| 00:18.15 | ``Erik | ah. |
| 00:18.16 | ``Erik | Choo. |
| 00:24.25 | brlcad | cobbaut: it's not a plugin, gimp supports it natively, it's a simple raw image -- interleaved R G B values dumped to a file |
| 00:24.36 | brlcad | but it has to be named right, .raw |
| 00:25.10 | brlcad | otherwise gimp tries to parse it as an sgi ".pix" which is something different |
| 00:28.35 | brlcad | ufraw is a different kind of raw afaik |
| 00:28.48 | brlcad | raw from digital cameras, which is encoded different |
| 00:30.16 | brlcad | unless gimp has changed in the past few years |
| 00:41.09 | brlcad | bah, just tried opening it in gimp, no longer see the option I remember |
| 00:41.29 | brlcad | so forgo the raw, convert to png with pix-png |
| 00:41.46 | brlcad | pix-png file.pix > file.png |
| 00:42.09 | brlcad | or convert to pnm pretty easily as is: |
| 00:42.20 | brlcad | echo "P6 |
| 00:42.28 | brlcad | 512 512 255" > file.pnm |
| 00:42.33 | brlcad | cat file.pix >> file.pnm |
| 01:19.54 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/BUGS: wavelet isn't working correctly on a simple image decomposition and reconstruction |
| 03:05.55 | justin_ | another long day >_< |
| 08:22.53 | *** join/#brlcad DTRemenak (n=DTRemena@c-67-187-164-66.hsd1.ca.comcast.net) | |
| 11:40.19 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) | |
| 12:25.03 | CIA-5 | BRL-CAD: 03d_rossberg * 10brlcad/src/librt/wdb_obj.c: allocate memory before RT_APPLICATION_INIT |
| 12:35.40 | *** join/#brlcad cad438 (n=d92be461@bz.bzflag.bz) | |
| 15:21.42 | brlcad | d_rossberg: nice catch, eek |
| 15:25.09 | d_rossberg | isn't it :-) |
| 15:36.58 | *** join/#brlcad cobbaut (n=paul@135.105-136-217.adsl.skynet.be) | |
| 17:28.23 | *** join/#brlcad birdmun (n=birdmun@ppp-70-134-51-45.dsl.spfdmo.swbell.net) | |
| 17:31.09 | birdmun | i get errors (many) after installing cygwin(complete) and after running configure and then make ... is there anywhere i can look that i havent looked? |
| 17:31.14 | birdmun | for help that is |
| 17:36.38 | birdmun | i guess it would be good if there was living being here |
| 17:40.09 | ``Erik | O.o |
| 17:40.29 | ``Erik | <-- doesn't know if it works on cygwin... what kinda errors? |
| 17:40.46 | ``Erik | (and spfdmo? springfield, mo? like, 6580X?) |
| 17:42.04 | birdmun | actually joplin mo |
| 17:42.07 | birdmun | but close enuf |
| 17:42.14 | birdmun | hour and a half |
| 17:42.18 | birdmun | maybe hour |
| 17:43.33 | birdmun | just a sec lemme see if i can get thru make and i will try posting ... it complains alot about hooks and some variables being defined more than once |
| 17:45.05 | birdmun | not that running make is a quick thing to do lol |
| 17:48.23 | birdmun | damn forgot to pipe the output to a txt file |
| 17:58.48 | birdmun | a great deal of the errors are prefaced w/ a tk |
| 18:01.41 | birdmun | am i to guess that the tk lib or some such item did not get installed properly? |
| 18:09.27 | *** part/#brlcad birdmun (n=birdmun@ppp-70-134-51-45.dsl.spfdmo.swbell.net) | |
| 18:09.34 | *** join/#brlcad birdmun (n=birdmun@ppp-70-134-51-45.dsl.spfdmo.swbell.net) | |
| 18:09.39 | birdmun | DOH! |
| 18:09.45 | birdmun | wrong window |
| 18:22.46 | ``Erik | libtk is built along with the system if necessary, in src/other/libtk |
| 18:23.09 | ``Erik | it might be that the libtk it's trying to build is the unix variant instead of the windows variant... I don't do windows, so'z I d'no |
| 18:23.20 | ``Erik | <-- lived in springfield for 8 yrs, went to smsu |
| 18:23.42 | ``Erik | actually, nixa for a bit, then springfield, near battlefield for most of it |
