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

Installing chromium in a Fedora-24-based DVM Template hijacks unrelated MIME types


Qubes OS version (e.g., R3.2):


Affected TemplateVMs (e.g., fedora-23, if applicable):


Expected behavior:

Files opened in DispVMs open in appropriate programs.

Actual behavior:

Files opened in DispVMs open in Chromium, even when Chromium cannot read them.

Steps to reproduce the behavior:

  1. Use a plain fedora-24-based DVM Template.
  2. Open, e.g., an “office” file (e.g., ODS, ODF, DOCX, XLS, etc.) file in a new DispVM. (Note: Trigger the opening from a non-DispVM in a DispVM. Do not qvm-copy the file then manually open it from the DispVM.)
  3. The file opens as expected in some appropriate “office suite” program, if installed.

By contrast:

  1. Install Chromium in the same DVM Template.
  2. Try to open the same file again in a new DispVM (after regenerating the DVM Template).
  3. The file opens in Chromium, but since Chromium can’t read it, it just triggers a useless and endless “download file” loop.

General notes:

Opening the file from, e.g., Nautilus inside the DispVM works as expected.

Updated 30/04/2017 16:47

Fix wrong operator in LinearAllocator


The linear allocator currently reallocates if the requested size i >= to MAX_SIZE, while it should only really need to reallocate if it is > MAX_SIZE. Make sure to run some tests to ensure that this still have correct behavior.

``` if (this->firstFree->used + size >= MAX_SIZE) { // Might be that we are reusing a block here, which can happen after we have done a reset. auto newBlock = (this->firstFree->next != nullptr) ? this->firstFree->next : new Block(); this->firstFree->next = newBlock; this->firstFree = newBlock; }

should be probably be: if (this->firstFree->used + size > MAX_SIZE) { // Might be that we are reusing a block here, which can happen after we have done a reset. auto newBlock = (this->firstFree->next != nullptr) ? this->firstFree->next : new Block(); this->firstFree->next = newBlock; this->firstFree = newBlock; }

Updated 30/04/2017 16:44

qubes-firewall service frequently fails


Qubes OS version (e.g., R3.2):



Network access is frequently lost


Manually restart qubes-firewall service in sys-firewall:


The status is qubes-firewall does not always say failed, even when the problem occurs.

systemctl restart qubes-firewall

Older discussion thread:

Updated 30/04/2017 16:40

Known Issues


The Extractor is in its alpha stage DO NOT attempt to use it for scholarly work. The closing of this issue would mark the beta stage. It is by then that you can attempt to use it for scholarly work.

  • No support for lexical texts for now. The documentation is literally absent on how they are encoded, and they differ considerably from the c-atf syntax.

  • Some determinatives have quirks like {d}en, this should have been conceived as {d}-en, but since determinatives are handled at a word level rather than sign level, sign separator method, doesn’t handle well signs that are not separated with -. This would be dealt with in very near future.

  • Signs and words that belong to Another Language Occurrence or determinative is not marked in the sign and word dictionaries. This will be done in very near future too.

  • Words and determinatives are handled at a line level. Whereas some words are most probably dispersed to multiple lines. They will be handled appropriately in very near future.

  • Very rarely in achaemenid royal inscriptions some complex signs, like ANSZE are dispersed into multiple lines. If they are presented as a specified compound sign, something like ANSZE(|AN.\newline SZE|), we will support it in the future, if they are not, if there is no documentation on it, we won’t be supporting it.

Updated 30/04/2017 16:48

Applications that require training never marked "completed"


Applications for facilities that require training currently only get as far as the application being accepted. This means that they are unable to start a new application because an accepted application is still classed as an open application.

Applications that require training and are marked as “accepted” need to have the functionality to mark as “training completed”, and therefore mark the application as complete.

Updated 30/04/2017 16:38

Novice at pass can be skipped via teleport


Quest “Hilfe für Dragomir”: Am Pass im Minental wartet ein Novize am Pass im Minental auf den Spieler. Es startet ein Dialog, dass der Novize den Spieler angegriffen hätte - das ist allerdings nicht passiert. Weiteres Problem an der Quest: man verpasst den Novizen eventuell, wenn man sich direkt nach Khorinis teleportiert (was nicht unwahrscheinlich ist)

Updated 30/04/2017 16:27

Logical error in swamp camp


