forceValue() were called after a value is copied effectively forcing only one of the copies keeping another copy not evaluated.
This resulted in its evaluation of the same lazy value more than once (the number of hits is not big though)
--
298d93fcb860116111611a8ab0662b409734227a by Alex Strelnikov <strel@google.com>:
Release ascii_benchmark.
PiperOrigin-RevId: 205275981
--
73a01469e5862eefbe5ef9d434f45b7073476272 by Abseil Team <absl-team@google.com>:
Internal cleanup
PiperOrigin-RevId: 205236717
--
53d6338bf49dab95bd0fbb5bbcd56970c6b868c2 by Abseil Team <absl-team@google.com>:
Removes InlinedVector's dependency on bitwise operators for size_type
PiperOrigin-RevId: 205236134
GitOrigin-RevId: 298d93fcb860116111611a8ab0662b409734227a
Change-Id: I754f5eea889567add2dbbdea358a533f54a912dd
I wasn't using the previous KBDs, which were modelled after Vim's answer
for Paredit KBDs.
Some of the mappings bothered me, so I needed that made more sense to
me.
Accidentally commited the version of this configuration that has this
variable set to false.
Since most of the time, this variable should be true, commiting the true
version will clean up my git status output.
Useful since the new configuration branches according to a boolean that
encodes whether or not the Ergodox is connected.
Since I'm frequently switching between programming at my desk to
programming AFK, I need a faster way to update my Slate configuration.
Ideally - some script would watch the USB port for whether or not the
Ergo was connected. Until then...
After writing this, I have a sneaking suspiscion that I'm reinventing
something like Tmuxinator. I guess this is just a poor man's Tmuxinator.
For now, however, haven't had enough time to look into Tmuxinator as a
viable alternative, so this will have to do for now.
Furthermore, moved root-level tmux files to a subdir.
Need to learn more about `evil-goto-definition` because it is possible
and even likely that there is a known, friendlier alternative way to
dispatch the goto-definition using Evil's infrastructure. Needed a
hotfix, however, so this will have to suffice for now.
After my hard drive fried, I lost important configuration settings.
Since then, I haven't been able to use `magit-gh-pulls` without it
incessantly prompting me for my GH username and password. Until I
remember how I previously configured this, I'm disabling it.
clo was difficult to use because `company-mode` suggested autocompletion
candidates like "close", "clone", etc. With clg, this is significantly
less likely.
Finally ported my up-to-date emacs configuration here. I was putting
this off for a long while, unsure of how to handle all of the work. All
it took was my laptop being fried to force me to do this. So... voila!
Not ready for this yet, causes the prompt to disappear in nix repl
and more generally can overwrite non-progress-bar messages.
This reverts commit 44de71a396.
--
60bc1e62580e0ff352a92c785f29550c2000447d by Xiaoyi Zhang <zhangxy@google.com>:
Import Github PR https://github.com/abseil/abseil-cpp/pull/143.
absl/synchronization/internal/kernel_timeout.h uses INFINITE macro that comes from windows.h that is included by winsock2.h that is included by absl/time/time.h. This internal header will be included by public header. It should not depend on windows.h.
PiperOrigin-RevId: 205109009
--
1617f0a333a8030e4e4c0bc1eef71f4a5fe9874d by Abseil Team <absl-team@google.com>:
Internal change
PiperOrigin-RevId: 205101804
--
8dce298b8c96c12c423943a366a4d92a554366c4 by Chris Kennelly <ckennelly@google.com>:
Define UNALIGNED_LOAD/STORE macros for UNDEFINED_BEHAVIOR_SANITIZER.
When using UBSan in trap mode (that is, without the ubsan runtime library), the
x86 macros can cause alignment errors as they assume unaligned loads/stores are
permitted. The macros defined in the presence of
{ADDRESS,THREAD,MEMORY}_SANITIZER require the runtime library.
PiperOrigin-RevId: 205096794
GitOrigin-RevId: 60bc1e62580e0ff352a92c785f29550c2000447d
Change-Id: I65a6cc86a711796c9d3a605310d67795b9f76ce9
