Contribute to Open Source. Search issue labels to find the right project for you!

Debug on "alternate location"


there is a bug related to the “alternate location”, in which the previous code only chose the “ ” and “A”. However, some residues with “alternate locations” that contain “ ”, “B” and “C”, without “A”. Therefore, the previous code missing some atoms in the residues.

Updated 20/08/2017 06:55

NullPointerException in Lucene indexer


The following is repeatable and only occurs during multiple simultaneous Jenkins jobs. The issue is repeatable using either a shared data directory or a per-workspace directory.

When multiple simultaneous jobs are occurring, one or more of the jobs may fail due to one or more items in the ExceptionCollection containing the following trace:

java.lang.NullPointerException at at org.owasp.dependencycheck.analyzer.CPEAnalyzer.searchCPE( at org.owasp.dependencycheck.analyzer.CPEAnalyzer.determineCPE( at org.owasp.dependencycheck.analyzer.CPEAnalyzer.analyzeDependency( at org.owasp.dependencycheck.analyzer.AbstractAnalyzer.analyze( at at at at java.util.concurrent.ThreadPoolExecutor.runWorker( at java.util.concurrent.ThreadPoolExecutor$ at

Other times, the message will be NullPointerException, but the trace will be empty.

I’ve attempted to reproduce this issue using the command-line without success. Therefore it’s my theory that the issue is a thread/race/timing issue that only occurs when multiple instances of the Engine are running in the same JVM - as is the case with Jenkins.

It’s been a long time since I’ve tested this scenario, so I’m not exactly sure when the issue was introduced.

Updated 20/08/2017 06:51

Get homepage working.


Heads up, since the content of client folder has been changed, the homepage is no longer working. Just serves up a Error: ENOENT: no such file or directory, stat '/Users/shahzeb/Dev/Chattr/react-ui/build/index.html' error.

Note: The homepage needs to have a <twitter> login button which is functioning. That’s not backend related.

Updated 20/08/2017 06:52

fix(FEC-7049): YT playback fails when ad load fails and there's autoplay


When YoutUbe player is loaded it takes time until YT player API is loaded. If we have ad configured and it fails to load then the player tries to continue the playback on the player. Due to timing issues the YT player API might not be ready yet and then the call to getCurrentTime on it fails and it blows up the player. Test page: fixes #3494

Updated 20/08/2017 06:30 1 Comments

`cargo make` in workspace does not work when members are not explicitly listed


<!– markdownlint-disable first-header-h1 no-multiple-blanks first-line-h1 required-headers –>

Problem Description

cargo make panics when workspace members are not explicitly listed.

Code Example

I have the following crate structure:

  • workspace_test is a root workspace crate
  • member1 and member2 are members
├── Cargo.lock
├── Cargo.toml
├── member
│  ├── member1
│  └── member2
└── src

In workspace_test I have the following Cargo.toml:

name = "workspace_test"
version = "0.1.0"
authors = ["Azriel Hoh <>"]

member1 = { path = "./member/member1" }
member2 = { path = "./member/member2" }

# note: no members explicitly listed

This is a valid set up based on Cargo’s workspace section docs:

All path dependencies residing in the workspace directory become members.

Error Stack/Info (if any)

When I run cargo make I get:

