profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/tomsmeding/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.
Tom Smeding tomsmeding https://git.tomsmeding.com Hacker of software, maths and music.

jonschlinkert/to-clipboard 17

Copy to the clipboard.

tomsmeding/circular-buffer 11

A NodeJS simple circular buffer implementation supporting indexing

tomsmeding/editor 9

My own text editor

simplyGits/simplyHomework 4

Education with a little more swag.

order-and-chaos/client 3

Order & Chaos client (and AI)

order-and-chaos/server 3

Order and Chaos server

tomsmeding/accelerate 1

Embedded language for high-performance array computations

tomsmeding/codetyper 1

Typeracer, for code

tomsmeding/Flow 1

My own programming language Flow. Work in progress.

push eventVMatthijs/CHAD

Tom Smeding

commit sha b4e2e93f27d1d337ff44cf9e7da42cb72a82d8ca

More WIP

view details

push time in 4 days

create barnchVMatthijs/CHAD

branch : lambdacalc

created branch time in 8 days

push eventVMatthijs/CHAD

Tom Smeding

commit sha ea70257aea965ff87af0b2d406f6423606857134

Documentation and formatting

view details

push time in 10 days

create barnchVMatthijs/CHAD

branch : factor-out-counting

created branch time in 10 days

push eventVMatthijs/CHAD

Tom Smeding

commit sha 4f124cf415a50536fee524a973ba6773f5d0d8bd

hfmt

view details

push time in 10 days

push eventVMatthijs/CHAD

Tom Smeding

commit sha 284ee6ecf1ef4945399a6f7efbd56165897d1b40

Fix minor grammar in README

view details

Tom Smeding

commit sha 47594bf18c889ba326c5320a793aaf60cdcaa387

Remove redundant default Setup.hs

view details

push time in 12 days

push eventVMatthijs/CHAD

Tom Smeding

commit sha 7d56c4741ff171594e25bd9fd5bd5c98ca521d8e

hie.yaml: Ensure HLS chooses stack for the main code directory Apparently the ordering matters, not sure why

view details

Tom Smeding

commit sha 4bb494a59227ad866b9e03f267f68584a21a1714

Update Tom's email address to personal

view details

push time in 12 days

delete branch tomsmeding/libarchive

delete branch : patch-1

delete time in 20 days

push eventtomsmeding/ADBench

Tom Smeding

commit sha 101508da93cd4fc80c0ea7a690c6de824c815ca6

Ensure stack.yaml works with GHC 8.10.6

view details

push time in a month

push eventtomsmeding/accelerate

Tom Smeding

commit sha 03a463380647a7b7b6e9085be0acc03ebd8724fd

Ensure stack-8.10.yaml is up to date

view details

push time in a month

push eventtomsmeding/ADBench

Tom Smeding

commit sha d790abb6602a39a9d1ba33d520f7ce75b22b3ca6

Allow haskell base version 4.15

view details

push time in a month

push eventtomsmeding/accelerate

Tom Smeding

commit sha e73f7131f8bcec6ecd2476352fc05438d3d9ff97

Actually add AD dependencies in stack-9.0.yaml

view details

push time in a month

push eventtomsmeding/accelerate

Tom Smeding

commit sha 3ad43c9c21bec2f2f8d61b580b92a212bda4068a

No CI on this fork

view details

push time in a month

push eventtomsmeding/accelerate

Trevor L. McDonell

commit sha 9e99fab2c1a592b9307eeff3f06f89942db240c8

update pkgconfig-depends

view details

Trevor L. McDonell

commit sha a5651a0eda0504ade8890a5957d4091f991ff536

ci: drop —fast

view details

Tom Smeding

commit sha 7a800e7a3812711db50f9420f7b8a456bc24b816

Fix typo in error message in app/Main.hs

view details

Trevor L. McDonell

commit sha ebbff92f5e0b3d1d0c18f79ea9fb8ebbf4e624d4

cabal: update description of ‘debug’ flag to include required dependencies

view details

Trevor L. McDonell

commit sha 649c9ff71a3298e56a52907ded0f6ca5ce7c28e6

Merge branch 'master' of github.com:AccelerateHS/accelerate

view details

Trevor L. McDonell

commit sha ad9ca9d7148866a7d1f42a1fa8f92d8eb2f92b8e

replace ‘text-format’ with ‘formatting’ The former seems unmaintained, and I really should have checked this before doing the initial work

