00:10:51 So, Windows on ARM is gonna be called Windows RT? 00:10:53 WTF? 00:11:04 Thaaaaats not confusing stuff. 00:12:33 * Afal part 00:24:30 so I can't have Windows RT on my desktop? 00:24:50 You can use WinRT on your desktop, but no, not Windows RT 00:24:54 *blink* 00:33:31 wwwwttttffff 00:33:38 * madewokherd quit (Ping timeout: 260 seconds) 00:33:51 I actually think they *should* have called it Windows TF 00:34:50 LOLOL 00:34:56 reminds me of Windows XP "built on Windows NT Technology" - NT stands for New Technology...New Technology technology... also who named NT? New Techology!?!? Great name! err.... 00:35:32 Well, NT was originally named for OS/2 3.0 NT (way back when MS and IBM were both working on OS/2) 00:37:21 An acronym for New Technology associated with a software product is a FAIL IMO 00:39:49 Also - what is the next version of Windows RT going to be called? Windows RT 2? (which is Windows 9)... 00:40:08 rt2d2? 00:40:11 probably Windows RT R2 00:40:15 lol 00:40:19 Windows RT 2 (built on Windows 9 Technology) 00:40:26 .NET 00:40:42 mobile? 00:40:49 Live 00:41:01 2013 00:41:13 windows rt r2 .NET mobile live 2014? 00:41:16 Wave 2.5 00:43:14 in all fairness it is a hard product to name - because ARM most likely won't only be used for mobile products going into the future - so calling it Windows 8 Tablet or Mobile or whatever wouldn't work... 00:43:51 i think 'Windows 8 Metro' could be ok... 00:44:17 I guess they already pissed away "Windows Mobile" 00:44:57 yeah...what is the diff between Windows 8 Mobile and Windows Phone 8?? confusing for the user 00:45:30 Well, and Windows Mobile was a version of Windows CE. 00:46:38 if they called Windows RT 'Windows 8 Mobile' - and then they had 'Windows Phone 8' - that would be confusing 01:55:49 * madewokherd joined 02:55:05 * Scotis is set as away : Reason(evening) 03:05:46 * gix quit (Ping timeout: 246 seconds) 03:11:23 * gix joined 03:41:26 what does WinRT stand for again? 03:45:02 RunTime 03:45:16 where does the Win come from? 03:45:22 Windows 03:45:55 so, Windows RunTime abbreviates to WinRT but not Windows RT, but something else abbreviates to the second thing? 03:46:09 Yep. 03:46:21 Hey, don't look at me, I think it's all retarded 03:46:22 do we know what the other RT is yet? 03:46:26 I didn't see it anywhere 03:46:42 It's just like XP I guess 03:46:45 *sigh* 03:48:17 I'm sure that stood for something 03:49:16 wikipedia says "eXPerience" 03:51:17 * gcs-develop joined 03:52:49 whatever it's called, I just really hope it's not locked down 03:52:56 good night 03:53:28 * madewokherd quit (Remote host closed the connection) 04:10:52 <@FearTheCowboy> Well, NT was originally named for OS/2 3.0 NT <--- Wasn't it much earlier from Cutler working on the N10 CPU, later moving to Microsoft who decided to call it NT? 04:11:36 * dmex is now confused about history 05:06:54 * TReKiE quit (Read error: Connection reset by peer) 05:09:05 * TReKiE joined 06:10:29 * ender joined 06:35:20 are you planning builder bots? 06:35:43 I would gladly donate if I didn't have to keep VMs with various Windowses and various MSVC versions 07:17:31 * dmex quit (Quit: internet implosion) 08:05:43 * ender quit (Read error: Connection reset by peer) 08:06:24 * ender joined 08:22:46 * gcs-develop quit (Read error: Connection reset by peer) 09:32:45 * ssam2 joined 09:56:41 * Jonny5 joined 09:56:41 * Jonny5 quit (Changing host) 09:56:41 * Jonny5 joined 10:24:20 * piscisaureus_ joined 12:06:48 http://www.theregister.co.uk/2012/04/17/windows_8_versions_announced_windows_rt_for_tablets/ 12:25:52 * dmex joined 12:45:28 mpreisler -> Yes, we have a build automation server where we have vc6 - vc10 and gcc. We can spin up as many build VMs as we need-- we've got two 64proc servers which can host one hell of a lot of VMs. 12:48:00 FearTheCowboy: any chance PH will be packaged in coapp? :P 12:49:24 (seeing as it's a required base package) 12:49:27 ;) 12:51:53 I wonder if there will commercials for Windows RT with a bunch of people saying ... "Hey, this wasn't my idea!" 12:52:01 PH? 12:58:55 ProcessHacker 12:59:08 I would agree abotu WinRT 12:59:09 Ah, right. 12:59:32 IMO Windows 8 will fail more than Vista... its sad. 12:59:56 Yes, I fully intend on building a PH package. Actually, if someone wanted to make a ptk build script and an .autopkg file, I'd publish it instantly. 13:00:15 * dmex thinks there was a autopkg somewhere 13:00:31 .autopkg is our format for making packages 13:00:36 ah 13:00:41 it's not terribly difficult. 13:01:03 I was thinking of the PortableApps.com format 13:01:14 ah. no, unrelated. 13:01:19 * dmex removed a bunch of those files from the server last year 13:01:47 lemme take a quick look at the source... 13:02:56 oh hell, this'll be easy to package 13:04:38 do you know if the .sys files are embedded in the EXE, or are they installed seperately? 13:05:01 seperate 13:05:18 it's in the zip 13:05:55 KPH.sys just needs to be in the same dir to load the kernel driver, doesn't actually matter if it's there or not but some features won't function 13:06:22 if UAC is enabled and PH isn't elevated, it's not loaded 13:06:38 Ok, that's a cakewalk. 13:06:49 * dmex looks at the 8 police in his driveway... 13:06:58 wait, make that 9 13:07:05 about to get busted? 13:07:07 lol 13:07:58 my brother was using his skateboard on the road........ 13:08:27 9 police were an appropriate response... 13:08:29 that's needs a 9 police ? 13:08:48 aparently 13:08:59 holy cow. If all stuff built as cleanly as PH, this project would be done already. 13:09:11 LOL 13:09:14 yes! 13:10:27 layout is simple 13:10:39 ph.exe and kph.sys in the install dir 13:10:56 folder called plugins, with plugins in that folder 13:11:15 the bin isn't setup for that 13:11:50 and the kph.sys just builds straight away with the ddk ? 13:11:54 so x86 and x64 need differnet scripts to ensure the X64 plugins are in the /plugins folder 13:11:55 yep 13:12:50 vice-versa X86 plugins need to be in the /plugins folder instead of the X64 plugins 13:13:05 (if you make sense from that) 13:13:08 :P 13:14:05 is there any reason to have x86 plugins installed in an x64 version? 13:14:14 no 13:14:25 if I make two packages, that should be correct, right? 13:14:29 yep 13:14:36 easy. as. pie. 13:14:59 although, now I gotta go make sure that ptk sets up the ddk build environment correctly. that's a 10 minute fiex. 13:15:07 fix 13:15:53 Hell, I should have this done & packaged before my third espresso this morning. 13:17:11 * dmex just broke the project page :| 13:31:32 * vpovirk joined 13:35:15 FearTheCowboy: are you going to build/sign the driver/exe/package? 13:38:20 Yeah, 13:38:20 all coapp packages be signed yes 13:38:24 we sign EVERYTHING 13:38:44 ack - speaking of 13:38:51 tim around? 13:38:56 not yet. 13:39:02 it's still EAARLY 13:39:05 can we add docs for how to generate a cert for self-signing? 13:39:10 it's almost 10 for me - LOL 13:39:25 and I have a 6am every day to make breakfast and do laundry - sigh 13:39:32 (9:30pm here) 13:40:04 it's pretty easy to generate a self-signed cert now; I'll have him insert that into tthe docs somewhere. 13:40:16 I wrote a script for that; there's actually a package! 13:40:23 hehe, I'm sure 13:40:35 I'm just trying to make sure the docs will do step by step for EVERYTHING 13:40:44 Yeah 13:40:44 so any idiot (me) can do a package easily 13:41:07 the packaging docs mention you need a cert and a self-signed one is fine... but not how to make one ;) 13:44:35 FearTheCowboy: heres a question, what happens when an app includes an updater? 13:45:14 well, since CoApp is handling updating of packages, updaters are really quite useless 13:45:34 we handle updates differently than most do anyway. 13:45:42 you can removed the updater.dll plugin from PH if thats the case 13:45:45 ok 13:45:52 removed/remove 13:45:53 will do 13:46:13 nice and handy :) 13:46:42 how will it work with other projects though 13:47:06 if they update themselfs, won't that break the package? 13:47:08 shallow fork 13:47:10 hack it out 13:47:11 sigh 13:47:15 well, we just won't include whatever update is in their project 13:47:24 most of the time, thats pretty easy 13:47:41 integrated updaters are a real pain in the ass 13:48:00 yup 13:51:35 it was one reason we made extras into plugins 13:51:55 some might only want the basics, some everything 13:52:15 it also helps if someone gets a crash, they can remove the plugins 13:53:33 well, in this case, I'll bundle in what's there. 13:54:04 except for the updater 13:54:49 ;) 13:56:03 oop. time to wake the wife and kids. Makin' waffles this morning... 13:57:11 heh 13:57:27 I do breakfast, husband does supper ;) works out good 14:01:36 * _Andrew joined 14:01:36 * _Andrew quit (Changing host) 14:01:36 * _Andrew joined 15:00:31 * virmitio joined 15:01:29 FearTheCowboy: cool that build bots are in the grand scheme of things :) I seriously need to try to build some packages, I was midway but had to postpone due to other obligations 15:01:45 building windows SDKs from Linux without Windows and MSVC licenses is the dream ;-) 15:12:24 hehe 15:12:32 push code to repo - wait for coapp build? 15:12:47 that's the likely scenario 15:13:18 might be smart to have testing builds/snapshots available then too though ;) 15:13:24 * virmitio is contemplating really crazy schemes for build automation 15:13:50 sweet 15:13:56 virmitio: did you see my earlier request? 