Add a preference option for the order that tracks will be loaded in 4-deck skins #6

ywwg · 2013-05-21T03:30:46Z

No description provided.

daschuer · 2013-05-21T06:01:51Z

Gihub complains:
"This diff is too big to show! We're showing status information only."

Any idea to solve it?

ywwg · 2013-05-21T13:33:56Z

I'll remove my new skins, that should help

ywwg · 2013-05-21T13:41:57Z

OK This should be readable now.

kain88-de · 2013-05-21T15:08:34Z

mixxx/src/playermanager.h

@@ -123,6 +147,8 @@ class PlayerManager : public QObject {
    // Used to protect access to PlayerManager state across threads.
    mutable QMutex m_mutex;

+    QList<QList<int> > ms_deck_orderings;


can you name this with camelCase

Hm, what's ms_?

oh it was originally a static member, so I was like, ms_!

rryan · 2013-05-21T15:36:05Z

I added some particular comments on the implementation, but in general I think we should do something more general than make a hard-coded fix for 4-decks.

rryan · 2013-05-21T15:40:10Z

mixxx/src/playermanager.cpp

@@ -34,17 +34,24 @@
        m_pAnalyserQueue(NULL),
        m_pCONumDecks(new ControlObject(ConfigKey("[Master]", "num_decks"), true, true)),
        m_pCONumSamplers(new ControlObject(ConfigKey("[Master]", "num_samplers"), true, true)),
-        m_pCONumPreviewDecks(new ControlObject(ConfigKey("[Master]", "num_preview_decks"), true, true)) {
+        m_pCONumPreviewDecks(new ControlObject(ConfigKey("[Master]", "num_preview_decks"), true, true)),
+        m_pCOSkinNumDecks(new ControlObject(ConfigKey("[Skin]", "num_decks"), true, true)) {


Need to create [Skin],num_samplers and [Skin],num_preview_decks as well.

rryan · 2013-05-21T15:57:03Z

The [Skin],num_decks thing is kind of unsettling the more I think about it.

If you can't load a track into a deck the skin didn't declare (e.g. decks the MIDI controller declares) then what's the point of letting a MIDI controller declare decks? This change means that in order to use N-decks your skin must support the N.

What issues are solved by the [Skin],num_decks change? I wonder if there's a better fix.

ywwg · 2013-05-21T16:29:31Z

I strongly disagree with allowing decks to exist that aren't visible in the skin. It's a recipe for user confusion -- it's even confused me when I have a third track that I can't see loaded in a 2-deck skin. The problem skin-decks solves is when I switch between 2-deck skins and 4-deck skins. I want to be able to load as many tracks as the skin has, period. Is there an example of a midi controller we support that has 4 decks that would be usable without a 4 deck skin? It would have to have waveform overview displays, text displays, something like a CDJ.

This change reveals the need for more 4-deck skins, but that's the point. We don't support 4 decks if you can't see them in the interface.

As for hard-coding 4-deck solutions, I think 4-decks is a special case. With more decks, like 8 or 16, the user is doing loop-based work and the decks should just be loaded in order as I've coded. With 2, it's obvious that they are left and right. 4 decks is where you see this CABD stuff, and when I've spun at middlesex they definitely had the "outside/inside" method of pairing decks. It's odd, but the way 4 decks are actually used justifies the code.

ywwg · 2013-05-21T17:56:25Z

Pushed changes. Untested, but it builds.

rryan · 2013-05-21T18:10:29Z

On May 21, 2013 9:29 AM, "Owen Williams" notifications@github.com wrote:

I strongly disagree with allowing decks to exist that aren't visible in
the skin. It's a recipe for user confusion -- it's even confused me when I
have a third track that I can't see loaded in a 2-deck skin.

Hmm you're right. I agree we shouldn't show decks that don't have a ui
representation.

But maybe we should instead fix the issue where mixxx can't remove a deck
that has already been created. That would fix the switching skins issue.

