Commit Graph

156 Commits

Author SHA1 Message Date
mmcwilliams 5d17925f9e Start v3. Render time breakthrough makes it the only viable design to iterate. 2020-05-27 15:10:13 -04:00
mmcwilliams e3c7913be2 Move stls from "dist" to "stl", makes more sense. 2020-05-27 15:09:11 -04:00
mmcwilliams 69eb64aae8 Ran benchmarking on spiral_7 on laptop just to confirm 2020-05-27 09:53:09 -04:00
mmcwilliams 81c87f9b7d Update comments on spiral code 2020-05-27 09:52:15 -04:00
mmcwilliams 1ff54374ae v3 work. Almost there! 2020-05-27 09:51:37 -04:00
mmcwilliams 84379317dd Add laptop benchmarks 2020-05-27 01:29:20 -04:00
mmcwilliams e05b6cfc8d Ran benchmark script on spiral_6 and spiral_7 on vps 2020-05-26 01:27:02 -04:00
mmcwilliams f7fe5b3c39 Tired of burning cpu hours on tests I know will fail, forced script to run spiral_7 tests. WOW. Also, using more recent OpenSCAD on Linux desktop I am not seeing limits on the $fn for 60 rotation spirals. No overflow errors with path_extrude tested up to $fn=450 (250 is limit on mac tests). Render times go up to 1min+ but that is over 100x speed improvements. Wow again. 2020-05-25 23:18:07 -04:00
mmcwilliams 29b7d7782f Another spiral 2 datapoint 2020-05-25 23:03:42 -04:00
mmcwilliams 93e4a3d1ed Next-gen spiral is ready to benchmark. Generates 60 rotation spiral in 7 seconds on laptop. Wow. Hitting limits at which the spiral module overflows in the path_extrude() module. Seems to be an issue where if length of path vector array exceeds ~8500, program throws. Will test this module on newer builds to see if the 2019.05 version is the issue. 2020-05-25 23:01:10 -04:00
mmcwilliams 56cee0fb3f Add verbose comments, trying to figure this spiral generation out. Ahhhh! 2020-05-25 22:28:12 -04:00
mmcwilliams d8479e33af Merge remote-tracking branch 'origin/master' 2020-05-25 19:07:48 -04:00
mmcwilliams 17afd6b5a6 Add comments to spiral generation code (surrounding v3, which is where the most significant performance is gained). Change gnal_v2 to use library gnal_v2.scad 2020-05-25 19:07:02 -04:00
mmcwilliams 5133266eae Rendered on desktop 2020-05-25 18:19:59 -04:00
mmcwilliams 4c0a5d33ff Merge remote-tracking branch 'origin/master' 2020-05-25 18:18:48 -04:00
mmcwilliams 7f07511b03 Render v2 2020-05-25 18:14:02 -04:00
mmcwilliams 69810b3d60 Rendered v1 and v2 images 2020-05-25 18:12:40 -04:00
mmcwilliams 90e7a173fd Benchmarks run. V2 script updated 2020-05-25 18:11:19 -04:00
mmcwilliams f09960f3d7 Name gnal library gnal_v2 2020-05-25 15:58:50 -04:00
mmcwilliams 6d23f0b0d3 Update readme 2020-05-25 12:21:11 -04:00
mmcwilliams 6f4eb31f92 V2 build process complete and run on laptop. Rerun on desktop with later OpenSCAD. Issue: render preview fails on spirals. Could be use of --preview flag in command but I don't want to just re-render an entire spiral to take a screenshot. Will test method of writing a scad file with an import statement referencing a rendered .stl 2020-05-25 01:35:51 -04:00
mmcwilliams ef5a29812a insert modules named correctly in gnal lib 2020-05-25 01:30:58 -04:00
mmcwilliams e9ea4d1993 V1 has been fully rendered, including 100ft 2020-05-25 01:30:32 -04:00
mmcwilliams 30a93e9c59 Add insert part flags to scad files 2020-05-25 01:29:28 -04:00
mmcwilliams 422533bef5 Spell check 2020-05-24 01:13:59 -04:00
mmcwilliams 002874282d Complete and render v2. Process is running 2020-05-23 21:47:47 -04:00
mmcwilliams 459bcecaea Publish v2. Work needs to be reconstructed and refactored from out-of-repo sources. Trying to maintain naming consistency, but work is needed to restore a complete render process for all parts 2020-05-23 19:45:35 -04:00
mmcwilliams c501ccba5c Complete section of benchmark script was broken 2020-05-23 16:58:01 -04:00
mmcwilliams c495ff6168 Render not compile (though they may mean the same thing in this case) 2020-05-23 15:15:03 -04:00
mmcwilliams ceb7580f8b Set filenames to variables in v1 script 2020-05-23 14:26:18 -04:00
mmcwilliams e2d4a3c63b complete was broken 2020-05-23 14:03:51 -04:00
mmcwilliams d6ec1cb1c0 Files rendered with latest Openscad 2020-05-23 12:09:13 -04:00
mmcwilliams 6d54a99960 Benchmark will only run on non-existent files to allow for the job to be discontinued. 2020-05-23 11:58:41 -04:00
mmcwilliams 8ca9992d48 Sample makefile for project. Could be useful for automating building and rendering images 2020-05-23 11:57:59 -04:00
mmcwilliams 5085b8bb48 Add linux cpu functionality to eponymous script 2020-05-23 00:12:28 -04:00
mmcwilliams 4e959df57b Run ALL benchmark tasks 2020-05-22 19:32:18 -04:00
mmcwilliams 2362ec56ec Notes about ADMesh (--help output). 2020-05-22 19:09:26 -04:00
mmcwilliams c677099742 Use ADMesh to calculate facets and volume of resulting models. Look for significant differences in render time and size between spiral approaches. 2020-05-22 19:09:02 -04:00
mmcwilliams 5f592661ba Run benchmark script on 6 spiral approaches 2020-05-22 19:00:32 -04:00
mmcwilliams 98467a4f6b Readme update 2020-05-22 15:47:02 -04:00
mmcwilliams 84e07d8e60 Version moved to its own script for code reuse 2020-05-22 14:54:38 -04:00
mmcwilliams cf32fcc856 Migrate libraries into ./libraries directory 2020-05-22 14:54:02 -04:00
mmcwilliams 6962744925 Log the version and cpu info when rendering models 2020-05-22 12:51:27 -04:00
mmcwilliams 0b1aaadb38 Merge remote-tracking branch 'origin/master' 2020-05-22 12:00:07 -04:00
mmcwilliams 70ddefcbb2 Merge branch 'master' of https://github.com/sixteenmillimeter/GNAL 2020-05-22 11:59:40 -04:00
mmcwilliams 958cae468d Store version for profiling 2020-05-22 11:58:23 -04:00
mmcwilliams 7972fe8d72 Updated script to run. 2020-05-22 11:09:15 -04:00
mmcwilliams 98b98cad3b Write stats out to file 2020-05-21 23:56:37 -04:00
mmcwilliams adc1581556 Readme initial draft. Need renders and photographs before publishing repo. 2020-05-21 23:53:48 -04:00
mmcwilliams bc1366c6ab A script for rendering all v1 models 2020-05-21 23:43:59 -04:00