15:14:04 the packaging docs mention you need a cert and a self-signed one is fine... but not how to make one 15:14:10 probably not, I only just got here 15:14:18 virmitio and I were just talking about making a feed for test/beta packages 15:14:31 auroraeosrose: yeah, I'm realizing that 15:14:46 and setting the threshold for packages on install to default to 0 (so that packages where stable < 0 would not be picked up automatically) 15:14:59 yeah.... 15:15:39 also - any examples of .buildinfo files that have to build and then run an .exe for testing? 15:15:48 just name a repo already doing it and I can look it up ;) 15:16:11 is it an exe made during the build? 15:16:21 not usually 15:16:35 it appears it builds as it's own test .exe and is then run 15:16:38 (weee!) 15:16:43 it depends on the build! 15:16:54 fair 15:16:56 this is for re2c 15:17:00 so you build re2c 15:17:07 then you run it to generate test .c files 15:17:12 then compile and run them to test 15:17:13 one sec, I'll find one where I needed to do something similar 15:17:37 ok, maybe not as similar as desired for that, but it can be managed 15:17:56 bison will have the same "fun" 15:24:49 * auroraeosrose1 joined 15:26:11 * auroraeosrose quit (Ping timeout: 240 seconds) 15:27:40 * ender quit (Read error: Connection reset by peer) 15:28:14 * ender joined 15:29:47 * auroraeosrose1 quit (Ping timeout: 250 seconds) 15:31:30 FearTheCowboy: searchable feed? 15:32:02 packages.xml?q=partialpackagename 15:35:20 not yet. 15:35:50 We'll be able to include that in the future; right now all filtering/searching is done on the client side. 15:36:50 * FearTheCowboy is just adding in wdk support for compiling with the wdk kit right now. 15:37:57 * auroraeosrose joined 15:43:03 http://windowsteamblog.com/windows/b/bloggingwindows/archive/2012/04/16/announcing-the-windows-8-editions.aspx 15:43:11 yay back to two versions 15:43:14 Four 15:43:24 ah, ARM 15:43:29 Windows, WindowsPro, Windows RT and Windows Enterprise 15:43:41 I love how the blog post skipped the 'enterprise' 15:43:43 FFS 15:44:19 like I said last night, I guess we should be thankful they didn't call it "Windows TF" 15:44:23 for the ARM version 15:44:36 XP is Win5.1 right? 15:44:40 speaking of, if there's a bug in the SDK, are you able to have it looked at? 15:44:41 yes 15:44:57 Is 2003 5.2 or just a 5.1 variant? 15:45:12 x64 is 5.2 15:45:25 not entirely sure about 2k3 x86 15:45:25 Truthfully, it's faster to log SDK bugs thru the visual studio section on connect.microsoft.com 15:45:32 yeah already have 15:45:35 that's where I actually end up putting mine 15:45:45 internally, I don't get near the response :S 15:45:54 Found an SDK bug did you? 15:46:00 * dmex isn't surprised 15:46:05 virmitio and I found one last week too. 15:46:22 yeah, the Filtering Platform headers have a bunch of spaces in the macros 15:46:40 spaces/newlines 15:46:55 niiice 15:47:17 works great, ensures you don't use them 15:47:55 ffs. 15:48:17 I've been working on a new PH plugin to show windows firewall/filtering platform activity like dropped/blocked/accepted connections 15:48:46 Last week we found that if you define "NOUSER" and "ISOLATION_AWARE_ENABLED" it still trys to declare functions you don't ask for (and didn't get the structs for) ... 15:49:00 ooh. that'll be cool 15:50:30 what I like the most is how C is the most popular programming language by far, and VS doesn't even properly support C98 15:50:58 I even have to #undef __cplusplus just so intellisence shows the C defines 15:51:13 C99? 15:51:21 lol 15:51:38 We're all a tad pissed about the lack of C99 support 15:52:05 no plans to support the next standard either (I think it's finalized this year) 15:52:21 "Hi: the feedback we receive from our users again and again is that we should focus on C++ (especially the new upcoming C++-0x Standard) - and that is what we do. Having said that we have been, slowly, adding C-99 features over time - and we may continue to add a few more features over time. 15:52:21 Jonathan Caves 15:52:21 Visual C++ Compiler Team" 15:52:40 http://visualstudio.uservoice.com/forums/121579-visual-studio/suggestions/2089423-c99-support 15:52:42 VOTE HERE 15:52:47 for you all to laugh at 15:52:48 http://news.php.net/php.internals/60058 15:53:48 http://news.php.net/php.internals/60075 15:54:06 auroraeosrose: "Less bitching, more fixing" embodied in stats? 15:54:14 http://en.wikipedia.org/wiki/C11_(C_standard_revision) 15:54:15 I love rasmus 15:54:25 wait, it was finalized last year? 15:54:53 Yeah, C11 is done 15:54:57 December 8, 2011 15:55:06 most of it's in Dev11 (vc2012?) 