Package Details: aseprite 1.2.39-2

Git Clone URL: (read-only, click to copy)
Package Base: aseprite
Description: Create animated sprites and pixel art
Upstream URL:
Keywords: animation art aseprite editor gamedev image pixel sprites
Licenses: custom
Submitter: None
Maintainer: ImperatorStorm (ISSOtm)
Last Packager: ImperatorStorm
Votes: 91
Popularity: 1.83
First Submitted: 2011-11-01 17:40 (UTC)
Last Updated: 2022-08-06 04:23 (UTC)

Required by (0)

Sources (8)

Pinned Comments

ImperatorStorm commented on 2022-08-06 04:31 (UTC)

A note that, per Aseprite's EULA, binaries created by this PKGBUILD CANNOT BE DISTRIBUTED.

I will comply with any upstream requests to remove this package.

ImperatorStorm commented on 2022-01-06 00:07 (UTC)

Big thanks to ISSOtm for rewriting the PKGBUILD!

ImperatorStorm commented on 2021-12-31 21:31 (UTC) (edited on 2022-01-03 04:42 (UTC) by ImperatorStorm)

Hosting this package's PKGBUILD at

Latest Comments

asday commented on 2022-08-06 15:23 (UTC)

Works perfectly now, thank you.

ImperatorStorm commented on 2022-08-06 04:31 (UTC)

A note that, per Aseprite's EULA, binaries created by this PKGBUILD CANNOT BE DISTRIBUTED.

I will comply with any upstream requests to remove this package.

ImperatorStorm commented on 2022-08-06 04:29 (UTC)

@asday Should be fixed, shared-libwebp.patch needed to be updated for current Aseprite.

asday commented on 2022-08-06 03:14 (UTC) (edited on 2022-08-06 03:14 (UTC) by asday)

I'm up to date with no updates available and base-devel is installed. prepare() fails with:

==> Starting prepare()...
patching file src/desktop/linux/aseprite.desktop
patching file src/desktop/linux/mime/aseprite.xml
patching file CMakeLists.txt
Hunk #1 succeeded at 48 (offset -6 lines).
Hunk #2 succeeded at 158 (offset -8 lines).
Hunk #3 succeeded at 198 (offset -8 lines).
patching file src/app/CMakeLists.txt
Hunk #1 succeeded at 672 with fuzz 1 (offset -26 lines).
patching file src/dio/CMakeLists.txt
Hunk #1 succeeded at 11 (offset 1 line).
patching file third_party/CMakeLists.txt
Hunk #1 succeeded at 113 (offset 7 lines).
patching file CMakeLists.txt
Hunk #1 succeeded at 52 (offset -8 lines).
Hunk #2 succeeded at 222 (offset 21 lines).
patching file src/app/CMakeLists.txt
Hunk #1 succeeded at 670 (offset 19 lines).
patching file third_party/CMakeLists.txt
Hunk #1 succeeded at 129 (offset 19 lines).
Hunk #2 succeeded at 148 (offset 24 lines).
patching file CMakeLists.txt
Hunk #1 succeeded at 58 (offset -8 lines).
Hunk #2 FAILED at 230.
1 out of 2 hunks FAILED -- saving rejects to file CMakeLists.txt.rej
patching file third_party/CMakeLists.txt
Hunk #1 FAILED at 32.
1 out of 1 hunk FAILED -- saving rejects to file third_party/CMakeLists.txt.rej
==> ERROR: A failure occurred in prepare().

The rejects files are as follows:

--- CMakeLists.txt
+++ CMakeLists.txt
@@ -230,9 +231,17 @@ add_definitions(-DPNG_NO_MMX_CODE) # Do not use MMX optimizations in PNG code

 # libwebp
-  set(WEBP_LIBRARIES webp webpdemux libwebpmux)
-  include_directories(${WEBP_INCLUDE_DIR})
+    find_library(WEBP_LIBRARY NAMES webp)
+    find_library(WEBPDEMUX_LIBRARY NAMES webpdemux)
+    find_library(WEBPMUX_LIBRARY NAMES webpmux)
+    find_path(WEBP_INCLUDE_DIRS NAMES decode.h PATH_SUFFIXES webp)
+  else()
+    set(WEBP_LIBRARIES webp webpdemux libwebpmux)
+    include_directories(${WEBP_INCLUDE_DIR})
+  endif()

 # tinyxml
CMakeLists.txt.rej (END)
--- third_party/CMakeLists.txt
+++ third_party/CMakeLists.txt
@@ -32,7 +32,7 @@ if(NOT USE_SHARED_GIFLIB)

   set(WEBP_BUILD_EXTRAS OFF CACHE BOOL "Build extras.")
third_party/CMakeLists.txt.rej (END)

N4ko commented on 2022-06-19 11:36 (UTC)

@ISSOtm: Ah... Sorry i thought i had it installed, this is new to me, thank you for your help, have a nice day ^^

ISSOtm commented on 2022-06-19 10:45 (UTC)

@N4ko: Make sure to install base-devel, as outlined in the "getting started" instructions (

N4ko commented on 2022-06-19 08:19 (UTC)

Hello, I'm getting this error

==> Retrieving sources...
  -> Found
  -> Found skia-m102.tar.gz
  -> Found desktop.patch
  -> Found shared-fmt.patch
  -> Found shared-libarchive.patch
  -> Found shared-libwebp.patch
  -> Found shared-skia-deps.patch
  -> Found optional-pixman.patch
==> Validating source files with sha256sums... ... Passed
    skia-m102.tar.gz ... Passed
    desktop.patch ... Passed
    shared-fmt.patch ... Passed
    shared-libarchive.patch ... Passed
    shared-libwebp.patch ... Passed
    shared-skia-deps.patch ... Passed
    optional-pixman.patch ... Passed
==> Extracting sources...
==> Starting prepare()...
env: ‘patch’: No such file or directory
==> ERROR: A failure occurred in prepare().

I've try re-clone the repository and even replacing the code with the previous one (aseprite-v1.2.34.1) my current system is manjaro 21.3 any idea about this problem ?

ryuukk commented on 2022-03-06 15:15 (UTC)

please clone skia with --depth=1 under the prepare, so it doesn't download the ENTIRE 500+mb repo...

such a waste

ImperatorStorm commented on 2022-01-06 00:07 (UTC)

Big thanks to ISSOtm for rewriting the PKGBUILD!

ImperatorStorm commented on 2021-12-31 21:31 (UTC) (edited on 2022-01-03 04:42 (UTC) by ImperatorStorm)

Hosting this package's PKGBUILD at

ISSOtm commented on 2021-12-31 13:13 (UTC)

I am currently working on a brand new PKGBUILD to fix the jank in question. If you are interested, is there a way to reach you so that we can collaborate on this more conveniently?

ImperatorStorm commented on 2021-12-30 21:59 (UTC)

I have half a mind to just use prebuilt skia and eliminate a bunch of jank from this package, but already exists, so I'm stuck with the jank.

Eremiell commented on 2021-12-30 21:56 (UTC)

OK, I build now. The first build error was coming from some kind of dirty state of the repo, the second one I got later was a compiler issue, that one is on me, need to resolve that properly later, using this walkaround for some weeks now. Seems to be working on my end, may need to make sure you have a clean repo, cloning it afresh should do the job.

Eremiell commented on 2021-12-30 21:11 (UTC)

Wiped the whole build directory and cloned it anew, seems to be building now. Will let you know in a bit how it went.

Eremiell commented on 2021-12-30 21:05 (UTC)

I only checked few random ones,the contents seem to be present, but it seems they were all created (the insides of them) at the moment I requested them. Possibly changing the hash? Can some file inside change? Timestamp definitely updates if I delete and redownload them. It's a bit odd given the PKGBUILD is passing specific fingerprints from what I've seen. Got a different build error now though, I'll try to poke it a bit more.

Eremiell commented on 2021-12-30 21:00 (UTC)

OK, now I crash on the build. :D Let's try to poke the archives a bit if I can figure anything interesting.

ImperatorStorm commented on 2021-12-30 20:53 (UTC)

Very strange. I'll nuke them.

Eremiell commented on 2021-12-30 20:50 (UTC)

Sadly nope, seem to be failing for me right now. Curious why they change this fast.

ImperatorStorm commented on 2021-12-30 19:58 (UTC) (edited on 2021-12-30 19:59 (UTC) by ImperatorStorm)

Alright, checksums should hopefully be fixed. Sorry about that.