Quest “Dem Mörder auf der Spur”: Geht man ins Sumpflager, um mehr über die magischen Zettel zu erfahren, wird man am Eingang von der Wache aufgehalten. Vorher konnte man sich allerdings unbehelligt frei im Lager bewegen. Entweder, das wird komplett entfernt, oder es wird eine andere Begründung gewählt, warum man alles abgeben muss

Updated 30/04/2017 16:26

`Env:-ar` and `:-kr` incorrectly use `mul` and `add` parameters


(Taken directly from @jamshark70’s comment in #2855.)

I just noticed that Env:ar and :kr don’t use mul and add correctly.

Everywhere else, mul and add can be modulated at any time.

These methods map mul and add onto levelScale and levelBias – meaning that changes to the control inputs are ignored until the next envelope segment begins. mul and levelScale are related but not interchangeable.

Perhaps change them like this?

    kr { arg doneAction = 0, gate = 1.0, timeScale = 1.0, mul = 1.0, add = 0.0;
        ^, gate, 1, 1, timeScale, doneAction).madd(mul, add)

    ar { arg doneAction = 0, gate = 1.0, timeScale = 1.0, mul = 1.0, add = 0.0;
        ^, gate, 1, 1, timeScale, doneAction).madd(mul, add)

I apologize that I hadn’t noticed before… But, if the help is going to start recommending Env(something).kr then I think mul and add should behave as they do with other UGens.

I would rather have it go the other way, and rename mul and add to levelScale and levelBias, because that is (1) more transparent (doesn’t also add a call to madd), (2) not a behavior change, and (3) doesn’t cut these functions off from access to levelScale and levelBias, which is probably the desired effect if the purpose of these calls is to forward arguments to EnvGen.

Updated 30/04/2017 16:49 1 Comments

Windows: Omnisharp plugin requires administrator permissions to download



In distribution builds, on Windows, the Omnisharp (C#) language service requires administrator privilege to download the windows client. The install location on my machine, for example, is here: “Downloading C:\Program Files (x86)\Oni\resources\app\node_modules\omnisharp-client\omnisharp-win-x64-net46\”}

Therefore, downloading this in a non-administrator context will result in the download failing.

Potential Fixes: - OPTION 1: See if there is a way to override the download / binary location for this. Potentially in the cache / userData folder of electron, or some of other writable place. - OPTION 2: Copy the default or select plugins to the cache / userData folder on startup. - OPTION 3: Package the downloaded file during windows install. This means the installation size is even larger, which would not be ideal.

Updated 30/04/2017 16:21

Improve flight of hunters during Ramsi quest


Quest “Jagd mit Ramsi”: Wenn der Jäger flüchtet, läuft er direkt in eine Schattenläuferhöhle, wo sein bereits getöteter Kumpel auf ihn wartet - um dann da vom Schattenläufer auch noch niedergestreckt zu werden

Updated 30/04/2017 16:20

Quest wie Mattheus doesn't work properly

  • Quest “Beschattung der Feuermagier”: Mattheus greift mich in Orlans Kneipe aus einem mir nicht bekannten Grund an. Schafft man es mit ein paar Tricks, doch mit ihm zu sprechen, kann man ihm ein Bier ausgeben. An diesem Bier stirbt er dann (warum auch immer) und man kann ihn looten, wird dabei jedoch von Orlan als Dieb beschimpft und angegriffen
Updated 30/04/2017 16:18

Too less empty bottles


es gibt [B]VIEL[/B] zu wenig Laborwasserflaschen zu kaufen sowie fast keine alchemistischen Ingredienzien. Beides sollte mindestens für einen Lehrling der Alchemie ausreichend vorhanden sein (spricht: unendlich Nachschub über Handel mit Constantino)

Updated 30/04/2017 16:17

Add more options to 13x6


Quest “13 mal die 6”: Alax nimmt nur eine ganz bestimmte Auswahl an Kräutern (Heilkräuter, Manakräuter, Zwillingsdorn); die meisten Sachen (Pilze, Beeren, Rabenkraut, Snapperkraut, Mondauge, Sumpfkraut, Blaufieder, Sumpfbeeren, etc etc) nimmt er nicht an, was die Quest unnötig verkompliziert

Updated 30/04/2017 16:16

parinfer--prepare: Invalid function: parinfer-save-excursion


After installing parinfer 0.4.10 from MELPA Stable in emacs 24.5 I cannot work on clojure files anymore. Cursor jumps to random place on backspace and in Messages I see lines: parinfer--prepare: Invalid function: parinfer-save-excursion.

It looks like version 0.4.9 works correctly. Previously I used version 0.4.5 which also worked for me.

Updated 30/04/2017 16:34 1 Comments

Header clipping issue in smaller screens


Select one … (check one with “x”) [x] bug [ ] feature request [ ] enhancement The header, with top-level navigation is clipped when the page is resized to mobile/small screens even though there’s still enough space to fit all the elements.

Small screen: image