view details

Trevor L. McDonell

commit sha 7f3faae1985b4673265d541df9eac13440f03e7a

remove Buildable instances

view details

Trevor L. McDonell

commit sha 2093102e299aad8a19ebfdef7cda19df23d48fdf

more type-safe formatting

view details

Trevor L. McDonell

commit sha f21cece645c26bc8d9f6e215d762077edb4880b8

TODO

view details

Trevor L. McDonell

commit sha ec7a6f915b5d82ec8ed3f34c475a0a6015f33bd4

more formatting

view details

Trevor L. McDonell

commit sha 62d4c74d970d92e1840a1a25ec096b99fcb3732a

bump version requirement for 'formatting'

view details

Trevor L. McDonell

commit sha 4c3ad356040bb858b3248c427c1a4e46988fdd3d

improve (?) trace formatting functions

view details

Trevor L. McDonell

commit sha 69ff12814acd14a1f5a37149e034ad858272fcad

Revert "ci: drop —fast" This reverts commit a5651a0eda0504ade8890a5957d4091f991ff536.

view details

Trevor L. McDonell

commit sha a6ef1c25c7381c045067a318a031f5785e203431

ci: build with haddocks

view details

Trevor L. McDonell

commit sha c9a74670f546ac0d92ce9e19ed6fc70b6159fa46

wibble

view details

Trevor L. McDonell

commit sha 7e5f0c1ef3dad9e006d2f58f527ea5245f4fb0d1

formatting for internalError

view details

Trevor L. McDonell

commit sha 8655a01f06812a243afcbdc7649737e54edbd548

build fix

view details

Trevor L. McDonell

commit sha 74274f426dfe6ac5d99cef31191ac514bd19a52f

wibble

view details

Trevor L. McDonell

commit sha 5d32f1710e39f8aa7596f6beffebd2e8369cc36d

Merge pull request #504 from tmcdonell/wip/formatting Replace 'text-format' package with 'formatting'

view details

Trevor L. McDonell

commit sha 2f4a64387cb55680d2e83cd56a1867dcf1b638c7

build fix for ghc-9

view details

push time in a month

push eventtomsmeding/ircbrowse

Tom Smeding

commit sha 197d241845f89cda5665b9f2b76903369ad1f176

Send correct mime type in outputRSS

view details

push time in a month

issue openedneovim/neovim

[request] LSP: Add config option to not filter completion results

Feature already in Vim?

No response

Feature description

Currently, in vim.lsp.util.text_document_completion_list_to_complete_items, the completion results from the language server are not kept as-is, but instead some preprocessing is performed:

  1. Only completions which exactly match the already-typed prefix are kept;
  2. The completions are sorted.

While this is probably nice for some language servers that return weird results, this is completely counterproductive when a language server tries to be helpful by returning results with:

  • case-insensitive matching
  • typo-fixing
  • intelligent result ordering so that most likely matches come first
  • (maybe other things too)

At least the first three are performed by the Haskell language server.

At least for the Haskell language server in particular, my completion results get significantly more useful when I remove the two lines linked above. However, this functionality is probably not in the LSP implementation for nothing. Therefore:

Request: A configuration option that allows me to turn off this result filtering and sorting. I don't know how granular this option should be, i.e. if it should be more than just one boolean. A logical option would seem to be two booleans: one to control exact-prefix-match filtering, and one to control sorting.

Thanks for the good work!

created time in a month

issue commentfjvallarino/monomer

Error when running example: glMakeCurrent

Seems to work! Awesome :)

tomsmeding

comment created time in a month

issue commentfjvallarino/monomer

Error when running example: glMakeCurrent

That indeed works for me. It would be kind of sad, though, if it wasn't possible to make a GUI with monomer for a Haskell app that uses the threaded runtime; there are more things that need the threaded runtime, for example certain FFI-using functionality I believe.

I'll be busy the coming weeks and might not have more time to look at this. In any case, thanks for the good work. :)

tomsmeding

comment created time in a month

issue commentfjvallarino/monomer

Error when running example: glMakeCurrent

@fjvallarino This is a modification of Main.hs of https://github.com/fjvallarino/sdl-continuous-resize: https://paste.tomsmeding.com/ai3lXJdj. With doTHREADED = True it should be equivalent to the original code; with doTHREADED = False it doesn't have the eventloop/renderloop thread separation and just runs the loops interlocked on one thread.