[cargo-make] info - cargo-make 0.3.58
[cargo-make] info - Using Build File: Makefile.toml
[cargo-make] info - Task: default
[cargo-make] info - External file not found, skipping.
[cargo-make] info - Setting Up Env.
[cargo-make] info - Setting Up Env.
[cargo-make] info - Running Task: init
[cargo-make] info - Setting Up Env.
[cargo-make] info - Running Task: workspace
[cargo-make] info - Setting Up Env.
thread 'main' panicked at 'index out of bounds: the len is 0 but the index is 0', /Users/travis/build/rust-lang/rust/src/liballoc/
stack backtrace:
   0:        0x1060c0af3 - std::sys::imp::backtrace::tracing::imp::unwind_backtrace::hea646f639f5724a9
   1:        0x1060c441a - std::panicking::default_hook::{{closure}}::h85c76fcfe66fd3c4
   2:        0x1060c40d2 - std::panicking::default_hook::h3eec3c203ceba436
   3:        0x1060c6692 - std::panicking::rust_panic_with_hook::h70dc815ddbea04a3
   4:        0x1060c6574 - std::panicking::begin_panic::h964e463242cdda22
   5:        0x1060c6442 - std::panicking::begin_panic_fmt::h5836acd93f32a20b
   6:        0x1060c63aa - rust_begin_unwind
   7:        0x1060ef373 - core::panicking::panic_fmt::hf5ab42af92a0e0c7
   8:        0x1060ef2e8 - core::panicking::panic_bounds_check::h0d8311933e2799f3
   9:        0x105ff9d6b - cargo_make::command::run_script::hb0271043a70e11b9
  10:        0x105ffd6a7 - cargo_make::runner::run_task::h1e124d98163bfa20
  11:        0x105fffca5 - cargo_make::runner::run_flow::h4fa2ea28900b8575
  12:        0x106002c4a - cargo_make::cli::run_for_args::hb238ff9e91649f98
  13:        0x10600480f - cargo_make::main::heaf0a96bae02441f
  14:        0x1060c776c - __rust_maybe_catch_panic
  15:        0x1060c6ae8 - std::rt::lang_start::h97e5b1c9e293b92d

No line numbers, because it’s probably release mode, and also I’m on OS X which rustc doesn’t give line numbers anyway. However we can at least see the panic happens in

Btw, if I add a glob like so:

members = ["member/*"]

which is also a valid configuration:

The path dependency can also use globs to match multiple paths.

Running cargo make from the workspace root will run the default task on workspace_test and member, but not member2

I’d’ve started looking into why, but then I had to type this twice because I accidentally clicked on a link and lost the first post ._.. Btw awesome crate!

Updated 20/08/2017 06:35 2 Comments

Moving note metabox


Since 1.1.4 moving the metabox is not allowed as CMB2 will break. Let’s see if we can get this working by either updating cmb2 and hoping it’s fixed, patching CMB2 and updating it, or fixing it some other way e.g. using our own drag/drop something.

Updated 20/08/2017 06:12

Event names are not limited


<!— If you’re submitting a feature request, you can delete the template (These are comments, these wont actually show in the preview) –>

Platform information

  • Operating system: Windows 10/Raspbian Jessie
  • Python version: 3.6
  • Repository branch: rewrite <!– If you don’t know which branch you’re using, you’re probably using master –>

Bug information

<!– Short description of the bug –> When creating an event or editing it’s name, there seems to be no character limit at all, letting the user create events with really long names that might break their messages.

Steps to reproduce:: 1. Create an event with a really long name 2. Event is created

Expected result: The event creation should be denied, notifying the user the event’s name was too long.

Actual result: Event gets created

Updated 20/08/2017 05:48

[RE_BGRU_2ATT] entity location embedding issue


Current entity location embedding is the first character of the entity, the rest of the entity characters are calculated as part of the sentence.

An improvement is to rewrite the original sentence, and make all characters within the entity as one unit only so that only entity location is embedded, and characters within the entity will not be involved into embedding calculation.

Updated 20/08/2017 05:40

Can't drag links to tab bar to open them


Platform: Windows 10 Version: In other browsers I’m able to drag a link to the tab bar to open it in an existing tab or a new one, I’m unable to do this on Brave. I’ve tried it on multiple websites. Here’s an example of a site that this doesn’t work on:

Updated 20/08/2017 06:09

cl--assertion-failed on malformed build cache


I get a stack trace starting with:

Debugger entered--Lisp error: (cl-assertion-failed ((cl-every (function stringp) list) nil))
  cl--assertion-failed((cl-every (function stringp) list))

This should never happen, as the calls to cl-assert are wrapped in an ignore-errors. I don’t know what’s happening here.

Updated 20/08/2017 05:01

sccz80: silently pushing 32-bit constant when result of computation is 16-bit

#include <stdio.h>

#define BUFFER_SIZE     16384
#define MAINBANK_ADDR  (65536 - BUFFER_SIZE*2)

void main(void)
   printf("M RAMTOP no good (%u)", MAINBANK_ADDR);

zcc +zx -vn -a test.c

    ld  hl,i_1+0
    push    hl
    ld  hl,32768    ;const
    ld  de,0
    push    de
    push    hl
    ld  a,3
    call    printf
    pop bc
    pop bc
    pop bc

MAINBANK_ADDR is being pushed as a 32-bit value when the result is 16-bits and a 16-bit value is expected. This is happening because there is a 32-bit value “65536” in the expression used to calculate MAINBANK_ADDR. If the constant is changed to 65535, 16-bits is pushed.

zsdcc pushes 16-bits for the value if it is in range of 16-bits.

