If you are wondering where the data of this site comes from, please visit 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.
Conrad Meyer cemeyer Seattle, Washington

alexsmith1612/hadoofus 44

HDFS client library for C

cemeyer/falsisign-mirror 14 mirrored (snapshot taken 2020-04-08)

cemeyer/avr-emu 9

An AVR emulator

cemeyer/advent-of-code-2017 1

Advent of Code 2017 puzzles

cemeyer/dynpk 1

rspanton's dynpk

cemeyer/APCSGridWorldChips 0

A clone of Chip's Challenge on the AP CS GridWorld framework (from highschool)

cemeyer/atf 0

Automatated Testing Framework

cemeyer/azure-storage-azcopy 0

The new Azure Storage data transfer utility - AzCopy v10

cemeyer/b-epsilon-tree 0

A simple implementation of the write-optimized Bε Tree 🌳 - for MIT 6.854 (Advanced Algorithms).

cemeyer/backtrace-rs 0

Backtraces in Rust

PR opened ranweiler/dent

Fix deprecation warnings (rust 1.51)

For Error, drop the deprecated description() method and move the logic into the Display impl.

In, use explicit 'dyn' annotation for Boxes of traits. This is functionally identical, but the older syntax is deprecated.

In the KAT support logic, the format string of an assert!() must be a string literal in Rust 2021. As the variable we were using for the format string was itself a formatted string, we can just simplify by doing the formatting in assert!() itself.

+16 -18

0 comment

3 changed files

pr created time in 2 months

create barnchcemeyer/dent

branch : fix-deprecated-warnings

created branch time in 2 months

fork cemeyer/dent

Library and tool for summarizing, comparing small data sets

fork in 2 months


started time in 2 months

issue commentezrosent/frawk

Is it possilbe to build frawk with musl?

The mallinfo dependency seems to come from LLVM's Process::GetMallocUsage() API, with perhaps buggy cross-build detection for HAVE_MALLINFO. Musl doesn't have mallinfo, but your system glibc (apparently) did.


comment created time in 2 months


started time in 2 months

pull request commentjedisct1/libsodium

Define randombytes_set_implementation argument to be const

If my memory serves me correctly, this was intentional, due to some compilers not supporting constant function pointers.

If we need to support broken compilers, we could conditionally define a macro that evaluates to const on conventional compilers and to nothing on broken compilers. Would that be an ok workaround for this concern? (Which compilers do we support that have this deficiency?)

There wouldn't be anything wrong with passing a function pointer and change it later. I can see applications doing this intentionally, hardly by accident.

So, I'd rather avoid a breaking change here

const in this position doesn't prevent applications from defining and modifying non-const vtables; it just promises that libsodium won't modify the vtable via the application-provided pointer. I don't believe this is a breaking change for applications using non-const storage for this vtable.


comment created time in 2 months