15:55:28 acutally, they did not a bad job getting a ton in VC10 for the time. 15:55:37 * wwahammy_ joined 15:55:50 but the lack of C99 makes it impossible to build stuff like VLC with Visual Studio 15:56:22 the intellisense issues with C/C++/MC++ was good too 15:56:35 nothing for mc, ctrl+space for C 15:56:43 Yeah, there's a few worth voting on. 15:56:52 * auroraeosrose1 joined 15:56:56 vs11 fixed all those 15:57:21 EVERYONE DRINK >> "auroraeosrose1 (...) has joined #coapp" 15:57:34 * FearTheCowboy likes to play the #coapp drinking game 15:58:00 FearTheCowboy: is this one that laughs at auroraeosrose's poor internets? 15:58:09 Yep 15:58:15 heh 15:58:37 * auroraeosrose quit (Ping timeout: 245 seconds) 15:58:45 * virmitio needs to go get another drink... 15:59:54 * auroraeosrose joined 16:00:31 auroraeosrose -> it's official now: https://github.com/coapp/coapp.org/wiki/Coapp-drinking-game 16:00:45 tee-hee 16:01:28 * auroraeosrose1 quit (Ping timeout: 276 seconds) 16:02:23 * auroraeosrose1 joined 16:02:37 this is getting ridiculous 16:02:39 * virmitio needs to go get yet another drink... 16:02:44 hate hate drivers 16:02:47 fucking ARGH 16:04:20 * auroraeosrose quit (Ping timeout: 252 seconds) 16:04:29 how come there's nothing from virmitio in there ;) 16:04:33 * auroraeosrose1 changed nick to auroraeosrose 16:05:02 * virmitio is too busy playing the drinking game to participate in writing it 16:06:10 Feel free to add more rules 16:07:22 but wait - there's more 16:07:27 oooh ooh -add that one 16:08:37 updated the rules 16:08:45 nice 16:08:58 #5 is going to keep me well hydrated 16:09:26 updated ;) 16:12:39 this cannot end well 16:12:50 oh it'll end well all right 16:12:52 on the floor! 16:18:59 the really sweet thing about ptk is how you never have to 'set' the environment ... it takes care of all that crap no matter what you're building. 16:19:43 If I'm building a house, does it also 'set' the environment? 16:19:47 yep. 16:19:54 Awsome 16:20:13 * dmex needs a spell check 16:20:17 Unfortunately, the SDK for building houses is broken :p 16:20:57 Damn, I wanted to set the environment to warm with seaside views... 16:22:06 you can't get mingw AND windows sdk at the same time 16:22:33 which is probably a good thing 16:23:37 Well, you can set the SDK, but I beleive that pretty much everything in the environment gets overwritten by mingw, so , yeah, you can't. 16:23:53 if there is something that I should do to allow that sort of thing, I can do that tho' 16:28:25 can't make something like the Windows SDK Configuration Tool? 16:28:54 (that hasn't been broken for almost two years) 16:29:13 ptk does everything we need. 16:29:22 you dont' need such things with ptk builds anymore at all 16:29:41 ah 16:30:24 ptk automatically discovers and configures the environment for VC6 -> VC10 and GCC, SDKs 6,7,7.1 and the WDK 17:04:53 has anybody tried to activate Remote Desktop Licensing Server on their MSDN subscription? :) 17:05:16 not yet 17:08:43 i've been trying for the last 2 days. yesterday no activations could be done at all, because they got some new systems, and they don't work, and today they were using the old systems again (because those do work), and it turns out that nobody knows about how to do such activation 17:23:35 dmex -> I've nearly got the PH package done. 17:23:47 o_O 17:24:20 while you've been doing that, I've been updating the rapps package manager for ReactOS: http://processhacker.sourceforge.net/test/index.php 17:24:22 :P 17:25:46 the sad thing, it downloads a cab file full of txt files and parses out package information using a custom delimiter format.... 17:26:01 eeeew 17:26:23 yeah, I decided to help em out and update it 17:27:27 they love enough to sign our kernel driver for x86, yet don't implement NT keyed events enabling it to run on ROS 17:27:34 PH^ 17:28:13 it would actually run under wine on linux if it also implemented keyed events 17:36:38 dmex -> so, for the x64 version, the plugins subdirectory is still called 'plugins'? 17:41:17 yes 17:43:37 In the zip\plugins directory, theres two folders, x86, x64... if you're using the x64 executable, you must extract the zip\plugins\x64 dll's into the \plugins folder 17:43:58 they love enough to sign our kernel driver for x86, yet don't implement NT keyed events enabling it to run on ROS <- which driver? 17:44:10 KProcessHacker.sys 17:44:14 oh 17:44:33 typo, not x86, it's for x64 17:44:35 * ender has a kmode-enabled certificate, too 17:45:04 with windows 8 we're considing one for the executable 17:45:07 (and i sign open-source drivers on request) 17:45:16 you don't sign the executable? 