It wouldn't fix midi controllers that will just break when they have been
written assuming a fixed number of decks. We could make mappings and script
that touches non existent controls just silently fail instead of throwing a
hissy fit. Also loading a preset that is for 4 decks should give the user a
warning if that many decks do not exist.

ywwg · 2013-05-21T18:14:14Z

I agree that we should be able to remove decks. In fact I tried to do that at first and ran into race conditions if I tried to delete a deck while it was playing. Maybe decks should be able to receive a "die()" signal so they can safely eject tracks and shut themselves down? I'm not sure what the proper multithreaded OO way of doing this is.

rryan · 2013-05-21T18:17:18Z

I think deck ordering is a good idea but I think the implementation should
be more general. What about 3 decks? Or 5? We'll surely get requests for it
even if it is not the main use case and if we went with the current
implementation it will box us in to adding a similar special case for every
number we want to support.
On May 21, 2013 9:29 AM, "Owen Williams" notifications@github.com wrote:

I strongly disagree with allowing decks to exist that aren't visible in
the skin. It's a recipe for user confusion -- it's even confused me when I
have a third track that I can't see loaded in a 2-deck skin. The problem
skin-decks solves is when I switch between 2-deck skins and 4-deck skins. I
want to be able to load as many tracks as the skin has, period. Is there an
example of a midi controller we support that has 4 decks that would be
usable without a 4 deck skin? It would have to have waveform overview
displays, text displays, something like a CDJ.

This change reveals the need for more 4-deck skins, but that's the point.
We don't support 4 decks if you can't see them in the interface.

As for hard-coding 4-deck solutions, I think 4-decks is a special case.
With more decks, like 8 or 16, the user is doing loop-based work and the
decks should just be loaded in order as I've coded. With 2, it's obvious
that they are left and right. 4 decks is where you see this CABD stuff, and
when I've spun at middlesex they definitely had the "outside/inside" method
of pairing decks. It's odd, but the way 4 decks are actually used justifies
the code.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/6#issuecomment-18220170
.

ywwg · 2013-05-21T18:22:35Z

true, I have a three-deck mixer actually, though the use-case is more for two-decks and an mp3 player. I don't know how to programatically generate sane orderings. e.g., no one is going to want DACB, right?

I guess all I can think of is a drag-reorderable list of decks. Which seems like overkill. But now that I think about use-cases, what if someone wants to use Mixxx as a kind of mixing board for a whole band? Maybe they have two turntables, and a mic, and a guitar, and a synth. Who knows where the turntables are.

rryan · 2013-05-21T18:31:53Z

Yea - The UI could be hard to make general. The preferences could be 4deck
hard coded for 1.12 but the underlying implementation should be general.
See what I mean? That makes future 3 deck support a UI issue alone.
On May 21, 2013 11:23 AM, "Owen Williams" notifications@github.com wrote:

true, I have a three-deck mixer actually, though the use-case is more for
two-decks and an mp3 player. I don't know how to programatically generate
sane orderings. e.g., no one is going to want DACB, right?

I guess all I can think of is a drag-reorderable list of decks. Which
seems like overkill. But now that I think about use-cases, what if someone
wants to use Mixxx as a kind of mixing board for a whole band. Maybe they
have two turntables, and a mic, and a guitar, and a synth. Who knows where
the turntables are.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/6#issuecomment-18227087
.

ywwg · 2013-05-21T18:50:34Z

How about making deckorderings a hash of int to lists of strings? QHash < QList < QString > > >. Or maybe a small class that has that structure internally, but externally just has convenience functions like .begin and .end for easy enumeration. (That would also solve the translation issue between the visual ordering, eg CABD, and the load ordering, eg 1,2,0,3.

the generic code would be:

if number_of_decks in deckorderings.keys:
return list at that key of possible orderings, where each ordering is a string "ABCDEFGH..." etc
else:
return numeric order only.

so for now we'd only prepopulate hashvalue of 4. later if we want to add something for 3 deck, we just push a string into the list for hash 3. There could even be a UI that would push those values in.

…-deck skins, since controllers vary

…views

daschuer · 2013-05-22T14:32:18Z

mixxx/src/playermanager.cpp

+    // Make sure the number of internal decks is in sync with the number of decks in the skin.
+    connect(m_pCOSkinNumDecks, SIGNAL(valueChanged(double)),
+            this, SLOT(slotNumDecksControlChanged(double)),
+            Qt::DirectConnection);


Qt::DirectConnection should be removed.
After our atomic changes this is always a Qt::DirectConnection because the signal is resend by the ControlObject itselve.
Direct connections are somehow not supported now because if the signal is from an other thread it is transfered into the current thread by the ControlObject before.