Updated 20/08/2017 04:38

Bug with custom default import name + component selector

import React from 'react';
import { create as render } from 'react-test-renderer';
import styledSc, { ThemeProvider } from 'styled-components';

const Ancestor = styledSc.div`
  color: red;

const Child = styledSc.div`
  background-color: ${p => p.theme.backgroundColor};
  font-size: ${p => p.fontSize}px;

const Parent = styledSc.div`
  border: 1px solid ${p => p.theme.borderColor};
  color: ${p => p.color};

  ${Child} {
    font-family: monospace;

const theme = {
  backgroundColor: 'cornflowerblue',
  borderColor: 'antiquewhite',

const themeVariant = {
  backgroundColor: 'mistyrose',
  borderColor: 'gainsboro',

const getVariant = useVariant => (
  <ThemeProvider theme={useVariant ? themeVariant : theme}>
      <Parent color={useVariant ? 'red' : 'black'}>
        <Child fontSize={useVariant ? 10 : 12} />

const instance = render(getVariant());

Results in the error "TypeError: Cannot call a class as a function".

Changing styledSc to styled makes the error go away.

Updated 20/08/2017 05:53 1 Comments

dir(FileDataset) and dir(DicomDir) return only Dataset methods



dir(FileDataset) and dir(DicomDir) don’t return their specific methods.

Steps/Code to Reproduce

from pydicom import read_file
>>> ds = read_file('DICOMDIR')
>>> type(ds)
<class 'pydicom.dicomdir.DicomDir'>
>>> hasattr(ds, 'parse_records')
>>> callable(ds.parse_records)
>>> 'parse_records' in dir(ds)

Occurs because Dataset.__dir__() only gets methods/properties from the Dataset class.

Updated 20/08/2017 04:02

Idiotic `print` implementation apparently nukes build cache


I found this in my build-cache.el today (the line-wrapping is not actually in the file; that’s just so you can read it on GitHub):

#s(hash-table size 145 test equal rehash-size 1.5 rehash-threshold 0.8 data ("s\
 traight" ("2017-08-16T12:42:51-0700" nil (:type git :host github :repo "raxod5\
 02/straight.el" :branch "develop" :files ("straight.el") :package "straight" .\
 ..)) "use-package" ("2017-08-16T12:40:00-0700" ("bind-key" "diminish") (:host \
 github :repo "raxod502/use-package" :upstream (:host github :repo "jwiegley/us\
 e-package") :package "use-package" :type git :local-repo "use-package")) "bind\
 -key" ("2017-08-16T12:40:00-0700" nil (:files ("bind-key.el") :package "bind-k\
 ey" :local-repo "use-package" :type git :repo "raxod502/use-package" :host git\
 hub ...)) "diminish" ("2017-08-16T12:40:00-0700" nil (:type git :host github :\
 repo "myrjola/diminish.el" :package "diminish" :local-repo "diminish.el")) "no\
 -littering" ("2017-08-16T12:40:00-0700" ("cl-lib") (:type git :host github :re\
 po "tarsius/no-littering" :package "no-littering" :local-repo "no-littering"))\
  "transpose-frame" ("2017-08-16T12:40:00-0700" nil (:type git :host github :re\
 po "emacsorphanage/transpose-frame" :package "transpose-frame" :local-repo "tr\
 anspose-frame")) "buffer-move" ("2017-08-16T12:40:00-0700" nil (:type git :hos\
 t github :repo "lukhas/buffer-move" :package "buffer-move" :local-repo "buffer\
 -move")) "iflipb" ("2017-08-16T12:40:00-0700" nil (:type git :host github :rep\
 o "jrosdahl/iflipb" :package "iflipb" :local-repo "iflipb")) "undo-tree" ("201\
 7-08-16T12:40:01-0700" nil (:type git :repo "\
 tree.git" :package "undo-tree" :local-repo "undo-tree")) "yasnippet" ("2017-08\
 -16T12:40:01-0700" ("cl-lib") (:type git :files ("yasnippet.el" "snippets") :h\
 ost github :repo "joaotavora/yasnippet" :package "yasnippet" :local-repo "yasn\
 ippet")) "el-patch" ("2017-08-16T12:40:01-0700" ("emacs") (:host github :repo \
 "raxod502/el-patch" :branch "develop" :package "el-patch" :type git :local-rep\
 o "el-patch")) "company" ("2017-08-16T12:40:02-0700" ("emacs") (:type git :hos\
 t github :repo "company-mode/company-mode" :package "company" :local-repo "com\
 pany-mode")) ...))