| 18:36.26 | brlcad | birdmun: cygwin should work, though I haven't tested in a few months |
| 18:37.11 | birdmun | i wish i knew how to make it work |
| 18:37.13 | brlcad | it might not be a clean build though shouldn't take much to get it |
| 18:37.19 | birdmun | currently going to mssu |
| 18:37.31 | brlcad | what's the error(s), pastebin it |
| 18:37.33 | birdmun | red headed step child to smsu i guess :P |
| 18:37.58 | birdmun | i should be able to pipe the make output to a txt file yes? |
| 18:38.11 | birdmun | just type make >> make.txt? |
| 18:38.19 | birdmun | or no? |
| 18:39.33 | brlcad | make > build.log 2>&1 |
| 18:40.12 | brlcad | or just copy/paste |
| 18:40.14 | brlcad | ~pastebin |
| 18:40.15 | ibot | hmm... pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste |
| 18:40.29 | brlcad | you can paste it there, then post the url here |
| 18:41.19 | birdmun | very odd ... i typed the make > ... and it seems to have hung it |
| 18:43.00 | birdmun | well ... just typed make and we are off |
| 18:43.06 | birdmun | i believe it takes about 20 mins |
| 18:44.10 | brlcad | did it fail to build, or fail to run? |
| 18:44.24 | birdmun | build ... i guess |
| 18:44.25 | brlcad | and are you just looking for a binary? or looking to build? |
| 18:44.39 | birdmun | i realize there is not a win32 binary yet |
| 18:44.47 | brlcad | actually there are |
| 18:45.05 | birdmun | ok ... |
| 18:45.09 | brlcad | we've had binaries for a couple years |
| 18:45.14 | brlcad | just don't distribute them :) |
| 18:45.17 | birdmun | ic |
| 18:45.21 | brlcad | there are issues to work out |
| 18:45.31 | brlcad | we're close to doing a full "first" release |
| 18:45.35 | birdmun | i see the topic says not stable |
| 18:46.27 | birdmun | i have dabbled in linux and am really trying to migrate to linux from windows ... just havent gotten that far yet |
| 18:47.20 | birdmun | i know i have enough knowledge to be dangerous |
| 18:47.48 | brlcad | :) |
| 18:47.58 | birdmun | i know this because i have managed to thrash my linux partition |
| 18:48.01 | brlcad | best way is to just take the plunge and install full ;) |
| 18:48.25 | birdmun | just enough that i cant figure out how to fix it or havent spent enough time to figure it out |
| 18:49.26 | birdmun | my problem so far is i use my system to play games as well and windows games on linux arent just install and run affairs |
| 18:50.55 | birdmun | although wine is coming along nicely from what i have been seeing |
| 18:56.42 | birdmun | http://pastebin.com/570668 |
| 18:56.53 | birdmun | i hope that is the link i needed to paste |
| 18:57.27 | brlcad | that's right |
| 18:57.28 | birdmun | that being after i ./configure --disable-opengl --enable-optimized |
| 18:58.42 | brlcad | hmm, it's trying to build the X11 interface to Tk |
| 18:59.05 | brlcad | do you have X11 installed in cygwin? |
| 18:59.10 | brlcad | headers and libs |
| 18:59.32 | birdmun | when i setup cygwin i installed everything except games and mail |
| 19:00.02 | birdmun | using cygwin 1.5.19-4 |
| 19:02.16 | birdmun | i have a shortcut on my desktop that is for Xserver |
| 19:02.27 | birdmun | the icon is in my system tray |
| 19:03.52 | brlcad | you can have the binary and not have the header or libs |
| 19:04.04 | brlcad | configure might have failed to find it as well |
| 19:04.15 | brlcad | can you mail your config.log? |
| 19:05.15 | birdmun | config.log will be in the brlcad dir that i was configuring in? |
| 19:05.45 | birdmun | found it |
| 19:06.34 | birdmun | sent |