Oops, GitHub did a really bad job formatting my reply-via-email. Re-posting:

That's not true -- an AutoConnection decides between direct invocation and queueing based on the emitting thread and the receiver object's thread property alone (if they match, direct. if not, queue). Nothing we did in atomic-co changes this -- a non main thread change made to this control (by looking it up with getControl or in the case of our many shared state multiple thread classes) would queue, not directly invoke. Also queueing is not acceptable here because other parts of Mixxx want this processing to be done immediately (eg so skin processing doesn't have to block while waiting for PlayerManager to get the signal) and this class is designed to be thread safe (locks) because of this.

2013/5/22 RJ Ryan notifications@github.com

In mixxx/src/playermanager.cpp:

@@ -46,6 +50,17 @@
this, SLOT(slotNumPreviewDecksControlChanged(double)),
Qt::DirectConnection);

// Make sure the number of internal decks is in sync with the number of decks in the skin.

connect(m_pCOSkinNumDecks, SIGNAL(valueChanged(double)),

this, SLOT(slotNumDecksControlChanged(double)),

Qt::DirectConnection);

Oops, GitHub did a really bad job formatting my reply-via-email.
Re-posting:

That's not true -- an AutoConnection decides between direct invocation and
queueing based on the emitting thread and the receiver object's thread
property alone (if they match, direct. if not, queue). Nothing we did in
atomic-co changes this -- a non main thread change made to this control (by
looking it up with getControl or in the case of our many shared state
multiple thread classes) would queue, not directly invoke. Also queueing is
not acceptable here because other parts of Mixxx want this processing to be
done immediately (eg so skin processing doesn't have to block while waiting
for PlayerManager to get the signal) and this class is designed to be
thread safe (locks) because of this.ect

Hi RJ,

in general, you are right, but in this case we have a double signal, one
with (QObject* sender), which is a queued connection in case of a different
sender and a direct connection else. The final signal without sender is a
direct connection in any case. But isn't the PlayerManager and the
LegacySkin parser in the same thread anyway?

—

Reply to this email directly or view it on GitHubhttps://github.com//pull/6/files#r4343623
.

Not sure what you mean by double signal? If you are referring to the chain of signals from ControlObject::set() -> ControlDoublePrivate::set() -> ControlDoublePrivate::valueChanged() -> ControlObject::privateValueChanged() -> ControlObject::valueChanged/valueChangedFromEngine then I think you are missing that the ControlNumericPrivate / ControlObject connection is also a direct connection (and must be -- otherwise Qt will take unintended roundtrips through the event queue if the current thread when set() is called does not match the thread property of the object attached to the signal).

Qt signals and slots do not care about the emitter of the signal at all, all they care about is the current thread when emit() is called. If the current thread when emit is called does not match the receiving objects 'thread' property then AutoConnection is queued, otherwise it is a DirectConnection.

isn't the PlayerManager and the LegacySkin parser in the same thread anyway

PlayerManager is thread safe and its methods are invoked by multiple threads in addition to the main thread. LegacySkinParser is totally in the main thread.

For example, MIDI controllers can (and do) set the number of decks. Then the valueChanged signal would come from the controller thread.

Oh, Yes, I have overlooked that ControlObject::initialize() sets up a
direct connection.

2013/5/23 RJ Ryan notifications@github.com

In mixxx/src/playermanager.cpp:

@@ -46,6 +50,17 @@
this, SLOT(slotNumPreviewDecksControlChanged(double)),
Qt::DirectConnection);

// Make sure the number of internal decks is in sync with the number of decks in the skin.

connect(m_pCOSkinNumDecks, SIGNAL(valueChanged(double)),

this, SLOT(slotNumDecksControlChanged(double)),

Qt::DirectConnection);

Not sure what you mean by double signal? If you are referring to the chain
of signals from ControlObject::set() -> ControlDoublePrivate::set() ->
ControlDoublePrivate::valueChanged() ->
ControlObject::privateValueChanged() ->
ControlObject::valueChanged/valueChangedFromEngine then I think you are
missing that the ControlNumericPrivate / ControlObject connection is also a
direct connection (and must be -- otherwise Qt will take unintended
roundtrips through the event queue if the current thread when set() is
called does not match the thread property of the object attached to the
signal).