With doTHREADED = True it errors for me with the glMakeCurrent error, just like the original sdl-continuous-resize project did. With doTHREADED = False it works fine for me. (Though that single waitEvent should really be pollEvents in this case to keep the app going when many events occur, but that's an unrelated, easily fixed issue.)

While this doesn't yet indicate what exactly goes wrong, it does strongly suggest that it's got to do with the separation of the render thread. I wonder if it's just the nvidia driver acting up with separate render threads, but maybe that's paranoia.

tomsmeding

comment created time in a month

issue commentfjvallarino/monomer

Error when running example: glMakeCurrent

@fjvallarino That project fails with the exact same error. I also tried using stack instead of cabal (which is made difficult by the fact that stack bundles an old version of c2hs that isn't compatible with my system's gcc 11, or something, meaning I have to laboriously convince it to use gcc 10), but that fails in the same way again.

I'll try debugging this some, but no guarantees.

tomsmeding

comment created time in a month

issue commentfjvallarino/monomer

Error when running example: glMakeCurrent

@fjvallarino glxinfo outputs lots of stuff and then exits with code 0; glxgears shows me some spinning gears. glxgears -info shows me that it's indeed running on my nvidia card. I had done a full system update recently but did it again to be sure, no change; I recompiled the world using GHC 8.10.6 and no change.

tomsmeding

comment created time in a month

push eventtomsmeding/ircbrowse

Tom Smeding

commit sha 953282240a9167bcab573bf29e4ae4547d99e613

Up to ghc 8.10.6

view details

push time in a month

push eventtomsmeding/ircbrowse

Tom Smeding

commit sha 455e7dd0353162adc2d619f3a91a658e97aae640

Up to ghc 8.10.6

view details

push time in a month

issue openedfjvallarino/monomer

Error when running example: glMakeCurrent

Trying to run the first tutorial example program by creating a new cabal project, adding nanovg-hs at https://github.com/cocreature/nanovg-hs/commit/ba8733f58da63e0934b5d2853ed84b71e872e6f1 and monomer at https://github.com/fjvallarino/monomer/commit/523355c389621330d3f026384b9673548a3f5f8d as in-tree dependencies. I also set the stb_truetype flag to true on nanovg because monomer's stack.yaml does that.

Everything compiles (with some unused-import warnings in nanovg-hs), but running the resulting executable produces the following output:

Renderer: NVIDIA GeForce GTX 1050 Ti/PCIe/SSE2
Version: 3.2.0 NVIDIA 470.57.02
test: SDLCallFailed {sdlExceptionCaller = "SDL.Video.OpenGL.glMakeCurrent", sdlFunction = "SDL_GL_MakeCurrent", sdlExceptionError = "Unable to make GL context current"}

(test is the name of the executable)

A window is created but nothing is ever drawn in it (only the window border gets drawn on my system). When I close the window, the application does print About to destroyWindow; I believe the startRenderThread thread just died, and nothing happens afterwards.

Any idea if this is a misconfiguration on my side? Can I debug this?

System:

  • Arch Linux (Linux 5.13.9) with i3
  • ghc 8.10.4, cabal 3.4.0.0

created time in a month

startedfjvallarino/monomer

started time in a month

issue commenthaskoin/secp256k1-haskell

[libsecp256k1] illegal argument: input != NULL

14:06 < tomsmeding> So that secp haskell lib should do some checking

I didn't mean to sound demeaning in any way!

I do believe that perhaps importSig should ensure that the pointer it passes to ecdsaSignatureParseDer is non-NULL. As I wrote on irc: Data.ByteString.empty and fromString "" :: ByteString are semantically equal in Haskell, so in order to make the API as foolproof as possible, I would say that library functions should preserve that property (regardless of whether it is sensible to pass empty bytestrings to those functions — whatever it should or shouldn't do, it should not crash :) ).

tim2CF

comment created time in a month

delete branch tomsmeding/streamly-posix

delete branch : streamly080-fix

delete time in a month

pull request commenthasufell/hpath

Make compatible with streamly 0.8.0

@hasufell Done :)

tomsmeding

comment created time in a month

push eventtomsmeding/pastebin-haskell

Tom Smeding

commit sha 25de4aa531a782ca8c34b409046b414f655f133c

Update changelog; 0.2.14.0

view details

push time in a month

push eventtomsmeding/pastebin-haskell

Tom Smeding

commit sha 6fd8e57b75f0b7a0f25560260e92b04920590a62

Add modification time to tar downloads

view details

push time in a month