| 19:06.41 | brlcad | okie dokie |
| 19:07.05 | birdmun | i wish i were more self sufficient about thi |
| 19:07.07 | birdmun | this |
| 19:17.58 | brlcad | yeah, it failed to find X11 headers and libs.. |
| 19:18.12 | brlcad | yet you do have a undefined r |
| 19:18.13 | brlcad | eference to `_XFillRectangle' |
| 19:18.23 | brlcad | er, you do have a /usr/X11R6 directory |
| 19:18.39 | brlcad | do you have /usr/X11R6/include/X11/Xlib.h ? |
| 19:18.59 | brlcad | and a /usr/X11R6/lib/libX11*something* |
| 19:20.29 | birdmun | i shall look |
| 19:21.11 | birdmun | yes on the Xlib.h |
| 19:23.19 | birdmun | no libX11*something* |
| 19:23.38 | brlcad | what's in your /usr/X11R6/lib dir? |
| 19:23.51 | birdmun | i realize that *something* was a 'wildcard' |
| 19:25.17 | birdmun | doh |
| 19:25.28 | birdmun | just a sec |
| 19:25.32 | birdmun | wrong dir |
| 19:26.15 | birdmun | i have a libX11-6.dll.a and a libX11.dll.a |
| 19:34.31 | brlcad | hmm |
| 19:34.35 | brlcad | try linking against them then |
| 19:34.52 | birdmun | beyond my minor knowledge |
| 19:34.56 | brlcad | make LIBS="/usr/X11R6/lib/libX11.dll.a" might work |
| 19:35.04 | birdmun | ok |
| 19:35.04 | brlcad | if it really is a .a |
| 19:35.10 | birdmun | it is |
| 19:35.16 | birdmun | listed anyway |
| 19:35.18 | brlcad | it's named that |
| 19:35.26 | birdmun | do i need to make clean first? |
| 19:35.32 | brlcad | i mean if it really is an archive, it'd be different if it was a dll |
| 19:35.34 | brlcad | no |
| 19:35.39 | birdmun | k |
| 19:36.41 | birdmun | its off ... i will let you know in a bit |
| 19:51.13 | birdmun | http://pastebin.com/570758 |
| 19:51.57 | birdmun | untrained eye says its something about opengl |
| 19:52.58 | birdmun | being as i used the --disable-opengl switch |
| 19:53.11 | brlcad | yeah |
| 19:53.40 | brlcad | wrong flag |
| 19:53.45 | brlcad | it's not --disable-opengl |
| 19:53.50 | brlcad | --without-opengl |
| 19:54.25 | brlcad | also when you paste errors, don't need to see the end of the errors, need to see the beginning of them ;) |
| 19:54.29 | brlcad | from the compile line down |
| 19:54.37 | brlcad | all the rest are just superfluous |
| 19:54.55 | brlcad | could be two listed, could be two thousand, the first is the only one that matters |
| 19:55.35 | birdmun | ic |
| 19:56.03 | birdmun | well i guess im going to have to work on this later ... gf is bein a pain |
| 19:56.07 | birdmun | hasta |
| 19:56.15 | birdmun | thanks for the help so far |
| 19:58.06 | *** join/#brlcad clock_ (n=clock@84-72-60-3.dclient.hispeed.ch) | |
| 20:31.16 | *** join/#brlcad cobbaut (n=paul@135.105-136-217.adsl.skynet.be) | |
| 20:39.05 | ``Erik | hmmmm |
| 20:39.56 | *** join/#brlcad Maloeran (n=maloeran@glvortex.net) | |
| 20:41.15 | ``Erik | pwn3d you like a car d00r, y0 |
| 20:41.18 | ``Erik | :D *duck* |
| 20:42.15 | Maloeran | BrlCad if you don't mind, or anyone else who would know, are the models you use generally/always made of distinct closed volumes? |
| 20:42.27 | Maloeran | Eheh erik :) |
| 20:44.36 | ``Erik | lemme go poke him for ya |
| 20:47.15 | *** join/#brlcad Twingy (n=justin@c-68-33-163-43.hsd1.md.comcast.net) | |
| 20:47.32 | Maloeran | Good afternoon Justin |
| 20:47.43 | Twingy | hey |
| 20:47.50 | Twingy | I keep forgetting to send mark something |
| 20:47.52 | Maloeran | A quick question for you, if you have 7 seconds, are the models you use generally/always made of distinct closed volumes? |
| 20:48.01 | Twingy | yep |
| 20:48.05 | Maloeran | What's the something, if I'm allowed to know? |