Desktop/full size: image

Environment details

  • Angular version: 2.0.X

  • Clarity version:

  • OS and version:

  • Browser: [all | Chrome XX | Firefox XX | IE XX | Safari XX | Mobile Chrome XX | Android X.X Web Browser | iOS XX Safari | iOS XX UIWebView | iOS XX WKWebView ]

Updated 30/04/2017 16:37

ERP test is broken


When I try to create an ERP test for CESM with the latest master I get the following error in the build:

2017-04-29 17:13:15: Exception during build:
  SMP_BUILD = c1a1l1i1o1r1g1w1e1
  SMP_VALUE = c1a1l1i1o1r1g1w1e1
  smpstr = c0a0l0i0o0r0g0w0e0
  A manual clean of your obj directories is strongly recommended
  You should execute the following:
    ./ --clean
  Then rerun the build script interactively
  ---- OR ----
  You can override this error message at your own risk by executing:
    ./xmlchange -file env_build.xml -id SMP_BUILD -val 0
  Then rerun the build script interactively

I have narrowed down the problem to the following hash:

`670cf97 Merge remote-tracking branch ‘origin/master’ into jgfouca/new_query_tools

I believe the problem is related to the fact that when we are halving the threads - (which in the case of yellowstone goes from 2->1) the build-threaded option should NOT change. That was always the intent of this test. You want to keep the threading compilation the same.

@jedwards noticed yesterday that this test was working on cheyenne - but this is a red herring since on cheyenne the threads are set to 1 in the initial pe-layout and so halving them keeps the threads at 1. On yellowstone the threads start at 2- so halving them gives the bad behavior we are seeing.

This is now a critical issue in our testing for the CESM2 release.

Updated 30/04/2017 16:43 1 Comments

Class loading by name is not supported for multidimensional arrays


Status Quo

Due to the investigative work performed in conjunction with #815, I noticed that the JUnit Platform currently has no support for loading classes by name for multidimensional arrays.

Consequently, it is currently impossible to create a MethodSelector (e.g., via DiscoverySelectors) which selects a method by fully qualified method name if the method accepts a multidimensional array as an argument.

A soon-to-be common use case for such a method selection would be a @ParameterizedTest that accepts int[][][] or String[][] as an argument.


As an intermediary step in #815, I added support for 2D arrays for primitives in ReflectionUtils. That solution is fine in terms of a hard-coded solution for common use cases; however, we need a generic solution to support multidimensional primitive arrays and object arrays.

Related Issues

  • 810

  • 815


  • [ ] Support class loading by name for multidimensional primitive arrays.
  • [ ] Support class loading by name for multidimensional object arrays.
  • [ ] Resolve all TODOs in ReflectionUtilsTests and DiscoverySelectorsTests.
Updated 30/04/2017 16:14

Building webpack in production mode will accidentally exit without finishing.

let pack = exec('yarn run pack') # yarn run pack:main && yarn run pack:renderer

pack.stdout.on('data', data => console.log(data))
pack.stderr.on('data', data => console.error(data))
pack.on('exit', code => build())

When we hit pack.on('exit', code => build()), webpack should finish yarn run pack:main && yarn run pack:renderer as well, however, process of yarn run pack:renderer seems to be interupted somewhere.

RIght now, the solution to this problem is to run yarn run pack manually before yarn run build.

Updated 30/04/2017 15:59

Wrap long button text


Issue Links with the .btn class have the text cut off if the text is longer than the width of the button. This is caused by the .btn class having the style white-space: nowrap. screen shot 2017-04-30 at 10 48 14 am

Proposed solution In the basbls theme, add white-space: wrap to the .btn class. screen shot 2017-04-30 at 10 48 47 am

Updated 30/04/2017 15:55

Take out remaining loklak references


Remaining references:

app.json: “logo”: “”, app.json: “website”: “”, app.json: “repository”: “”, app.json: “image”: “mariobehling/loklak”, app.json: “BUILDPACK_URL”: “” docker-cloud.yml: name: “loklak” docker-cloud.yml: image: mariobehling/loklak docker-compose.yml: name: “loklak” Dockerfile:# hack until loklak support no-daemon Dockerfile:# start loklak manifest.yml: buildpack: “” manifest.yml: image: “mariobehling/loklak” manifest.yml: logo: “” manifest.yml: repository: “” manifest.yml: website: “” Pulls join our mailing list to discuss questions regarding the project:!forum/loklak, this is the tool for that! Just put into your rss reader. Oh wait.. you will get a lot of information about tasty Cambodian food with that as well. Alternatively you may also read the authors timeline using or just follow @0rb1t3r (that’s a zero after the at sign) and tutorials are also on our blog at README.rst:!forum/loklak README.rst: into your rss reader. Oh README.rst: or just follow @0rb1t3r README.rst:Articles and tutorials are also on our blog at README.rst:.. |Docker Pulls| image:: README.rst: :target: scalingo.json: “logo”: “”, scalingo.json: “website”: “”, scalingo.json: “repository”: “”, scalingo.json: “image”: “mariobehling/loklak”, scalingo.json: “value”: “”