Melon_Bread commented on 2021-12-30 00:28 (UTC)

Yeah an now none of the *.tar.gz are passing validity checks. Those checksums need updated as well.

ImperatorStorm commented on 2021-12-28 22:34 (UTC)

Adopted package. I have no idea what the original maintainers were doing, so I just bumped aseprite to v1.2.30 and skia to aseprite-m96

warmwaffles commented on 2021-12-28 19:42 (UTC)

There is also but the compile process will peg your CPU to 100%. Consider this if you need to heat your home for 10 minutes.

ImperatorStorm commented on 2021-12-16 03:04 (UTC)

This package is 3 releases out of date. Mind updating it?

majamin commented on 2021-11-03 00:23 (UTC)

I managed to compile aseprite by changing -DWITH_WEBP_SUPPORT=ON \ to -DWITH_WEBP_SUPPORT=OFF \ in the PKGBUILD. See

Hope this helps someone M

HurricanePootis commented on 2021-09-17 00:13 (UTC)

Hey, I don't know why no one has updated the package in 2 months now, but I am going to copy and paste a git diff of changes I made to the PKGBUILD that allowed me to compiled 1.2.29. Now, admittedly, I didn't look into any changes into Cmake flags, but this does get the program compiled and running on my system.

diff --git a/.SRCINFO b/.SRCINFO
index d98c6ca..825fecf 100644
--- a/.SRCINFO
+++ b/.SRCINFO
@@ -1,6 +1,6 @@
 pkgbase = aseprite
        pkgdesc = Create animated sprites and pixel art
-       pkgver = 1.2.27
+       pkgver = 1.2.29
        pkgrel = 1
        url =
        arch = x86_64
@@ -57,7 +57,7 @@ pkgbase = aseprite
        noextract = aseprite-swiftshader.tar.gz
        noextract = aseprite-wuffs.tar.gz
        noextract = aseprite-zlib.tar.gz
-       source =
+       source =
        source = aseprite.desktop
        source = git+
        source = git+
@@ -91,7 +91,7 @@ pkgbase = aseprite
        source = aseprite-wuffs.tar.gz::
        source = aseprite-zlib.tar.gz::
        source = gn::
-       sha256sums = 541a5c561fe7481cfb6aec80ecc9a8435d8387325ba30c9e3116c3c76ef539d7
+       sha256sums = ed37e12495bf81ad1126076ec2634b132976c40dd56d4f3924e6fddad4a1e2e4
        sha256sums = deaf646a615c79a4672b087562a09c44beef37e7acfc6f5f66a437d4f3b97a25
        sha256sums = SKIP
        sha256sums = SKIP
@@ -127,4 +127,3 @@ pkgbase = aseprite
        sha256sums = c8c2d617f1a33d6eb27f25ebcc30bd8ba1e6a0aa980cada21dda2ad1401fa4a2

 pkgname = aseprite
diff --git a/PKGBUILD b/PKGBUILD
index 90e5bc9..dc7fd3d 100644
@@ -5,9 +5,10 @@
 # Contributor: Benoit Favre <>
 # Contributor: Alexander Rødseth <>
 # Contributor: Kamil Biduś <>
+# Contributor: HurricanePootis <>

 pkgdesc='Create animated sprites and pixel art'
@@ -85,7 +86,7 @@ ea3ba903faac98b64b2bf8de5e98cd97b335a474.tar.gz"

codesteban commented on 2021-09-01 23:56 (UTC) (edited on 2021-09-02 00:00 (UTC) by codesteban)

It screwed up my rofi configuration. everything worked just fine, tried reinstalling, deleting cached packages, installing rofi-git, still nothing, can't really figure out what to do. it shows me this options are missing --> ~/.config/rofi/launcher/

(process:1172724): WARNING : 18:39:56.333: /home/shaggy/.config/rofi/launcher/style.rasi:10:16: Option: lines is not found.

(process:1172724): WARNING : 18:39:56.338: /home/shaggy/.config/rofi/launcher/style.rasi:11:20: Option: columns is not found.

(process:1172724): WARNING : 18:39:56.338: /home/shaggy/.config/rofi/launcher/style.rasi:13:38: Option: bw is not found.

(process:1172724): WARNING : 18:39:56.338: /home/shaggy/.config/rofi/launcher/style.rasi:15:23: Option: padding is not found.

(process:1172724): WARNING : 18:39:56.338: /home/shaggy/.config/rofi/launcher/style.rasi:19:29: Option: separator-style is not found.

(process:1172724): WARNING : 18:39:56.340: /home/shaggy/.config/rofi/launcher/style.rasi:20:33: Option: hide-scrollbar is not found.

(process:1172724): WARNING : 18:39:56.340: /home/shaggy/.config/rofi/launcher/style.rasi:21:30: Option: fullscreen is not found.

(process:1172724): WARNING : 18:39:56.340: /home/shaggy/.config/rofi/launcher/style.rasi:22:28: Option: fake-transparency is not found.

(process:1172724): WARNING : 18:39:56.340: /home/shaggy/.config/rofi/launcher/style.rasi:26:30: Option: show-match is not found.

betalars commented on 2021-08-24 13:10 (UTC) (edited on 2021-08-24 13:12 (UTC) by betalars)

Package seems to be broken right now, I am getting this error:

/usr/include/c++/11.1.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](std::vector<_Tp, _Alloc>::size_type) [with _Tp = long unsigned int; _Alloc = std::allocator<long unsigned int>; std::vector<_Tp, _Alloc>::reference = long unsigned int&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__n < this->size()' failed.
[1]    97097 abort (core dumped)  aseprite

Typology commented on 2021-07-22 17:59 (UTC)


that's right, there was an issue in github about this and the dev said it's a problem caused by trying to support the cmark dynamic linking on all distros, I compiled aseprite without the USE_SHARED_CMARK flag to avoid this issue. the dev is thinking about removing all USE_SHARED_ flags and link everything statically.

titanlock commented on 2021-07-22 16:11 (UTC)

There is an issue with how you compile this, it doesn't end up pointing to the symlink and points to the libcmark.0.xx.x. So whenever I update my system and that library is updated it will fail to run.

Typology commented on 2021-06-17 23:00 (UTC) (edited on 2021-06-17 23:00 (UTC) by Typology)

@150Kirill150 @jigebi6613

This issue was solved, It will be included in the next release (1.2.28). you can use that solution as a workaround until the next release, or use the aseprite-git package instead.

jigebi6613 commented on 2021-06-17 17:41 (UTC)


In PKGBUILD before cmake line add

export CXXFLAGS="-march=x86-64 -mtune=generic -O2 -pipe -fno-plt -fexceptions \ -D_FORTIFY_SOURCE=2 \ -Wformat -Werror=format-security \ -fstack-clash-protection -fcf-protection"

It work

150Kirill150 commented on 2021-06-14 05:29 (UTC)

➜ ~ aseprite /usr/include/c++/11.1.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator [with _Tp = long unsigned int; _Alloc = std::allocator<long unsigned int>; std::vector<_Tp, _Alloc>::reference = long unsigned int&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__n < this->size()' failed. [1] 360108 abort sudo aseprite WHAT???

clarfonthey commented on 2021-06-03 00:53 (UTC)

I figured out the issue and submitted a PR to the devs to fix the issue; see for details.

In the meantime, the flag that is causing the issue is -Wp,-D_GLIBCXX_ASSERTIONS in CXXFLAGS. If you want to temporarily override this flag, you can add export CXXFLAGS="$CFLAGS" before the cmake line in the PKGBUILD until a new fix comes out.

ISSOtm commented on 2021-05-31 16:50 (UTC)

Overriding all the build flags doesn't seem like a great solution; which flag(s) cause the crash?

warmwaffles commented on 2021-05-31 01:25 (UTC)

Still having the same issues as @IZ5K, @Typology, and @aseoane. Is this an issue upstream with aseprite or is it how this project is compiled?