Qt signals and slots do not care about the emitter of the signal at all,
all they care about is the current thread when emit() is called. If the
current thread when emit is called does not match the receiving objects
'thread' property then AutoConnection is queued, otherwise it is a
DirectConnection.

isn't the PlayerManager and the LegacySkin parser in the same thread anyway

PlayerManager is thread safe and its methods are invoked by multiple
threads in addition to the main thread. LegacySkinParser is totally in the
main thread.

For example, MIDI controllers can (and do) set the number of decks. Then
the valueChanged signal would come from the controller thread.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/6/files#r4351458
.

daschuer · 2013-05-22T15:21:46Z

I did not get the point why we need [Master] and [Skin] deck_num.
We have also m_preview_decks.size() for a internal count of decks.

I would like to stay with [Master] deck_num, Because there is IMHO no need for renaming it.

I like the Idea to let the skin control the Deck count. Is there a way to avoid others to control the deck count?
Should we implement a ro control object for this purpose?

Will the deck lable be shown in the Skin?
I wonder why the Mixer orientation is no issue here?

ywwg · 2013-05-22T15:34:48Z

The reason I added [Skin] deck_num is because [Master] deck_num can never be reduced. So if I go from a skin with 4 decks to a skin with 2 decks, [Master]deck_num still says 4. So I wanted a CO that would keep track of how many decks are actually usable.

A better solution to this would be to support the removing of decks so that [Master]deck_num is accurate in that use-case.

Banshee merge

Fix compile on WIN, use GetSystemTimePreciseAsFileTime() on Windows 8

use QEvent::WindowDeactivate to reset the press state

fix scripts for Qt 5.12

Mastersyc double / half Bpm Fix.

Hercules DJControl Inpulse 200: Fix some code stye issues

Ddj400

…h sync When loading a track that is not yet present in the library (and thus doesn't have any BPM because it hasn't been analyzed yet) while another deck is playing and both decks have sync enabled, a debug assertion is triggered: DEBUG ASSERT: "isValid()" in function double mixxx::Bpm::value() const at src/track/bpm.h:53 Aborted (core dumped) The backtrace looks as follows: #0 0x00007f175c87234c in __pthread_kill_implementation () at /usr/lib/libc.so.6 #1 0x00007f175c8254b8 in raise () at /usr/lib/libc.so.6 #2 0x00007f175c80f534 in abort () at /usr/lib/libc.so.6 #3 0x00007f175cf05ee4 in qt_assert(char const*, char const*, int) () at /usr/lib/libQt5Core.so.5 #4 0x000055deb2e67e1c in mixxx::(anonymous namespace)::handleMessage(QtMsgType, QMessageLogContext const&, QString const&) (type=<optimized out>, context=<optimized out>, input=<optimized out>) at src/util/logging.cpp:355 #5 0x00007f175cf47128 in () at /usr/lib/libQt5Core.so.5 #6 0x00007f175cf3fd8a in () at /usr/lib/libQt5Core.so.5 #7 0x00007f175cf06526 in QMessageLogger::critical(char const*, ...) const () at /usr/lib/libQt5Core.so.5 #8 0x000055deb2e5c720 in mixxx_debug_assert(char const*, char const*, int, char const*) (assertion=assertion@entry=0x55deb39bd0db "isValid()", file=file@entry=0x55deb39bbf30 "src/track/bpm.h", line=line@entry=53, function=function@entry=0x55deb39bbf08 "double mixxx::Bpm::value() const") at gsrc/util/assert.h:9 #9 0x000055deb2ee7e7e in mixxx_debug_assert_return_true(char const*, char const*, int, char const*) (function=0x55deb39bbf08 "double mixxx::Bpm::value() const", line=53, file=0x55deb39bbf30 "src/track/bpm.h", assertion=0x55deb39bd0db "isValid()") at gsrc/util/assert.h:18 #10 mixxx::Bpm::value() const (this=<synthetic pointer>) at src/track/bpm.h:53 #11 mixxx::operator*(mixxx::Bpm, double) (multiple=1, bpm=...) at src/track/bpm.h:160 #12 SyncControl::setLocalBpm(mixxx::Bpm) (this=<optimized out>, localBpm=...) at src/engine/sync/synccontrol.cpp:567 #13 0x000055deb34c7ba3 in EngineBuffer::postProcess(int) (this=0x55deb56eb060, iBufferSize=2048) at src/engine/enginebuffer.cpp:1318 #14 0x000055deb3139023 in EngineMaster::processChannels(int) (this=0x55deb5449440, iBufferSize=<optimized out>) at src/engine/enginemaster.cpp:383 #15 0x000055deb31394f7 in EngineMaster::process(int) (this=0x55deb5449440, iBufferSize=iBufferSize@entry=2048) at src/engine/enginemaster.cpp:410 #16 0x000055deb2f91d0b in SoundManager::onDeviceOutputCallback(long) (this=<optimized out>, iFramesPerBuffer=iFramesPerBuffer@entry=1024) at src/soundio/soundmanager.cpp:596 #17 0x000055deb32dd794 in SoundDevicePortAudio::callbackProcessClkRef(long, float*, float const*, PaStreamCallbackTimeInfo const*, unsigned long) (this=0x55deb553e6b0, framesPerBuffer=1024, out=<optimized out>, in=<optimized out>, timeInfo=<optimized out>, statusFlags=<optimized out>) at src/soundio/sounddeviceportaudio.cpp:965 This happens because `newLocalBpm` is invalid when `localBpm` is invalid. Trying to do sync decks while no tempo information is available does not make sense, so we only synchronize decks if the local BPM is available.