("0blayout" "popup" "ensime" "sbt-mode" "scala-mode" "prescient" "with-feature"\
 "ssh-config-mode" "shpec-mode" "pug-mode" "julia-mode" "ess" ...)

Sigh. We’ll have to use a different function that doesn’t abbreviate using ....

Updated 20/08/2017 03:57 2 Comments

Checkbox state is rendered incorrectly


The pattern used for data persistence across form submission attempts doesn’t really work for checkboxes. Specifically, checkboxes will always be set to the default value passed into the renderer.checkbox() call, regardless of their state in the form.

Updated 20/08/2017 05:20 1 Comments

Make the website work on mobile, or disable it on mobile


I’m not sure about the best way to layout this page for mobile. For the moment, it is only designed to work on desktop, where I think it will be most useful.

I think that the options are: 1. Make the website keep its desktop appearance, and have the user zoom and scroll (low effort). 2. Use bootstrap (already used) and make the website more mobile friendly (lots of work).

I think that option 1 will work best. Websites with notifications that run in the background (e.g., when a user switches to another app) will probably be unreliable on mobile.

The better way to make it run on mobile would be to make an app that can properly run in the background and present notifications in a meaningful way.

Updated 20/08/2017 05:33

Investigate HipChat integration test failures


HipChat integration tests appear to be failing in Travis CI. The following is representative of the failing output.

     ** (MatchError) no match of right hand side value: {:error, {%RuntimeError{message: "Error: Timeout waiting for bot to join room:"}, [{Cog.Test.Support.HipChatClient, :wait_for_join, 1, [file: 'test/support/hipchat_client.ex', line: 72]}, {Cog.Test.Support.HipChatClient, :"-init/1-fun-0-", 3, [file: 'test/support/hipchat_client.ex', line: 55]}, {Enum, :"-each/2-lists^foreach/1-0-", 2, [file: 'lib/enum.ex', line: 651]}, {Enum, :each, 2, [file: 'lib/enum.ex', line: 651]}, {Cog.Test.Support.HipChatClient, :init, 1, [file: 'test/support/hipchat_client.ex', line: 51]}, {:gen_server, :init_it, 6, [file: 'gen_server.erl', line: 328]}, {:proc_lib, :init_p_do_apply, 3, [file: 'proc_lib.erl', line: 247]}]}}
       test/integration/hipchat_test.exs:18: Integration.HipChatTest.__ex_unit_setup_2/1
       test/integration/hipchat_test.exs:1: Integration.HipChatTest.__ex_unit__/2
     The following output was logged:
     2017-08-20T02:52:54.693 module=Romeo.Roster line=16 [info]  Getting roster items
     2017-08-20T02:52:54.806 module=Romeo.Transports.TCP line=30 [info]  Established connection to
     2017-08-20T02:52:54.866 module=Romeo.Transports.TCP line=95 [info]  Negotiating secure connection
     2017-08-20T02:52:54.928 module=Romeo.Transports.TCP line=100 [info]  Connection successfully secured
     2017-08-20T02:52:54.957 module=Romeo.Auth line=47 [info]  Authenticating with PLAIN
     2017-08-20T02:52:55.098 module=Romeo.Auth line=81 [info]  Authenticated successfully
     2017-08-20T02:52:55.161 module=Romeo.Transports.TCP line=133 [info]  Bound to resource: none||proxy||5252
     2017-08-20T02:52:55.190 module=Romeo.Transports.TCP line=149 [info]  Session established
     2017-08-20T02:52:55.232 module=Romeo.Roster line=16 [info]  Getting roster items
Updated 20/08/2017 03:03

js npe error when using nativeScroller:true and expanding grid row


Hi there, I set nativeScroller:true in my grid because the scrolling is much better. However now I get a js error whenever I doubleclick a row to expand it. There is no error when I set it to false so I believe its related to this. Please see the attached debugger screenshot for all the details. Thx & regards!

