profile
viewpoint
Justin Spahr-Summers jspahrsummers London, United Kingdom https://jspahrsummers.com

joshaber/Few.swift 1073

Views as functions of their state.

ashfurrow/FunctionalReactivePixels 710

A demonstration of how to use FRP with ReactiveCocoa in an iOS context using the 500px API.

jspahrsummers/enemy-of-the-state 209

My talk explaining what state is and why it's so harmful

jspahrsummers/Crust 204

Reimagining Mantle in Swift (experimental)

ArbiterLib/Arbiter 113

Dependency manager library that supports decentralization

jspahrsummers/adt 105

Algebraic data types for Python

jspahrsummers/Clairvoyant 61

Append-only key-value database

jspahrsummers/carthage-talk 37

Ruthlessly Simple Dependency Management with Carthage

bankroll-py/bankroll 36

Ingest portfolio and other data from multiple brokerages, and analyze it

jspahrsummers/documentalist 27

A multi-language, extensible documentation generator

startedendless-sky/endless-sky

started time in 2 months

pull request commentwebbertakken/unity-builder

Rename custom -version flag to -customProjectVersion

Sure, makes sense, thanks! Now renamed to -buildVersion.

jspahrsummers

comment created time in 2 months

push eventjspahrsummers/unity-builder

Justin Spahr-Summers

commit sha ddb8b949a2b54c1af8b92bcd6be5e617693b801c

Rename flag to -buildVersion

view details

push time in 2 months

issue commentwebbertakken/unity-builder

2019.4.8f1 builds are empty, but reported as success

Ah, I understand now that -version is a custom flag used by this code, and not something that was baked into Unity (until now, when it's causing problems!).

I've submitted a possible fix as #137.

jspahrsummers

comment created time in 2 months

PR opened webbertakken/unity-builder

Rename custom -version flag to -customProjectVersion

-version is now used (as of 2019.4.8f1 or earlier) by the Unity CLI to print its own version, then exit with code 0. This interferes with unity-builder's version passing, so I propose renaming the argument to something less likely to collide.

I was unable to run Yarn as requested in the contributing guidelines, as I'm missing tons of prerequisites. Besides whatever CI checks this repo can provide, I can confirm that this fixed #136 in my game's workflow.

+2 -2

0 comment

2 changed files

pr created time in 2 months

create barnchjspahrsummers/unity-builder

branch : rename-version-flag

created branch time in 2 months

push eventjspahrsummers/unity-builder

Justin Spahr-Summers

commit sha 3c52f96d9c3f544381481c2ac7c164bd2f51cc2a

Revert "HACK workaround for webbertakken/unity-builder#136" This reverts commit 4483025abd59c222b97ab1473ef7184edfb79c17.

view details

Justin Spahr-Summers

commit sha 1491168c8c072050745f066a5d349c4739f07556

Rename custom -version flag to -customProjectVersion

view details

push time in 2 months

push eventjspahrsummers/unity-builder

Justin Spahr-Summers

commit sha 4483025abd59c222b97ab1473ef7184edfb79c17

HACK workaround for webbertakken/unity-builder#136 The -version flag is causing issues, so skip it for now. This probably messes up project versioning, but it should make builds work again.

view details

push time in 2 months

issue openedwebbertakken/unity-builder

2019.4.8f1 builds are empty, but reported as success

After upgrading my Unity version (with no other changes to my use of unity-actions), now build output looks like this:

###########################
#    Building platform    #
###########################

2019.4.8f1 (60781d942082)
Build succeeded

###########################
#     Build directory     #
###########################

total 8.0K
drwxr-xr-x 2 root root 4.0K Aug 15 13:25 .
drwxr-xr-x 3 root root 4.0K Aug 15 13:25 ..

I encountered this by using unityVersion: 2019.4.8f1, but you can reproduce the underlying issue more directly by trying the -version flag:

$ docker pull gableroux/unity3d:2019.4.8f1-linux-il2cpp
$ docker run -it --rm gableroux/unity3d:2019.4.8f1-linux-il2cpp
# /opt/Unity/Editor/Unity -batchmode -version "v1.0" -logfile /dev/stdout
2019.4.8f1 (60781d942082)

It appears that -version is interpreted by the Unity CLI to print its own version, rather than provide a version for the build.

When it's not provided, Unity continues to start up as normal:

# /opt/Unity/Editor/Unity -batchmode -logfile /dev/stdout
[LicensingClient] ERROR Failed to connect to local IPC
[Licensing::Module] Failed to connect to channel: LicenseClient-root
[Licensing::Module] Successfully launched the LicensingClient
[SignatureVerifier] Application signature verification not supported on this platform.
[Licensing::Module] Successfully connected to LicensingClient on channel: LicenseClient-root
[Licensing::Module] Successfully launched and connected to LicensingClient
Entitlement-based licensing initiated
[LicensingClient] Licenses Updated successfully in LicensingClient
Error opening default X display
(Filename: ./Editor/Platform/Linux/X11Quarantine.cpp Line: 33)

(This failure is expected, as I didn't set up an X server for this test.)

I couldn't find any mention of a -version flag even in the 2019.3 docs, so I'm not sure what the best solution is here.

created time in 2 months

startedbadges/shields

started time in 2 months

startedrjstone/discord-webhook-notify

started time in 3 months

issue commentjspahrsummers/adt

Function "bind" on Result adt doesn't typecheck

Does it work if you try to write something like this without the @adt sugar? The only thing that I can think of at this point is that there is a genuine mypy bug or limitation.

ianliu

comment created time in 3 months

issue closedantequant/ib-gateway-docker

How to change the restart time?

Thank you for developing this container. My question is not an issue of the container. I was actually wondering how I can change the restart time of ibgateway?

closed time in 3 months

afalahat

issue commentantequant/ib-gateway-docker

How to change the restart time?

Sorry, I don't know. This project is just a simple wrapper.

afalahat

comment created time in 3 months

issue commentjspahrsummers/adt

Function "bind" on Result adt doesn't typecheck

I haven't had a chance to try this sample out myself, but my guess is that the typechecker is failing to unify the types of ok and err. If you try the same code, but with lambdas wrapping fun and Result.ERR, does it work?

ianliu

comment created time in 3 months

startedopenai/spinningup

started time in 3 months

starteddavid-cortes/contextualbandits

started time in 3 months

more