Silence some float conversion warnings by making conversion explicit

VersionStore: Remove Apple fallback

kain88-de reviewed May 21, 2013
View reviewed changes

rryan reviewed May 21, 2013
View reviewed changes

ywwg added 5 commits May 21, 2013 21:35

Add a preference option for the order that tracks will be loaded in 4…

eb0fd22

…-deck skins, since controllers vary

Address most notes. Still need to add [Skin] COs for samplers and pre…

57ea2f1

…views

Make sure we have enough samplers and previewdecks to match the skin

aa98cdd

put definition back in header

f1af434

not getting number of decks on load, but otherwise working ok

8977f64

daschuer reviewed May 22, 2013
View reviewed changes

clean up, and fix loading config at startup

b4923bc

daschuer mentioned this pull request Dec 28, 2014

Spinny updates #440

Merged

daschuer mentioned this pull request Jan 8, 2015

New SoundSource/AudioSource API #411

Merged

uklotzde mentioned this pull request Jul 21, 2015

CachingReader: Reduce log spam for corrupt files #652

Merged

ferranpujolcamins referenced this pull request in ferranpujolcamins/mixxx Jul 22, 2015

Merge pull request #6 from daschuer/LibraryConcurencyRefactoring

8138f46

Banshee merge

ywwg pushed a commit that referenced this pull request Oct 28, 2015

Merge pull request #6 from daschuer/1.12-shoutcast-fixes

2ea06af

Fix compile on WIN, use GetSystemTimePreciseAsFileTime() on Windows 8

daschuer mentioned this pull request Jul 27, 2016

Follow-up: Verify properties of audio signals #985

Merged

daschuer mentioned this pull request Aug 13, 2017

Multi-broadcasting #1300

Merged

19 tasks

daschuer mentioned this pull request Jun 3, 2018

[WIP] Livemetadata PR #1675

Closed

daschuer mentioned this pull request Jun 11, 2018

Library Layout Redesign Single pane #1117

Closed

11 tasks

uklotzde mentioned this pull request Jan 16, 2020

Replace deprecated QSignalMapper with lambda expressions #2427

Merged

daschuer pushed a commit that referenced this pull request Mar 4, 2020

Merge pull request #6 from daschuer/pushbutton-pressed-state

31ac96d

use QEvent::WindowDeactivate to reset the press state

Holzhaus referenced this pull request in Holzhaus/mixxx Apr 9, 2020

Merge pull request #6 from Be-ing/qjsengine_migration

99f9f36

fix scripts for Qt 5.12

Holzhaus pushed a commit that referenced this pull request Apr 27, 2020

Merge pull request #6 from daschuer/mastersyc_onemaster2

b0e0f37

Mastersyc double / half Bpm Fix.