I have it working on one of my machines and don't know how, but on my other dev machine it won't compile. I fear a larger update will cause my main workstation to no longer work.

aseoane commented on 2021-05-20 12:51 (UTC)

@IZ5K, @Typology I'm having the same issue

Typology commented on 2021-05-18 08:52 (UTC)

@IZ5K Same issue here

IZ5K commented on 2021-05-18 01:35 (UTC)

the fix that @CrashKilljoy worked but now im getting the error

/usr/include/c++/10.2.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator [with _Tp = long unsigned int; _Alloc = std::allocator<long unsigned int>; std::vector<_Tp, _Alloc>::reference = long unsigned int&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__builtin_expect(__n < this->size(), true)' failed. Aborted (core dumped)

and can't launch the application

CrashKilljoy commented on 2021-05-16 02:22 (UTC)

The Common package doesn't exist anymore, all the necessary functions are now inside skia tools. Fixing the PKGBUILD is quite easy. Simply remove the line containing, remove one 'SKIP' from the sha256sums and finally, in line 129, remove common.

sanket143 commented on 2021-05-15 06:10 (UTC)

I'm also facing this issue

==> ERROR: Failure while downloading
error downloading sources: aseprite

IZ5K commented on 2021-05-13 15:15 (UTC)

I getting the same error as @Avahe. Is there anyway to fix this?

Necco commented on 2021-05-10 01:57 (UTC) (edited on 2021-05-10 02:03 (UTC) by Necco)

I'm getting the same error as @Avahe - depot_tools is working, but the skia download returns a 404 error. Looking at indicates the common.git folder has been moved or deleted. The aseprite-git package downloads skia from, could it be used for this package as well?

YahArt commented on 2021-05-05 16:48 (UTC)

Sadly still no luck even after removing the package, doing a sudo pacman -Syu and reinstalling it.

Avahe commented on 2021-05-04 20:33 (UTC)

Tried this, but am now getting ==> ERROR: Failure while downloading - Looks like it was moved, or no longer exists?

Lukalot commented on 2021-05-04 17:47 (UTC)

Ah, @Joe084, after rebooting (after running pacman -Syu and reinstalling aseprite) it runs fine again! I should have rebooted, my mistake.

@YahArt, you might want to give pacman -Syu; reboot a try.

YahArt commented on 2021-05-03 16:18 (UTC)

I am also having issues with this version of aseprite. After successfully building and installing aseprite via pamac I get the following error message after trying to start aseprite via the terminal:

/usr/include/c++/10.2.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator [with _Tp = long unsigned int; _Alloc = std::allocator<long unsigned int>; std::vector<_Tp, _Alloc>::reference = long unsigned int&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__builtin_expect(__n < this->size(), true)' failed. fish: Job 1, 'aseprite' terminated by signal SIGABRT (Abort)

joe084 commented on 2021-05-02 17:04 (UTC)

Weird, it worked on my notebook too... No hang or error message. Built with trizen on VM and my notebook.

Lukalot commented on 2021-05-02 16:36 (UTC)

Hm, I've just tested and after running pacman -Syu and reinstalling Aseprite, I still have no luck getting it to run. Interesting result from your VM.

joe084 commented on 2021-05-02 15:36 (UTC)

Hello, it worked for me today with no error or hang on VM, after a fresh pacman update. I will test on my notebook later

Lukalot commented on 2021-04-28 16:59 (UTC)

I'm also having this error, just using Libresprite for now.

cafce25 commented on 2021-04-24 16:36 (UTC)

I'm getting the same error but only after installing libiconv it probably belongs in the depends. The error I got without it: aseprite: error while loading shared libraries: cannot open shared object file: No such file or directory

rouhannb commented on 2021-04-23 14:43 (UTC)

Tried building the last version in the PKGBUILD (v1.2.25), exactly the same as before aside from python being added to the makedepends.

Same error. Probably caused by some Arch update?

rouhannb commented on 2021-04-23 14:34 (UTC)

After rebuilding the package, the error occurred for me as well. I’m positive I tested it when I first updated the PKGBUILD (I checked my command history just in case), so I’m not exactly sure what went wrong.

I’ll go ahead and try building it myself as well.

timjk36 commented on 2021-04-23 13:59 (UTC)

@rouhannb I was able to build successfully from the asperite git main branch after doing a slight workaround to a known issue with multiple definitions for libewbp constructs. I ended up using a prebuilt Skia package described in the install instructions.

Avahe commented on 2021-04-23 13:51 (UTC)

@rouhannb I end up with the same issue - it compiles successfully, but gives that error when I try to run aseprite. Maybe an issue with a dependency? I don't know c++

rouhannb commented on 2021-04-23 02:46 (UTC)

@Avahe, @timjk36 Sorry, but I can’t reproduce the error. Could either of you try building Aseprite manually?

timjk36 commented on 2021-04-22 23:11 (UTC)

@Avahe I am getting the same error after building today.

Avahe commented on 2021-04-22 03:37 (UTC)

Having an error after building today:

/usr/include/c++/10.2.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](std::vector<_Tp, _Alloc>::size_type) [with _Tp = long unsigned int; _Alloc = std::allocator<long unsigned int>; std::vector<_Tp, _Alloc>::reference = long unsigned int&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__builtin_expect(__n < this->size(), true)' failed.
zsh: abort (core dumped)  aseprite

Not sure if this is an upstream issue. Anyone else having this problem?

rouhannb commented on 2021-04-20 13:32 (UTC)

@quaqqer I’ve added it

quaqqer commented on 2021-04-18 22:03 (UTC)

I think python3 should be added to the make dependencies.

warmwaffles commented on 2021-03-29 20:16 (UTC)

@rokku no it's the only way I know that I am getting my moneys worth out of these two Xeon processors.

rokku commented on 2021-03-29 20:15 (UTC)

You should think about limiting ninja to a half of the available cpu cores. my computer was frying while the build process. IMOH it is enough to use a few cores but not all.

rouhannb commented on 2021-02-02 23:04 (UTC)

@HealerLFG Skia clones normally for me. Could you provide more details? Just so we’re clear, the “listed URL” is, right? What URL are you redirected to? And just in case I misunderstand something, could you post a copy of your terminal log when trying to build this package?

HealerLFG commented on 2021-02-02 14:38 (UTC)

Hangs when cloning skia. Manually cloning skia from the listed URL fails with a redirect; Manually cloning the redirect and rerunning makepkg successfully builds the package.

rouhannb commented on 2021-01-09 22:17 (UTC)

@Wrexes Have you tried building the package in a chroot, with makepkg directly instead of yay? The package builds fine for me, and I evidently don’t have ninja installed with pip (running pip show ninja says that the package is not found).

Wrexes commented on 2021-01-09 14:46 (UTC)

When installing/updating this package with YAY, I kept getting an error about ninja being an unsatisfied dependency. To be precise I had a traceback from python saying "No module named ninja" even though it is part of the dependencies. Installing ninja via pip install --user ninja seems to have fixed the issue.

osc commented on 2021-01-06 23:23 (UTC)

@rouhannb, yeah my bad. thanks!

rouhannb commented on 2021-01-06 19:15 (UTC) (edited on 2021-01-06 19:15 (UTC) by rouhannb)

@osc You did clone the whole repository, right? You can check the “Sources” list; aseprite.desktop is there.

osc commented on 2021-01-06 17:14 (UTC)

I am getting this error...

==> ERROR: aseprite.desktop was not found in the build directory and is not a URL.

rouhannb commented on 2020-11-18 08:24 (UTC) (edited on 2020-11-18 08:51 (UTC) by rouhannb)

@ISSOtm I don’t think that’s the problem. The tarballs are not part of the git repo, they’re downloaded just like any other source. However, the MD5s seem to be completely different each time they are downloaded.

EDIT: It was the timestamps. These tarballs seem to be generated on the fly, and each file is given a modify time of when the request was made, so they’ll be different every time. For now, I think I’ll just set all the tarballs’ checksums to SKIP, but I’ll try to see if there’s some way to perform a checksum on the file contents, not the metadata.

EDIT 2: The checksums are now all SKIP, so there should be no more problems.

ISSOtm commented on 2020-11-17 13:19 (UTC)

The .gitignore is set to ignore the tarballs, so git pull doesn't fetch them. makepkg finds the MD5s invalid, and aborts.

rouhannb commented on 2020-11-16 23:31 (UTC) (edited on 2020-11-17 02:50 (UTC) by rouhannb)

I’ll see if I can replace some of them with non-VCS dependencies, but as far as I know I can’t remove any of them.

EDIT: All the git dependencies, save for skia and depot_tools, are now tarballs.

treeshateorcs commented on 2020-11-13 10:20 (UTC)

it pulls an insane amount of dependencies from git

rouhannb commented on 2020-10-19 00:41 (UTC)

Is that all there is? Please post the entire log in a pastebin because I can’t exactly tell what’s going on here.

