00:00:00 --- log: started forth/07.11.17 00:25:10 --- quit: Quartus (Read error: 110 (Connection timed out)) 00:25:29 --- join: Quartus (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 00:25:29 --- mode: ChanServ set +o Quartus 01:05:35 ah, figured out the gc bug 01:08:25 cool 01:08:40 after moving the blocks up, i wasn't updating all data->code references 01:08:46 recently i introduced a new data type for captured callstacks 01:08:52 and i wasn't updating those 01:09:12 yeah, gc bugs usually tend to be of this form. you forget to update a pointer somewhere 01:09:54 oh, that's cool it was introduced recently 01:10:53 yeah 01:29:50 ok my fix works 01:29:50 yay 01:38:30 --- join: Raystm2 (n=Ray@unaffiliated/raystm2) joined #forth 01:44:18 --- join: Raystm2_ (n=Ray@c-24-8-127-241.hsd1.co.comcast.net) joined #forth 01:49:08 --- nick: Raystm2_ -> Raystm2- 01:50:21 --- quit: Raystm2 (Nick collision from services.) 01:50:48 --- nick: Raystm2- -> CosmicRays 01:51:45 --- join: ygrek_ (i=user@gateway/tor/x-68b3b8ed3ecc201d) joined #forth 01:51:50 --- quit: ygrek (Remote closed the connection) 01:51:53 --- nick: CosmicRays -> abnoRmAlYSane 01:53:39 --- nick: abnoRmAlYSane -> Raystm2 02:17:09 --- nick: arke_ -> arke 02:17:15 --- mode: ChanServ set +o arke 02:25:28 --- join: arcus (n=ajt@203.173.187.131) joined #forth 05:06:26 --- join: maht__ (n=maht@85.189.31.174.proweb.managedbroadband.co.uk) joined #forth 05:16:13 --- quit: arcus ("Ex-Chat") 05:17:54 --- quit: maht_ (Read error: 110 (Connection timed out)) 06:10:14 --- join: maht_ (n=maht@85-189-31-174.proweb.managedbroadband.co.uk) joined #forth 06:19:40 --- quit: maht__ (Read error: 110 (Connection timed out)) 07:21:16 --- join: foxchip (n=fox@166.129.11.122) joined #forth 07:52:09 grrrrr 07:52:29 verizon sells 3 or 4 phones with WiFi, but they all run *&@#^$@ windoze 07:53:32 don't think I'm willing to buy a device (or even get one for xmas) that only runs windoze 07:53:47 what a crock 07:53:50 what happened to palm os? 07:54:14 and why is apple being such an asswipe about locking down their devices 07:55:46 to get the $18/phone/month from at&t 08:01:44 http://www.sparkfun.com/commerce/categories.php?cPath=66_69 08:02:14 or http://www.sparkfun.com/commerce/categories.php?cPath=66_68 08:38:16 --- join: crc_ (n=crc@pool-70-110-171-211.phil.east.verizon.net) joined #forth 08:38:19 --- quit: crc (Read error: 101 (Network is unreachable)) 08:38:27 --- nick: crc_ -> crc 08:40:10 --- mode: ChanServ set +o crc 09:06:39 --- quit: ygrek_ (Remote closed the connection) 09:12:11 --- join: ecraven (i=nex@eutyche.swe.uni-linz.ac.at) joined #forth 09:20:34 --- quit: foxchip () 09:28:04 --- join: ygrek_ (i=user@gateway/tor/x-05ec49a81b8fe126) joined #forth 09:29:20 --- quit: ecraven ("bbl") 09:35:26 --- join: mem4tim (n=timlarso@user-12l37rb.cable.mindspring.com) joined #forth 09:47:14 --- quit: mem4tim (Read error: 145 (Connection timed out)) 11:14:24 --- join: deform (n=joe@c-68-61-240-49.hsd1.mi.comcast.net) joined #forth 11:21:20 --- join: foxchip (n=fox@166.129.136.37) joined #forth 11:22:07 --- quit: foxchip (Client Quit) 11:24:32 --- join: foxchip (n=fox@166.129.25.247) joined #forth 11:28:10 --- quit: ygrek_ ("Leaving") 11:39:09 --- quit: Off_Namuh (Remote closed the connection) 11:41:49 Hello 11:43:31 I am at Silicon Valley Forth Day and have setup a skypecast starting at 12:45 local time, that's one hour from now 11:44:39 I haven't done it before and I hope it works. If you have skype you can join in at 12:45 and we have 15 minutes to test it before the first presentation begins after lunch 11:46:36 I will move the microphone and computer up to the front before we start and hopefully the presenters will be loud enough to be heard with this setup. 