Btw: Happens in 1.6.3 and 1.6.5 also

bildschirmfoto 2017-08-20 um 03 57 22

Updated 20/08/2017 06:59 1 Comments

server crash at line validateJwt(req, res, next) in server/auth/auth.service.js:25:7


Below is the stack trace . Issue is with server/auth/auth.service.js:25:7.

Will hit the issue sometime during the code change and server restart

UnauthorizedError: No authorization token was found at middleware (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express-jwt/lib/index.js:80:21) at Middleware_Common_Object.<anonymous> (/Users/ragmurug/ws/raghucode/fullstack/meanshop/server/auth/auth.service.js:25:7) at next (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/composable-middleware/lib/composable-middleware.js:59:18) at Middleware_Common_Object.middleware (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/composable-middleware/lib/composable-middleware.js:76:7) at middleware (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/composable-middleware/lib/composable-middleware.js:31:25) at Layer.handle [as handle_request] (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/layer.js:82:5) at next (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/route.js:100:13) at Route.dispatch (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/route.js:81:3) at Layer.handle [as handle_request] (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/layer.js:82:5) at /Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:234:24 at Function.proto.process_params (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:312:12) at /Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:228:12 at Function.match_layer (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:295:3) at next (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:189:10) at /Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:214:18 at Function.match_layer (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:295:3) at next (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:189:10) at /Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:191:16 at Function.match_layer (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:295:3) at next (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:189:10) at Function.proto.handle (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:165:3) at router (/Users/ragmurug/ws/raghucode/fullstack/meanshop/node_modules/express/lib/router/index.js:34:12) GET /api/users/me 401 15ms

Updated 20/08/2017 01:20

Issue selecting value sin on boarding flow


Steps to reproduce:

  1. Go to recruiter portal
  2. Click “Ingresar”
  3. Sing up with a new user
  4. Click “Continuar”
  5. user clicks on “Cuantos empleados tiene su empresa” select box and the options are displayed <img width=“537” alt=“screenshot 2017-08-19 20 04 51” src=“”> 6.User select one of the options but the select box gets a red border keeping the value “seleccionar” <img width="544” alt=“screenshot 2017-08-19 20 04 55” src=“”>
  6. User select a an option again <img width=“541” alt=“screenshot 2017-08-19 20 04 59” src=“”>
  7. The select box is final update with the selected value <img width=“537” alt=“screenshot 2017-08-19 20 05 04” src=“”>

Environment: development Browser: chrome Os: Mac

Updated 20/08/2017 01:17

Polyline/Polygon issues with the dateline


Loading a data set containing borders of conus. I get this when zoomed out image

I’ve seen this issue come up before when polygons cross the dateline, but never with conus.

Video scrolling when soon with some strange behaviors.

Updated 20/08/2017 01:39 2 Comments

Audio Streamer Stacktrace

[ERROR] Error in playback of eTku5QDpbak Something broke when playing the track.
        at com.sedmelluq.discord.lavaplayer.track.playback.LocalAudioTrackExecutor.execute(
        at com.sedmelluq.discord.lavaplayer.player.DefaultAudioPlayerManager.lambda$executeTrack$1(
        at java.util.concurrent.ThreadPoolExecutor.runWorker(
        at java.util.concurrent.ThreadPoolExecutor$
Caused by: java.lang.RuntimeException: org.apache.http.conn.HttpHostConnectException: Connect to [,] failed: Network is unreachable (connect failed)
        at com.sedmelluq.discord.lavaplayer.container.matroska.MatroskaAudioTrack.loadMatroskaFile(
        at com.sedmelluq.discord.lavaplayer.container.matroska.MatroskaAudioTrack.process(
        at com.sedmelluq.discord.lavaplayer.track.DelegatedAudioTrack.processDelegate(
        at com.sedmelluq.discord.lavaplayer.track.playback.LocalAudioTrackExecutor.execute(
        ... 4 common frames omitted
Caused by: org.apache.http.conn.HttpHostConnectException: Connect to [,] failed: Network is unreachable (connect failed)
        at org.apache.http.impl.conn.DefaultHttpClientConnectionOperator.connect(
        at org.apache.http.impl.conn.PoolingHttpClientConnectionManager.connect(
        at org.apache.http.impl.execchain.MainClientExec.establishRoute(
        at org.apache.http.impl.execchain.MainClientExec.execute(
        at org.apache.http.impl.execchain.ProtocolExec.execute(
        at org.apache.http.impl.execchain.RetryExec.execute(
        at org.apache.http.impl.execchain.RedirectExec.execute(
        at org.apache.http.impl.client.InternalHttpClient.doExecute(
        at org.apache.http.impl.client.CloseableHttpClient.execute(
        at com.sedmelluq.discord.lavaplayer.container.matroska.format.MatroskaEbmlReader.readEbmlInteger(
        at com.sedmelluq.discord.lavaplayer.container.matroska.format.MatroskaFileReader.readNextElement(
        at com.sedmelluq.discord.lavaplayer.container.matroska.MatroskaStreamingFile.readFile(
        at com.sedmelluq.discord.lavaplayer.container.matroska.MatroskaAudioTrack.loadMatroskaFile(
        ... 9 common frames omitted
Caused by: Network is unreachable (connect failed)
        at Method)
        at org.apache.http.conn.ssl.SSLConnectionSocketFactory.connectSocket(
        at org.apache.http.impl.conn.DefaultHttpClientConnectionOperator.connect(
        ... 27 common frames omitted
Updated 20/08/2017 01:03

Problema ao adicionar lançamento em grupo


Por alguma razão desconhecida não se está adicionando os lançamentos da conta nos grupos de lançamentos. Na correção anterior o problema aparentemente havia sido resolvido. Ao que parece tem alguma questão não resolvida que está causando o problema ao adicionar.

Updated 20/08/2017 00:57 1 Comments

Verify from example code why some test vectors are not aligned


Some test vectors proposed by the RFC document are not producing the same result in tests. This seems to be related to the numeric representation used. The RFC document report a PoC Java implementation that can be used to verify the question content and why it’s different by the golang produced one.

Updated 20/08/2017 00:25

Missing pop after unused values.


Since the existence of CodeAPI, generated bytecode does have pop after unused expression values. This leads to an uncommon bug with control flow. Example:

if (a) run() else run2()

Assuming that run returns void and run2 returns Object (but occurs with any value type), the generated bytecode will be incorrect, because when the flow reach the end, a value of type Object may or may not be present, so the Frame types cannot be calculated correctly and even if it could be, JVM will fail to load the class.

CodeAPI-BytecodeWriter should track carefully when expression is used as a value or not, so the bytecode will be only incorrect if the user was made a mistake. The current workaround is to use BytecodeWriter Pop CodeInstruction after run2 invocation:

if (a) run() else run2(); pop

Or to call another method on if body which returns Object, so frame types are calculated correctly:

if (a) runWithReturn() else run2(); pop

Assuming that both runWithReturn and run2 returns either an Object instance or values of same primitive type.

Updated 20/08/2017 00:13

Hotfix for issue with stories panel


Issue: once you hide the stories (“left”) panel, you can’t open it again. In fact, it doesn’t even hide: <img width=“69” alt=“screen shot 2017-08-20 at 02 48 03” src=“”> Related: #1614

Unfortunately, I overlooked those lines when reviewing #1598. Unfortunately, I can’t figure out where this error get caught so that the failure is silent.

@igor-dv, I think it’s better to use an additional state field instead of input ref there

Updated 20/08/2017 05:12 1 Comments

Setting time in past near-instantly removes raid


Every minute the system checks for old raids to clean up, so if someone makes a mistake on setting the time, it’s going to cause a lot of headaches for everyone.

We definitely should put some smarts into it, and make sure times are always set in the future, perhaps even restricting to a minimum of 10 minutes in the future to give people time to arrive?

Updated 20/08/2017 01:11 4 Comments

Fork me on GitHub