Updated 30/04/2017 15:55

Topics: Incorrect "Understanding the Measure" headings


There are incorrect headings (“Youth fruit and vegetable consumption - Weighted Number” and “Youth fruit and vegetable consumption - Weighted Percent”) listed under the “Understanding the measure” section for the following indicators:

  • Bullying
  • School fights
  • School safety
  • Youth condom use
  • Youth depression

<img width=“938” alt=“screen shot 2017-04-30 at 10 46 47 am” src=“”>

Updated 30/04/2017 15:47



切换桌面时发现菜单栏的歌词不显示了,切回到之前的桌面也不会显示,只有关闭菜单栏显示功能再重新打开才可以。 后来发现好像是有chrome的桌面,然后滚动下chrome的页面,歌词就不显示了。

Updated 30/04/2017 16:41

Fix campaign encoding


Preserve entities in repairString method of Tidy

$body = $tidy->repairString($html, array('output-xhtml' => true, 'preserve-entities' => 1), 'utf8');

Updated 30/04/2017 15:39

Footer: Use local license button and provide correct link

  • [ ] Please use the small Creative Commons buttons instead of large buttons.
  • [ ] Store buttons on the itself. Do not embed external buttons.
  • [ ] Use the correct link to the actual license
  • [ ] There is a link to the organizer, e.g. “© 2017 FOSSASIA and Science Centre Singapore.” with a wrong link. Take out that link.
  • [ ] Behind “The website and it’s contents are licensed under Creative Commons Attribution 4.0 International License.” or similar add the following: “The site was generated using the Open Event format on the eventyay site generator. Please submit issues here.”
Updated 30/04/2017 16:13

Topics: Error bars appear broken on some disparity charts


I only reviewed the /teen-birth-rate indicator topic and found this issue with the error bars. My recommendation is to review all indicators for the same issue.

When reviewing different years on the “disparities” the error bars seem to be incorrect. Here is a screenshot: <img width=“921” alt=“screen shot 2017-04-30 at 10 26 02 am” src=“”>

Here are the years when I am finding this same issue for teen birth rate: 2001, 2003, 2005, 2007, 2009

Updated 30/04/2017 15:28

Use service url for botstate in emulator

OAuth2::Error: {"code"=>"BadArgument", "message"=>"Invalid channel ID"}: 
  "error": {
    "code": "BadArgument",
    "message": "Invalid channel ID"

Botstate storage for emulator is to be taken from activity.service_url in emulator. Related code in node version

Updated 30/04/2017 15:28

Topics: Race/Ethnicity labels overlap on disparities graph


Similar to issue #279, we are seeing that the race/ethnicity labels on the disparities chart are overlapping making this text difficult to read.

Below is a screenshot from my MacBook Pro 13-inch and this would be an even bigger issue viewing on smaller screens. <img width=“875” alt=“screen shot 2017-04-30 at 10 16 27 am” src=“”>

Updated 30/04/2017 15:21

Maps not displaying all records


Just tried to produce a map of Odontomyia argentata (Silver Colonel soldierfly), and I can only get a partial set of records to display, and it is partitioned differently in different places.

Via Summary details (


Via activity page - explore activity data (


Via Explore - all records:


Updated 30/04/2017 15:16

Can't click this comment


Then load the full thread and find that comment. You can’t click it. However, if you load the link I posted, you can click it.

Updated 30/04/2017 15:12

Quicksilver Fountain


This card has a few problems:

  • it allows you to select lands that you do not control (i.e. other players' lands) to apply its effects
  • any player with only islands as lands will get stuck in their upkeep phase as the card insists on them selecting a non island land


Updated 30/04/2017 15:53 1 Comments

Multiple TreeModel update broadcasts when height changed


Check AbstractModel.variableChangedEvent:

    public final void variableChangedEvent(Variable variable, int index, Parameter.ChangeType type) {
        handleVariableChangedEvent(variable, index, type);

        // todo AR - I am not sure this is required and may be overruling modelChange events on parts of the
        // model. If a parameter changes it should be handleVariableChangedEvent() job to fireModelChanged
        // events
        listenerHelper.fireModelChanged(this, variable, index);

Is the fireModelChanged needed?

Updated 30/04/2017 15:07

Fork me on GitHub