| 20:48.17 | Twingy | paperwork stuff that allows us to hire you |
| 20:48.30 | Twingy | been really busy >_< |
| 20:48.39 | ``Erik | making cardboard xserves |
| 20:48.43 | Twingy | and that |
| 20:48.49 | Twingy | like I said, busy |
| 20:48.50 | Maloeran | So always closed volumes... You know, keeping track of rays being inside/outside volumes as they traverse the geometry would be more reliable than discarding hits excessively close to the previous intersection |
| 20:49.21 | Twingy | and that's the conclusion I came to about a year ago |
| 20:49.21 | Maloeran | Thanks for the answer, I have to answer Mark's concerns on this |
| 20:49.40 | Maloeran | Doesn't adrt use the distance method? |
| 20:49.41 | Twingy | when I was complaining to the mgmt at a bnd meeting about this |
| 20:49.46 | Twingy | for what? |
| 20:49.56 | Maloeran | For discarding multiple "faulty" hits near edges |
| 20:50.04 | Twingy | for overlaps? |
| 20:50.07 | Maloeran | Yes |
| 20:50.40 | Maloeran | Maybe I'm just not up-to-date, it has been a while since we discussed that particular issue |
| 20:50.56 | Twingy | if you have a chunk of steel and a chunk of copper and they overlap by 1mm, then 1mm steel + 1mm copper get counted while traversing 1 mm |
| 20:51.06 | Twingy | there is no "correct" way, there is simply a "consistent" way |
| 20:51.22 | Maloeran | Okay, I'm refering to overlapping triangles, when a ray hits their common boundary |
| 20:51.40 | Twingy | an edge? |
| 20:51.43 | Maloeran | Yes |
| 20:52.04 | Twingy | um, I'm not sure what I do actually, I don't have any algorithms that bork up because of that issue |
| 20:52.19 | Twingy | I think my algorithms picks one or the other though |
| 20:52.25 | Twingy | -s |
| 20:52.35 | Twingy | triangle 'A' -or- 'B' |
| 20:52.41 | Twingy | whichever is closest |
| 20:52.42 | brlcad | Maloeran: generally, they are all solid volumes |
| 20:52.59 | Maloeran | And you detect "faulty" hits from rays hitting both triangles from the very rapproched distance of the two intersections |
| 20:53.11 | Maloeran | Thanks brlcad, so there are exceptions? |
| 20:53.17 | Twingy | I avoid stupid algorithms that try to grasp a definitive answer from like 10 rays |
| 20:53.36 | Twingy | I prefer numerically integrating over the problem to converge |
| 20:53.54 | Twingy | that completely eliminates small errors |
| 20:54.19 | Twingy | firing 10 trillion rays where 3 rays are degenerate on an edge simply doesn't matter |
| 20:54.32 | Maloeran | Well, if all volumes are definitely closed, there's another way to solve this, avoiding playing with the range of numerical errors |
| 20:54.52 | Maloeran | Ah, quite ;). Just trying to address Mark's concerns |
| 20:55.06 | Twingy | a good tesselator doesn't create overlaps |
| 20:55.33 | Maloeran | Triangles overlap because of numerical error in the ray-triangle intersections, it's unavoidable |
| 20:55.34 | Twingy | but we have crappy geometry usually |
| 20:55.35 | ``Erik | too bad we don't use good tesselators (yet) |
| 20:56.00 | Twingy | because of numerical instability in IEEE 754 actually |
| 20:56.05 | Maloeran | Right |
| 20:56.24 | Twingy | like I said, my stuff is built around not caring about LIVE -or- DIE based on 3 rays |
| 20:56.38 | Twingy | I think that's a retarded way of thinking about things |
| 20:56.42 | brlcad | it's not usually numerical error that results in the overlaps |
| 20:56.46 | Twingy | based on todays technology available |
| 20:56.52 | brlcad | it's either bonefide modeling overlaps |
| 20:57.07 | brlcad | or it's a result of inadequate or at least inconsistent tessellation |
