summary refs log tree commit diff
Commit message (Collapse)AuthorAge
* Clean up gfxx and add default paletteJune McEnroe2018-02-28
| | | | | | | | | | | | | | The palette was generated in Rust: extern crate hsl; fn main() { for i in 0..256 { let h = i as f64 * 360.0 / 256.0; let hsl = hsl::HSL { h, s: 1.0, l: 0.5 }; let (r, g, b) = hsl.to_rgb(); println!("0x{:02X}{:02X}{:02X},", r, g, b); } }
* Print message in gfx/noneJune McEnroe2018-02-27
|
* Check revents with &June McEnroe2018-02-27
|
* Remove jrpJune McEnroe2018-02-25
| | | | It was fun.
* Tweak watch exit codesJune McEnroe2018-02-25
|
* Clean up klonJune McEnroe2018-02-25
|
* Replace cmcenroe.me with causal.agencyJune McEnroe2018-02-24
|
* Clean up xxJune McEnroe2018-02-23
| | | | There are probably like 12 commits with this title.
* Clean up wakeJune McEnroe2018-02-23
|
* Avoid unnecessary FDs in errorsJune McEnroe2018-02-23
|
* Clean up pbdJune McEnroe2018-02-23
| | | | Seems probably a good idea to CLOEXEC sockets.
* Clean up fbclockJune McEnroe2018-02-23
| | | | | | | You can't use the return value of gzerror to check if an error occurred or not. Its implementation actually checks if the internal error is NULL and returns the empty string if it is! This is stupid and unhelpful, so check gzeof first since its return value actually means something.
* Don't bother checking result of close(2) in dtchJune McEnroe2018-02-23
|
* Tweak fbatt exit codesJune McEnroe2018-02-23
|
* Clean up dtch (again, probably)June McEnroe2018-02-23
| | | | | Mostly changes to exit codes and renaming len to size. Also no more writeAll.
* Return NOINPUT on execvp failureJune McEnroe2018-02-23
|
* Return EX_OSFILE in briJune McEnroe2018-02-23
|
* Rewrite briJune McEnroe2018-02-22
|
* Exit cleanly from hnel on LinuxJune McEnroe2018-02-22
| | | | Okay the change that actually does that is checking revents == POLLIN.
* Abort on other values of enums in pngoJune McEnroe2018-02-22
|
* Fix zlib types on 32-bitJune McEnroe2018-02-22
|
* Git config merge.conflictStyle diff3June McEnroe2018-02-21
| | | | | This is probably a good idea, though I haven't run into a merge conflict yet.
* Switch to and vendor Go Mono fontJune McEnroe2018-02-21
| | | | | | | Go Mono is almost identical to Luxi Mono (same designers) but MIT-licensed. From https://go.googlesource.com/image font/gofont/ttfs.
* Add unistd.h include to pngo.cJune McEnroe2018-02-20
| | | | Where getopt properly lives.
* Take multiple inputs to pngoJune McEnroe2018-02-20
|
* Reduce bit depth in pngoJune McEnroe2018-02-20
|
* Always use None filter for indexed or bit depth < 8June McEnroe2018-02-20
|
* Add pngo command line optionsJune McEnroe2018-02-19
|
* Clean up pngo optimization functionsJune McEnroe2018-02-19
|
* Index color if possible in pngoJune McEnroe2018-02-19
|
* Read and write palette in pngoJune McEnroe2018-02-19
|
* Add pngo to READMEJune McEnroe2018-02-18
|
* Always dump truecolor PNG in gfxxJune McEnroe2018-02-18
| | | | pngo can handle reducing it to grayscale.
* Eliminate redundant color in pngoJune McEnroe2018-02-18
|
* Just globalize all the things in pngoJune McEnroe2018-02-18
|
* Tweak pngo codeJune McEnroe2018-02-17
|
* Eliminate redundant alpha in pngoJune McEnroe2018-02-17
|
* Support all bit depths in pngoJune McEnroe2018-02-17
|
* Implement filter heuristic from PNG specJune McEnroe2018-02-17
|
* Fix pngo glitchJune McEnroe2018-02-16
| | | | | Basically changing the filter type without actually recalculating the bytes results in some interesting stuff.
* Refactor scanlines and filter bytes, more glitchJune McEnroe2018-02-16
|
* Fix paethPredictor comparisonJune McEnroe2018-02-16
|
* Handle more color types, depths, IDATs in pngoJune McEnroe2018-02-16
| | | | | Also fixed reconData so that it can at least round-trip the glitches it creates.
* Add WIP pngo which produces glitch artJune McEnroe2018-02-15
|
* Encode grayscale PNGs in gfxxJune McEnroe2018-02-14
|
* Avoid accessing past last byte in gfxxJune McEnroe2018-02-14
|
* Print PNG path on gfxx dumpJune McEnroe2018-02-14
|
* Add gfxx 4444 presetJune McEnroe2018-02-14
|
* Fix gfxx row scrolling for BITS_TOTAL < 8June McEnroe2018-02-14
|
* Write at least 1 for each component of sBIT in gfxxJune McEnroe2018-02-14
| | | | PNG requires each to be 0 < x <= depth. ImageMagick actually checks.
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-11-27 05:01:29 +0000