17:45:33 that fullscreen smartscreen thing on windows 8 is bloody annoying to run unsigned executables 17:45:37 my build for processhacker is signing everything now. 17:45:40 no, we don't have our own cert 17:45:53 at least get certum's certificate, even though the process is slightly annoying 17:46:07 I had trouble with the german 17:46:14 (if thats the same site) 17:46:21 it's polish 17:47:10 I wonder what happened to verisign's annoucement just over a year ago about offering free certs for OSS 17:47:29 verisign got bought by symantec. 17:47:36 it's a dead-end 17:47:43 verisign? yes. 17:48:21 yep 17:48:26 wait, was it your or someone else from Microsoft who mentioned something for OSS certs? 17:48:33 Yeah, that's me 17:48:45 We're going to add in support for the CACert root fairly soon 17:49:00 so apps can be signed with certs you can get for free with CACert 17:49:07 via coapp or the rootcert cab update? 17:50:17 We're going to integrate their root cert into our bootstrapper 17:51:27 ah, I was talking about the one CAPI silently downloads on windows every so often 17:52:02 no, we're not going to get it into the root cert update process. 17:54:02 auroraeosrose: got the doc up for making self-signed certs and added links to the forking instructions 17:55:02 i've been using certum's free certificates for my open-source stuff since XP SP2 was released. it's slightly annoying to sign up, since pages randomly switch to polish, but other than that, i never had any problems 17:55:47 I think the other thing was their certs showing our full name 17:55:57 (i've also been sending them the same scan of my passport for the past 7 years) 17:59:09 btw, speaking of Process Hacker, would it be possible to add setting of text color, too? 18:03:50 to where? 18:04:16 (there's alot of text elements) 18:05:49 I have been playing with text color though: http://i.imgur.com/Ztga8.png 18:06:22 dmex -> what is one of the functions that needs the kph driver? 18:06:28 right now only the background can be set for hilighting, and colour seems to be automatically set to black or white, depending on background. i like dark backgrounds, but white text it a bit too bright for me 18:06:56 ( this is my colour scheme: http://eternallybored.org/imgs/thebat.png ) 18:07:10 * ender waits for "my eyes!" screams 18:07:12 FearTheCowboy: if KPH is available, all NT functions are routed via KPH 18:07:26 if it's not available, we use the NTAPI 18:07:31 how can I tell if it's available? 18:07:38 (just checking my build...) 18:07:43 ender: looks pretty pleasant to me 18:08:07 virmitio: the only two reactions i get are "that's nice" and "my eyes!" 18:08:11 there's a few ways, depends what is appropriate 18:08:40 does the 'disk' tab work without the driver? 18:10:53 Yes, also to detect KPH you could CreateFile > \Device\KProcessHacker2, enum the service list or hit the services key for the count(??) HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\services\KProcessHacker2\Enum 18:11:04 (speaking of colours, i wouldn't mind being able to change all other bright backgrounds to something dark, but since i rarely have to look at those lists, they're nowhere nearly as annoying) 18:11:22 (note, I think thats the wrong ControlSet key) 18:11:31 CurrentControlSet 18:11:52 (though AFAIK, CCS is usually just a symlink to ControlSet001) 18:12:21 or, I could just look in the services tab in ph and find the kprocesshacker device driver :) 18:12:51 it appears my PH package works :) 18:12:56 I'l still surprised regedit doesn't include symlink support... god help you if you delete a simlink key mapped to something critical... 18:13:09 hah 18:13:13 dmex -> found a bug in PH... 18:13:21 o_O? 18:13:30 btw, why don't you extract the driver from resource on demand? 18:13:48 too bloaty 18:14:25 when searching for the relative path for the driver + plugins, you need get the actual path of the module itself ... if I run it from a symlink, it's not finding the driver/plugins 18:14:58 1mb exe becomes > 4mb, also no guarantee the extraction will work properly on some setups 18:15:13 interesting 18:16:06 it's a common bug when running something from a symlink 18:16:13 dmex: but then i'd be able to just put the executable on my server and run it (almost) directly :) 18:16:19 (i love live.sysinternals.com) 18:16:43 it's also smb accessible 18:16:46 dmex: why wouldn't it work reliably? just write to %TEMP% 18:16:47 ender -> you'll have a coapp package in a few minutes 18:17:27 ender: situations like the symlink bug 18:17:50 hmm, let's see what happens if i symlink lspci 18:18:10 btw, you can enter \\live.sysinternals.com in explorer and browse the netbios (WebDAV??) share/run the executables directly 18:18:22 ...works fine 18:18:51 (though with lspci, the driver is loaded by a dll from it's own resource) 18:20:07 hah, Service DirectIO has been deleted 18:20:27 FearTheCowboy: what are repo steps? 