| 20:57.23 | Maloeran | I'm refering to triangles sharing a common edge, brlcad. They are considered distinct entities, and a ray hitting right on the edge will hit both triangles |
| 20:57.33 | Twingy | btw, you got any news with you an survice? |
| 20:57.37 | ``Erik | proe->iges->proe->brlcad-bot-> ... |
| 20:57.39 | ``Erik | ? |
| 20:57.51 | Twingy | proe->iges->proe->brlcad->bot->adrt :) |
| 20:58.12 | brlcad | ``Erik: the proe converter itself does fine to tesselate overlaps all by itself |
| 20:58.13 | Maloeran | Apparently, they are almost ready to hire me to work from Canada until the visa is ready |
| 20:58.23 | Twingy | on their stuff? |
| 20:58.29 | brlcad | on their stuff |
| 20:58.51 | brlcad | they tessellate based on alignment |
| 20:58.57 | brlcad | modelers rarely align |
| 20:59.06 | Twingy | modelers rarely think |
| 20:59.10 | brlcad | so you end up with bolts in holes that don't match up |
| 20:59.18 | Maloeran | Ahah. Nice |
| 20:59.30 | brlcad | that's something I wouldn't want to care about while modeling |
| 20:59.36 | brlcad | it's a bolt, it goes in the hole |
| 21:00.08 | brlcad | the fact that rotating it 37.382 degrees might have helped the tessellator is a detail that is just a pita |
| 21:00.11 | Maloeran | brlcad, any notable exceptions on the volumes of models you work with not being closed? |
| 21:00.26 | brlcad | none come to mind |
| 21:00.44 | brlcad | unless we're just playing around with something |
| 21:00.45 | Maloeran | So this is pretty much a constant, okay. |
| 21:00.50 | brlcad | pretty much |
| 21:00.54 | brlcad | solid modeling ;) |
| 21:01.05 | Maloeran | I know :), I was wondering if there were exceptions |
| 21:01.07 | Maloeran | Thanks |
| 21:01.07 | brlcad | otherwise we could just use a generic modeler |
| 21:01.29 | Twingy | for a strong defense! |
| 21:01.49 | Twingy | you coming in today? |
| 21:01.55 | Twingy | mike keeps wandering in here |
| 21:02.02 | ``Erik | cattleprod, dude |
| 21:02.03 | brlcad | working on quantum stuff |
| 21:02.11 | Twingy | I got no cattleprod |
| 21:02.26 | Twingy | maybe I'll throw that broken hard drive at him |
| 21:02.36 | brlcad | get that jpeg encoding working? |
| 21:03.08 | Twingy | nyet, but I do have an algorithm I'm going to try coding up tonight |
| 21:03.22 | Twingy | I didn't get home till past 10 last night |
| 21:03.52 | Twingy | working on chucks algorithm |
| 21:06.07 | brlcad | heh, that'd make for a nice book/paper |
| 21:06.13 | brlcad | "chuck's algorithm" |
| 21:06.18 | Twingy | yep |
| 21:06.37 | brlcad | "bob's postulate" |
| 21:06.49 | brlcad | "ed's conjecture" |
| 21:06.57 | Twingy | "justin's color maker thingy" |
| 21:07.14 | brlcad | woowwwoo woo knyuck knyuck knyuck |
| 21:07.21 | Twingy | pjy would love reading those |
| 21:07.24 | Twingy | pjt |
| 21:07.31 | Twingy | pidgjit |
| 21:08.09 | ``Erik | I wonder what he'd do if we started calling him pidjit |
| 21:08.31 | Twingy | good question |
| 21:08.36 | Twingy | he seems to fire people when they do good stuff |
| 21:08.42 | Twingy | maybe you'd get a raise? *scratches head* |
| 21:09.23 | ``Erik | heh, opposed to "you're fired!" it'd be "you're mgmt!" |
| 21:09.47 | Twingy | haha |
| 21:09.57 | Twingy | I might get division chief for that one |
| 21:10.26 | Maloeran | Eheh. By the way Justin, for a manager, Mark did surprise me in his reply regarding the ray-tracer's specifications |
| 21:10.42 | Twingy | hrm? |
| 21:11.43 | Maloeran | He's just more knowledgeable on the technical stuff than I expected, he saw the potential flaw of rays hitting both triangles sharing an edge |