I have just built this package in a chroot, and it never gave an error like this one. Have you tried the same, or built Aseprite manually, and does it give the same error?

joilnen commented on 2020-10-15 21:57 (UTC)

b/ /usr/lib/ lib/libjpeg.a /usr/lib/ /usr/lib/ lib/libwebpdemux.a lib/libwebpmux.a lib/libwebp.a -lpthread -lm /usr/lib/ /usr/lib/ lib/libjson11.a lib/libarchive.a /usr/lib/ /usr/lib/ /usr/lib/ /usr/lib/ /usr/lib/ /usr/lib/ /usr/lib/ lib/libfmt.a lib/libtinyexpr.a lib/libnet-lib.a lib/liblaf-base.a lib/libmodpbase64.a /usr/lib/ /usr/lib/ lib/liblauxlib.a lib/liblua.a lib/liblualib.a && : /usr/bin/ld: lib/libapp-lib.a(browser_view.cpp.o): in function app::BrowserView::CMarkBox::processNode(cmark_node*)': browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x8b): undefined reference tocmark_iter_new' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0xd5): undefined reference to cmark_iter_next' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0xea): undefined reference tocmark_iter_get_node' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0xf6): undefined reference to cmark_node_get_type' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x175): undefined reference tocmark_node_get_literal' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x20d): undefined reference to cmark_node_get_literal' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x37d): undefined reference tocmark_node_get_literal' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x665): undefined reference to cmark_node_get_literal' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0x912): undefined reference tocmark_iter_free' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node[_ZN3app11BrowserView8CMarkBox11processNodeEP10cmark_node]+0xada): undefined reference to cmark_node_get_url' /usr/bin/ld: lib/libapp-lib.a(browser_view.cpp.o): in functionapp::BrowserView::CMarkBox::loadFile(std::__cxx11::basic_string<char, std::char_traits\<char>, std::allocator\<char> > const&)': browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x163): undefined reference to cmark_parser_new' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x213): undefined reference tocmark_parser_feed' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x254): undefined reference to cmark_parser_feed' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x25d): undefined reference tocmark_parser_finish' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x279): undefined reference to cmark_node_free' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0x28b): undefined reference tocmark_parser_free' /usr/bin/ld: browser_view.cpp:(.text._ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE[_ZN3app11BrowserView8CMarkBox8loadFileERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE]+0xcef): undefined reference to `cmark_parser_feed' collect2: error: ld returned 1 exit status ninja: build stopped: subcommand failed. ==> ERROR: A failure occurred in build(). Aborting...

rouhannb commented on 2020-08-22 18:16 (UTC)

Readlink shows /usr/bin/python2 for me, so it should be fine.

ISSOtm commented on 2020-08-22 18:11 (UTC)

Oh yeah, symlinking makes more sense for folders, less so for files. Got caught up in symlinking all the things!

As for which -p, it seems to be a zsh thing, actually, so I guess it'll be fine on bash?

rouhannb commented on 2020-08-22 17:37 (UTC)

I’ve decided to take out the first loop and just use it as one statement, and to use which for finding python2 (although I’ve never heard of a -p flag for which).

But what exactly is the point of symlinking gn? Makepkg already symlinks files to src/, so I can’t see any performance benefit, and the program works and looks fine without doing so.

ISSOtm commented on 2020-08-22 16:57 (UTC)

I had not refreshed the page so I didn't see your comment, but I added a Gist link to the PKGBUILD I used.

It does not loop over the deps, though that loses the ability to split them across lines, and whatever performance gain is probably irrelevant. I does symlink gn as well, though. It also doesn't hardcode the path to python2, but maybe using which isn't a great idea without -p.

rouhannb commented on 2020-08-22 15:47 (UTC)

Yeah, using symlinks would work better. I’ve changed the PKGBUILD so rebuilding won’t be necessary anymore.

ISSOtm commented on 2020-08-22 15:11 (UTC) (edited on 2020-08-22 16:52 (UTC) by ISSOtm)

Runing makepkg while directories are already present under skia causes errors (mv: cannot move 'buildtools' to 'skia/buildtools': Directory not empty). This means that everything needs to be re-downloaded and re-built every time... :/

Wouldn't symlinking the folders (only if they aren't already present) work better?

[EDIT] This PKGBUILD seems to work fine

co1ncidence commented on 2020-07-25 18:54 (UTC)

Can confirm that this builds and works 100% fine on manjaro 5.7, just by running `yay -S aseprite`, it takes a while to build (around 25 minutes for me) so be prepared for a hot laptop.

Luke commented on 2020-05-22 21:58 (UTC) (edited on 2020-05-22 21:59 (UTC) by Luke)

yeah, you're totally right rouhannb, sorry for my assumptions, I was wrong, I was like a crazy trying to get aseprite to work again after nettle update and just wanted to share what have worked with me

About nettle7, it could just be me inexperience with arch, when I've tried it using yay it just don't import the key, no matter what I do, again, sorry for my wrong assumptions

rouhannb commented on 2020-05-21 20:04 (UTC) (edited on 2020-05-21 20:23 (UTC) by rouhannb)

maybe try this first before trying to recompile everything.

Correct me if I’m wrong, but by “everything”, you’re referring to this whole package, right? Apparently, all standard Arch Linux packages are rebuilt after a library update, and the wiki explicitly says “If the system has locally built packages (such as AUR packages), users will need to rebuild them when their dependencies receive a soname bump,” so it looks like rebuilding the package is actually the correct course of action.

I wouldn’t exactly recommend using an old version of nettle—it sort of puts a dent into the whole rolling release aspect of Arch—but I guess it’s technically fine too. You might want to put it into /usr/local/lib though; /usr/lib should be reserved for packages.

(Edit: nettle7 does not have a key-sign problem. You need to import the key in the validpgpkeys array.)

Luke commented on 2020-05-21 18:18 (UTC) (edited on 2020-05-22 21:59 (UTC) by Luke)

I've tried to compile manually from source, but it still doesn't works, compiling ( manually and copying it to /usr/lib/ solves the problem (at least to me). Just letting it here, maybe try this first before trying to recompile everything.

Maybe using nettle7 package ( could be an option, but they're having a key-sign problem.

edit: not recommended

rouhannb commented on 2020-05-15 12:02 (UTC)

It seems libnettle was updated recently, from version 7 to 8. Maybe you just need to rebuild the package?

Besides, libnettle isn’t even showing up as a dependency on my copy. Can you run ldd $(which aseprite) for me and tell me if nettle shows up?

Sorry it took so long, I need to check my email more frequently :P

k4rakara commented on 2020-05-11 19:23 (UTC)

Broke after I updated recently. Errors with "aseprite: error while loading shared libraries: cannot open shared object file: No such file or directory", even if reinstalled.

rouhannb commented on 2020-05-01 02:43 (UTC)

I’m not sure what you mean when you say the package is using the latest commit. The source list clearly says “”. Have I misunderstood something?

Aseprite needs both the source and compiled versions of Skia, and unless I’m mistaken, it’s hardcoded to statically link it. What benefit could come out of separating the packages, especially since only Aseprite uses this particular fork of Skia?

shaybox commented on 2020-04-30 14:12 (UTC)

This is the aseprite package, it should be using the tagged release source files, not the the latest commit, that's what the -git package is for. It should also not be compiling dependencies like skia, those need to be split into a seperate package to be depended on.

rouhannb commented on 2020-04-13 01:37 (UTC)

I’ve just published that package under the name aseprite-skia-bin. Thanks for the help with the PKGBUILD!

silverwyrda commented on 2020-04-12 18:18 (UTC)

Hi! It seems the author now provides the appropriate pre-built skia library, which makes building aseprite much quicker... You can find it here: Would it be possible to support that (eventually in a different package)?

You probably don't need this but the relevant changes are the following (except I changed the -D order to match CMakeLists.txt and removed the ones that are the same as defaults):

makedepends=('cmake' 'ninja' 'git' 'freeglut')

prepare() {

build() {
  cd "${srcdir}"

  # Build aseprite
  cd "${srcdir}"
  mkdir --parents --verbose build && cd build

  cmake \
    -DLAF_BACKEND=skia \
    -DSKIA_DIR="${srcdir}" \
    -DSKIA_LIBRARY_DIR="${srcdir}/out/Release-x64" \
    -DCMAKE_BUILD_TYPE=Release \
    -G Ninja \

  ninja aseprite


rouhannb commented on 2020-04-12 17:01 (UTC)

The package you needed was harfbuzz-icu, not harfbuzz. I’ve updated the dependencies.

kronaemmanuel commented on 2020-04-12 10:45 (UTC)

I'm constantly getting this error. Any idea on how to solve this? I've even tried installing harfbuzz separately but still the same error.