18:20:41 symlink a dir, extract and run PH from the symlink? 18:20:53 no, 18:21:01 mkdir c: 18:21:02 er 18:21:04 one sec 18:22:17 you would have more problems than PH if you needed to mkdir C: 18:22:19 :P 18:22:26 1. create a directory c:\foo 18:22:26 2. put PH in a folder 'c:\bar' 18:22:26 3. mklink c:\foo\processhacker.exe c:\bar\processhacker.exe 18:22:26 (make sure that c:\bar\ is not in the path) 18:22:26 4. run processhacker c:\foo\processhacker.exe 18:23:25 wouldn't you need to symlink the driver, too? 18:23:32 for reference, if you inspect the process properties for PH, what is the command line and current directory? 18:24:11 (e.g. is NT telling us we're in foo or bar) 18:24:13 dmex: btw, one reason why i put the driver for lspci in a resource is to be able to run the thing directly off the network (usually you can't access the files on network anonymously) 18:24:27 http://fearthecowboy.com/stuff/2012-04-17-112339-21a00007.png 18:24:52 in my example, processhacker was run from c:\programdata\bin\processhacker.exe 18:24:53 well thats an off command line :O 18:24:57 off/odd 18:25:06 it's in my path 18:25:16 windows is weird when it comes to both symlinks and hardlinks 18:25:21 (when you run programs) 18:25:26 * auroraeosrose1 joined 18:25:26 so if c:\foo was in the path, this is what you'd see 18:26:32 eg. i have PH in O:\Drivers\Software\ProcessHacker\x64\ProcessHacker.exe, and i hardlinked that to O:\Drivers\Software\ProcessHacker\ProcessHacker.exe. Image File is still "O:\Drivers\Software\ProcessHacker\x64\ProcessHacker.exe" 18:26:54 plugin.c, line 267 18:26:57 :) 18:27:42 i first noticed that with Process Explorer and nvidia control panel, when i tried hardlinking some programs to run them with different settings 18:27:57 had to actually copy them to be able to get the settings to apply 18:28:34 * auroraeosrose quit (Ping timeout: 260 seconds) 18:28:35 we read the path/commandline direct from NtQuerySystemInformation 18:29:36 hmm, what happens if i hardlink the hardlink and then delete the original source of the first hardlink 18:29:36 for plugins, we use CreateFile with a relative path (or if it's not relative, we call RtlDetermineDosPathNameType and check) 18:30:27 hardlinks are just refcounted links to the same physical file 18:30:34 if you delete one, nothing happens to the second one. 18:30:50 (unless its a reg symlink) 18:31:11 iiinteresting 18:31:54 it seems that it always assigns one of the files as "primary", and Image Path will always return that file (if you delete it, the first remaining hardlink that you run will be then always Image Path) 18:33:08 it would probably be easier to implement your own FS driver that does the mappings 18:33:35 even more interesting: if i move (rather than delete) the file that was shown in Image File, the old path remains there when I run the program again, and Process Hacker doesn't show the file information (including icon) 18:34:09 http://eternallybored.org/Image1.png 18:34:19 there's actually a corner case with NTOS there 18:35:30 I forget what triggers it but ntos will 'cache'?? an incorrect image filepath 18:35:42 though Process Explorer does show the correct paths 18:35:50 it won't update even if you exit the process 18:36:01 yep 18:36:20 procxp doesn't use this class to enumerate process paths 18:36:56 if you notice, PH shows the image filepath even without elevation, while procxp doesn't 18:37:11 but I remember PE showing the original hardlink name with other hardlinks in the past (and nvidia's control panel similarly didn't see hardlinked executables as distinct programs) 18:41:31 http://www.oldify.net/ 18:41:49 * dmex looks for photo's of friends 18:42:26 Process Hacker is now packaged: http://coapp.org/pages/packages.html 18:47:04 http://i.imgur.com/53YHU.png 18:47:22 for some reason, it's showing some invisible dialog 18:47:58 * dmex kills the process 18:48:05 dmex: patience 18:48:15 HA - someone else can break it! 18:48:19 * auroraeosrose1 changed nick to auroraeosrose 18:48:21 if it's the first time you install a coapp package, it can sometimes take several minutes 18:48:41 thats silly... 18:48:47 heh 18:48:51 I've still got some debugging to do on the bootstrapper 18:48:53 what takes so long the first time? 18:49:08 I think it's spending a lot of time scanning the installed MSIs 18:49:15 might be nice to have a log people can look at "I can't contact the place to get the msi" 18:49:17 I need to be able to do that to know the state of the system 18:49:32 *sigh* 18:50:04 I'd take patches : 18:50:06 :) 18:50:09 Temp\Installer.5600.exe is just sitting on %5 cpu, doing...something... 18:50:26 http://eternallybored.org/imgs/cats/P1040052.jpg 18:50:27 5% ...sound like it's doing *something* ... 18:50:50 do you have a lot of software installed? 18:50:59 (and possibly, a slow drive?) 18:51:06 45, CoApp.Bootstrapper.SingleStep.InstallerStageTwo(System.AppDomain) + 0x260 <-- 0x7ff001488d0 18:51:22 SSD, I don't install much on my dev machine 18:51:25 HMM. 18:51:36 introduces variables during development 18:52:09 * ender uses VMs with snapshots for dev 18:52:28 some retarded app does something, changes X, later when I'm implementing or testing something, I find i've done it wrong thanks to X 18:52:46 * dmex looks at registry cleaners 18:53:05 is it still stuck running? 18:53:10 yes 18:53:16 grrrr. 18:53:18 kill it. 18:53:31 http://coapp.org/cleaner 18:53:32 run that 18:53:38 as admin 18:53:40 that invisible window prevents me from clicking cancel... 18:54:02 just kill the process 18:54:15 from the task manager 18:54:59 http://i.imgur.com/mLvTm.png 18:55:17 that installer window is the invisible one 18:55:26 really shouldn't be parented 18:56:23 http://i.imgur.com/uLaCM.png 18:56:27 ^crash 18:56:42 dammit 18:56:50 * FearTheCowboy hates the bootstrapper. 18:57:34 take two drinks 18:57:41 LOL 18:58:22 InvokeFuture ---> Crash 18:59:09 I think someone jinxed it ;) 18:59:24 it's future garrett! 19:00:31 Well, this is honestly a new one. 19:00:38 Hasn't failed *here* before. 19:00:59 look in the eventviewer 19:01:07 look in the coapp category 19:01:40 (or run the whole thing with dbgview running) 19:01:50 there should be an error in the eventviewer 19:02:02 from coapp.service 19:02:09 System.IndexOutOfRangeException - Index was outside the bounds of the array. 19:02:16 WHAT? 19:02:17 FFS 19:02:23 is there a stacktrace ? 19:02:35 thats was 9s after the event: ==[Trying to start Win32 Service]== 19:02:49 nope, thats the entire log message 19:02:55 check the details for a stacktrace 19:03:30 ggrrrrr 19:03:57 http://pastebin.com/s1AXgVsT 19:04:00 * Jonny5 quit (Quit: Leaving.) 19:04:10 its 'In Bytes' 19:04:41 dunno why 19:05:46 Engine.DoPackageComposition 19:05:53 Yeah, I can see this 19:06:35 did you convert the stack into integers? 19:07:01 btw, FearTheCowboy, the cleaner has (nearly) white background, but uses system text colour 19:08:07 Yeah, I just dump the stack into a bytearray; I don't usually need to see this... 19:08:18 oh, before I forget again, no the disk tab runs without KPH 19:08:34 it uses ETW which requires elevation 19:08:37 Yeah, I noticed, it was the missing plugin that caused it to not find it :) 19:09:01 I ran the cleaner, same thing happened 19:09:15 * _Andrew quit (Quit: sleep) 19:09:27 Yeah, I've got a fault in the composition rules ... just trying to see how to find it. 19:10:00 also, should i reboot after using the cleaner? 19:10:06 no, it's not needed 19:10:16 one handy thing with PH, it'll show .net stacks with the native stacks while inspecting threads 19:10:19 well, in that case i can't install PH 19:10:43 if you have some symbols, I can dump the threads 19:11:30 (dunno if you can just throw them up on the MS symbol server, probably windows only) 19:11:45 not handy... 19:11:49 :P 19:13:01 hang on, I'm gonna try and see if I can replicate this. 19:14:51 does it help knowing the CoApp.Service created two \Device\NamedPipe\CoAppInstaller pipes? 19:15:01 that's common 19:15:45 can you install http://coapp.org/install 19:15:51 or does that do the same thing? 19:16:39 same thing but nothing 19:16:55 I just see two UAC prompts, installer becomes idle 19:17:04 (and invisible) 19:17:15 it's failing the same way. That's *good* I guess 19:18:07 if I delete the service, terminate the process, the installer recreates it but does nohting 19:18:28 terminate the service process^ 19:18:43 is it a good thing if i tell you it's doing the same for me? 19:19:25 I's say the installer is stuck trying to communicate with the service 19:19:33 Yes; I see where It's going awry... 19:19:52 it's stuck because the service returns an exception during an unlikely spot 19:20:12 and the UI is actually not quite finished... I haven't put in hardly anythng off the happy path 19:20:24 one reason why we rewrote PH from C# into C was exceptions 19:20:36 LOL 19:20:51 there was also the mashalling problems 19:21:07 I'd like to have written this in C++, but I really need Linq and the TPL. 19:21:12 well, on alot of poeples machines, the runtime would be corrupted for one reason or another 19:21:15 you don't NEED it 19:21:17 ;) 19:21:36 you could probably write the service as native 19:21:38 oh, it actually went through 19:21:46 service would probably be nice native 19:21:59 it's using 41mb atm 19:22:02 [and it's now offering me to (X) Remove (X) Cancel] 19:22:05 I think I've found it. 19:22:58 Well, I can't see how this didn't blow up before. dag nabbit 19:23:45 every time garrett says "how did this work before" 19:24:23 hmm. no, that's not it. 19:24:51 I would've gone with "run coapp cleaner" 19:27:00 nice text effect: http://eternallybored.org/Image1.gif :) 19:29:10 actually 19:29:15 dmex -> do you have VC11 or .NET 4.5 installed? 19:29:19 no 19:29:49 I did run into a problem last week with coapp 19:30:15 when I tried it last month, I disabled the service 19:30:26 when I tried it again last week, it wanted to update 19:30:36 that's fine. the cleaner deletes the service anyway 19:30:50 the update worked but the toolkit itself was crashing 19:31:06 it may be that some stuff is still on here 19:31:20 when it updates, the command line stops; but that's not really a functional problem. 19:32:18 I guess it depends of the cleaner also removes anything from older versions 19:32:21 hmm. I've never seen a case where a List<> can throw an IndexOutOfRangeException as a result of going to far in a foreach 19:32:30 the cleaner removes *everything* 19:33:08 the C# compiler makes the anonymous .ForEach a for loop of sorts 19:34:59 oh 19:35:25 if your anon function throws an exception, it can throw on the for 19:35:39 http://pastebin.com/5KkR14nG 19:35:57 thats the reflect for System.Collections.Generic.List 19:36:41 it also depends how you access the item 19:37:23 if you inherit the list class and access the item using [i], it can throw there 19:38:02 http://pastebin.com/6RZrM6FA 19:39:22 so if you enherit the class, you also need to preform the index >= count check and bail before accessing the item if you don't want it throwing 19:39:34 I'm not doing anything like that at all 19:39:37 I'm not even close to that 19:42:20 Win7/x64 19:42:21 ? 19:42:52 yes 19:43:21 try running sysinterals' dbgview and running it again 19:43:32 there's gotta be soemthing else I can't see 19:43:44 the future? 19:43:48 lol 19:44:24 * piscisaureus_ quit (Quit: ~ Trillian Astra - www.trillian.im ~) 19:45:10 why the x86/x64 msi packages? 19:45:32 what the 19:45:35 it worked this time 19:45:38 FFS. 19:46:01 that's not nice. 19:46:04 :S 19:46:30 server connection issue? :| 19:46:36 yeah, it worked the second time for me, too 19:46:54 (although i did get a funny text effect: http://eternallybored.org/Image1.gif ) 19:47:03 no; there's not any network communication going on at that pojnt 19:47:27 Yeah, that's my poor-mans-forcing-wpf-to-dowhat_i-want. 19:48:24 http://pastebin.com/X7xJ5Nz8 19:49:29 I ran the cleaner after the first crash when you suggested it, not the second or third but again this time 19:50:28 I also deleted the coapp service before running the cleaner this time 19:51:41 o_O you signed all the plugins too 19:51:46 Yeah, why not? 19:52:13 * ssam2 quit (Remote host closed the connection) 19:52:14 we could implement secure plugin loading with that 19:52:32 dont load if it's not signed etc... 19:52:54 hmmm..... how old of version of CoApp did you have instlled before? 19:54:02 I gotta stop for some lunch... I'll be back in a bit. 19:54:11 december/january before it updated last week 19:57:23 ah 19:57:26 another thing 19:57:47 I removed a bunch of symlinks in ProgramData a few days ago 19:58:01 now I know those were from coapp :| 19:59:18 * piscisaureus_ joined 20:02:40 * remy_o joined 20:05:44 if I run the PH setup again, the installer outright crashes 20:06:56 still shows a UAC prompt and displayes the package information 20:12:05 if I run the cleaner then try installing the package again, it hangs on starting the service 20:13:05 the event is also logged again: System.IndexOutOfRangeException - Index was outside the bounds of the array. 20:22:53 yep, crashed with the same exception 20:23:13 yay for consistancy 20:23:28 [3300] »RawData: at CoApp.Toolkit.Engine.Client.PackageManager.<>c__DisplayClass3.b__1() 21:01:41 * vpovirk quit (Remote host closed the connection) 21:32:37 * madewokherd joined 22:02:58 * wwahammy_ quit (Ping timeout: 252 seconds) 22:22:28 * ender` quit (Quit: I have always wished that my computer would be as easy to use as my telephone. My wish has come true. I no longer know how to use my telephone. -- Bjarne Stroustrup) 22:41:13 * piscisaureus_ quit (Ping timeout: 276 seconds) 22:42:25 * piscisaureus_ joined 22:44:16 * remy_o quit (Quit: WeeChat 0.3.7) 22:53:58 * virmitio quit (Quit: Leaving.) 23:39:38 * cH40z-Lord quit (Ping timeout: 260 seconds) 23:39:46 * cH40z-Lord joined 23:52:40 * piscisaureus_ quit (Quit: ~ Trillian Astra - www.trillian.im ~)