#utah lukfugl (~lukfugl@ip65-46-62-74.z62-46-65.customer.algx.net) |
| 15:22.50 | *** part/#utah RyanE (rberick@shell.xmission.com) |
| 15:23.59 | *** join/#utah RyanE (Piadas@rberick.dsl.xmission.com) |
| 15:24.23 | *** join/#utah jalbretsen (~jalbretse@70.102.57.206) |
| 15:39.12 | *** join/#utah craigj (~craigj@neutroninteractive-xmission.xmission.com) |
| 15:44.59 | *** join/#utah sjansen (~sjansen@ip65-44-231-178.z231-44-65.customer.algx.net) |
| 16:06.04 | *** join/#utah larrys (~larry@c-76-27-113-134.hsd1.ut.comcast.net) |
| 16:18.42 | sjansen | ~lart Fedora for moving X to tty1 |
| 16:18.42 | ibot | follow's Fedora with a gauntlet and ... scratch ... HUMILIATION for moving X to tty1 |
| 16:20.14 | goozbach | that is annoying |
| 16:20.51 | eightyeight | heh |
| 16:22.01 | shalkie | I am sure sjansen could spend the 5 minutes and correct it on his system if that is the issue. :-P |
| 16:22.35 | redbeard2 | yeah, it would probably take him at least 5 minutes |
| 16:22.37 | redbeard2 | ducks |
| 16:23.01 | sjansen | shalkie: You know better than that. Would that life were as easy as always working with the same small set of systems. |
| 16:23.03 | shalkie | Though I find it annoying that with RHEL6 there are the occasional hiccups and you end up with an uncontrolled terminal1. |
| 16:23.54 | sjansen | shalkie: As it happens, I'm complaining because the change slowed me down while troubleshooting your student's system. |
| 16:24.01 | shalkie | sjansen: Oh you mean this is on more then just your laptop... |
| 16:24.09 | shalkie | Fair enough. |
| 16:28.41 | shalkie | I say we move to runlevel 3 as default. Problem solved. :-) |
| 16:32.37 | eightyeight | or just ditch runlevels all together |
| 16:44.14 | *** join/#utah bitsmart (~bitsmart@97.75.165.50) |
| 16:48.02 | *** join/#utah jasona (~jasona@nat-out.fw01.slc.dechocorp.net) |
| 18:54.33 | *** join/#utah giffordrb1 (~giffordrb@216.49.181.128) |
| 19:00.24 | *** join/#utah jasona (~jasona@nat-out.fw01.slc.dechocorp.net) |
| 19:25.20 | *** join/#utah mabes_ (~mabes@24-104-67-234-ip-static.hfc.comcastbusiness.net) |
| 19:36.39 | *** join/#utah rsimpkins (~rsimpkins@unaffiliated/rsimpkins) |
| 19:43.53 | *** join/#utah d3c4f-work (~d3c4f-wor@64.245.158.66) |
| 19:44.19 | *** part/#utah bitsmart (~bitsmart@97.75.165.50) |
| 19:44.35 | *** join/#utah bitsmart (~bitsmart@97.75.165.50) |
| 20:00.29 | fadein | How many here are familiar with atool? http://www.nongnu.org/atool/ It's the one archive program to rule them all. |
| 20:00.31 | brac | [atool home] |
| 20:00.44 | fadein | My reaction was "why didn't I think of that?" |
| 20:03.22 | *** join/#utah mabes (~mabes@24-104-67-234-ip-static.hfc.comcastbusiness.net) |
| 20:03.40 | sjansen | pax was an earlier step in that direction |
| 20:07.44 | fungus | fadein, that's not very new, or complete. |
| 20:07.46 | fungus | http://obsoleet.org/code/unfoo/ |
| 20:12.07 | fadein | accordioning to the manpage for a recent atool, it handles more formats than unfoo |
| 20:12.16 | fadein | but the main point is, why didn't *I* think of it |
| 20:12.24 | fadein | or, why didn't I think to look for such a tool |
| 20:14.05 | fozzmoo | And atool is written in Perl. ;-) |
| 20:14.42 | fadein | fozzmoo: so it is! atool++ |
| 20:15.00 | sjansen | ~hug git |
| 20:15.00 | ibot | ACTION hugs git tightly until git turns slightly blue |
| 20:15.57 | sjansen | I just ran "git stash apply" to reapply the same edits for the third time in the last 30 minutes. |
| 20:16.24 | sjansen | Keep coming acrros minor fixes that belong in their own patches. |
| 20:16.48 | sjansen | Doing the same with SVN or *shiver* CVS would have been _so_ much harder. |
| 20:16.53 | sjansen | ~git++ |
| 20:18.22 | goozbach | sjansen: I concur |
| 20:20.42 | levi | Yay for powerful SCM tools! |
| 20:20.59 | levi | Any node.js hackers here? |
| 20:21.43 | eightyeight | not yet. i've been meaning to check it out |
| 20:21.57 | levi | It's pretty interesting. |
| 20:22.47 | levi | interesting in the sense that it's fun to play with, anyway. |
| 20:23.40 | sjansen | I've played with it, but I'm not anywhere near hacker. |
| 20:23.47 | sjansen | Why? |
| 20:23.59 | levi | Just curious what people thought about it. |
| 20:24.42 | levi | I'm writing a little web chat thing, because apparently the web chat is the new blog engine. |
| 20:25.28 | levi | Plus, a site I help to admin has an absolutely terrible PHP chat thing that I want to replace. |
| 20:26.15 | sjansen | When I was playing with it, the node.js developers didn't go out of their way to mention it's limits. |
| 20:27.26 | sjansen | V8 is very much aimed at desktop usage. The current limit is slightly better, but at the time V8 could't use more than 1GB RAM. |
| 20:28.39 | sjansen | If you're willing to accept the added coding overhead, node.js let's you create buffers outside V8's garbage collected memory pool. |
| 20:30.03 | levi | Why would you be creating such huge buffers? |
| 20:30.37 | sjansen | If you want to take advantage of multiple cores, you have to run multiple node.js instance. Thankfully there are third-party tools to start and manage clusters of node.js processes. |
| 20:32.54 | sjansen | Three's some really cool new libraries and tools for node.js, for example Mongoose and Stylus. |
| 20:33.35 | sjansen | But some on the things you might take for granted on other platforms haven't made it to node.js yet. For example a robust postgresql driver. |
| 20:33.42 | levi | Yeah, I've been looking through those. I'm using socket.io, which is nice, although some of the more advanced stuff is broken. |
| 20:34.34 | sjansen | To respond to the memory question, it helps to illustrate my larger thesis: |
| 20:35.25 | sjansen | node.js makes it easier to create things that used to be much harder to create, but node.js is not yet a good general purpose platform. |
| 20:35.30 | goozbach | has heard of node.js |
| 20:35.35 | goozbach | looks up what it is |
| 20:36.00 | sjansen | Unless you bend over backwards, you'll probably end up combining node.js with something else like Rails or Django. |
| 20:36.40 | levi | Nah, none of the cool kids use postgreql anymore, and there are a billion node.js bindings to NoSQL services. :P |
| 20:37.02 | sjansen | indeed |
| 20:37.13 | goozbach | node.js isn't a javascript framework? |
| 20:37.16 | goozbach | wow |
| 20:37.31 | levi | Anyway, if you are trying to create super-huge buffers in node.js, you're probably doing it wrong. |
| 20:37.48 | goozbach | I assumed it was something like prototype or jquery |
| 20:37.59 | levi | Oh, no, not at all like that. |
| 20:38.11 | sjansen | levi: If you've already started writing node.js code, I'm sure you've already noticed its biggest advantage: writing javascript callbacks for everything can be a giant PITA. |
| 20:39.13 | goozbach | is still confused |
| 20:39.22 | goozbach | this guy is sayig that it's javascript |
| 20:39.26 | goozbach | ok |
| 20:39.31 | levi | It's V8, the google Chrome javascript engine, with a thin layer of network libraries around it and a small standard library that makes the javascript environment kind of like a browser's. |
| 20:39.34 | goozbach | listens to the video on node.js |
| 20:39.55 | levi | So you run it on your server, or desktop, or whatever. |
| 20:40.20 | levi | And then you can do stuff with javascript aside from manipulating DOM nodes on web pages. :) |
| 20:40.40 | goozbach | so it's a standalone javascript vm? |
| 20:40.47 | levi | Yes. |
| 20:40.54 | sjansen | levi: My point exactly. node.js is good at sending small text snippets in many directions. Working with binary data is possible, but exponentially painful. And compliance with Javascript's broken Unicode means node.js is less Unicode-capable than a modern platform should be. |
| 20:40.56 | levi | But not *just* that. |
| 20:41.15 | fungus | node.js is an attempt to make javascript a general purpose scripting language. Beyond browsers. |
| 20:41.58 | goozbach | fungus: that's a good description |
| 20:41.59 | levi | sjansen: I would say that it's also good at dealing with humongous streams of data, but only because it doesn't let you do it in the obvious, easy, and wrong way. |
| 20:42.11 | goozbach | closes the hour-long video |
| 20:42.40 | sjansen | Anything you might do in Perl, Python, or Ruby you can bet someone is trying to do it with node.js. In some cases node.js is the best tool, it other cases it's much worse than Perl, Python, or Ruby because it's much less general purpose. |
| 20:42.53 | levi | But yes, writing all your code in continuation-passing-style is kind of painful. |
| 20:43.12 | levi | I don't think it's any less general purpose, it's just younger. |
| 20:46.28 | levi | It also has the advantage of having simple, well-understood, but powerful semantics. |
| 20:46.34 | *** join/#utah fugue88 (~dsowen@173-165-137-65-utah.hfc.comcastbusiness.net) |
| 20:47.37 | levi | I mean, it does have quite a few warts, but they're fairly isolated and not too hard to pick up with a linting tool. |
| 20:48.25 | fungus | node.js is designed primarily for asynchronous programming, event-driven. Similar to Twisted or EventMachine. |
| 20:49.10 | levi | CoffeeScript is a little syntactic sugar layer on top of Javascript that eliminates the warts and gives you cleaner syntax too, but translates in a very straightforward way to pure Javascript. |
| 20:49.10 | fungus | Instead of creating a evented library for a general purpose language, they are bring more general purpose to an evented language. |
| 20:49.55 | levi | There's nothing inherently evented or asynchronous about Javascript. It's just a very common usage. |
| 20:50.35 | sjansen | But there is with node.js |
| 20:50.53 | levi | Well, the core libraries are all programmed in that style, yes. |
| 20:51.51 | sjansen | It might be technically correct to seperate the language from the API for basics like opening a file, but it isn't very helpful in my opinion. |
| 20:52.58 | sjansen | My original interest in node.js was reducing duplication between server and client-side. "Same language everywhere!" Eventually I came to the conclusion that was easier to consider node.js a different language. |
| 20:53.26 | sjansen | s/easier/safer/ |
| 20:53.58 | levi | I guess if you're used to single-implementation batteries-included languages, that's a natural way to look at things. |
| 20:54.40 | fugue88 | Even C's standard library tends to be viewed as somewhat "integral" in many ways, |
| 20:54.42 | levi | I mostly write in languages like C, Scheme, etc. There are a ton of implementations of them, and very little is provided in the core language. |
| 20:54.49 | fugue88 | even though it's plenty easy to never use it or link to it. |
| 20:55.12 | levi | C's standard library is pretty integral, but it doesn't *do* much. |
| 20:55.20 | fugue88 | I think it's just because that's what so many C programmers use (obviously not the embedded guys), |
| 20:55.22 | levi | And you often replace bits of it anyway. |
| 20:55.35 | sjansen | Almost every language is heavily influenced by its I/O model. Javascript is one of the few that were created without I/O as a basic concern. |
| 20:55.36 | fugue88 | and they never really think about doing it differently, though it's really easy with C. |
| 20:55.39 | levi | Everyone uses the standard C library, even embedded guys. |
| 20:56.04 | fugue88 | Well, yeah, but the embedded guys tend to have a better picture of how the lib relates to the language. |
| 20:56.49 | fugue88 | Library too big? Find a smaller alternative. bcopy broken? Write your own. |
| 20:57.03 | fugue88 | Apps programmers are probably less likely to do that. |
| 20:57.12 | fugue88 | For the standard lib, anyway. |
| 20:57.53 | fugue88 | At least, that's all my opinion. I reserve the right to a distorted world view, blah, blah, blah. :) |
| 20:58.23 | levi | Apps programmers probably have some library layer on top of the standard library anyway. |
| 21:00.15 | levi | sjansen: I'm not sure how the I/O model plays so heavily into most languages. Care to elaborate? |
| 21:03.16 | levi | If you mean more of a processing model, i.e. whether you expect your program to be doign batch-style work or interactive-style work, I see what you mean, although I don't think many languages have really done a good job of addressing that at the language level. |
| 21:04.43 | fugue88 | I can see the existence of closures having a big affect on the I/O style of the library... |
| 21:05.42 | sjansen | I'm thinking older langauges with record-based I/O baked in, and experimental languages like Haskell. Although I also count things like Python's generators. |
| 21:05.45 | levi | Well, callbacks have existed for ages without closures, though closures can certainly make using them more convenient. |
| 21:09.49 | levi | The thing is, there wasn't actually a lot of domain-specific stuff programmed into the Javascript language itself. That was all library-level stuff, although it's pretty standard library stuff. |
| 21:10.12 | levi | node.js mimics a lot of the standard library stuff from the browser, at least as far as it makes sense to. |
| 21:10.23 | *** join/#utah lukfugl (~lukfugl@ip65-46-62-74.z62-46-65.customer.algx.net) |
| 21:11.09 | levi | I mean, there's no point in assuming there's a window and DOM in node, since there typically isn't, but you still have a global namespace object that has some of the same library functions on it. |
| 21:13.04 | levi | Anyway, it makes me kind of sad that the node.js guys are so minimalistic about their standard library when their mechanisms for managing all the callbacks that get generated are not very strong. |
| 21:13.34 | levi | The used to have an implementation of futures, but apparently it was buggy so they took it about and switched to straight-up callbacks instead. |
| 21:16.37 | levi | Anyway, it turns out that if you can implement your programs in node.js in a way that doesn't require too much complexity in the callback structure, node.js is pretty awesome. Very lightweight and scalable. But it quickly devolves into spaghetti-code if you have complex logic that has to wait for callbacks in the middle. |
| 21:18.16 | levi | It got me interested in looking at lua again, though. |
| 21:18.58 | sjansen | I assume you're aware of mongrel2? |
| 21:19.07 | *** join/#utah eightyeight (~88@oalug/member/pdpc.supporter.monthlybronze.eightyeight) |
| 21:21.07 | *** join/#utah lukfugl (~lukfugl@ip65-46-62-74.z62-46-65.customer.algx.net) |
| 21:25.07 | *** join/#utah eggyknap_ (~jtolley@74-37-185-142.dsl1.dta.ut.frontiernet.net) |
| 21:26.07 | levi | sjansen: Yes, but that's not specifically what got me interested in looking at it again. Or at least not that alone. |
| 21:27.42 | levi | mongrel2 does sound interesting, though. |
| 21:39.56 | *** join/#utah eggyknap (~jtolley@74-37-185-142.dsl1.dta.ut.frontiernet.net) |
| 21:39.56 | *** join/#utah eggyknap (~jtolley@unaffiliated/eggyknap) |
| 21:55.39 | *** join/#utah findlay (~justin@155-99-201-71.uconnect.utah.edu) |
| 21:56.10 | findlay | how do you prove that the sum of two monatomic functions can have only one local extremum? |
| 21:57.20 | fugue88 | monotonic? |
| 21:57.34 | findlay | yes |
| 21:58.02 | findlay | I think I started thinking 'monoatomic' halfway through |
| 21:58.02 | fugue88 | Seems like such a some could have no local extrema. |
| 21:58.12 | fugue88 | d'oh |
| 21:58.13 | fugue88 | :) |
| 21:58.17 | fugue88 | "such a sum" |
| 21:59.17 | findlay | but they can. Consider 1/x^2 - e^(-x^2) |
| 21:59.51 | findlay | both terms are monatomic on (0,\infty) and they have opposite curvature |
| 22:00.34 | fugue88 | Consider f(x) = x. Monotonic, right? The sum of f and f is 2x. No local extrema, right? |
| 22:00.48 | fugue88 | So, isn't this a proof by counter-example? |
| 22:01.21 | fugue88 | Maybe I've forgotten what monotonic means. |
| 22:02.01 | lukfugl | fugue88: keyword "local" |
| 22:02.09 | findlay | monotonic means strictly {increasing|decreasing} |
| 22:02.21 | findlay | well |
| 22:02.29 | findlay | monotonic means strictly non-{increasing|decreasing} |
| 22:02.56 | fugue88 | lukfugl: Yes, I got that part. A cubic can have a local extrema that's not global. |
| 22:03.30 | fugue88 | Or are you saying that the edge of some given bounds could count as local extrema? |
| 22:03.36 | lukfugl | yes |
| 22:04.41 | fugue88 | Oh, see, because he didn't qualify "over some bound" I assumed no bounds. |
| 22:05.25 | fugue88 | begins thinking of the problem... |
| 22:05.33 | lukfugl | the qualification "local" implies some bounded subset of the domain |
| 22:05.40 | lukfugl | at least to me it does |
| 22:05.54 | lukfugl | and without it, you're right that it doesn't make sense |
| 22:06.08 | fugue88 | Not to me. Like I said, a cubic can have a local extrema even when the domain is unbounded. |
| 22:07.25 | lukfugl | it has local extrema because there exist subsets where there are extrema, even if the full domain has no extrema |
| 22:07.51 | lukfugl | but I think I see what you're saying now, we're defining extrema differently |
| 22:08.43 | fugue88 | We drove findlay away. :~( |
| 22:08.45 | lukfugl | so I think to go back to original question, I'd reword it thus (diff emphasized): "how do you prove that the sum of two monotonic functions can have *at most* one local extremum" |
| 22:09.19 | fugue88 | Is f(x) = 0 monotonic? |
| 22:09.28 | lukfugl | yes |
| 22:09.43 | lukfugl | monotonic non-decreasing (and non-increasing) |
| 22:09.46 | fugue88 | How many extrema does it have? |
| 22:09.48 | lukfugl | but not monotonic increasing |
| 22:09.51 | lukfugl | 0 |
| 22:10.04 | lukfugl | if by extrema you mean point greater than all other points |
| 22:10.05 | fugue88 | I'm too rusty, I think. |
| 22:10.37 | lukfugl | maybe findlay should come back with a more precise version of his question :) |
| 22:10.37 | fugue88 | What about f(x)=sin(x), over a sufficiently large interval to include multiple cycles? |
| 22:10.48 | lukfugl | sin(x) is not monotonic |
| 22:10.59 | lukfugl | unless you restrict the domain |
| 22:11.08 | fugue88 | No, I'm just wondering about extrema here. |
| 22:11.28 | fugue88 | Say, over [0, 10]. |
| 22:11.40 | fugue88 | Would the correct answer be 0? |
| 22:12.24 | lukfugl | sin(x) (x in radians) over [0, 10] has 6 local extrema |
| 22:12.39 | lukfugl | if x is in degrees, it has no local extrema |
| 22:13.15 | fugue88 | ~lart lukfugl for *ever* considering sin's arg to be degrees |
| 22:13.15 | ibot | lowers lukfugl's priority for *ever* considering sin's arg to be degrees |
| 22:13.23 | fugue88 | :) |
| 22:13.30 | lukfugl | hey, I just threw it out there for completeness :) |
| 22:13.58 | fugue88 | Boy, even mathworld isn't so great at defining this. |
| 22:14.19 | fugue88 | From both "extremum" and "maximum" you'd be lead to believe that f(x)=0 has extrema. |
| 22:14.38 | fugue88 | Wait... |
| 22:14.54 | fugue88 | http://mathworld.wolfram.com/ExtremumTest.html |
| 22:14.57 | brac | [Extremum Test -- from Wolfram MathWorld] |
| 22:15.30 | fugue88 | Under "necessary" it says that f''(x) >= 0, etc. |
| 22:15.44 | fugue88 | Under "sufficient" it says f''(x) > 0, etc. |
| 22:15.49 | lukfugl | my definition of extremum: c is a local maximum of f if there exists a â > 0 such that for all x in (c - â, c + â) - {c}, f(x) < f(c) |
| 22:16.04 | lukfugl | (where f(x) is defined) |
| 22:16.13 | fugue88 | So now I wonder at an example that has extrema without satisfying all the "sufficient conditions" |
| 22:16.22 | lukfugl | similar for minimum, and extremum = minimum or maximum |
| 22:16.56 | fugue88 | lukfugl: Okay, that's a nice definition. |
| 22:17.13 | lukfugl | put in laymans terms, is there a non-empty neighborhood around c where c is larger than all neighbors |
| 22:17.21 | lukfugl | *layman's |
| 22:17.40 | fugue88 | laymen's |
| 22:17.45 | fugue88 | Or laypersons' |
| 22:17.45 | lukfugl | laymens' :) |
| 22:18.18 | lukfugl | but I'm pretty sure the saying is based of layman singular, with laymen being an acceptable alternative ;) |
| 22:18.22 | lukfugl | but we digress... |
| 22:18.47 | *** join/#utah bgerber (~gerberb@suse103.zenez.com) |
| 22:20.18 | lukfugl | going back to findlay's problem, he must be talking about a different definition of extrema, because using my definition and assuming monotonic, there will *never* be a local extremum except at the boundaries of the domain (if any) |
| 22:20.43 | fugue88 | Right. |
| 22:20.54 | fugue88 | Maybe the point is to prove that though. |
| 22:21.22 | fugue88 | Then break out cases for the two addends, having their extrema on the same side versus on different sides. |
| 22:22.22 | lukfugl | oh, I feel silly. I didn't read the follow up on findlay's question. the two functions in the sum don't have to be monotonic in the same direction |
| 22:22.29 | lukfugl | ok, I understand the question perfectly now :) |
| 22:22.53 | lukfugl | yeah, I'd probably do proof by cases here |
| 22:22.53 | fugue88 | I'm sure I felt sillier at not remembering how to define extrema. ;) |
| 22:23.13 | fugue88 | ~whalelart paypal |
| 22:23.17 | fugue88 | ~whaleslap paypal |
| 22:23.17 | ibot | ACTION beats paypal upside and over the head with a freakishly huge killer whale named Hugh |
| 22:23.31 | lukfugl | both increasing/non-decreasing? no extrema. same for both decreasing/non-increasing |
| 22:24.23 | fugue88 | Hmm. I think it depends on whether the bounds are open or closed. |
| 22:24.24 | lukfugl | the meat is in f monotonic non-decreasing and g monotonic non-increasing |
| 22:24.40 | lukfugl | true |
| 22:24.49 | lukfugl | s/no extrema/trivial extrema/ |
| 22:25.02 | fugue88 | Gotchya. |
| 22:25.12 | lukfugl | anyways, back to work |
| 22:25.23 | lukfugl | I have to get my math geek on sometimes :) |
| 22:46.09 | *** join/#utah bgerber (~gerberb@suse103.zenez.com) |