==> Starting build()...
Done. Made 60 targets from 43 files in 389ms
ninja: Entering directory `out/Release-x64'
[361/1593] compile ../../src/opts/SkOpts_sse41.cpp
In file included from ../../src/sksl/SkSLInterpreter.h:10,
                 from ../../src/opts/SkRasterPipeline_opts.h:13,
                 from ../../src/opts/SkOpts_sse41.cpp:12:
../../src/sksl/SkSLByteCode.h: In constructor ‘SkSL::ByteCode::Vector<width>::Vector(skvx::Vec<width, int>) [with int width = 8]’:
../../src/sksl/SkSLByteCode.h:91:9: note: the ABI for passing parameters with 32-byte alignment has changed in GCC 4.6
   91 |         Vector(skvx::Vec<width, int32_t> i)
      |         ^~~~~~
[587/1593] ACTION //third_party/icu:make_data_assembly(//gn/toolchain:gcc_like)
Generated gen/third_party/icu/icudtl_dat.S
[779/1593] compile ../../src/core/SkOpts.cpp
In file included from ../../src/sksl/SkSLInterpreter.h:10,
                 from ../../src/opts/SkRasterPipeline_opts.h:13,
                 from ../../src/core/SkOpts.cpp:43:
../../src/sksl/SkSLByteCode.h: In constructor ‘SkSL::ByteCode::Vector<width>::Vector(skvx::Vec<width, int>) [with int width = 8]’:
../../src/sksl/SkSLByteCode.h:91:9: note: the ABI for passing parameters with 32-byte alignment has changed in GCC 4.6
   91 |         Vector(skvx::Vec<width, int32_t> i)
      |         ^~~~~~
[1064/1593] compile ../../modules/particles/src/SkParticleEffect.cpp
In file included from ../../src/sksl/SkSLInterpreter.h:10,
                 from ../../modules/particles/include/SkParticleEffect.h:19,
                 from ../../modules/particles/src/SkParticleEffect.cpp:8:
../../src/sksl/SkSLByteCode.h: In constructor ‘SkSL::ByteCode::Vector<width>::Vector(skvx::Vec<width, int>) [with int width = 8]’:
../../src/sksl/SkSLByteCode.h:91:9: note: the ABI for passing parameters with 32-byte alignment has changed in GCC 4.6
   91 |         Vector(skvx::Vec<width, int32_t> i)
      |         ^~~~~~
[1074/1593] compile ../../modules/skshaper/src/SkShaper_harfbuzz.cpp
FAILED: obj/modules/skshaper/src/libskshaper.SkShaper_harfbuzz.o
c++ -MD -MF obj/modules/skshaper/src/libskshaper.SkShaper_harfbuzz.o.d -DSKSHAPER_IMPLEMENTATION=1 -DNDEBUG -DSK_GAMMA_APPLY_TO_A8 -DSK_SHAPER_HARFBUZZ_AVAILABLE -DSK_GL -DSK_SUPPORT_PDF -DSK_HAS_JPEG_LIBRARY -DSK_HAS_PNG_LIBRARY -DSK_CODEC_DECODES_RAW -DSK_HAS_WEBP_LIBRARY -DSK_XML -DSK_R32_SHIFT=16 -DU_USING_ICU_NAMESPACE=0 -DSK_USING_THIRD_PARTY_ICU -I../../modules/skshaper/include -I../.. -I/usr/include/harfbuzz -fstrict-aliasing -fPIC -O3 -fdata-sections -ffunction-sections -Wno-unused-parameter -I /home/krona/builds/aseprite/src/skia/third_party/externals/icu/source/common -I /home/krona/builds/aseprite/src/skia/third_party/externals/icu/source/i18n -I /home/krona/builds/aseprite/src/skia/third_party/icu -std=c++17 -fno-exceptions -fno-rtti -c ../../modules/skshaper/src/SkShaper_harfbuzz.cpp -o obj/modules/skshaper/src/libskshaper.SkShaper_harfbuzz.o
../../modules/skshaper/src/SkShaper_harfbuzz.cpp:33:10: fatal error: hb-icu.h: No such file or directory
   33 | #include <hb-icu.h>
      |          ^~~~~~~~~~
compilation terminated.
[1079/1593] compile ../../modules/skshaper/src/SkShaper.cpp
ninja: build stopped: subcommand failed.
==> ERROR: A failure occurred in build().

rouhannb commented on 2020-04-11 15:27 (UTC)

I’ve added it in.

WaffleLapkin commented on 2020-04-11 09:34 (UTC)

The xorgproto package is missing from makedepends array. This is what currently happens when trying to install aseprite without xorgproto installed:

==> Starting build()...
Done. Made 60 targets from 43 files in 307ms
ninja: Entering directory `out/Release-x64'
[258/1593] compile ../../src/gpu/gl/glx/GrGLMakeNativeInterface_glx.cpp
FAILED: obj/src/gpu/gl/glx/gpu.GrGLMakeNativeInterface_glx.o
c++ -MD -MF obj/src/gpu/gl/glx/gpu.GrGLMakeNativeInterface_glx.o.d -DNDEBUG -DSK_R32_SHIFT=16 -DSK_GAMMA_APPLY_TO_A8 -DSKIA_IMPLEMENTATION=1 -DSK_GL -I../.. -fstrict-aliasing -fPIC -O3 -fdata-sections -ffunction-sections -Wno-unused-parameter -std=c++17 -fno-exceptions -fno-rtti -c ../../src/gpu/gl/glx/GrGLMakeNativeInterface_glx.cpp -o obj/src/gpu/gl/glx/gpu.GrGLMakeNativeInterface_glx.o
In file included from /usr/include/GL/glx.h:30,
                 from ../../src/gpu/gl/glx/GrGLMakeNativeInterface_glx.cpp:15:
/usr/include/X11/Xlib.h:44:10: fatal error: X11/X.h: No such file or directory
   44 | #include <X11/X.h>
      |          ^~~~~~~~~
compilation terminated.
[267/1593] compile ../../src/sksl/SkSLSPIRVCodeGenerator.cpp
ninja: build stopped: subcommand failed.
==> ERROR: A failure occurred in build().
Error making: aseprite

rouhannb commented on 2020-04-04 02:34 (UTC)

Seems that the problem only occurs when running git clone --mirror as makepkg does. I tried running that on another machine (running Ubuntu) and it worked fine.

rouhannb commented on 2020-04-01 04:31 (UTC) (edited on 2020-04-01 04:32 (UTC) by rouhannb)

I decided to try it out for myself, and the same thing is happening to me. I’m not really sure why it’s happening, but my guess is that it has something to do with makepkg. I’ve tried building the AUR package depot-tools-git and I ran into the same problem, but cloning the repository by hand works fine.

I’ll just leave this pinned until the problem goes away: if the package is hanging when downloading depot_tools, run git clone --bare depot_tools and rerun makepkg.

sramey40 commented on 2020-04-01 02:37 (UTC)

I'm having the same issue as Vincentca.

rouhannb commented on 2020-03-31 15:10 (UTC)

If you’re referring to the Git repos, shallow cloning would be hard, if not impossible, as AFAIK Git doesn’t have a way to shallow clone specific commits.

maxcahill commented on 2020-03-31 05:09 (UTC)

Nice to have this package but it sure does hammer your machine and data while installing! Would it be possible to get a shallow clone on the dependencies?

rouhannb commented on 2020-03-30 20:57 (UTC) (edited on 2020-03-30 21:04 (UTC) by rouhannb)

I’ve had that happen to me quite a few times (though only with angle2 or swiftshader). It has nothing to do with the 3/13 update; it’s happened long before. I’m not really sure why it happens, but deleting the repo and rerunning makepkg fixes the problem.

EDIT: I forgot to mention that I think I ran into the same problem while compiling skia manually when I first updated this package. I’m almost certain it’s something funky with either Git or Google’s repos, and I can’t even begin to think of a workaround.

Vincentca commented on 2020-03-30 14:14 (UTC) (edited on 2020-03-30 14:15 (UTC) by Vincentca)

The aur package still doesn't seem to build. I don't know if that was updated on 3/13? It keeps hanging at the last part of the log I'm pasting below.

Cloning aseprite build files...
Checking aseprite dependencies...
Resolving dependencies...
Checking inter-conflicts...

