Package Details: slic3r-git a-20

Git Clone URL: https://aur.archlinux.org/slic3r-git.git (read-only, click to copy)
Package Base: slic3r-git
Description: Slic3r is an STL-to-GCODE translator for RepRap 3D printers, aiming to be a modern and fast alternative to Skeinforge.
Upstream URL: http://slic3r.org/
Licenses: GPL
Conflicts: slic3r, slic3r-xs, slic3r-xs-git
Provides: slic3r, slic3r-xs, slic3r-xs-git
Submitter: swiftgeek
Maintainer: swiftgeek (nickoe)
Last Packager: swiftgeek
Votes: 21
Popularity: 0.006672
First Submitted: 2012-06-23 10:45
Last Updated: 2019-09-13 15:28

Dependencies (27)

Required by (2)

Sources (3)

Pinned Comments

swiftgeek commented on 2016-04-19 00:37

If something breaks on PKGBUILD side, please bother me directly on freenode (usually available at #slic3r) XS compilation requires more than 512MiB of installed RAM, adding same amount of swap makes it build (tested on armv7h, Allwinner A13)

Perl update guide: https://lists.archlinux.org/pipermail/arch-dev-public/2015-June/027236.html

or since 5.26:

pacman -Qqo '/usr/lib/perl5/5.26' where 5.26 is OLD version of perl

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 7 8 ... Next › Last »

swiftgeek commented on 2015-01-02 16:49

1. Cannot fix that, binary repo's packaging of perl package is ultra weird, perl-*-aur packages are provided as temporary workarounds (until resolved/updated in binary repo)
2. I need to fix awk line separately, i procrastinated on this issue for way too long (31C1 happened and…)
I cannot reproduce »anymore« that warning happening in tests - 69da8b09997c58c3f47d1e676fe578f68cbb96e9

pzl commented on 2015-01-02 16:26

two things I currently have to do to get master (69da8b0) to build correctly:

1: I installed perl-threads-aur
2: I added another sed line to the PKGBUILD, just below the others, to remove the line that parrots out the "threads.pm >= 1.96 required" warning. ( sed -i '/^\s*warn \"threads\.pm/d' ./lib/Slic3r.pm )


There seems to be a logic problem in lib/Slic3r.pm that makes this warning print out no matter what. I believe I have the threads dependency satisfied, but the warning prints out anyway, and disturbs the version number checking, which leads to a corrupt package. The warning had to be silenced for makepkg to build correctly, and end up with a tar package that didn't have the warning itself in the filename.

Here's a small gist I created that prints each of the tests used in the Slic3r check before it prints the warning: https://gist.github.com/pzl/b53064b335457c417dbd

If anyone else can shed some light on what's wrong here, I'd love to hear. I'm not too familiar with perl

swiftgeek commented on 2014-11-09 20:49

My packages are quite different - if human user is detected it provides means to configure package by `dialog` ways, as human interaction opt-in. Otherwise behaves like any other PKGBUILD or guesstimates configuration :]

Interactive is needed when dealing with software in weird licenses (like nvidia installer sh/binary bundle - accept license to unpack image; requiring in license conscious agreement to license)

Sorry for assuming that you were not ok with missing interactivity.
And ofc thanks for semi-pointing out missing way to change commit/branch from PKGBUILD itself

pzl commented on 2014-11-09 19:38

"Pacaur is ... aiming at speed and simplicity, designed to minimize user prompt interaction" -- I believe this is a design decision.

I'm also fine with non-interactive package installation. A google for "aur interactive pkgbuild" gives this forum link as the first result: https://bbs.archlinux.org/viewtopic.php?id=141372 which discourages interactive pkgbuilds anyway. it's easy enough to just edit the pkgbuild, so that's why I left the note for people who edit as well (I didn't realize other AUR helpers accept interactive prompts)

swiftgeek commented on 2014-11-04 16:50

I would consider it as a wrapper bug - tty responds with "not a tty" - which clearly isn't the case as you are using terminal
(yaourt responds there with value from which was spawned eg. /dev/pts/5)

Seriously, this is a huge bug - one of the easiest way to debug (or create package one-time-only) is to insert `bash -i` inside package/build etc
pacaur/cower just floods std input with "Y"... That's retarded

pzl commented on 2014-11-04 15:49

I use pacaur as an AUR helper, it skips interactive mode entirely

swiftgeek commented on 2014-10-29 18:44

Interactive works everywhere (tested with yaourt/standalone) [if not describe bug]
TODO: Make sure to not break "fragment" in source line

I introduced _gitfragment variable (line ~36) to pass either branch or commit (I have to make it work with fragment vcs nicely)

swiftgeek commented on 2014-10-29 18:23

Interactive works everywhere (tested with yaourt/standalone) ;) [if not describe bug]
TODO: Make sure to not break "fragment" in source line

pzl commented on 2014-10-29 13:40

for anyone else who's been building off of master instead of stable (this being a git package and all), the last working commit for me is 33edda0. After that, the retraction tests fail because of the retraction refactoring immediately after this commit. commits around ed17c22 fail because the config test fails (fixed in 01133fd) but the newest commits immediately segfault for me.

To install at this commit: if you use the interactive installer, just specify commit 33edda0. If you use an AUR helper, edit/view the pkgbuild before installing and edit line 105 from "git checkout stable -f" to be "git checkout 33edda0 -f" and resume installing

swiftgeek commented on 2014-07-13 11:12

Now gui also works, which looks like this now:
https://www.dropbox.com/s/8tdhk8poehk9ovl/3d_plater.png