11:47:46 the agenda for the afternoon is at http:\\www.forth.org/svfig/next.html 11:49:43 if you have skype the 'live' section should show the session if all goes well. 12:00:43 If you have any questions you would like asked of someone please provide them before 12:45 12:06:11 --- join: FunG07 (i=Unix@74-132-210-197.dhcp.insightbb.com) joined #forth 12:40:24 --- join: Off_Namuh (i=GPS@gateway/tor/x-90e60ec41e714309) joined #forth 12:44:54 --- join: iano (n=iosgood@sub26-46.member.dsl-only.net) joined #forth 12:46:53 --- quit: iano (Client Quit) 12:54:12 --- quit: foxchip () 12:54:49 --- join: nighty^ (n=nighty@sushi.rural-networks.com) joined #forth 13:05:41 --- quit: Quartus (Read error: 110 (Connection timed out)) 13:05:59 --- join: Quartus (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 13:05:59 --- mode: ChanServ set +o Quartus 13:43:36 --- join: swestres (n=Dev@Doc-14-47.veberod.com) joined #forth 13:45:24 So, here's the deal. I'm implementing a forth system for an AVR32 system running Linux. I'm thinking of keeping the TOS in the r12 reg. What's the best way to know if the parameter stack is empty? 13:46:05 if the stack pointer is < stack bottom, then the stack is empty and r12 has random data 13:46:06 --- quit: Deformative (Nick collision from services.) 13:46:25 --- nick: deform -> Deformative 13:48:22 Hmm, putting it that way... I kept my mind locked on not letting PSP get beneath S0. 13:48:38 thx 14:10:16 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 14:10:16 --- mode: ChanServ set +o tathi 14:18:07 did anybody see the skypecast, or jeff in person? 14:19:17 fie on skype. use open protocols 14:19:55 i'll use whatever protocols i need to to communicate with the people i want to talk to 14:20:05 although of course i'm happier to talk to them if they're using open protocols ;) 14:20:14 I won't; non-open protocols don't run on my Linux PPC box, generally. :) 14:20:32 ditto 14:20:49 sounds like your choice of linux/ppc is the limiting factor here 14:21:05 If I'm even capable of using proprietary protocols on my computer it's generally because they've been reverse-engineered 14:21:06 yeah...I'm too cheap to buy another computer; this one works just fine. 14:21:10 tathi: you could run linux/86 with wine inside qemu ;) 14:21:40 KragenSitaker: that's a tad slow... 14:22:04 KragenSitaker: might be able to emulate a 20MHz pentium 14:22:32 wine is actually faster than many versions of windows 14:22:39 i used to care about 100% open source but then i realized the linux guys will never get their shit together when it comes to usability, so i may as well not cripple my computing experience 14:23:14 i still care about 100% open source, and it gets closer to practicality every year 14:23:21 slava: neither will the microsoft guys 14:23:30 yeah, but i've been waiting for 10 years 14:23:33 started using linux in 1997 14:23:45 people predicted every year since then that the next year would be the year of "linux on the desktop" 14:24:02 meanwhile distros are getting more bloated and buggier every year and core problems are never fixed 14:24:10 slava: what's your definition of "usability"? 14:24:12 you can go to dell.com and buy a desktop with ubuntu on it 14:24:37 tathi: well in the case of linux, even basic stuff, like dialog boxes which don't have spelling mistakes 14:24:39 yeah, it's true that it's taking a fucking long time 14:24:45 or documentation that's up to date 14:24:55 or not having to edit config files because your gui printer setup tool segfaults 14:25:01 speling isn't usability 14:25:03 ah 14:25:26 documentation might be, but i'm not sure what proprietary desktop system has better documentation than linux 14:25:31 mac os and windows 14:25:32 maybe sunos 4? 14:25:36 freebsd 14:25:45 freebsd does 14:25:55 macos and windows are very poorly documented in my experience 14:26:14 well mac os and windows have good api docs and mac os has good end user docs too 14:26:23 linux programming apis are practically undocumentd 14:26:29 they don't even have man pages for all syscalls 14:26:51 maybe you shouldn't use them all 14:26:57 no, he's right about that 14:27:03 linux also has poor driver support 14:27:10 recently i was tyring to get wireless working on a laptop with ubuntu 14:27:26 there are no gui tools, and the command line 'iwconfig' tools are not documented except for a README file buried in /usr/share/doc 14:27:39 there is no excuse to not have automatic configuration for wireless networks in 2007 14:27:45 i shouldn't have to write a shell script which executes 5 'iwconfig' commands 14:27:52 i should pick my network from a list, and type a password 14:28:18 that's weird. on my debian system, iwconfig has a man page 14:28:19 this isn't about reverse-engineering, its about the people who wrote the tools not caring about usability 14:28:31 i shouldn't have to read a man page or a readme file to configure wireless 14:28:42 on mac os x, i have an icon in the menu bar, i click it and it shows a menu of networks 14:28:44 i choose one and type the password 14:28:45 that's it 14:29:01 agreed. you shouldn't even have to know that there's a program called iwconfig 14:29:12 yup 14:29:15 I'm supprised ubuntu makes you do this 14:29:24 maybe it doesn't anymore 14:31:20 ubuntu wiki says you go to Applications->add/remove and grab something called "Wireless Assistant" 14:31:57 another thing i had trouble with recently was using a usb harddisk that was formatted with ntfs 14:32:12 first, i couldn't mount it as non-root even though i had a 'user' line in fstab, because the ntfs driver is a user-space fs driver and it doesn't support that 14:32:17 also installing the driver in the first place was a pain 14:32:51 i understand ntfs is proprietary, but in this case the code already works, its just not integrated or well-designed 14:33:06 for instance its not integrated with hotplug so if you hotplug an ntfs usb disk it didn't mount 14:33:07 etc 14:35:03 yep, interfacing with proprietary stuff is a PITA 14:35:13 its not even the fact that its proprietary 14:35:39 it's a major contributing factor 14:36:00 if it were and open, well-documented standard then I bet more people would be into hacking on it 14:37:12 Linux does suffer from the "what the uber-geeks don't care about is in bad shape" thing 14:39:18 hardware compatibility is an issue. But it's also impressive in a way 14:39:47 eg I'm running linux on a computer that cannot run windoze (my old mac) and one that cannot run mac os (on old x86 laptop) 14:40:31 but it doesn't support the hardware as well as either native OS 14:40:39 http://www.forth.org/svfig/next.html doesn't seem to have a link to a skypecast, and there is no "live" section 14:41:17 slava: right, because it doesn't have the market share to bully the hardware manufacturers into supporting it 14:41:55 there are an impressive array of topics though 14:42:02 even leaving out the intellasys stuff 14:42:05 i can't get to that above url 14:42:26 afaik macos doesn't have good hardware support because they wrote a ton of drivers, it's because vendors make mac drivers 14:42:31 because mac has a decent market share 14:43:03 linux is fighting an uphill battle, and I think it's making headway 14:43:33 if linux "wins" its really no better than windows 14:43:35 macos has good hardware support because it mostly runs on laptops designed by apple 14:43:44 i strongly disagree 14:43:48 I dissagree 14:43:53 windows is more advanced in many ways too 14:43:59 if linux wins, you'll have the fredom to copy and modify it 14:44:11 if linux gets a market share, than hardware vendors will be sure that their hardware is well supported in linux 14:44:12 i don't really need to be able to modify my kernel 14:44:13 or copy it 14:44:42 i want to write software, not maintain the OS that I use 14:45:05 windows is not trustworthy 14:45:08 the point isn't that you'll spend your weekends maintaining your OS 14:45:19 the point is that nobody can slip things into it against your will 14:45:34 there'd be a hell of a lot less spam in the world if people used linux instead of windows 14:45:38 well, unless your will is so bizarre that you can't find anybody else who shares it 14:45:52 it's true, microsoft treats security as a PR problem 14:46:02 linux is not trustworthy either 14:46:03 but on linux any programmer can patch a hole 14:46:14 afaik most spam comes from machines runing unlicensed coppies of windows that microsoft won't give security updates to 14:46:20 look at php 14:46:25 most insecure piece of shit in the world 14:46:28 linux is buggy, but the linux community's attitude toward security is totally different 14:46:28 php is crap 14:46:44 php is not crap, nor is it insecure 14:46:48 KragenSitaker: its not the attitude that matters its the end result 14:46:49 it just makes it easy to write insecure code 14:46:52 which lots of people do 14:46:59 i think both linux and windows are fundamentally insecure just because they use too much C 14:47:01 but that's because lots of people are writing code 14:47:11 php itself has/had security flaws in the library 14:47:19 yes, occasionally 14:47:22 its not just the fact that it makes it easy to write insecure code 14:47:36 but its level of security flaws in the library is hardly unusual 14:47:40 php has probably had more security problems than anything, except maybe firefox 14:48:03 yes, if you include the security problems in code written in php 14:48:16 linux is far more secure than windows 14:48:26 php is flawed by design 14:48:31 even if it's only because security updates are released within hours or days, not months 14:48:40 because so much of it is written in C, even stuff like date formatting, if the author of that code fucks up, you get a buffer overflow 14:48:41 or years 14:48:42 instead of an exception 14:49:04 have you ever heard of a security flaw in a java application server? almost never 14:49:08 slava: that's true, but a lot of that code is only exposed to people writing PHP code on your server 14:49:08 and certainly never a remote execution hole 14:49:10 or a buffer overflow 14:49:24 rather than to random people on the internet 14:49:27 yeah, but java is slow 14:49:31 that's because the java web stack is written in java and java is a pointer-safe language 14:49:33 java is not slow 14:49:36 java is faster than php 14:49:39 and its better for web development 14:49:48 better designed APIs, more scalable, more robust 14:50:00 java isn't great at all but its much better than php 14:50:03 after 11 years it is now to the level of C in microbenchmark performance --- see the shootout 14:50:10 I thought you were compairing java and C 14:50:15 java is way way slower than C 14:50:25 not according to the shootout 14:50:31 unless you mean startup time 14:51:06 factor should be much better than forth along this axis, since factor is almost entirely pointer-safe 14:52:01 JasonWoof: nobody writes web applications in C though 14:52:08 they use 'slow' languages like ruby, python, perl, and php 14:52:10 except Advogato 14:52:12 and java is much faster than all of those 14:52:16 slava: you were talking about the OS being written in C I believe 14:52:20 java is on part with C these days 14:52:24 for speed 14:52:29 JasonWoof: i'm not suggesting we write OSes in java 14:52:34 KragenSitaker: BS 14:52:41 just that we minimize the amount of code written in C to the bare minimum needed for performance and hardware integration 14:52:55 JasonWoof: for servre-side apps java *is* on par with C/C++ 14:53:05 distributed trading applications used to be mostly C++ and people have migrated to Java for those 14:53:07 JasonWoof: i'll say it a third time --- look at the objective measurements from the shooutout 14:53:35 do they include startup time? 14:53:46 server side apps are only started once a month 14:53:48 startup time is one measurement, and java does horribly on that 14:53:49 startup time is irrelevant for them 14:54:18 I have a hard time believing that one language that has runtime type checking is as fast as one that doesn't 14:54:57 also, I'm not interested in ignoring startup time 14:55:04 C programs also have runtime type checking 14:55:07 except the type system is not part of C 14:55:12 its hand-rolled by programmers over and over again 14:55:15 so it might be even slower 14:55:25 JasonWoof: you should read the Self dissertations 14:55:38 KragenSitaker: perhaps 14:55:49 JasonWoof: for network-bound server side apps, 20% performance hit makes no dfiference anyway 14:55:52 JasonWoof: most of the time Java doesn't have to do runtime type checking except for method dispatch 14:56:02 JasonWoof: if your app spends most of its time waiting for the network... 14:56:12 JasonWoof: so in these situations people choose languages that mximize developer productivity and minimize bugs 14:56:13 I want the programs on my computer that were written in C to startup faster 14:56:14 because it statically proves that the rest of the code is typesafe 14:56:31 method dispatch is done with polymorphic inline caches and dynamic specialization 14:56:31 JasonWoof: also high level languages don't have to run slow 14:56:34 i mean 14:56:36 startup slow 14:56:45 so it's typically an extra couple of instructions per method call 14:56:56 JasonWoof: with factor i can deploy apps as minimal images and the only startup time is the time to load the image from disk 14:57:03 for a development image startup time isn't that bad either, maybe half a second at worst 14:57:17 also sun just GPLed all this shit, so we should be able to look at them 14:57:32 sun just announced that they would 14:57:38 did they actually do it yet? 14:57:45 i think so; i haven't looked 14:57:52 yes they did 14:58:02 you can download the source to the class library and vm and its gpl'd 14:58:06 but its not java 6, its a java 7 prerelease 14:58:16 even systems that use JIT to get good run-time performance don't have to start slow, although sun's java definitely does 14:58:18 but you've been able to get the source for years 14:58:19 cool, wonder how long it'll be before I can install it on my box without licensing bullshit 14:58:28 well you'd have to wait for a linux/ppc port 14:58:34 which is unlikely to happen since nobody will want to write a ppc codegen 14:58:40 there's already a ppc port, isn't there? 14:58:43 nope 14:58:45 or did apple write that? 14:58:47 only apple's proprietary port 14:58:49 which is not open source 14:58:51 that sucks 14:58:56 all linux/ppc java's i've seen were interpreted 14:58:57 hence java being slow in my world... 14:59:01 which makes them unsuitable for real use 14:59:14 hmm... maybe ibm has a ppc java for aix, not sure about linux 14:59:27 I've never had an x86 box as my primary computer 14:59:38 I have my reasons for hating java 14:59:51 JasonWoof: in that case you are correct --- my apologies for assuming you were using x86 14:59:59 you probably don't do the kind of programming java is good at 15:00:10 nope 15:00:16 distributed server-side applications that deal with thousands of connections per second 15:00:28 java sucks for client-side 15:00:44 it also sucks for small-scale web apps because its not as concise as eg, ruby, and a bit harder to get started wwith an app server as opposed to cgi 15:00:45 that's been my impression 15:01:15 slava: it's getting to be reasonable for client-side --- Eclipse shows that 15:01:29 eclipse avoids sun's gui toolkit though 15:01:31 it uses swt 15:01:36 and swt is actually slower than swing on mac os 15:01:53 only client side java apps i use are jedit and limewire 15:02:17 "eclipse uses swt" is hardly an argument against "java is getting to be reasonable for client-side" 15:02:23 it's still in java 15:02:40 in my experience with java programs in the last 10 years: about 3/4 didn't work at all, 3/4 of what's left were horribly broken, but did manage to display something, and a tiny percentage actually functioned correctly, although a little laggy 15:02:50 JasonWoof: is this all on linux/ppc? 15:02:57 no 15:03:01 mostly ppc though 15:03:09 KragenSitaker: the point is that sun's own gui libraries have been crappy historically 15:03:12 mac os 8, 9, 10 beta, linux 15:03:15 KragenSitaker: if you roll your own, you can be fast, sure 15:03:20 mac os 8 and 9 had bad java support 15:03:26 they were stuck with java 1.1 even when 1.4 was out 15:03:30 so did every OS I've ever used much 15:03:35 you use weird OSes 15:03:51 windows, linux/x86 and os x have good java implementations these days 15:03:54 so does solaris 15:04:09 yes, and one of my major issues with proprietary software is that it tends to be shitty for people like me who don't have the normal crap 15:04:17 well, one of my goals with factor is to be more portable than java :) 15:04:35 JasonWoof: its hard to make a business case for paying people to maintain linux/ppc software 15:04:42 free software tends to support everybody it can. proprietary software tends to support whoever makes it the most money, which generally doesn't include me 15:04:45 because almost nobody uses linux/ppc except for hobbyists, who are not in java's target audience 15:04:50 every gui thing from sun has always been crappy, although not necessarily slow 15:04:58 nobody runs high-volume production web sites on linux/ppc, hence no java 15:05:05 KragenSitaker: yes, excpt for openwindows back in the day 15:05:06 sometimes proprietary software is purposely exclusive. monopoly-like situations are good for making money, bad for me 15:05:20 personally i only test the linux/ppc port of factor once a month or so 15:05:23 i hope it works still 15:05:29 slava: you didn't think openwindows sucked? 15:05:37 openwindows was awesome! so was NeWS 15:05:44 i never actually used it --- just read the programming manual 15:05:59 heh 15:06:04 i did actually use xnews once. it was kind of cool but it wasn't a great gui 15:06:08 it was way better than motif at the time 15:06:51 being more portable than java isn't hard ;) 15:06:56 i'm sure you've achieved that already 15:07:03 well, i don't have solaris/sparc yet 15:07:11 but on the other hand i have an arm compiler, which sun's java doesn't 15:07:21 sun's java vm is hard to port 15:07:23 that's a fact 15:07:37 in part because of complexity and in part because of bad coding 15:07:59 there are maybe 1000 ARM chips in the world for every SPARC 15:08:15 heh 15:08:43 once factor 1.0 is out, i'm going to work on 2.0 where the main feature is native threading support. then a sparc port will be interested, since sun now sells 32-core machines for ~$3000 15:08:50 right now its totally pointless 15:09:21 --- join: crest_ (n=crest@p5489E9CF.dip.t-dialin.net) joined #forth 15:09:22 yeah, i think sun's hardware is interesting 15:10:03 neat, you can run server-side web apps in python on a series 60 phone under approximately apache: http://radar.oreilly.com/archives/2006/06/nokia_s60_apache_open_sourced.html 15:10:13 heh 15:10:28 i could run factor-httpd on my phone if i wanted to, but it doesn't have a plan right now so no network connection 15:11:00 --- quit: gnomon ("food") 15:11:12 is windows actually getting enough better that you see it as being better than linux in some way besides compatibility? 15:11:25 ubuntu linux that is 15:11:47 well, win32 has some interesting capabilities not found in linux 15:12:08 i prefer linux programming because its simpler 15:12:17 but win32 doesn't totally stink 15:12:42 to me a windows machine (without a super geek maintaining it) sooner or later becomes nearly or completely disfunctional 15:12:52 yup 15:12:53 for one thing, win32 is supported both on windows and on linux ;) 15:13:14 last time i tried running factor/win32 in wine, it crashed because wine didn't support iocp's 15:13:45 that sucks. wine should be fixed. 15:14:36 maybe it supports io completino ports by now 15:14:40 i tested a few years ago 15:14:53 pretty fundamental feature 15:14:58 its like windows select() 15:15:56 for some reason people find some of my views really controversial 15:16:07 java people think i'm a lisp/forth fanboy and ruby people think i'm a java fanboy 15:16:16 linux people think i'm a linux hater who loves windows and windows people think i'm a mac os zealot 15:20:07 --- quit: crest__ (Read error: 110 (Connection timed out)) 15:22:42 slava: everybody wants you to join their fundamentalism? 15:22:47 yes 15:23:07 only fundamentalism i subscribe to these days is reflective, extensible stack-based programming 15:23:09 the good stuff 15:25:39 :) 15:26:59 hehe 15:28:07 you don't even believe in fundamentalism about pointer-safety? :) 15:28:54 depends on the task at hand 15:29:19 even factor has pointer-unsafe words exposed through low-level vocabularies 15:29:22 for interfacing with C data 15:29:34 we build pointer-safe abstractions on top 15:37:05 --- join: iano (n=iosgood@sub26-46.member.dsl-only.net) joined #forth 15:37:18 --- quit: FunG07 (Remote closed the connection) 15:46:22 Python is a little more fundamentalist about that --- and I think loses a lot thereby 15:46:54 there's a ctypes module for it that provides pointer-unsafe words for interfacing with C libraries, but the standard method is to write a Python extension in C 15:47:11 and ctypes is not in the main distribution 15:48:39 sbcl takes the approach of ctypes and factor 15:58:14 yup 15:58:19 factor's ffi is even better than sbcl's though 15:58:29 sbcl is missing some features, like passing and returning structs by value 15:58:59 also i'm not sure if sbcl supports callbacks (passing lisp code to c as a function pointer) and indirect calls (c gives you a function pointer, you call it via ffi) 15:59:01 factor does 15:59:26 i highly value interoperability 15:59:37 even if the standards involved stink (C ABI's, XML, etc) 16:12:35 --- quit: Off_Namuh (Remote closed the connection) 16:18:22 --- join: Off_Namuh (i=GPS@gateway/tor/x-bbc9d75bffc8c2b0) joined #forth 16:26:21 --- quit: tathi ("leaving") 16:56:36 --- quit: iano (Nick collision from services.) 17:08:34 --- quit: Quartus () 17:08:49 --- join: Quartus (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 17:08:49 --- mode: ChanServ set +o Quartus 17:15:48 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #forth 17:15:48 --- mode: ChanServ set +o tathi 17:26:07 Hey tathi. 17:26:37 hi Quartus 17:27:57 How's the world treating you? 17:28:11 Pretty well, and you? 17:28:16 Can't complain. 17:28:40 What are you up to these days? 17:29:16 Some contract work; some writing. Not enough downtime. :) 17:30:16 The lack of downtime has been going on for a while, no? 17:30:18 --- quit: Off_Namuh (Remote closed the connection) 17:31:33 Yes; have to remedy that soon. 17:31:36 How about you? 17:32:18 Not much downtime either; getting the farm ready for the real cold weather, and working on the house so we'll have space to put guests. 17:32:28 (US Thanksgiving coming up next week) 17:32:43 Big occasion down there. 17:32:48 After that it will slow down a bit, probably. 17:33:27 Yeah, my family sort of ignores a lot of holidays, but not this one. 17:33:34 A shame to skip the feasts. :) 17:33:53 That's right; any holiday involving a big feast is OK by me. :) 17:34:05 :) 17:36:23 Right now I'm slacking off though; mucking about with Forth. 17:36:34 Good deal. Anything interesting? 17:37:43 Nothing useful at all. :) 17:37:47 :) 17:37:52 Playing around with writing a text editor again, actually. 17:40:50 I was talking to JasonWoof about something akin to that the other day; using FSMs for the purpose. 17:41:58 I was thinking about doing that. Only a couple of small modes, but it seemed like it might be worth it anyway. 17:43:17 Gives you endless ability to extend it. 17:43:44 Add extra event types, etc. 17:43:52 Yeah. 17:49:30 --- join: Quartus____ (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 17:54:37 --- join: CyberKid (i=HydraIRC@c-68-63-244-236.hsd1.ky.comcast.net) joined #forth 17:57:50 wish I understood the concept better 18:00:42 I'll show you when you get here. :P 18:00:43 Bit of experimentation should get you there. 18:01:34 :) I was just going to say I should get you to explain when I'm up there for turkey day 18:02:11 thanksgiving isn't the biggest US holliday, it's just my favorite because of our tradition of going to tathi's house 18:02:37 very cool family 18:02:44 and we don't see them much since they live 4-5 hours away 18:03:24 Sounds like a good Thanksgiving coming up, then. :) 18:03:32 I pretty much ignore the hollidays except for thanksgiving and xmas 18:03:47 and I have to pay at least a little attention to newyears or I sign my checks wrong :) 18:03:51 yeah 18:04:04 heh 18:04:12 4 full days :) 18:04:19 I better get some work done soon! 18:04:25 I slacked off this week 18:05:23 ok, believe it or not, I'm going to sleep 18:05:29 --- join: nighty__ (n=nighty@sushi.rural-networks.com) joined #forth 18:05:29 ttyl 18:06:45 ciao :) 18:06:45 --- quit: Quartus (Read error: 110 (Connection timed out)) 18:06:48 --- quit: Quartus____ () 18:07:03 --- join: Quartus (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 18:07:03 --- mode: ChanServ set +o Quartus 18:12:29 --- quit: CyberSpace (Read error: 110 (Connection timed out)) 18:12:29 --- nick: CyberKid -> CyberSpace 18:15:05 --- quit: tathi ("Goodnight all") 18:17:14 --- quit: nighty^ (Read error: 113 (No route to host)) 18:46:50 --- join: nighty^ (n=nighty@sushi.rural-networks.com) joined #forth 19:00:17 --- quit: cmeme (Excess Flood) 19:01:17 --- quit: nighty__ (No route to host) 19:02:25 --- join: cmeme (n=cmeme@boa.b9.com) joined #forth 19:07:12 --- quit: cmeme (Client Quit) 19:07:46 --- join: cmeme (n=cmeme@boa.b9.com) joined #forth 19:12:33 --- quit: cmeme (Client Quit) 19:13:10 --- join: cmeme (n=cmeme@boa.b9.com) joined #forth 19:27:03 --- join: arke_ (n=arke@p54A7EE9A.dip.t-dialin.net) joined #forth 19:37:05 --- join: immortial (n=nomail@85.97.128.200) joined #forth 19:37:11 --- part: immortial left #forth 19:44:46 --- quit: arke (Read error: 110 (Connection timed out)) 20:07:11 --- join: Off_Namuh (i=GPS@gateway/tor/x-49ab4db57e7d4580) joined #forth 20:14:36 --- quit: Quartus (Read error: 110 (Connection timed out)) 20:14:46 --- join: Quartus (n=neal@CPE0001023f6e4f-CM001947482b20.cpe.net.cable.rogers.com) joined #forth 20:14:46 --- mode: ChanServ set +o Quartus 20:35:28 here is a complete factor program to convert a file for 32-bit words from little to big endian or vice versa: 20:35:29 USING: io.mmap io.files math.functions splitting sequences ; 20:35:29 : byte-swap ( file -- ) 20:35:29 dup file-length 4 align [ 20:35:30 4 [ reverse-here ] each 20:35:30 ] with-mapped-file ; 21:07:51 slava: does it work for files that are bigger than virtual memory? 21:08:06 no 21:08:14 you can use sequential read/write and it just makes it a bit longer 21:08:30 but the each only makes one pass over the file? 21:08:48 (forgive my ignorance of factor) 21:08:51 yes 21:09:06 { 1 2 3 4 5 6 } 2 [ reverse ] map . 21:09:17 => { { 2 1 } { 4 3 } { 6 5 } } 21:09:28 { 1 2 3 4 5 6 } dup 2 [ reverse-here ] each . 21:09:33 => { 2 1 4 3 6 5 } 21:09:49 oh! so it's actually updating the file in place 21:10:12 yup 21:10:19 reverse-here reverses a sequence in place 21:10:28 takes a sequence and outputs a 'virtual' sequence of slices 21:10:36 wiht-mapped-file presents a file as a sequence 21:12:11 well, that's pretty nifty 21:12:19