Building aseprite...
==> Making package: aseprite 1.2.17-2 (Mon 30 Mar 2020 09:55:21 AM EDT)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Found
  -> Found aseprite.desktop
  -> Cloning depot_tools git repo...
Cloning into bare repository '/var/tmp/pamac-build-corey/aseprite/depot_tools'...

rouhannb commented on 2020-03-13 22:21 (UTC)

Also my bad. The PKGBUILD uses a custom PATH with depot_tools, and I assumed that the which command would use that PATH. It didn’t.

Building the package worked on my system because I set the PATH when building the package by hand, and for some reason, it never reverted.

And now it looks like just entering the expanded value of which gives a completely different error that never showed up before. I’m just going to restore the old hack again. Maybe I should just use the depot_tools AUR package. Or maybe not, if that’s going to open a whole new can of worms.

bboogie commented on 2020-03-13 21:58 (UTC)

Thanks! But it doesn't build... There is something with and python2. Is there a missing dependency?

which: no in (/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/bin/site_perl:/usr/bin/vendor_perl:/usr/bin/core_perl)

rouhannb commented on 2020-03-13 21:28 (UTC) (edited on 2020-03-13 21:39 (UTC) by rouhannb)

My bad. Looks like I forgot to redownload it, and updpkgsums doesn’t redownload automatically either.

bboogie commented on 2020-03-13 20:36 (UTC)

gn validity check fails. When skipping it aseprite won't build.

rouhannb commented on 2020-02-19 23:31 (UTC)

That's... odd. I bumped the pkgrel, so hopefully that fixes things.

Gaspi commented on 2020-02-19 12:06 (UTC)

I still have the issue that Ev1lbl0w commented a week ago (package always classified as outdated). Thanks for maintaining this package.

rouhannb commented on 2020-02-11 20:57 (UTC)

Fixed, and pandoc removed

Ev1lbl0w commented on 2020-02-11 13:16 (UTC)

You have updated the pkgrel version to 2 on the SRCINFO file, but not on the PKGBUILD file. Thus package is always classified as outdated, and it constantly being updated. Please fix this ASAP.

mmermerkaya commented on 2019-12-28 17:24 (UTC)

+1 to removing pandoc.

deko commented on 2019-11-30 12:19 (UTC)

Can the package do without pandoc? For me it's 500mb of haskell dependencies just to convert to html. And there's really no reason to convert it.

rouhannb commented on 2019-11-22 22:51 (UTC) (edited on 2019-11-22 22:55 (UTC) by rouhannb)

Whoops. I fumbled up my git repo and I guess I forgot to reupdate the sums before pushing :/ The AUR's authentication system seems to be down (the last pushed package was a bit over two hours ago), so I can't push the fix right now. For now, just run makepkg with `--skipchecksums`, or replace the second element of the `sha256sums` array with `deaf646a615c79a4672b087562a09c44beef37e7acfc6f5f66a437d4f3b97a25`. I'll push the fix when I find the AUR back online.

birdsnotberries commented on 2019-11-22 10:12 (UTC)

The hash for aseprite.desktop is outdated

rouhannb commented on 2019-08-31 14:57 (UTC) (edited on 2019-08-31 18:51 (UTC) by rouhannb)

I made a PKGBUILD that does what @jeancf suggested without needing a separate aseprite-skia-git package. I have a diff at if anyone wants to check it out

Also it looks like libwebp is statically linked so it's not needed as a dependency.

EDIT: Looks like hastebin is down...? I've posted it at if so.

jeancf commented on 2019-07-10 19:04 (UTC)

the AUR package aseprite-git builds the aseprite dependency, compiles fine and runs. maybe this package can use the same method.

warmwaffles commented on 2019-07-03 15:19 (UTC) (edited on 2019-07-03 15:27 (UTC) by warmwaffles)

@jusw85 the ticket that dacap is tracking for skia stuff is here I recently added a comment about trying to get it to compile for arch

jusw85 commented on 2019-07-03 14:10 (UTC)

I wasn't able to build with a current version of skia, in case anyone's thinking of fixing the current skia-git to build aseprite with. Even though aseprite-skia is only six months old, there are too many backward compatibility breaking changes. Some of them can be trivially hacked around e.g. patching os/CMakeLists.txt in laf to point to new SKCMS location, adding base directory for includes, but APIs have been wholesale changed or removed

warmwaffles commented on 2019-07-03 04:00 (UTC) (edited on 2019-07-03 04:01 (UTC) by warmwaffles)

All right I suck at making PKGBUILD files but I made a real quick and dirty aseprite-skia-git and then altered the current PKGBUILD for aseprite to use the following

    -DCMAKE_BUILD_TYPE=RelWithDebInfo \
    -DLAF_OS_BACKEND=skia \
    -DSKIA_DIR=/usr/src/aseprite-skia/ \
    -DSKIA_OUT_DIR=/usr/lib \

Here is my glorious mess

After I ran the PKGBUILD.aseprite-skia-git and added those defines, it runs fine for me.

The AUR skia-git currently fails to compile for me. Aseprite also requires some of the source code from skia as well. This dependency is a pain in the ass to build, and it won't be going away and will be required from here on out.

warmwaffles commented on 2019-07-02 22:03 (UTC)

I'm getting the same error as @grevius. I'm going to checkout the older version of the PKGBUILD until something is figured out. I'm sure that separate fork of skia is gonna be a pain. You could make a new package aseprite-skia and add it as a dep here. Just seems like a PITA.

jusw85 commented on 2019-07-02 19:29 (UTC)

@grevius I successfully compiled and built it manually. I had to use a custom skia provided by aseprite at The PKGBUILD provided here won't work with skia until it's updated. I'm not sure if the latest aseprite can work without skia.

grevius commented on 2019-07-02 11:12 (UTC)

after the update for the version 1.2.12-2 i can't open aseprite and i have the following error: Unable to create a user-interface display. Details: Unknown error Error initializing module: gui Error initializing module: gui

abueide commented on 2019-07-02 10:39 (UTC)

Hey sorry I pushed the update to 1.2.12 because I thought I had it working but it turns out it was an older version of aseprite. I tried a lot of things to get the new version to compile compile with skia (the new dependency that broke this package in newer versions). If there's a way to roll back to an older version until I get this figured out please tell me.

warmwaffles commented on 2019-06-28 13:43 (UTC)

Any way we can get a bump to version 1.2.12?

belara commented on 2019-06-15 19:08 (UTC)

Would be great if abueide would push his update.

fbrennan commented on 2019-05-22 03:12 (UTC)

I've added abueide as a co-maintainer. They can now push their work on to this AUR package.

phunni commented on 2019-05-15 13:31 (UTC) (edited on 2019-05-15 13:44 (UTC) by phunni)

Latest version of cmark is 0.29, but this package seems to be looking for 0.28 and, therefore, it won't load.

Edit: a reinstall seems to fix it.

fbrennan commented on 2018-12-27 03:16 (UTC)

I suppose when it comes out I'll just have to create a skia-bin or skia package. :-)

Not sure what you mean by compiling it being a monumental task. Do you mean Skia is fiddly to compile? If so not to worry, I maintain many packages that are long past their prime and have difficult compilations. If you mean it requires a fast PC, I can't help you there.

iirelu commented on 2018-12-18 16:47 (UTC)

The upcoming 1.2.10 release of aseprite is going to start depending on skia. How is that going to be resolved here? As far as I can tell, there aren't any prepackaged skia binaries anywhere, and compiling it is a monumental task. My computer can't even handle building skia-git.

scippie75 commented on 2018-10-19 11:49 (UTC) (edited on 2018-10-19 12:50 (UTC) by scippie75)

After doing all steps without errors and no important warnings, after tidying, checking for issues, creating package, leaving fakeroot, finishing making, it comes to the Installing package step and give me: WARNING: Failed to install built package(s).

Can someone help me (or tell me what extra info I should add to this comment)?

EDIT: Was able to install it with yay (never needed an aur manager before), so I guess something is missing but is not checked for?

bb010g commented on 2018-09-27 03:22 (UTC)

This has a dependency on libx11 that is not listed.

Rubonnek commented on 2018-08-14 00:11 (UTC)

@godofgrunts Thank you for reporting it. The issue should now be fixed.

godofgrunts commented on 2018-08-10 17:35 (UTC)

Using aurman, my package has failed to build.