| 21:11.52 | Twingy | ah |
| 21:12.05 | ``Erik | probably cuz it's been an issue of great discussion for a while now? :) |
| 21:12.16 | Twingy | for oh say, the last 25 years? |
| 21:12.26 | ``Erik | the dreaded "overlap" spew |
| 21:12.36 | Maloeran | Oh? :) Anyway it's solved, it was just not present in the specs draft |
| 21:12.37 | Twingy | lee likes to babble about that |
| 21:12.47 | Twingy | okie |
| 21:13.07 | ``Erik | and the final line of the spec is "it ain't rocket science" |
| 21:13.08 | ``Erik | ? |
| 21:13.08 | ``Erik | :D |
| 21:13.28 | Twingy | hehehe |
| 21:13.33 | Twingy | *kaboom* |
| 21:17.01 | Maloeran | It's there if you ever want to take a look, Justin : http://www.rayforce.net/manpages/ A first draft |
| 21:18.01 | ``Erik | heh, kinda GL-ey |
| 21:18.48 | Maloeran | Quite |
| 21:19.10 | ``Erik | being able to deal with infiniband or myrinet without using the tcp/ip wrapper would probably be useful, no? |
| 21:19.52 | ``Erik | strawberry sundaes? I think raspberry is a better choice :D |
| 21:20.07 | Maloeran | That's what I thought, yes. It involves throwing all the low-level packets into the hands of the user, but I suppose that's the best solution |
| 21:20.17 | Maloeran | Oh, absolutely :) |
| 21:20.53 | ``Erik | those generated from roff? |
| 21:21.36 | ``Erik | rtBuild, RT_HINT_FASTEST_BUILD, 'performance' should probably be singular |
| 21:22.31 | ``Erik | heh, sed -i.bak s/erformances/erformance/g * |
| 21:22.31 | ``Erik | o.O |
| 21:22.32 | Maloeran | Ah yes, I still make this error. Thanks |
| 21:22.44 | ``Erik | np :) |
| 21:23.14 | ``Erik | "zero pointer"? a NULL? |
| 21:24.19 | Maloeran | I prefer the use of zero, personal preference. People tend to forget the shortcuts from the fact that NULL is zero, bitwise arithmetics on multiple pointers and so on |
| 21:24.29 | Maloeran | In the C standard, it's defined as the same thing anyway |
| 21:25.46 | ``Erik | hm, is NULL actually guaranteed to be 0x0? I didn't think it was |
| 21:25.46 | ``Erik | just like EXIT_SUCCESS isn't guaranteed to be 0, though I've never seen it otherwise |
| 21:26.46 | Maloeran | NULL isn't garanteed to be 0x0, but assigning 0x0 to a pointer garantees it to be the equivalent of NULL... Go figure :p |
| 21:30.51 | Twingy | nil |
| 21:31.19 | Twingy | ``Erik, I feel like resuming my work on BAD |
| 21:31.21 | ``Erik | nil is a good language agnostic term :) |
| 21:31.29 | ``Erik | ok, do you need help interpretting? |
| 21:31.34 | Twingy | nah |
| 21:31.38 | ``Erik | aight |
| 21:31.42 | Twingy | actually, I don't recall at this point |
| 21:31.51 | Twingy | but I probly won't look at it today |
| 21:32.01 | ``Erik | well, you need 3 algorithms in place to do it "for real" |
| 21:32.05 | Twingy | yep |
| 21:32.25 | ``Erik | and I have documentation and now reference code (java) for all 3 |
| 21:33.26 | ``Erik | neo over there coded it up... heh, he's been wearing big muff earphones with his ipod all freakin' day :) |
| 21:34.04 | Twingy | haha |
| 21:34.32 | *** join/#brlcad PrezKennedy (n=Apathy@c-68-33-243-45.hsd1.md.comcast.net) | |
| 21:34.45 | ``Erik | adrt vl stuff was listed as "cancelled" at the ttm |
| 21:34.57 | ``Erik | fyi |
| 21:35.05 | Twingy | so? :) |
| 21:35.39 | ``Erik | okie, just a minor detail, anything you do will be unexpected, so you have some cover from the magnifying glass |
| 21:41.23 | Twingy | nah |
| 21:41.27 | Twingy | that's why I do it off the clock :) |
| 21:55.41 | Twingy | hrm |
| 21:55.49 | Twingy | mike was explaining how his trigger guard gets cold |