Holzhaus pushed a commit that referenced this pull request Jun 12, 2020

Merge pull request #6 from Holzhaus/Hercules-Inpulse-200

276da56

Hercules DJControl Inpulse 200: Fix some code stye issues

jusko pushed a commit to jusko/mixxx that referenced this pull request Dec 19, 2020

Merge pull request mixxxdj#6 from dj3730/ddj400

4648400

Ddj400

jusko pushed a commit to jusko/mixxx that referenced this pull request Dec 22, 2020

Merge pull request mixxxdj#6 from dj3730/ddj400

12ce9fe

Ddj400

jusko pushed a commit to jusko/mixxx that referenced this pull request Jan 30, 2021

Merge pull request mixxxdj#6 from dj3730/ddj400

a5b5d85

Ddj400

ronso0 mentioned this pull request Sep 4, 2022

Effects: allow clearing chains with empty '---' preset #10859

Merged

ronso0 mentioned this pull request Sep 13, 2022

adding WVuMeterGL as alternative for WVuMeter #10893

Merged

ronso0 mentioned this pull request May 9, 2023

qopenglwindow-based replacement of qglwidget, full glsl shader based waveforms, vumeters and spinnies #10989

Merged

ywwg mentioned this pull request Jun 2, 2023

Build with Qt6 and optionally with QML #11608

Merged

daschuer pushed a commit that referenced this pull request Jun 3, 2023

Merge pull request #6 from ywwg/build-with-qt6-optional-qml

3137224

Silence some float conversion warnings by making conversion explicit

Swiftb0y mentioned this pull request Aug 7, 2023

GNOME 44 (Wayland): Context menu on tree items in side bar closes immediately #11576

Closed

Frotus mentioned this pull request Jan 27, 2024

Mixxx hang (infinite) at startup, when a secondary monitor is the application window target and RunAs Admin is NOT used #12655

Open

daschuer pushed a commit that referenced this pull request Jan 28, 2024

Merge pull request #6 from fwcd/fixversionstore

cdf87b9

VersionStore: Remove Apple fallback

acolombier mentioned this pull request May 22, 2024

Simplify waveform preferences panel #13226

Open

JoergAtGithub mentioned this pull request May 22, 2024

Add latest changes to CHANGELOG (main) #13193

Merged

m0dB mentioned this pull request Nov 10, 2024

thread sanitizer warning BrowseThread / BrowseTableModel #13861

Open

ronso0 mentioned this pull request Nov 14, 2024

2.5 and main crash on shutdown #13885

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add a preference option for the order that tracks will be loaded in 4-deck skins #6

Add a preference option for the order that tracks will be loaded in 4-deck skins #6

ywwg commented May 21, 2013

daschuer commented May 21, 2013

ywwg commented May 21, 2013

ywwg commented May 21, 2013

kain88-de May 21, 2013

rryan May 21, 2013

ywwg May 21, 2013

rryan commented May 21, 2013

rryan May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

daschuer May 22, 2013

rryan May 22, 2013

daschuer May 22, 2013

rryan May 22, 2013

daschuer May 23, 2013

daschuer commented May 22, 2013

ywwg commented May 22, 2013

Add a preference option for the order that tracks will be loaded in 4-deck skins #6

Add a preference option for the order that tracks will be loaded in 4-deck skins #6

Conversation

ywwg commented May 21, 2013

daschuer commented May 21, 2013

ywwg commented May 21, 2013

ywwg commented May 21, 2013

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rryan commented May 21, 2013

Choose a reason for hiding this comment

rryan commented May 21, 2013

ywwg commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

rryan commented May 21, 2013

ywwg commented May 21, 2013

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

daschuer commented May 22, 2013

ywwg commented May 22, 2013