[ 52%] Building CXX object src/gen/CMakeFiles/gen.dir/ui_class.cpp.o [ 52%] Building C object third_party/libarchive/tar/CMakeFiles/bsdtar.dir/__/libarchive_fe/passphrase.c.o [ 52%] Linking C executable ../../../bin/bsdtar [ 52%] Built target bsdtar [ 52%] Linking CXX executable ../../bin/gen [ 52%] Built target gen make: *** [Makefile:141: all] Error 2 ==> ERROR: A failure occurred in build(). Aborting... 2018-08-10 13:30:11,669 - wrappers - makepkg - ERROR - makepkg query ['makepkg', '-cf', '--noconfirm'] failed in directory /home/godofgrunts/.cache/aurman/aseprite ~~ the following packages are now orphans :: community/cmark

Rubonnek commented on 2018-04-12 15:04 (UTC)

It should not be added. Most packages from the AUR are expected to be built from source code unless they include '-bin' in their pkgname.

Also AUR users must ensure the base-devel package group is installed as stated in the wiki:

pkg-config is included in that package group.

KP1533TM2 commented on 2018-04-12 14:51 (UTC)

Hi there. I think pkg-config should be added as a dependecncy too to enable CMake to locate harfbuzz.

yahbu commented on 2018-03-19 01:54 (UTC)


Thank you for your quick reply, seems to be working perfectly now.

Rubonnek commented on 2018-03-19 01:30 (UTC)

@yahbu Thank you for reporting this. I forgot to drop the configuration for using the shared libloadpng library. You should be able to build it now.

This library is included with the allegro4 package from the official repository, a dependency that was dropped since it was causing weird aseprite glitches. It's better to use the bundled allegro4 and libloadpng libraries instead, which is what the PKGBUILD does now.

yahbu commented on 2018-03-19 00:53 (UTC) (edited on 2018-03-19 01:17 (UTC) by yahbu)

Hiya! @Rubonnek

I can't build the latest update of Aseprite Keep getting this CMake Error:


CMake Error: The following variables are used in this project, but they are set to NOTFOUND.

Please set them or make sure they are set and tested correctly in the CMake files: LOADPNG_INCLUDE_DIR


-- Configuring incomplete, errors occurred!

See also "/tmp/pamac-build-username/aseprite/src/build/CMakeFiles/CMakeOutput.log".

See also "/tmp/pamac-build-username/aseprite/src/build/CMakeFiles/CMakeError.log".

==> ERROR: A failure occurred in build(). Aborting...

I checked both error logs but they are very long and i am not a programmer.

Bionov commented on 2018-03-12 00:35 (UTC)

@Rubonnek Oh thank you for the reply ! I gonna try using another terminal emulator then.


Rubonnek commented on 2018-03-11 01:46 (UTC)

I'm not able to reproduce that warning, but it seems to be unrelated to building this package, and related to the terminal emulator that you are using.

Bionov commented on 2018-03-11 01:29 (UTC)

Hello, when trying to build this I get thris error :

WARNING: terminal is not fully functional

Any idea why ?

aperez commented on 2017-11-16 21:34 (UTC)

@Rubonnek: Wow, that was indeed a very quick fix. Package builds and installs fine now for me also. Thanks a lot!

kiswa commented on 2017-11-16 20:34 (UTC)

Thanks for the quick turnaround! Just finished updating and everything works great. Interestingly, Aseprite identifies itself as v1.3-dev, but that's an Aseprite issue.

Rubonnek commented on 2017-11-16 14:37 (UTC)

@aperez thanks. It's now fixed. @aperez @ronjouch @Brottweller @kiswa thanks a lot for the help. After hurricane Maria, what little power I could get for my laptop was used only for necessary things, but thanks to all of you I was able to get this done and still have some power left. aseprite should now build and install successfully.

aperez commented on 2017-11-16 13:51 (UTC)

@Rubonnek: Version 1.2.4-2 builds, but it won't install because it tries to overwrite files. Pacman will say: ... error: failed to commit transaction (conflicting files) aseprite: /usr/bin/bsdcat exists in filesystem aseprite: /usr/bin/bsdcpio exists in filesystem aseprite: /usr/bin/bsdtar exists in filesystem Errors occurred, no packages were upgraded.

Rubonnek commented on 2017-11-16 13:06 (UTC)

@ronjouch @Brottweller, thanks. PKGBUILD updated. Please let me know if you find any other issues. I'll work on those as soon as I can.

ronjouch commented on 2017-11-16 12:50 (UTC)

Confirming @Brottweiler's build failure below

Brottweiler commented on 2017-11-16 12:45 (UTC)

Fails to build;

Rubonnek commented on 2017-11-16 12:35 (UTC)

I'm currently unable to thoroughly test the PKGBUILD, but it has been updated to 1.24 and package compilation starts at the very least. Let me know if you find any issues.

kiswa commented on 2017-11-13 21:18 (UTC)

Is there some way to help update this to the latest release (v.1.2.4)?

Rubonnek commented on 2017-03-03 20:45 (UTC)

aseprite.desktop file has been updated.

shy13 commented on 2017-03-03 13:05 (UTC)

Please, add StartupWMClass=allegro in .desktop file. That fixes the bug described here Thanks.

Rubonnek commented on 2017-03-03 05:30 (UTC)

Thanks, I was able to reproduce the error. The PKGBUILD has been fixed.

ronjouch commented on 2017-03-02 22:47 (UTC)

1.1.13-1 fails to build with this error: ==> Continue building aseprite ? [Y/n] ==> ---------------------------------- ==> ==> Building and installing package ==> Making package: aseprite 1.1.13-1 (Thu Mar 2 17:45:25 EST 2017) ==> Checking runtime dependencies... ==> Checking buildtime dependencies... ==> Retrieving sources... -> Downloading % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 604 0 604 0 0 3471 0 --:--:-- --:--:-- --:--:-- 3491 100 12.2M 100 12.2M 0 0 2211k 0 0:00:05 0:00:05 --:--:-- 2739k -> Found aseprite.desktop ==> Validating source files with sha256sums... ... Passed aseprite.desktop ... Passed ==> Extracting sources... -> Extracting with bsdtar ==> Starting build()... Do you accept the EULA? yes/NO (set ASEPRITE_ACCEPT_EULA=yes to skip this message) yes fatal: Not a git repository (or any parent up to mount point /tmp) Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set). ==> ERROR: A failure occurred in build(). Aborting... ==> ERROR: Makepkg was unable to build aseprite. ==> Restart building aseprite ? [y/N] ==> --------------------------------- ==> ==> ERROR: unable to update

hollunder commented on 2016-12-20 14:01 (UTC)

This package installs and works if you add git to makedepends and libx11 to depends.

commented on 2016-12-20 12:51 (UTC)

The dependencies from allegro should be added to the PKGBUILD. I have no time to do it now, but you can install allegro first to make sure they are in.

hollunder commented on 2016-12-20 11:33 (UTC)

This one is a bit weirder: -- Found Threads: TRUE CMake Error at /usr/share/cmake-3.6/Modules/FindX11.cmake:439 (message): Could not find X11 Call Stack (most recent call first): CMakeLists.txt:356 (find_package)

hollunder commented on 2016-12-20 10:57 (UTC)

I get a build error: ==> Validating source files with sha256sums... aseprite ... Skipped aseprite.desktop ... Passed ==> ERROR: Cannot find the git package needed to handle git sources. ==> ERROR: An unknown error has occurred. Exiting... ==> ERROR: Build failed, check /var/lib/aurbuild/x86_64/murks/build

commented on 2016-12-19 20:41 (UTC)

Removed conflict with cmark and gtest (thx prurigro). As for the EULA, please point me to a discussion where TUs consider that building from AUR equals accepting a license. One more thing, I am willing to orphan the package if someone wants to take over maintaining it.

napa3um commented on 2016-12-19 14:44 (UTC)

conflicts 'cmark' and 'gtest' is correct?

Darkhogg commented on 2016-12-17 17:58 (UTC)

Ok, the env var works, but... Why not skip the EULA entirely? Just drop it in the licenses folder, installing the package just means accepting the EULA.

prurigro commented on 2016-12-12 19:21 (UTC)

Aseprite loads without any errors if you move cmark and gtest from conflicts to depends and delete their files like I've done here: It might be worth going through and creating an array of only files that conflict with either of those two packages instead of deleting anything with cmark or gtest in their filenames in an actual update.

nlowe commented on 2016-12-11 03:56 (UTC)

@benob: what about the install makes it conflict with cmark and gtest? We should be opening an issue upstream (, and at a quick glance I don't see one opened yet.

commented on 2016-12-10 15:11 (UTC)

Updated to 1.1.11 Note that upstream messed up the install so for now aseprite conflicts with cmark and gtest. You can also skip the EULA prompt by setting ASEPRITE_ACCEPT_EULA=yes

yarrok commented on 2016-11-06 20:28 (UTC)

I'm getting an error when running makepkg, "CMake Error at third_party/CMakeLists.txt:52 (message): gtest/gtest.h is missing. Initialize submodules: git submodule update --init --recursive" running `git submodule update --init --recursive` in src/aseprite doesn't seem to help, any pointers? thanks for keeping this up to date, btw

commented on 2016-10-29 11:32 (UTC)

What about having the user set an environment variable like ACCEPT_EULA=y to bypass the interactive step?

Darkhogg commented on 2016-10-29 11:25 (UTC)

PLEASE, remove the interactive EULA acceptance, it completely breaks unattended updates via AUR helpers and it's plain annoying. By installing a package it's assumed you accept its license, there's no need for an extra step.

commented on 2016-07-31 13:10 (UTC)

Updated 1.1.7, sorry about the delay.

saverio commented on 2016-07-29 10:47 (UTC) (edited on 2016-07-29 10:47 (UTC) by saverio)

Hi benob, don't you update anymore this package? Looking at its github site (, it's alive and the version up to now is 1.1.7. Thanks for the information you can provide.

commented on 2016-05-16 15:40 (UTC)

Updated to

commented on 2016-04-19 08:31 (UTC)

Updated to

commented on 2016-04-12 18:23 (UTC)

Updated to 1.1.4

commented on 2016-03-29 13:04 (UTC)

Added webp support and included mime types in .desktop file.

rautesamtr commented on 2016-03-26 12:21 (UTC)

Could you enable webp support? It is enabled in the commercial distribution. -DWITH_WEBP_SUPPORT=ON -DUSE_SHARED_LIBWEBP=ON this would add the extra/libwebp dependency. Also it would be nice if you add the MimeType key to the desktop file: MimeType=image/bmp;image/gif;image/jpeg;image/png;image/x-pcx;image/x-tga;image/;video/x-flic;image/webp; Maybe you could already also add image/x-aseprite to the list. I am trying to get the upstream project to include a mimetype file. And maybe also modify the exec key to: Exec=aseprite %U

pixelsyntax commented on 2016-03-23 22:09 (UTC) (edited on 2016-03-24 00:17 (UTC) by pixelsyntax)

After a few minutes of trying to work out why Aseprite wasnt responding to mouse clicks, it seems the left/right mouse buttons are swapped, only in this application. Am I the only one experiencing this issue? Edit: Fixed in aseprite-git

gururise commented on 2016-03-04 19:36 (UTC)

Nevermind about my last comment. After rebooting, the compile worked fine. I guess yaourt must have updated some other libraries on my system that required a reboot.

commented on 2016-03-04 19:15 (UTC)

Do you have a custom install of gcc or libc? Are you running on an exotic architecture? Triggering bugs in gcc is rather unusual.

commented on 2016-02-29 10:44 (UTC)

Updated to 1.1.2

cosarara97 commented on 2015-11-21 15:05 (UTC)

It seems my cmake had somehow broken - reinstalling cmake fixed the issue.

commented on 2015-11-21 13:46 (UTC)

No problem compiling it through yaourt. Please retry and try to find where the issue comes from.

cosarara97 commented on 2015-11-21 13:08 (UTC)

The package seems broken

aaronp commented on 2015-11-09 00:17 (UTC)

The latest update seems to be including freetype in the package.

commented on 2015-09-09 14:12 (UTC)

Updated to (thanks sirblackheart)

rautesamtr commented on 2015-09-09 13:40 (UTC)

with this patch for the PKGBUILD, version should build and run

commented on 2015-09-03 06:36 (UTC)

No time to update to right now, bumping the version number does not compile. Feel free to contribute a working PKGBUILD

chanster commented on 2015-09-03 01:14 (UTC)

Is this package still maintained?

commented on 2015-02-25 16:53 (UTC)

Updated to 1.0.9

commented on 2015-02-02 17:10 (UTC)

Updated to

commented on 2014-12-05 07:59 (UTC)

Updated to 1.0.8

commented on 2014-11-25 09:15 (UTC)

Updated to 1.0.7

commented on 2014-11-14 14:57 (UTC)

1.0.6-2: disable use of system libpng which seems to not work for everybody

gmenguez commented on 2014-11-14 14:20 (UTC)

Hi, I wasn't able to open aseprite, it failed with this error: $ aseprite libpng warning: Application built with libpng-1.5.9 but running with 1.6.14 Error loading skins/default/sheet.png file Segmentation fault (core dumped) libpng 1.6.14 is the one I have installed. However, modifying the cmake command line in PKGBUILD from -DUSE_SHARED_LIBPNG=ON \ to -DUSE_SHARED_LIBPNG=OFF \ has made it work again. Does anybody know why is this happening? Is something wrong in my system setup? Thanks.

commented on 2014-11-13 12:57 (UTC)

Updated to 1.0.6 (thanks Kamil)

commented on 2014-11-12 20:07 (UTC)

Upstream v 1.0.6 does not compile with the following error: CMake Error: File src/aseprite/third_party/pixman-cmake/../pixman/pixman/ does not exist. CMake Error at third_party/pixman-cmake/CMakeLists.txt:13 (configure_file): configure_file Problem configuring file I'll investigate as soon as I have time, but let me know if you have a fix.

commented on 2014-09-22 12:59 (UTC)

Updated to 1.0.5 (thanks bidek)

commented on 2014-09-09 09:34 (UTC)

Updated to 1.0.4, with a small patch to fix the compilation issue until upstream come up with a proper fix (

commented on 2014-09-09 08:29 (UTC)

Having trouble compiling 1.0.4. aseprite/src/aseprite-1.0.4/src/she/alleg4/she_alleg4.cpp:98:28: error: expected unqualified-id before numeric constant event.setType(Event::None); Will update as soon as I have time to look into this.

Etzos commented on 2014-08-31 23:35 (UTC)

@benob: I don't know if you have this slated for the next release, but PKGBUILDs support git repositories directly now (as well as branch and tag fragments) so the repo can be added to the source instead of doing it separately. i.e. source=("git+${pkgver}")

commented on 2014-06-28 19:56 (UTC)

Updated to v1.0.2 -- Note that sources are now downloaded directly from tagged git repository.

xyproto commented on 2013-10-14 12:40 (UTC)

Updated PKGBUILD that includes a desktop shortcut (the source package already includes an icon):

commented on 2013-08-21 14:05 (UTC)

@lintz: do you have an icon in mind?

lintz commented on 2013-08-18 17:33 (UTC)

Would be nice if it installed a .desktop file in /usr/share/applications/ (Would make it show up in menus like for example xfce4) Basic template would be [Desktop Entry] Type=Application Encoding=UTF-8 Name=Aseprite Comment=Pixel-art paint program Exec=aseprite Terminal=false Categories=Graphics

cookies commented on 2012-11-08 14:22 (UTC)

The package gets currently installed into "/usr/local", you should change "cmake --prefix=/usr ." into "cmake -DCMAKE_INSTALL_PREFIX:STRING=/usr ." Thanks.

commented on 2012-10-10 08:15 (UTC)

> Works for me on 64bit, but the package installs as asesprite-git instead of aseprite-git. > Thanks :) Fixed package name and submitted to aur. By the way, 0.9.5 works on x86_64 again.

commented on 2012-10-10 03:49 (UTC)

<<Comment by: benob on Sun, 15 Jul 2012 21:55:17 +0000 I made a PKGBUILD for aseprite-git at, but it currently does not build.>> Works for me on 64bit, but the package installs as asesprite-git instead of aseprite-git. Thanks :)

tomasabril commented on 2012-07-16 00:59 (UTC)

Thanks, I have 0.9.2 for now.

commented on 2012-07-15 21:55 (UTC)

I made a PKGBUILD for aseprite-git at, but it currently does not build. Your best option is probably to revert to 0.9.2

tomasabril commented on 2012-07-15 20:34 (UTC)

[url][/url] is fixed, i wonder when there will be a new release. Or can you patch it?

commented on 2012-07-13 17:24 (UTC)

Updated to 0.9.3, but it does not compile on x86_64 because of that upstream issue: