Re: [opensource-dev] Slightly off topic but need ideas how to fix

[Cinder Roxley]

On Thu, 13 Jan 2011 13:13:32 -0700, Altair Sythos Memo syt...@gmail.com
wrote:

 On Thu, 13 Jan 2011 11:17:27 -0500
 Erin Mallory angel_of_crim...@hotmail.com wrote:


 This is slightly off topic but as the ramifications of what will
 happen when the old profiles are taken down is starting to get out
 there, merchants and content creators are starting to panic about
 what these new profiles mean for them. There is a a ton of content
 that is going to be broken by this, much of it content that
 businesses rely on heavily. a more full list can be found in web-3509
 but a number of rewards systems, venders, and security devices found
 within hundreds if not thousands of SL stores are built heavily
 around the old profile pages and cannot be easily updated or replaced
 before the old profiles will likely be taken down... Even if
 businesses begin to switch over, that involves changing out sometimes
 hundreds of venders, everything associated with the venders (like
 gift cards etc), and during all of that the ability to make new
 content is affected. Many SL businesses are teetering on the edge of
 bankrupcy already, including many where this is the SOLE or largest
 percentage of income to those that simply are incapable of getting an
 income any other way. So I'm pleading with the lindens and devs on
 this list to PLEASE start putting some thought to how to mitigate
 this...

 i'm sorry but don't understand how web interface to show profile data
 can involve business. scripts still work in both directions
 (groups, key2name, the new istr about display names), and web interface
 is just a re-design of the old profile in sidebar (upper piece
 SL-photo +SL desc, just under RL photo +info, on a sie groups and
 picks), is just a interface design... you can stil ask for friendship,
 open an IM, offer a teleport, invite to a group (and bot can still
 invite to groups), so how shops, security systems and other can be
 involded by web profiles?

Poor design that involves grepping world.secondlife.com for information
which is pushed to outworld databases. (something Erin seems
vehemently against in her dozen other recent JIRA bug reports on
webprofiles). I don't see how it's a critical bug that content
creators wouldn't be able to mine world.secondlife.com anymore for
userkeys, groups, profile pics, etc on their outworld servers.


-- 
Sent from my Tracfone!
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] SSL Handshake Error

[Cinder Roxley]

No, but I get it regularly with other ssl secured websites using the built-in 
browser.

Trilo Byte trilobyte5...@gmail.com wrote:

Is anybody else getting 'SSL Handshake Error' whenever they attempt to
look up a resident's profile in the Viewer?  Is this something that's
happening Viewer-side, or (web) Server-side?  I know the new profile
system was working fine not long ago, but sometime during the course of
the last week they stopped.

Any fixes or workarounds would be greatly appreciated.

TriloByte Zanzibar

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges

-- 
Sent from my Tracfone!
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Testers needed to repo storm-1002

[Cinder Roxley]

I CAN delete broken links in Viewer 2.5.0. I could not reproduce the issue.

--
Cindy

-Original Message-
From: opensource-dev-boun...@lists.secondlife.com
[mailto:opensource-dev-boun...@lists.secondlife.com] On Behalf Of Jonathan
Welch
Sent: Sunday, February 20, 2011 3:59 PM
To: OpenSource Mailing List
Subject: [opensource-dev] Testers needed to repo storm-1002

I started to work on Storm-1002 (Cannot delete broken links) today and
_could_ delete a broken link.  I tried reproducing the non-deleting issue on
viewer 2.1.2 (it was reported as affecting v2.1.1) and did not have trouble
there, either.

If you have a few minutes could you follow the steps in
https://jira.secondlife.com/browse/STORM-1002 and see if you can or cannot
delete a broken link?  At this point I can only surmise that a server change
has fixed this issue somehow.

Thank you,

-jonathan

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Camera UI in Second Life beta viewer is broken

[Cinder Roxley]

Hi,

 

There's a little Refresh button (the itty bitty arrow that makes a circle.)
You need to click that to refresh the picture, 

which will then make the Save button clickable again. Probably not ideal,
but it works. In previous viewers, you needed

to refresh the snapshot the same way anytime a setting was changed.

 

Kind regards,

Cinder Roxley

 

From: opensource-dev-boun...@lists.secondlife.com
[mailto:opensource-dev-boun...@lists.secondlife.com] On Behalf Of Laurent
Bechir
Sent: Saturday, November 19, 2011 8:00 PM
To: opensource-dev@lists.secondlife.com
Subject: Re: [opensource-dev] Camera UI in Second Life beta viewer is broken

 


It seems my answer has not arrived. I send a new link to show what happens
when I try to take pictures. The version of the viewer is  3.2.2 (244666)
Nov  8 2011 17:17:35 :

http://youtu.be/93JYlOwYjY0

the Jira is here :

https://jira.secondlife.com/browse/VWR-27652





 mailto:cin...@cinderblocks.biz Cinder Roxley

19 novembre 2011 03:55

Hi,

The video linked in the jira doesn't work in my browser. I'm able to take
snapshots in all modes using the new snapshot workflow.

Kind regards,
Cinder Roxley



 mailto:laurent.bec...@madonie.org Laurent Bechir

19 novembre 2011 03:29

Hello,

I've opened a Jira about the camera in Second Life beta viewer since it 
is impossible to take pictures with it. I don't know if this jira has 
already been submitted, but there is something to do about that :

https://jira.secondlife.com/browse/VWR-27652


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges

image001.jpgimage002.jpg___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Odd disappearing vehicle

[Cinder Roxley]

Greetings,

The creator is Yasen Tomcast, so you may try asking her for details.

You can see the object profile if you click it as it enters the region.
(She's placed it on a border to prevent autoreturn.)

It uses a MOAP hack to disappear. On right clicking:

2012-01-30T15:11:20Z INFO: LLViewerMediaImpl::navigateInternal: media id=
9c8badee-5539-3447-4a2d-9ca9d5ae3202 url=about:blank mime_type=
2012-01-30T15:11:20Z INFO: LLViewerMedia::getCurrentUserAgent:
SecondLife/3.2.8.2 (Milkshake; default skin) 2012-01-30T15:11:20Z INFO:
LLViewerMediaImpl::loadURI: Asking media source to load URI: about:blank
2012-01-30T15:11:20Z WARNING: LLUICtrl::initCommitCallback: No callback
found for: 'Media.ResetCurrentUrl' in control: current_url_reset_btn
2012-01-30T15:11:20Z INFO: LLViewerMediaImpl::navigateInternal: media id=
9c8badee-5539-3447-4a2d-9ca9d5ae3202 url=data:image/svg+xml,%3Csvg
xmlns=%22http://www.w3.org/2000/svg%22 width=%22100%%22 height=%22100%%22
%3E%3Cdefs%3E%3Cpattern id=%22checker%22 patternUnits=%22userSpaceOnUse%22
x=%220%22 y=%220%22 width=%22128%22 height=%22128%22 viewBox=%220 0 128
128%22 %3E%3Crect x=%220%22 y=%220%22 width=%2264%22 height=%2264%22
fill=%22#ff%22 /%3E%3Crect x=%2264%22 y=%2264%22 width=%2264%22
height=%2264%22 fill=%22#ff%22 /%3E%3C/pattern%3E%3C/defs%3E%3Crect
x=%220%22 y=%220%22 width=%22100%%22 height=%22100%%22
fill=%22url(#checker)%22 /%3E%3C/svg%3E mime_type= 2012-01-30T15:11:20Z
INFO: LLViewerMediaImpl::loadURI: Asking media source to load URI:
data:image/svg+xml,%3Csvg%20xmlns=%22http://www.w3.org/2000/svg%22%20width=%
22100%%22%20height=%22100%%22%20%3E%3Cdefs%3E%3Cpattern%20id=%22checker%22%2
0patternUnits=%22userSpaceOnUse%22%20x=%220%22%20y=%220%22%20width=%22128%22
%20height=%22128%22%20viewBox=%220%200%20128%20128%22%20%3E%3Crect%20x=%220%
22%20y=%220%22%20width=%2264%22%20height=%2264%22%20fill=%22#ff%22%20/%3
E%3Crect%20x=%2264%22%20y=%2264%22%20width=%2264%22%20height=%2264%22%20fill
=%22#ff%22%20/%3E%3C/pattern%3E%3C/defs%3E%3Crect%20x=%220%22%20y=%220%2
2%20width=%22100%%22%20height=%22100%%22%20fill=%22url(#checker)%22%20/%3E%3
C/svg%3E
2012-01-30T15:11:20Z INFO: LLMediaCtrl::handleVisibilityChange: visibility
changed to false 2012-01-30T15:11:21Z INFO:
LLPluginProcessParent::receiveMessage: plugin version string: Webkit media
plugin, Webkit version 2.02.47681 (QtWebKit version 4.7.1)
2012-01-30T15:11:21Z INFO: LLPluginProcessParent::receiveMessage: message
class: base - version: 1.0 2012-01-30T15:11:21Z INFO:
LLPluginProcessParent::receiveMessage: message class: media - version: 1.0
2012-01-30T15:11:21Z INFO: LLPluginProcessParent::receiveMessage: message
class: media_browser - version: 1.0 2012-01-30T15:11:21Z INFO:
LLFloater::openFloater: Opening floater toolbox floater 2012-01-30T15:11:21Z
WARNING: LLFloaterTools::refresh: Failed to get selected object
2012-01-30T15:11:22Z WARNING: LLFloaterTools::refresh: Failed to get
selected object 2012-01-30T15:11:22Z WARNING: LLFloaterTools::refresh:
Failed to get selected object

--
Kind regards,
-Cinder

-Original Message-
From: opensource-dev-boun...@lists.secondlife.com
[mailto:opensource-dev-boun...@lists.secondlife.com] On Behalf Of Jonathan
Welch
Sent: Monday, January 30, 2012 7:26 AM
To: OpenSource Mailing List
Subject: [opensource-dev] Odd disappearing vehicle

Someone just reported to me that they had found a vehicle at
http://slurl.com/secondlife/Fortuna/5/243/26

If you try to select it or if you try to right click on it it disappears.
How can this be possible?  I used the object scanner in Radegast to do a bit
more sleuthing, but even in there if you click on the name in the list
(which is usually Loading...) it disappears right away.

Can someone investigate and report back?  I am curious what is going on
here.  Also, maybe drop a note to Laetizia Coronet, who discovered this.

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Odd disappearing vehicle

[Cinder Roxley]

That should actually be Tomcat, not Tomcast. :/

-Original Message-
From: Cinder Roxley [mailto:cin...@cinderblocks.biz] 
Sent: Monday, January 30, 2012 8:23 AM
To: 'OpenSource Mailing List'
Subject: RE: [opensource-dev] Odd disappearing vehicle

Greetings,

The creator is Yasen Tomcast, so you may try asking her for details.

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] upload settings file for mesh upload?

[Cinder Roxley]

There's an upload config file the viewer downloads to send snapshots to
profile feeds, it's not related to mesh, but I don't know any other upload
settings files the viewer grabs.

Kind regards,
-Cinder

-Original Message-
From: Lance Corrimal
Sent: Wednesday, February 15, 2012 11:28 PM
To: opensource-dev@lists.secondlife.com
Subject: [opensource-dev] upload settings file for mesh upload?

Hi

has anyone here ever heard of a upload settings file for second life,
somehow related to mesh uploads?


bye,
LC

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Policy Changes

[Cinder Roxley]

Yes, you're mistaken. The key phrase there is alters the shared 
experience of the virtual world.  A tpv can alter individual user's 
experiences, (UI, build tools, controls, graphics enhancements) but not 
the shared experience of the world.  IE, exposing information such as 
the friend online visibility of *other users*.


Kind regards,
-Cinder

On 2/24/2012 4:44 PM, Nalates Urriah wrote:
Does this new policy essentially eliminate the reason for the 
existence of 3rd party viewers:


2.k : You must not provide any feature that alters the shared 
experience of the virtual world in any way not provided by or 
accessible to users of the latest released Linden Lab viewer.


http://community.secondlife.com/t5/Second-Life-Viewer/Third-Party-Viewer-Policy-Changes/m-p/1399141

This seems to say all changes can be submitted to LL but not 
implemented until and unless LL approves them and adds them to the SL 
viewer. Am I mistaken?


--
Nalates Urriah (SL AV)


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Help testing needed on STORM-1812

[Cinder Roxley]

On 4/1/2012 8:56 AM, Oz Linden (Scott Lawrence) wrote:
 QA reports that https://jira.secondlife.com/browse/STORM-1812 failed on
 Windows and Linux.

 This is a fix for music fading when teleporting within a parcel... the
 music should either be continuous or at most fade out and back in.

 I can't repro the problem on my Mac... if you can on one of the above
 platforms, please attach a DEBUG level log to the issue.
 ___
 Policies and (un)subscribe information available here:
 http://wiki.secondlife.com/wiki/OpenSource-Dev
 Please read the policies before posting to keep unmoderated posting privileges
I attached a debug level log, Second Life 3.3.0.251182 Windows7(x86). 
Reproduced the issue. Music fades out after teleport when on the same 
parcel from around 30m to 3000m. (Does not repro at shorter distances, 
like 10m to the left)

-- 
kind regards,
Cinder
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Tutorial needed on TPV viewer-side AOs

[Cinder Roxley]

On 4/10/2012 5:01 PM, Henri Beauchamp wrote:
 On Tue, 10 Apr 2012 09:01:24 -0400, Oz Linden (Scott Lawrence) wrote:

 I'd like to get a tutorial on how the AOs built into viewers work - what
 inputs do they use, and how do they set the animations they set.

 Would someone who's got deep know-how on this either write up one for me
 (or point me to one if it exists), or make some time to go over it with
 me interactively?
 It would be better implementing a server-side AO (with the viewer only
 transmitting the replacement animation UUIDs to the server, for example
 via a capability), because the current viewer-side AOs simply duplicate
 what scripted AOs are doing (so they are not really better regarding
 animations priority conflicts, etc) but lack the capability offered by
 (good) scripted AOs to be auto-switched on and off via the Lockmeister
 booton/bootoff commands which allow for cooperation between AOs and
 with device you sit onto and that want to play their own anim instead of
 the AO's.

 I really hope a proper server side AO feature is to be implemented...
I concur. Client-side ao's are also plagued by issues with sim crossings 
and not releasing animations when they ought to that scripted ao's do 
not have. Personally, I'd always dreamed for a way to wear an 
animation on an animation slot (like with attachments) to replace or 
override the default.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Tutorial needed on TPV viewer-side AOs

[Cinder Roxley]

On 4/13/2012 10:16 AM, Oz Linden (Scott Lawrence) wrote:
 On 2012-04-12 17:50 , glen wrote:
 On Thu, 2012-04-12 at 14:09 -0700, Ann Otoole wrote:
 Thankfully the previously bad aos are not so bad now. If a client side
 AO cannot perform what Oracul and/or Vista AOs do then it is a total
 waste of time to bother with the client side code. In order to do
 client side AOs requires AO expertise. Period. Don't even bother if
 you don't have it. Because it will be a waste and people will still
 use AOs.
 Agree. I'm one of the ones who's written a scripted AO. I tried the
 client-side AO in Firestorm and went back to my own because of the
 feature set. A server-side AO would like be even worse.

 Ok... so those are nice opinions to have, but you're not succeeding in
 educating me... what is it that makes these better or worse?

 What do they do or not do that differentiates one from another?
For the most part, the only advantages of client-side ao are those that 
bypass lsl limitations such as working in areas where Run Scripts is 
disabled and being able to create very large animation sets with almost 
zero load time to switch (after their initial creation of course.) For 
example, I have an AO set with 115 stands which cycle, 15 walks, 48 
sits, 22 ground sits. This configuration is far greater than what can be 
configured in a scripted AO because of the lsl's memory limitation. This 
is probably not a typical use case, but it is an advantage.

An advantage to scripted AO's as has already been stated is that 
bundling the config, AO, and animations into one object that can be worn 
and added to specific Outfits is simpiler and conserves Inventory count. 
AFAIK, any client-side AO relies on animations being unpacked and in the 
user's inventory, sometimes creating additional object links. AFAIK, the 
only client-side AO implientation that doesn't do that relied on an 
external xml config saved to hard disk and was abused by users plugging 
UUID's of animations in which they did not own. It was subsequently 
replaced with another implementation.

To my knowledge all client-side ao's also suffer from region crossing 
bugs which, for whatever reason, cause them to misinterpret the avatar's 
state or don't release the animation they were currently in while 
crossing sims (being stuck in a falling flight state being the most 
common that I've encountered.) Scripted ao's, of course, don't suffer this.

Scripted AO's can listen for commands from other objects in the region 
more easily and interact better when they've been scripted to do so. 
There are patches floating around that add compatibility with 
lockmeister and whatnot, but there is no standard viewer api for such cases.

kind regards,
   -Cinder
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

[opensource-dev] Adding a library

[Cinder Roxley]

Hello!

I've been working on a patch for VWR-2125 Include EXIF data in snapshot 
downloads and it's just about working. However, I've done this without 
properly adding the library to use (I used libexif 0.6.20 found in 
cygwin, and just included it that way.) I'm not that familiar with cmake 
and I'm not even sure including the entire library would be best since 
the viewer doesn't really need to decode exif tags, just write them. I 
was also not sure whether or not there's already a library included with 
the viewer to write exif tags. This just seemed to be easiest.

So I guess I'm just looking for feedback on how to best include the 
library or rewrite functionality into the viewer itself. Apologies if 
this e-mail seems a bit scattered. It's the weekend and I've had a few 
drinks.

-- 
Kind regards,
-Cinder
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Adding a library

[Cinder Roxley]

Yes, exactly.

On 4/14/2012 9:06 PM, Nicky Perian wrote:

Assuming you are on windows and linking
cyg*exif-*12.dll into the viewer. Is that correct?


*From:* Cinder Roxley cin...@cinderblocks.biz
*To:* opensource-dev@lists.secondlife.com
*Sent:* Sunday, April 15, 2012 10:39 AM
*Subject:* [opensource-dev] Adding a library

Hello!

I've been working on a patch for VWR-2125 Include EXIF data in
snapshot
downloads and it's just about working. However, I've done this
without
properly adding the library to use (I used libexif 0.6.20 found in
cygwin, and just included it that way.) I'm not that familiar with
cmake
and I'm not even sure including the entire library would be best
since
the viewer doesn't really need to decode exif tags, just write
them. I
was also not sure whether or not there's already a library
included with
the viewer to write exif tags. This just seemed to be easiest.

So I guess I'm just looking for feedback on how to best include the
library or rewrite functionality into the viewer itself. Apologies if
this e-mail seems a bit scattered. It's the weekend and I've had a
few
drinks.

-- 
Kind regards,

-Cinder
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated
posting privileges




___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] 'Frame load interrupted by policy change' error for youtube video playback on a prim

[Cinder Roxley]

On 4/24/2012 12:19 PM, Moriz Gupte wrote:
 Hello,
 I just bought a new laptop and found that I could no longer playback
 video a prim. Everything works fine on my older laptops. I tried to
 search the community forums and so on ... and found that some people
 are experiencing similar issues which appear to be unresolved.
 Tried turning off firewalls etc..., clearning cache, reinstalling
 Flash player plugin and still no luck. I now also recall that a few of
 my colleagues had similar issues but they too could  not fix the
 problem, but they lost interest interest in Second Life so not a big
 issue for them anyway.
 Please help with pointers to solutions if you can.
 R
Hello! Which version of the SL viewer are you running and which platform 
are you running it on? I'm not sure the status of this as far as the SL 
viewer, but this is a bug with QTWebkit which the viewer depends on. 
https://bugs.webkit.org/show_bug.cgi?id=35656 If you have control of the 
webpage you are trying to display the video from, you can try renaming 
the page with an .html extension, and it should work. Jira seems to be 
down at the moment, but you may want to give some more detail about 
which viewer version you're using and the URL you are trying to display 
on the prim so it can be reproduced, understood, and fixed.

-- 
Kind regards,
Cinder
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Using Boost the viewer

[Cinder Roxley]

Hi!

The warnings in your build log relate to the error_code library in the 
Boost.System library. They're a side effect of the design decision made 
in the library which I will spare you details of (more info on it here: 
http://www.boost.org/doc/libs/1_41_0/libs/system/doc/reference.html#Class-error%5Fcategory 
)


If you aren't using the definitions you're getting the warnings about, 
try including the following lines of code before including the header file:


|#ifndef  BOOST_SYSTEM_NO_DEPRECATED
#define  BOOST_SYSTEM_NO_DEPRECATED1
#endif|

Kind regards,
Cinder Roxley

On 7/12/2012 5:09 AM, CJ Davies wrote:

I'm trying to get the viewer to read data from an Arduino via serial
(long story...). I'm using Boost for this.

So far I have changed the Boost prebuilt in autobuild.xml to this one
after a conversation with LightDrake on #opensl on freenode;

https://bitbucket.org/LightDrake/public-libs/downloads/boost-1.45.0-linux-20120213.tar.bz2

then added a boost_thread line to indra/cmake/Boost.cmake  added
${BOOST_SYSTEM_LIBRARY}  ${BOOST_THREAD_LIBRARY} to
indra/newview/CMakeLists.txt.

This allows the .h  .cpp files that implement the serial functionality
to be added to the build  successfully built, however when I try to
include one of these header files in llviewerjoystick.h the build fails
spectacularly.

This is the file that I am trying to #include in llviewerjoystick.h --
http://paste2.org/p/2071086

This is the output of the build -- http://paste2.org/p/2071087

I assume there isn't actually anything wrong with the code  I'm just
still missing part of Boost that I need. Can anybody help me identify this?

Regards,
CJ Davies
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges



___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] MOSES Office Hours Screenshot(s) 26-Oct-2012

[Cinder Roxley]

From the Wordpress.com Terms of Service:

Responsibility of Contributors. By making Content available, you represent and 
warrant that:…

...your blog is not getting advertised via unwanted electronic messages such as 
spam links on newsgroups, email lists, other blogs and web sites, and similar 
unsolicited promotional methods…

http://en.wordpress.com/report-spam/

On Oct 26, 2012, at 3:30 PM, Sitearm site...@gmail.com wrote:

 
 http://sitearm.wordpress.com/2012/10/26/moses-office-hours-screenshots-26-oct-2012/

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] BUG-1610: Current development source does not build with glibc 2.17

[Cinder Roxley]

Lance,

I use a workaround on OSX Mountain Lion which fails with the same error, 
and it might work for you too. Patch attached and as follows:


START---

# HG changeset patch
# User Cinder Roxley cin...@cinderblocks.biz
# Date 1360441583 25200
# Node ID e3673fb4cbfafe3c0a7c2a88851316fb60422254
# Parent  7ccbd027c9b098d42632ae17f281804a0f8a5cb5
Yacc fix for Mountain Lion

diff --git a/indra/lscript/lscript_compile/indra.l 
b/indra/lscript/lscript_compile/indra.l

--- a/indra/lscript/lscript_compile/indra.l
+++ b/indra/lscript/lscript_compile/indra.l
@@ -62,9 +62,9 @@
 #define ECHO do { } while (0)

 #if defined(__cplusplus)
-extern C { int yylex( void ); }
-extern C { int yyparse( void ); }
-extern C { int yyerror(const char *fmt, ...); }
+extern int yylex( void );
+extern int yyparse( void );
+extern int yyerror(const char *fmt, ...);
 #endif

 %}
diff --git a/indra/lscript/lscript_compile/indra.y 
b/indra/lscript/lscript_compile/indra.y

--- a/indra/lscript/lscript_compile/indra.y
+++ b/indra/lscript/lscript_compile/indra.y
@@ -3,7 +3,7 @@
#include lscript_tree.h

 #ifdef __cplusplus
-extern C {
+extern //C {
 #endif

int yylex(void);
@@ -21,7 +21,7 @@
#endif

 #ifdef __cplusplus
-}
+//}
 #endif
 %}

--END---

Kind regards,
Cinder Roxley

On 9 Feb 2013, at 12:40, Lance Corrimal wrote:


This is where I'm stuck now:


[  363s] /home/abuild/rpmbuild/BUILD/viewer-development/build-linux-
i686/lscript/lscript_compile/indra.y.hpp:176: error: previous 
declaration of

'int yyparse()' with 'C++' linkage
[  363s] /home/abuild/rpmbuild/BUILD/viewer-
development/indra/lscript/lscript_compile/indra.l:61: error: conflicts 
with

new declaration with 'C' linkage



any ideas?


cheers,
LC


Am Samstag, 9. Februar 2013, 13:56:56 schrieb Lance Corrimal:

Am Samstag, 9. Februar 2013, 04:41:10 schrieb Nicky Perian:

Would you mark BUG-1610 as a duplicate of OPEN-164?


definitely not. the problem that I had is because glibc 2.17 brings 
the

siginfo_t definition by itself... definitely something else.

What I'd think suitable would be a meta bug that says Current 
source
fails to build on newer linux distributions that would depend on 
BOTH

open-164 and bug-1610 and possibly others.


besides, I found two more build breakages on openSUSE 12.3 by now...
One might be fixable by just adding the right build requirement tho.

cheers,
LC




From: Lance Corrimal lance.corri...@eregion.de

To: opensource-dev@lists.secondlife.com
Sent: Saturday, February 9, 2013 5:00 AM
Subject: Re: [opensource-dev] BUG-1610: Current development source 
does

not
build with glibc 2.17

Thanks... I haven't even gotten that far, my build died way 
earlier, with

a
conflicting declaration for siginfo_t, but I got that tackled now.


cheers,
LC

Am Samstag, 9. Februar 2013, 02:34:09 schrieb Nicky Perian:

https://bitbucket.org/NickyP/kokua-3.4.4/commits/22012aa8388a38fb657113
b0
db3 effcfdaa3a8d9




From: Lance Corrimal lance.corri...@eregion.de

To: opensource-dev@lists.secondlife.com
Sent: Saturday, February 9, 2013 4:13 AM
Subject: [opensource-dev] BUG-1610: Current development source 
does

not
build with glibc 2.17

hi,

does anyone have anything for me about fixing up the source to 
build

with
glibc 2.17? I'm getting ready for openSUSE 12.3 here, and that 
uses

the
newer glibc...

for details, see BUG-1610

cheers,

LC

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated 
posting

privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting 
privileges
# HG changeset patch
# User Cinder Roxley cin...@cinderblocks.biz
# Date 1360441583 25200
# Node ID e3673fb4cbfafe3c0a7c2a88851316fb60422254
# Parent  7ccbd027c9b098d42632ae17f281804a0f8a5cb5
Yacc fix for Mountain Lion

diff --git a/indra/lscript/lscript_compile/indra.l 
b/indra/lscript/lscript_compile/indra.l
--- a/indra/lscript/lscript_compile/indra.l
+++ b/indra/lscript/lscript_compile/indra.l
@@ -62,9 +62,9 @@
 #define ECHO do { } while (0)
 
 #if defined(__cplusplus)
-extern C { int yylex( void

Re: [opensource-dev] BUG-1610: Current development source does not build with glibc 2.17

[Cinder Roxley]

Here's a better fix than the other one I posted earlier… Don't use 
extern C.

START-

diff --git a/indra/lscript/lscript_compile/indra.l 
b/indra/lscript/lscript_compile/indra.l
--- a/indra/lscript/lscript_compile/indra.l
+++ b/indra/lscript/lscript_compile/indra.l
@@ -61,11 +61,9 @@

  #define ECHO do { } while (0)

-#if defined(__cplusplus)
-extern C { int yylex( void ); }
-extern C { int yyparse( void ); }
-extern C { int yyerror(const char *fmt, ...); }
-#endif
+int yylex( void );
+int yyparse( void );
+int yyerror(const char *fmt, ...);

  %}

diff --git a/indra/lscript/lscript_compile/indra.y 
b/indra/lscript/lscript_compile/indra.y
--- a/indra/lscript/lscript_compile/indra.y
+++ b/indra/lscript/lscript_compile/indra.y
@@ -2,10 +2,6 @@
#include linden_common.h
#include lscript_tree.h

-#ifdef __cplusplus
-extern C {
-#endif
-
int yylex(void);
int yyparse( void );
int yyerror(const char *fmt, ...);
@@ -19,10 +15,6 @@
#pragma warning (disable : 4702) // warning C4702: unreachable code
#pragma warning( disable : 4065 )   // warning: switch statement 
contains 'default' but no 'case' labels
#endif
-
-#ifdef __cplusplus
-}
-#endif
  %}

  %union

--END---
Kind regards,
Cinder Roxley

 On 9 Feb 2013, at 12:40, Lance Corrimal wrote:

 This is where I'm stuck now:


 [  363s] /home/abuild/rpmbuild/BUILD/viewer-development/build-linux-
 i686/lscript/lscript_compile/indra.y.hpp:176: error: previous 
 declaration of
 'int yyparse()' with 'C++' linkage
 [  363s] /home/abuild/rpmbuild/BUILD/viewer-
 development/indra/lscript/lscript_compile/indra.l:61: error: 
 conflicts with
 new declaration with 'C' linkage



 any ideas?


 cheers,
 LC


 Am Samstag, 9. Februar 2013, 13:56:56 schrieb Lance Corrimal:
 Am Samstag, 9. Februar 2013, 04:41:10 schrieb Nicky Perian:
 Would you mark BUG-1610 as a duplicate of OPEN-164?

 definitely not. the problem that I had is because glibc 2.17 brings 
 the
 siginfo_t definition by itself... definitely something else.

 What I'd think suitable would be a meta bug that says Current 
 source
 fails to build on newer linux distributions that would depend on 
 BOTH
 open-164 and bug-1610 and possibly others.


 besides, I found two more build breakages on openSUSE 12.3 by now...
 One might be fixable by just adding the right build requirement tho.

 cheers,
 LC

 

 From: Lance Corrimal lance.corri...@eregion.de

 To: opensource-dev@lists.secondlife.com
 Sent: Saturday, February 9, 2013 5:00 AM
 Subject: Re: [opensource-dev] BUG-1610: Current development source 
 does
 not
 build with glibc 2.17

 Thanks... I haven't even gotten that far, my build died way 
 earlier, with
 a
 conflicting declaration for siginfo_t, but I got that tackled now.


 cheers,
 LC

 Am Samstag, 9. Februar 2013, 02:34:09 schrieb Nicky Perian:
 https://bitbucket.org/NickyP/kokua-3.4.4/commits/22012aa8388a38fb657113
 b0
 db3 effcfdaa3a8d9

 

 From: Lance Corrimal lance.corri...@eregion.de

 To: opensource-dev@lists.secondlife.com
 Sent: Saturday, February 9, 2013 4:13 AM
 Subject: [opensource-dev] BUG-1610: Current development source 
 does
 not
 build with glibc 2.17

 hi,

 does anyone have anything for me about fixing up the source to 
 build
 with
 glibc 2.17? I'm getting ready for openSUSE 12.3 here, and that 
 uses
 the
 newer glibc...

 for details, see BUG-1610

 cheers,

 LC

 ___
 Policies and (un)subscribe information available here:
 http://wiki.secondlife.com/wiki/OpenSource-Dev
 Please read the policies before posting to keep unmoderated 
 posting
 privileges

 ___
 Policies and (un)subscribe information available here:
 http://wiki.secondlife.com/wiki/OpenSource-Dev
 Please read the policies before posting to keep unmoderated 
 posting
 privileges

 ___
 Policies and (un)subscribe information available here:
 http://wiki.secondlife.com/wiki/OpenSource-Dev
 Please read the policies before posting to keep unmoderated posting
 privileges
 ___
 Policies and (un)subscribe information available here:
 http://wiki.secondlife.com/wiki/OpenSource-Dev
 Please read the policies before posting to keep unmoderated posting 
 privileges

 [yacc-hacc.diff]
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 4, 2015 at 2:18:55 PM, Nicky Perian (nickyper...@gmail.com) wrote:
I know I'm not the sharpest knife in the draw especially when it come to 
python. But, if I am reading this code snip from SingularityViewer develop.py
correctly a command line build using devenv.com is executed each time it finds 
visual studio and does not use it for Express because it isn't present in 
Express.
{code}
        environment = self.find_visual_studio()
        if environment == '':
            environment = self.find_visual_studio_express()
            if environment == '':
                environment = self.find_visual_studio_express_single()
                if environment == '':
                    print  sys.stderr, Something went very wrong during 
build stage, could not find a Visual Studio?
                else:
                    build_dirs=self.build_dirs()
                    print  sys.stderr, \nSolution generation complete, it 
can can now be found in:, build_dirs[0]    
                    print  sys.stderr, \nAs you are using an Express Visual 
Studio, the build step cannot be automated
                    print  sys.stderr, \nPlease see 
https://wiki.secondlife.com/wiki/Microsoft_Visual_Studio#Extra_steps_for_Visual_Studio_Express_editions
 for Visual Studio Express specific information
                    exit(0)
    
Either you’ve intentionally removed msbuild from this block of code or your 
sources are well out of date. -_-

https://bitbucket.org/lkalif/singularityviewer/src/f0b18e52a3d977c2c19f2a6b4e50292acb5941f2/indra/develop.py?at=master#cl-675

-- 
Cinder Roxley
Sent with Airmail


        # devenv.com is CLI friendly, devenv.exe... not so much.
        return ('%sdevenv.com %s.sln /build %s' % 
                (self.find_visual_studio(), self.project_name, self.build_type))
{code}___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 6, 2015 at 4:54:10 PM, Nicky Perian (nickyper...@gmail.com) wrote:

The issue was in fact paths. Operating with in the Developer Command Prompt for 
VS2013 requires a path of:
C:\cwgwin64\bin;C:\cygwin64\usr\bin

Reason: coreutils which contains printf and a host of other bash commands 
resides in C:\cygwin64\bin. Flex and Bison are installed in 
C:\cygwin64\usr\bin. 

So from Developer Command Prompt for VS2013 you can reach printf and flex and 
bison.

Now the kicker if you now open the Cygwin64  Terminal the aforementioned 
C:\cygwin64\bin and C:\Cygwin64\usr\bin are mangled by cygwin64's remake of the 
path into the form /cygdrive/c/...
and writes usr/bin;usr/bin in the path which breaks the path in
Cygwin64  Terminal.

If you don't include C:\cygwin64\bin and C:\Cygwin64\usr\bin in the path then 
cygwin64 is smart enough to figure it out and printf works in the Cygwin64  
Terminal but, now it is broken in the Developer Command Prompt for VS2013. 

I'm thinking a workaround for this is to begin each windows batch file with a 
path command the appends C:\cwgwin64\bin;C:\cygwin64\usr\bin but do not put 
those in the system path.
Hi,

It would simply things greatly to remove lscript from the viewer completely, 
thereby removing the dependency on flex/bison for building. lscript is unused 
other than a few const values. Several tpvs have dropped lscript from the 
build. (I’ve personally done that in two different tpv projects.)

Is this something Linden Lab would be willing to consider? I’d be happy to do 
the work to remove it if it is.

-- 
Cinder Roxley
Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 7, 2015 at 3:13:31 PM, Oz Linden (Scott Lawrence) 
(o...@lindenlab.com) wrote:
On 2015-02-06 16:12 , Cinder Roxley wrote:

It would simply things greatly to remove lscript from the viewer completely, 
thereby removing the dependency on flex/bison for building. lscript is unused 
other than a few const values. Several tpvs have dropped lscript from the 
build. (I’ve personally done that in two different tpv projects.)
Is this something Linden Lab would be willing to consider? I’d be happy to do 
the work to remove it if it is.


What would the end user effect be?


The change would be transparent for the user other than saving them a small 
amount of space on their hard drive. I have no idea what the final size of 
lscript is when compiled, but it’s basically all dead and broken code and has 
been for several years.

-- 
Cinder Roxley
Sent with Airmail___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 7, 2015 at 6:09:54 PM, Argent Stonecutter (secret.arg...@gmail.com) 
wrote:
On 2015-02-07, at 16:24, Cinder Roxley cin...@alchemyviewer.org wrote: 
 On February 7, 2015 at 3:13:31 PM, Oz Linden (Scott Lawrence) 
 (o...@lindenlab.com) wrote: 
 On 2015-02-06 16:12 , Cinder Roxley wrote: 
 It would simply things greatly to remove lscript from the viewer 
 completely, thereby removing the dependency on flex/bison for building. 
 lscript is unused other than a few const values. Several tpvs have dropped 
 lscript from the build. (I’ve personally done that in two different tpv 
 projects.) 
 Is this something Linden Lab would be willing to consider? I’d be happy to 
 do the work to remove it if it is. 

 What would the end user effect be? 

 The change would be transparent for the user other than saving them a small 
 amount of space on their hard drive. I have no idea what the final size of 
 lscript is when compiled, but it’s basically all dead and broken code and has 
 been for several years. 

Doesn't the script editor use it for syntax coloring? Or has it dropped back to 
using a more simplistic algorithm? 
Syntax highlighting is handled by LLSyntaxID now. This allows it to be 
dynamically updated by a region cap and not hardcoded into the viewer.

-- 
Cinder Roxley
Sent with Airmail___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer-tools-update build environment setup procedures

[Cinder Roxley]

On February 11, 2015 at 5:33:19 PM, Nicky Perian (nickyper...@gmail.com) wrote:
I've not seen specific procedures to set the build environment for Mac. May be 
that I just missed it along the way.

Is there a build environment wiki for Mac?

Nothing has changed in that regard except upgrading to Xcode 6. 
https://wiki.secondlife.com/wiki/Compiling_the_viewer_(Mac_OS_X)

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

Hi,

All the tests pass for me using the open source build. If they aren’t passing 
it would be good to file a jira or post the failures here so they can be fixed.

-- 
Cinder Roxley
Sent with Airmail

On January 30, 2015 at 2:41:09 PM, Oz Linden (Scott Lawrence) 
(o...@lindenlab.com) wrote:

On 2015-01-30 11:51 , Tank Master wrote:
You are getting an error because that isn't a valid command.  Use autobuild 
configure -c RelWithDebInfoOS OR autobuild configure -c ReleaseOS  Also, 
This will still fail do to tests.  I used: autobuild configure -c ReleaseOS -- 
-DPACKAGE:BOOL=OFF -DLL_TESTS:BOOL=OFF to turn off both packaging and tests.

The tests should work - they do for us, though I suppose some may depend on 
libs not in open source.

--
Oz Linden (Scott Lawrence) | Engineering Director, Second Life
Email or Hangouts o...@lindenlab.com | Second Life Oz Linden
Linden Lab | Makers of Shared Creative Spaces
Check out what we're working on!
___  
Policies and (un)subscribe information available here:  
http://wiki.secondlife.com/wiki/OpenSource-Dev  
Please read the policies before posting to keep unmoderated posting privileges___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

Hi,

I thought the plan was to only support VS2013 Community Edition and support for 
Express since Community Edition is, for all intents and purposes, free and 
without any further restrictions than the LGPL already lays out? There’s really 
absolutely no reason for using Express since all qualifications are met for 
both individuals and organizations building the viewer because the source code 
has already been released under an OSI-approved license that we all must adhere 
to anyways.

In short, why support both methods when one is vastly superior with no 
disadvantages?

-- 
Cinder Roxley
Sent with Airmail

On January 31, 2015 at 9:39:44 AM, Nicky Perian (nickyper...@yahoo.com) wrote:

[#OPEN-290] Windows Toolchain update - Second Life Bug Tracker
 
 
 
 
 
 
[#OPEN-290] Windows Toolchain update - Second Life Bug Tracker
Document and offer patches for Windows Platform open source building.
View on jira.secondlife.com
Preview by Yahoo
 
Two  methods for command line building have been tested each starting with a 
fresh build-vc120 directory, with tests and package on.
Machine spec is Intel i7 all 8 cores running with 6GB memory.

autobuild configure -c ReleaseOSCE and autobuild build -c ReleaseOSCE
Calls Devenv.exe which is the build program for VS2013 Community Edition
Build completed in 53 minutes.

autobuild configure -c ReleaseOS and autobuild build -c ReleaseOS
Calls MSBuild.exe which is the build program available in VS2013 Express 
Edition AND VS2013 Community Edition.
Build completes in 77 minutes.

Should we provide open source developers both methods?

Patch is here:

NickyP / viewer-tools-update-OPEN-290 / commit / 96461e3be52e — Bitbucket
 
 
 
 
 
 
NickyP / viewer-tools-update-OPEN-290 / commit / 96461e3be52e — Bitbucket
keyname/key stringRelWithDebInfoOS/string
View on bitbucket.org
Preview by Yahoo
 



From: Oz Linden (Scott Lawrence) o...@lindenlab.com
To: opensource-dev@lists.secondlife.com
Sent: Friday, January 30, 2015 3:41 PM
Subject: Re: [opensource-dev] Viewer Tools Upgrades - with a call for help



On 2015-01-30 11:51 , Tank Master wrote:
You are getting an error because that isn't a valid command.  Use autobuild 
configure -c RelWithDebInfoOS OR autobuild configure -c ReleaseOS  Also, 
This will still fail do to tests.  I used: autobuild configure -c ReleaseOS -- 
-DPACKAGE:BOOL=OFF -DLL_TESTS:BOOL=OFF to turn off both packaging and tests.


The tests should work - they do for us, though I suppose some may depend on 
libs not in open source.

--
Oz Linden (Scott Lawrence) | Engineering Director, Second Life
Email or Hangouts o...@lindenlab.com | Second Life Oz Linden
Linden Lab | Makers of Shared Creative Spaces
Check out what we're working on!

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges


___  
Policies and (un)subscribe information available here:  
http://wiki.secondlife.com/wiki/OpenSource-Dev  
Please read the policies before posting to keep unmoderated posting privileges___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Jonathan's attempt to rearrange the graphics preferences

[Cinder Roxley]

Hi,

Is there any justification at all for rearranging everything into a single 
scrolling panel, or is this just change for the sake of change? There doesn’t 
appear to be any benefit to the user in making the change, and I haven’t seen 
any report that the current layout is hindering usability. It has certainly 
worked for people, as is, for the past ten years. 

There are also plenty of bugs that have been described in detail that have been 
given no acknowledgement on the jira issue itself. It doesn’t surprise me at 
all that Martin doesn’t want to spend time elaborating on something that very 
likely will be ignored especially given that a fair amount of them are 
regressions directly related to UI shuffling.

-- 
Cinder Roxley
Sent with Airmail

On January 24, 2015 at 6:04:20 AM, Jonathan Welch (jhwe...@gmail.com) wrote:

Martin, I am sorry you are unhappy with LL's new UX design but this is  
one of those cases where no matter what is done someone will be  
unhappy.  

I have an idea on how to address the I want to see everything at  
once issue and will send it to Oz for evaluation.  

How about describing what those bugs are (I am sure there are some to be 
found)?  

-jonathan  
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Raspberry Pi 2?

[Cinder Roxley]

On February 10, 2015 at 12:19:09 PM, Jack Lang (j...@raspberrypi.org) wrote:
 

Anyone porting a viewer to Raspberry Pi 2 (quad core Arm7)? Any hints ?

Jack Lang

Raspberry Pi Foundation



If they’d stay in stock long enough for me to get through checkout, I’d have 
one to try it out. ;)

Gonna take a lot of GL changes to get it running on ARM though.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Raspberry Pi 2?

[Cinder Roxley]

On February 10, 2015 at 12:57:11 PM, Dana Moore (dana.virt...@gmail.com) wrote:
I have a slightly different question which is does anyone know of any high 
barriers to porting the _server_ (for example Doug Maxwell’s MOSES / OpenSim 
implementation) to a PI 2.
I have a feeling it ought to be a handsome server, especially in a server grid 
(granted no heavyweight GPU capabilities, but still … ). BTW, I _think_ the 
spec is an ARM11 quad core for the 2 IIRC.
I’ve run opensim from a Raspberry Pi B+ and seen others do it as well, so yes, 
it should be able to run MIAB. At least the simulator portion. I’m not entirely 
sure what MIAB consists of, (like if it includes the voice server or not).

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 25, 2015 at 6:49:56 AM, Henri Beauchamp (sl...@free.fr) wrote:
On Tue, 24 Feb 2015 19:03:29 -0800, Darien Caldwell wrote: 

 I'm pretty sure the reasoning behind LL moving all the multimedia to a 
 plugin system was so that 3rd parties could write their own plugins. 

I doubt it... 
Yeah, that was kind of the whole point of the change.

http://wiki.secondlife.com/wiki/Linden_Lab_Official:Media_Rendering_Plugin_System

Though it doesn’t appear that work in this was ever completed. (You still have 
the build the entire viewer to build and test a media rendering plugin. You 
need to stay relatively in sync upstream so releasing these plugins can’t 
really be independent from viewer-release and can’t realistically be released 
and work across the board with tpv’s based on Snowstorm.)

Outside of that, there are many other factors as to why nobody has developed a 
media plugin. For one, licensing costs money, and us lowly, non-commercial 
developers cannot afford or do not wish to shell out money to pay for codec 
licenses like mp3.

-- 
Cinder Roxley
Sent with Airmail___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Tools Upgrades - with a call for help

[Cinder Roxley]

On February 21, 2015 at 5:50:42 PM, Jonathan Welch (jhwe...@gmail.com) wrote:
I just started up VS2013 and was exploring some of the windows and 
menus and discovered that my trial license is only good for 30 days. 
The instructions should cover what to do to obtain a long-term 
license. 

Warning window: http://i.imgur.com/RwTzJxH.png 
The warning window explains what to do to obtain a license. (You need to sign 
in with a valid Microsoft account.) In my opinion, acquiring licenses for the 
build tools falls out of scope. Don’t want it to get too long or everyone 
starts skipping steps.

-- 
Cinder Roxley
Sent with Airmail___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac build instructions completely broken

[Cinder Roxley]

On April 10, 2015 at 5:45:46 PM, Chris Nelson (cnelson1...@gmail.com) wrote:
Has anyone in the dev community attempted to build the sl viewer from scratch 
using the instructions from  
http://wiki.secondlife.com/wiki/Compiling_the_viewer_(Mac_OS_X)?
I have been going through the process over the last few days and it seems 
pretty much completely broken.
I have managed to overcome most of the hurdles up until now, but I seem to have 
hit a show stopper with the ogg_vorbis package download problem I posted 
earlier.


I build the viewer on a regular basis on OS X. The problem you describe is 
likely because you are using Autobuild 1.0 but building the viewer-release 
viewer. Autobuild 1.0 is not compatible with viewer-release as Autobuild 1.0 
has a stricter package versioning scheme. Hence, your failure.

Your options are these:

1) Remove the version of autobuild you have and download the older autobuild 
version 0.9 from https://bitbucket.org/lindenlab/autobuild using setup.py to 
install it and build the viewer.

2) Build the viewer-tools-update viewer which is configured to use the newer 
version of autobuild. That can be found here: 
https://bitbucket.org/lindenlab/viewer-tools-update 

Best of luck.

— 
Cinder Roxley
Sent with Airmail___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Fwd: Mac build instructions completely broken

[Cinder Roxley]

On April 11, 2015 at 6:36:14 PM, Chris Nelson (cnelson1...@gmail.com) wrote:

-- Forwarded message --
From: Chris Nelson cnelson1...@gmail.com
Date: Sun, Apr 12, 2015 at 9:16 AM
Subject: Re: [opensource-dev] Mac build instructions completely broken
To: Cinder Roxley cin...@alchemyviewer.org


Thanks Cinder,

That did the trick!
Looks like I am now stuck with not having the OS X 10.8 sdk - (I only have 
10.9) and gcc 4.2(?)
Getting the message:
error: There is no SDK with the name or path 
'/Users/cnelson/Documents/GitRepositories/SecondLifeViewer/indra/macosx10.8'

I am using xcode 6.1.1. Thought I read that sl viewer would compile with this 
version.

Cheers,

Chris
If you’re building viewer-release, you’re in trouble building current 
viewer-release, you still need llvm-gcc 4.2 which isn’t included in Xcode 5 or 
6. Apple went to great lengths to make it hard to replace, and although it is 
possible to readd it (see my Xcode compiler plugin collection at 
https://bitbucket.org/cinderblocks/xcode-compiler-plugins/wiki/Home) It’s a lot 
more effort than just installing Xcode 4.6.3 and using that instead for 
viewer-release.

If you want to use Xcode 6, you need to build out of the viewer-tools-update 
repository, (and from what I’ve heard, that will be promoted to the canonical 
viewer-release soon.) Xcode 5 still contains 10.8 SDK iirc, but I just checked 
and viewer-tools-update targets 10.9 now. So for modern tooling, you’re better 
off just building from https://bitbucket.org/lindenlab/viewer-tools-update with 
the newer autobuild.

-- 
Cinder Roxley
Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac OS X Metal support

[Cinder Roxley]

MetalGL is an OpenGL ES 2.0 implementation that uses the Metal framework, so 
you’d be trading off performance and rewriting a pretty big chunk of the 
rendering code for not much gain. Furthermore, MetalGL’s license is 
incompatible with the LGPL.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On December 13, 2015 at 7:41:02 PM, Gistya Eusebio (gis...@gmail.com 
<mailto:gis...@gmail.com> ) wrote:

 
I heard that there is a project called MetalGL that aims to compile OpenGL 
shaders using Metal... might make a port easier. Either way it would be awesome 
to see an SL client written from the ground up in Swift running Metal ... :D 
Maybe for SL 2 eh?

On Fri, Oct 30, 2015 at 6:43 PM, Oz Linden (Scott Lawrence) <o...@lindenlab.com 
<mailto:o...@lindenlab.com> > wrote:

On 2015-10-24 05:14 , Laurent Bechir wrote:

Hello

Do you have some plan concerning the support of Metal in the viewer ?



Not at this time.

--
Oz Linden (Scott Lawrence) | Engineering Director, Second Life
Email or Hangouts o...@lindenlab.com <mailto:o...@lindenlab.com> | Second 
LifeOz Linden
Linden Lab | Makers of Shared Creative Spaces <http://lindenlab.com/> 
Check out what we're working on! <http://lindenlab.com/products> 

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev 
<http://wiki.secondlife.com/wiki/OpenSource-Dev> 
Please read the policies before posting to keep unmoderated posting privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Quicktime on Mac

[Cinder Roxley]

On June 15, 2016 at 7:22:27 AM, Oz Linden (Scott Lawrence) (o...@lindenlab.com 
<mailto:o...@lindenlab.com> ) wrote:
 
On 2016-06-13 18:18 , Cinder Roxley wrote:
Hello,

As the replacement of quicktime was being discussed here recently, I’m just 
checking out Xcode 8 and Quicktime has been completely removed from macOS 10.12 
SDK. Haven’t installed Sierra yet to see, but there’s a good chance the plugin 
is broken in it. (You certainly can’t compile.)

This not only affects the quicktime plugin, but the mac volume catcher in the 
viewer.


Our plan is to address the quicktime plugin for the Mac in the 64bit project; 
the focus for the moment with respect to Quicktime is replacing it on Windows
Shouldn’t be an issue then, I would think. macOS Sierra isn’t due out until 
fall.

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

[opensource-dev] Quicktime on Mac

[Cinder Roxley]

Hello,

As the replacement of quicktime was being discussed here recently, I’m just 
checking out Xcode 8 and Quicktime has been completely removed from macOS 10.12 
SDK. Haven’t installed Sierra yet to see, but there’s a good chance the plugin 
is broken in it. (You certainly can’t compile.)

This not only affects the quicktime plugin, but the mac volume catcher in the 
viewer.

-- 
Cinder Roxley
Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Quicktime

[Cinder Roxley]

On April 21, 2016 at 9:39:31 AM, Henri Beauchamp (sl...@free.fr 
<mailto:sl...@free.fr> ) wrote:
 
On Thu, 21 Apr 2016 13:18:37 +, Cinder Roxley wrote: 

>> Like I already wrote earlier, the way to go is to use the gstreamer SDK  
>> for Windows and get a gstreamer plugin compiled for the latter.  
> 
> With all due respect, gstreamer is a major pain to build on Windows 

It's a pain to build under Linux too :-D But CEF is even more painful 
to build, so... 

Plus that "pain" is to go through only once and for all, to build the 
pre-built package that will then be used to build all the viewers 
(painlessly). 
True.

> and runs afoul of dozens of patents and licenses. It’s fine if you’re 
> building from source for linux, but it’s a lawsuit waiting to happen for 
> commercial software if you want to play any “standard” media format like 
> h264 or mp3. 

Please, elaborate and give pointers. There are dozens (and probably closer 
to dozens of dozens) of Linux and *BSD distributions (i.e. binary, ready 
to run OS+software, commercial or not) providing and using gstreamer, and 
I never heard about any lawsuite related to this fact... 
These distributions either do not include the patented video formats or provide 
them in source form for building. This is addressed in gstreamer’s 
documentation:

https://gstreamer.freedesktop.org/documentation/licensing.html (Licenses of 
applications using gstreamer)

> Something like libvlc might work if you want a cross platform library 
> (but again, there are per-install royalties to use h264 so you’d still 
> be screwed on mp4.) 

Software patents are a US thingy... I'm glad the UE rejected them. MP4, 
H264 and anything involving patented protocols and formats are free to 
play in the whole world, but in the US... *If* such patents prevent to 
provide a full set of CODECs, I guess LL would have to restrict their 
number in their pre-built library package (gstreamer is not monolithic, 
it is fully modular), but TPV developers outside the US won't have to 
bother with such restsictions... 
US and Australia, not to mention Canada, United Kingdom, Germany, Japan, and 
any other country whose legal codes frown upon violation of software patents. 
Distributing a viewer that decodes the patented video compression puts the end 
user in legal jeopardy. (Albeit, there is slim to no chance of anyone being 
prosecuted for decoding h264 in a viewer, but we can’t condone breaking any 
country’s law on an official Second Life communications channel.)

> Platform-specific plugins could take advantage of the OS’s media playback 
> capabilities, without license and patent headaches. 

Perhaps for Windows. But again, Linux don't have any patent for playing 
media and no lawsuite whatsoever prevent Linux distros to be distributed 
(including in the US)... I'm still extremely doubtful about such patent 
issues for software that are only meant to *play* *existing* media files 
(i.e. media files you acquired legally and already paid any patent for). 
Linux distributions, for example Ubuntu, don’t ship with proprietary codecs, 
you must install them after the fact. (You can buy legitimate gstreamer plugins 
from the Canonical store.) Even if you build the gstreamer-plugins-ugly package 
from their repo, you’re greeted with a popup the first time you try and play a 
video with it in the US that what you’re doing may be illegal.

As far as patent infringement lawsuits, MPEG-LA, the patent holder, is one of 
the biggest patent trolls in the industry. Going even as far as billion dollar 
lawsuits against paid license holders like Microsoft.

MPEG LA sues Audiovox 
http://www.twice.com/article/257658-Audiovox_Disputes_MPEG_LA_Lawsuit.php

MPEG LA sues Alcatel Lucent 
http://www.businesswire.com/news/home/20100329006257/en/MPEG-LA-Lawsuit-Alcatel-Lucent-Settled

MPEG LA sues Apex 
https://www.allbusiness.com/legal/legal-services-litigation/5917877-1.html

MPEG LA has sued Google over V8 (for being too similar to h264, MPEG LA lost.)

-- 
Cinder Roxley
Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 23, 2016 at 1:13:54 PM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:

 
For OS X save yourself all the trouble in the world by using the OS X media 
frameworks and QuickTime. 

To create a 32-bit OS plugin based on LibVLC just to covert it to 64-bit is 
asking for trouble and weeks of headaches when you already have everything your 
need both for 32 and 64 bit QuickTime available and virtually just need to set 
a flag in the compiler. 

If you are ever to be serious developing on OS X or iOS, snap out of it and 
start to use what is available, supported and tested through and trough. The 
time you free up can be used for the rest of your projects! ;-)

The QuickTime C APIs can’t be used directly for 64-bit applications…. See 
Apple’s 64-Bit Guide for Carbon Developers. 

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 23, 2016 at 12:42:12 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:
 
This is what I propose for moving forward with the "Remove QuickTime from the 
viewer" work:

(TLDR;  Replace QuickTime plugin with one based on LibVLC and use it to play 
MPEG-4 and MP3 media URLS plus anything else we get for free. Additionally, 
turn on flags in Chromium->CEF->CEF-bin->LLCEFLib->media_plugin_cef builds that 
enable embedded media support.)
*   Remove QuickTime entirely from the viewer.
*   Replace it with a new plugin:
*   
*   Version for Windows (32 bit) using LibVLC  
*   Version for OS X (32 bit) using LibVLC
*   Ask for help from open source developer community to create a 
version for Linux using LibVLC
*   Update mime_types.xml (etc) to point old QuickTime handled media at new 
version (plus any others we think should not go to the default, CEF plugin)
*   Ask for help from the open source developer community to flip Linux 
GStreamer output since we flipped the prim media texture coordinates
*   
*   I hope this is possible - reason it was done is that both CEF 
and LibVLC need to be flipped so it seems foolish to flip everything twice.
*   Inhibit the "This file needs to be downloaded" message in CEF for media 
types we are unable to handle - replace with Alert?
Then as a separate task maybe since it's more of a feature vs. a replace 
QuickTime issue:
*   Assuming legal gives us the go ahead to turn on the CEF embedded media 
support, go ahead and update the Windows/OS X 32 bit CEF media plugins 
accordingly.
Then, once this is finished and we resume the 64 bit conversion work:
*   Create 64 bit versions of the LibVLC plugin for Windows and OS X
*   Create 64 bit versions of the media-enabled CEF plugin for Windows and 
OS X


​Unless ​ 
​anyone has any significant objections, I'll go ahead and clean up the existing 
LibVLC plugin, get it working on OS X and make a version of the CEF plugin with 
embedded media.

I have no ability to do anything for the Linux side of things so would 
appreciate help from someone with a contributor agreement.

Cheers!​
 

Sounds good to me. 

I wonder if doing vlc on darwin32 is worthwhile though. Quicktime for Mac is 
deprecated, but doesn’t exhibit the security holes Windows does. Once 64-bit is 
building, Quicktime has got to go as there is no 64-bit support anyway. Guess 
it depends on how soon darwin64 could be out the door.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 18, 2016 at 5:40:18 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:
 
Digesting all the suggestions here - thank you.

Intrigued by Nicky's suggestion, I am currently trying to build CEF directly 
via Chromium - first attempt is without the extra flags (proprietary_codecs=1 
ffmpeg_branding=Chrome).  Building the branch in use in the viewer failed with 
a bunch of errors - fixable but there were just too many. Still not sure *why* 
it fails - I would expect specifying a branch would chckout and build a tagged 
point in the repo that built.  Maybe because I'm on a slightly older Xcode and 
on 10.10 vs 10.11? 

Now trying the tip for OS X /64bit (only have my OS X box with me today) - if 
this works (on 9856 of 15438 files) then I have high hopes we can build it with 
the flags switched on for the platforms and bit widths we care about.

Do people agree that this would be the best solution?  It would, I think play 
media URLs directly in the CEF plugin like Chrome does and of course, allow us 
to support embedded media.


I think it’s a little heavy to run a browser instance to play a video.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

Gstreamer is probably your best bet just based on the friendliness of the 
developer community when asking questions and gstreamer being somewhat more 
focused on being an intuitive framework to integrate into an application. 
However, libvlc has a much simpler api which makes it faster to integrate, and 
you don’t need to worry about plugins as much, as only the codecs are modular 
(which has benefits and drawbacks, of course.) As Henri pointed out though, 
getting a functional framework to link against is a job that only happens once 
(or once every library update anyway.)

If licensing isn’t an issue, I’d go with gstreamer.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On May 16, 2016 at 2:41:20 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:

 
As many of you know, support for playing media in Second Life using QuickTime 
is being removed after Apple announced in April 2016 that they have no further 
plans to provide security updates for QuickTime for Windows. This email is 
intended to be a solicitation for feedback and review of a proposed replacement 
and hopefully some discussion around its' merits, inadequacies and possible 
alternatives.

A replacement media plugin for QuickTime needs to be created based on an 
existing media playback SDK and mime_types.xml (etc.) updated to direct 
appropriate media types (MPEG-4, MPG, MP3) at this new plugin.

The two technologies that seemed like they might work were identified as 
GStreamer and LibVLC. The former was an attractive option since that was 
already used in the Linux viewer. The existing GStreamer media plugin code was 
somewhat complex and unknown to me whereas playback of a stream in LibVLC 
seemed very straightforward and I had toyed with it in the past.

In order to get something working and provide something to base discussion on, 
I forked viewer-release and made a new viewer that had QuickTime removed and a 
new media plugin based on LibVLChere. Only the Win32 implementation is filled 
out currently until we can collectively decide if this is the right approach.

I made a Linden autobuild VLC binary package that the new media code 
(media_plugin_libvlc.cpp) consumes - currently using the latest version (2.2.3).

The limitations with this solution are:
*   I don't know how to resize the playback buffer once a stream starts 
playing so if a media plugin resize message come in, I restart the stream from 
the beginning. It is regrettable but not as big a concern as it would be for 
say, web media where resizing is much more common.
*   It doesn't appear to play back the QuickTime MOV files I have tried. 
I'm still investigating whether that is expected behavior or not.
*   If you click on a (say) link to an MPEG-4 movie from a web page that is 
rendered using the CEF plugin, the media system is (currently) not able to 
switch out the plugin implementation and play it - at the moment I think the 
CEF plugin reports that this type of file must be downloaded. As far as I know, 
this wasn't possible before though (clicking on link to QuickTime movie would 
play in same media instance).
Our legal department have given us clearance to use either GStreamer or LibVLC 
as we see fit so that is not a concern.

So what do we all think?  Given the limited resources we have to throw at this, 
is this approach good enough?  If not, what are are the alternatives, what are 
their advantages versus this one and how complex will they be to implement?

Many thanks in advance.

-- 
CALLUM PRENTICE | Software Engineer

LINDEN LAB | Create Virtual Experiences <http://www.lindenlab.com/> 


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 16, 2016 at 3:41:47 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:
 

Gstreamer is probably your best bet just based on the friendliness of the 
developer community when asking questions and gstreamer being somewhat more 
focused on being an intuitive framework to integrate into an application. 
However, libvlc has a much simpler api which makes it faster to integrate, and 
you don’t need to worry about plugins as much, as only the codecs are modular 
(which has benefits and drawbacks, of course.) As Henri pointed out though, 
getting a functional framework to link against is a job that only happens once 
(or once every library update anyway.)

​Understood and thanks for the insight.  Does the GStreamer plugin in the LL 
viewer-release branch still build?
I’m not sure. You’d definitely want to update to gstreamer 1.x though and 
probably ditch the old repo and build with Cerbero.

  Do you know if anyone has made a Windows or OS X version of it ?
​
  I made an attempt three years ago, got it working on OS X. Got frustrated 
with mingw and moved on to something else. I know the Imprudence team had some 
success with replacing both FMOD and Quicktime with gstreamer earlier than 
that, but this issue from their tracker shows that they then debated moving to 
FFMpeg or libvlc instead: 
https://sourceforge.net/p/team-purple/imprudence/tickets/340/

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 16, 2016 at 3:39:53 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:
 

​​

 My knowledge about libVLC is rather outdated I would say, however it should be 
able to play quicktime/MOV files so I'm pretty sure it's a bug somewhere.

​
 ​Investigating now - the source of test material when I wrote the original 
QuickTime plugin was the Apple movie trailers site but they all seem to be in 
Apple's own m4v​ format which doesn't play in SL on my system with the latest 
version of QuickTime for windows installed.
M4v is like the video container version of mp4. Here are the official quicktime 
sample files:

https://support.apple.com/en-gw/HT201549

​
  No experience with gstreamer but if i remember when we were arguing for 
replacements a while back that there was a concern about licensing codecs wise. 
If legal cleared it then it should in theory be fine though. Building it for 
windows might be atrocious though.
 ​
 ​That was my experience a long time ago when we briefly discussed using 
GStreamer everywhere. However, I see now there is a page ​- 
 https://gstreamer.freedesktop.org/download/​ - with binary downloads for all 
platforms.
 ​
  Windows building support has come a long way as long as you’re using 
gstreamer 1.x (no more nasty mingw/msys) They claim that it builds with 
Microsoft compilers now.


-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Replacement for QuickTime media plugin - a straw man proposal

[Cinder Roxley]

On May 16, 2016 at 4:03:29 PM, Callum Prentice (Callum) (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:
 
  Do you know if anyone has made a Windows or OS X version of it ?
​
   I made an attempt three years ago, got it working on OS X. Got frustrated 
with mingw and moved on to something else. I know the Imprudence team had some 
success with replacing both FMOD and Quicktime with gstreamer earlier than 
that, but this issue from their tracker shows that they then debated moving to 
FFMpeg or libvlc instead: 
https://sourceforge.net/p/team-purple/imprudence/tickets/340/

​

Good info - I'll go take a look - thanks Cinder.​
 ​
Looks like this is the last efforts on gstreamer for the viewer from Inworldz. 
Autobuild 3p for building gstreamer 1.0 win32: 
https://bitbucket.org/mccabe/3p-gstreamer-sdk-x86-iw

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Quicktime

[Cinder Roxley]

On April 21, 2016 at 7:02:37 AM, Henri Beauchamp (sl...@free.fr 
<mailto:sl...@free.fr> ) wrote:
Plus, frankly, is it reasonnable to launch a new instance of CEF (i.e. 
a *full* embedded web browser instance, using over 80Mb of memory while 
the QuickTime plugin uses 100 times less) for *each* playing media on 
surrounding prims ? That's a bit like using a hammer and an anvil to 
squash a bug, don't you think so ?... Quite inelegant ! YUCK !!! 

Agreed.

Like I already wrote earlier, the way to go is to use the gstreamer SDK 
for Windows and get a gstreamer plugin compiled for the latter. 
With all due respect, gstreamer is a major pain to build on Windows and runs 
afoul of dozens of patents and licenses. It’s fine if you’re building from 
source for linux, but it’s a lawsuit waiting to happen for commercial software 
if you want to play any “standard” media format like h264 or mp3. Something 
like libvlc might work if you want a cross platform library (but again, there 
are per-install royalties to use h264 so you’d still be screwed on mp4.)

Platform-specific plugins could take advantage of the OS’s media playback 
capabilities, without license and patent headaches.

-- 
Cinder Roxley

Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] About memory management on macOS 10.12 (Sierra) potentially affecting all viewers

[Cinder Roxley]

None of the examples you pasted have anything to do with Apple’s garbage 
collector, hun. They’re just examples of advanced MRR. 

As I already stated, the viewer has never used Apple’s garbage collector. It 
uses a combination of autorelease pools and MRR which is perfectly fine to do, 
especially in Objective-C++ where we are not only managing objective-c objects, 
but C++ objects as well.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On July 8, 2016 at 1:47:16 PM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:

 
Cinder Roxley said:
The viewer has never used the garbage collector so it?s not an issue.


Why is it that in llopenglview-objc.mm, as en example, we have statement like:

NSOpenGLPixelFormat *pixelFormat = [[[NSOpenGLPixelFormat alloc] 
initWithAttributes:attrs] autorelease];

or 

- (void)dealloc
{
[[NSNotificationCenter defaultCenter] removeObserver:self];
[super dealloc];
}

while Apples ARC migration guidelines states:

You cannot explicitly invoke dealloc, or implement or invoke retain, release, 
retainCount, or autorelease.

You can’t invoke dealloc.

Custom dealloc methods in ARC do not require a call to [super dealloc] (it 
actually results in a compiler error). The chaining to super is automated and 
enforced by the compiler.


Or in llwindowmacosx-objc.mm we have code such as:

bool copyToPBoard(const unsigned short *str, unsigned int len)
{
NSAutoreleasePool *pool = [[NSAutoreleasePool alloc]init];
NSPasteboard *pboard = [NSPasteboard generalPasteboard];
[pboard clearContents];

NSArray *contentsToPaste = [[NSArray alloc] initWithObjects:[NSString 
stringWithCharacters:str length:len], nil];
[pool release];
return [pboard writeObjects:contentsToPaste];
}

Apple’s documentation specifically says how to rewrite this construct for ARC. 

https://developer.apple.com/library/mac/documentation/Cocoa/Reference/Foundation/Classes/NSAutoreleasePool_Class/index.html#//apple_ref/occ/cl/NSAutoreleasePool


So all this code needs to be rewritten to support ARC or it will not run or 
fail on 10.12. 


Geir Nøklebye,

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] About memory management on macOS 10.12 (Sierra) potentially affecting all viewers

[Cinder Roxley]

Like, I said, just wants to argue. Not interested in understanding. ;)



On July 9, 2016 at 7:58:33 AM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:

 Having worked in Apple Product Management 
heh.

The GPU crash has been both acknowledged by Apple product development and by 
LL, but the LL code needs to be fixed for it to work. This crash is clearly 
related to memory management both on the GPU and on the main system.
Except that it isn’t related to memory management at all. It’s an alpha sorting 
bug in Apple's OpenGL implementation. The more you know! 

One popular viewer compiled with the old GCC could not start at all. 
This isn’t new, and it’s not related to the viewer as much as “One popular 
viewer” has an apr lib with some really janky black magic to workaround Apple’s 
failure at POSIX adherence.


-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] About memory management on macOS 10.12 (Sierra) potentially affecting all viewers

[Cinder Roxley]

All of that is fixed in STORM-2116. Just waiting for Oz to review it again and 
merge.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On July 10, 2016 at 3:12:46 AM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:

 
The viewer does not build unless this flag is turned on. It hasn’t for a long 
time. 


On 10. jul. 2016, at 11.09, Nicky Perian <nickyper...@gmail.com 
<mailto:nickyper...@gmail.com> > wrote:

Does LL client also crash when built with suppress warnings of deprecations 
flag on?  

On Sun, Jul 10, 2016 at 1:07 AM, Geir Nøklebye <geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> > wrote:
There is no need to have any detailed build and test cases. Just turn off the 
compiler flag that LL set to suppress warnings of deprecations and you have 
your work cut out before you. ;-)

The GPU crash is already in JIRA. 

Cheers,
Geir



On 10. jul. 2016, at 04.33, Nicky Perian <nickyper...@gmail.com 
<mailto:nickyper...@gmail.com> > wrote:

It will likely be after 64 bit is fully in place before anything would be done.
I suggest you send an email to Oz and feel him out on opening a Storm jira. But 
be warned it would have to be detailed and include build and test cases. 

Nicky
-- Forwarded message --
From: Argent <secret.arg...@gmail.com <mailto:secret.arg...@gmail.com> >
Date: Sat, Jul 9, 2016 at 2:52 PM
Subject: Re: [opensource-dev] About memory management on macOS 10.12 (Sierra) 
potentially affecting all viewers
To: Geir Nøklebye <geir.nokle...@dayturn.com <mailto:geir.nokle...@dayturn.com> 
>
Cc: opensource-dev@lists.secondlife.com 
<mailto:opensource-dev@lists.secondlife.com> 


I've been following Apple for over 35 years now, and they have introduced and 
abandoned technologies on a regular basis. You can't dismiss the fact that they 
introduced and abandoned GC in only a few years and confidently declare that 
Swift won't go the same way. It might stick. It might not. It might be like FFS 
and ZFS and 64 bit Carbon and Blue Box and Metal. Or it may be like Yellow Box 
and turn into the next Cocoa. Same with their new memory management. Best keep 
on doing what works instead of diving into the churn.


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev 
<http://wiki.secondlife.com/wiki/OpenSource-Dev> 
Please read the policies before posting to keep unmoderated posting privileges





___
Policies and (un)subscribe information available here:

http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] About memory management on macOS 10.12 (Sierra) potentially affecting all viewers

[Cinder Roxley]

On July 7, 2016 at 1:57:23 PM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:
 
Apple posted the first public beta of macOS 10.12 (Sierra) today, meaning some 
users will install it and find there are issues with just about every viewer.

>From the Xcode 8 beta 2 release notes, please take note that:

macOS 10.11 was the last major release of macOS that supported the previously 
deprecated garbage collection runtime. Applications or features that depend 
upon garbage collection may not function properly or will not launch in macOS 
Sierra. Developers should use Automatic Reference Counting (ARC) or manual 
retain/release for memory management instead. (20589595) 


The viewer has never used the garbage collector so it’s not an issue.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer and Apple maintained opensource libraries

[Cinder Roxley]

On February 2, 2017 at 2:57:13 PM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:
 

@Cinder 

With reference to Xcode 8 release notes, section Deprecations it reads: 

OS X 10.11 was the last major release of macOS that supported the previously 
deprecated garbage collection runtime. Applications or features that depend 
upon garbage collection may not function properly or will not launch in macOS 
Sierra. Developers should use Automatic Reference Counting (ARC) or manual 
retain/release for memory management instead. (20589595) 

https://developer.apple.com/library/content/releasenotes/DeveloperTools/RN-Xcode/Introduction.html
 

This was enforced in the Xcode 8 beta program, but was eased at release, but I 
believe it will be reintroduced in the forthcoming 8.3 release as it again 
appears in the 8.3 beta release notes. 

Referring to Apple developer notes on deprecations and changes as koolaid 
pretty much says it all. 


Please see last July when you were already banging this drum and it was 
explained multiple times that the viewer never used GC so your claim was 
misinformed and the deprecation irrelevant to sl viewer:
https://lists.secondlife.com/pipermail/opensource-dev/2016-July/thread.html#10248

I won’t waste any further time on this. Enjoy spamming the list and pulling our 
changes. :)

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer and Apple maintained opensource libraries

[Cinder Roxley]

On January 31, 2017 at 4:55:26 AM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:
 
Are you all aware of the Apple maintained opensource libraries already included 
in macOS that you also maintain and use in the viewer?  Examples are libexpat, 
pcre, openAL, hunspell and openssl.
Most of which are woefully out of date.

-- 
Cinder Roxley
Sent with Airmail

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer and Apple maintained opensource libraries

[Cinder Roxley]

On February 2, 2017 at 9:08:44 AM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:
 
Henri said:

> Meaning you actually didn't fix anything...

> Stupid question: what happens if you disable the Objects occlusion
setting in the graphics preferences ?

The root cause of the problem is not fixed, no. It put some bandaids on it to 
reduce the crash frequency. The autoreleasepool issues are “real” fixes, as it 
is both deprecated and won't even compile on Xcode 8. 


It actually isn’t, as was already discussed on this list several months ago at 
length… but keep sipping that koolaid from the company who can’t even implement 
fork() properly and removes features from its own OS because it crashes their 
3D driver.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer and Apple maintained opensource libraries

[Cinder Roxley]

On January 31, 2017 at 2:40:44 PM, Monty Brandenberg (mo...@lindenlab.com 
<mailto:mo...@lindenlab.com> ) wrote:
 
On 1/31/2017 3:24 PM, Cinder Roxley wrote:

> ~ % otool -L /usr/lib/libssl.dylib
>
> /usr/lib/libssl.dylib:
>
> /usr/lib/libssl.0.9.8.dylib (compatibility version 0.9.8, current
> version 0.9.8)
>
> /usr/lib/libcrypto.0.9.8.dylib (compatibility version 0.9.8, current
> version 0.9.8)

Ouch. I was ashamed of how long we stayed at 0.9.8. (Would kinda like
to use LibreSSL...)


For what it’s worth, Apple did warn developers to stop using it and switch to 
Cocoa’s crypto frameworks or ship their own. I’ve been waiting for a peer 
review of BearSSL myself.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac viewer and Apple maintained opensource libraries

[Cinder Roxley]

On January 31, 2017 at 2:12:24 PM, Geir Nøklebye (geir.nokle...@dayturn.com 
<mailto:geir.nokle...@dayturn.com> ) wrote:
 
Some are outdated and some are more up to date than the current viewer 
libraries. 
They usually get serious amount of public flack if security issues are not 
fixed, meaning they are automatically patched when users upgrade their systems. 
~ % otool -L /usr/lib/libssl.dylib

/usr/lib/libssl.dylib:

/usr/lib/libssl.0.9.8.dylib (compatibility version 0.9.8, current version 0.9.8)

/usr/lib/libcrypto.0.9.8.dylib (compatibility version 0.9.8, current version 
0.9.8)

I rest my case.

Plus, the majority of libraries consumed by the viewer don’t offer api 
compatibility between versions making cross-platform development, not to 
mention QA, all the more painful.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] 64 bit viewers build instructions

[Cinder Roxley]

We’ve had OpenJPEG 1.5.1 running solidly on Alchemy for years now. I can put 
together some patches to get it working.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On January 17, 2017 at 9:03:11 AM, Nat Goodspeed (n...@lindenlab.com 
<mailto:n...@lindenlab.com> ) wrote:

 
Nat Goodspeed <n...@lindenlab.com> erian <nickyper...@gmail.com> wrote:

> https://bitbucket.org/lindenlab/p64_3p-openjpeg/pull-requests/2
>
> Please take this in and then provide for the updated archives in the viewer. 
> I also received a report that the missing texture issue is in macOS when 
> using openjpeg-1.5.1.

There are admittedly painful aspects to the two-step autobuild
mechanism: (1) rebuild the 3p package and (2) update every consumer.

However, one of the benefits of that approach is that we can adjust
the version of a given 3p package consumed by (e.g.) the viewer
without actually having to revert the 3p repository source. We can
just change back the package URL specified in the viewer's
autobuild.xml.

> Or, if the problem with 1.5.1 is easily corrected

Please forgive me if this should already be on my plate, but I don't
recall a Jira about the openjpeg 1.5.1 issues? Does 1.5.1 fail to load

texture files that are correctly handled by 1.5.0?

I think Cinder has a point: if we can move forward with 1.5.1, we should.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Question about BUG-41029 and 64 bit usage

[Cinder Roxley]

Mine has 6GB and was relatively inexpensive ($211 USD)

As far as the viewer, I think the best way to go would be to bite the bullet 
and rework those to use size_t. 
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On December 15, 2016 at 7:03:58 PM, Callum Prentice (Callum) 
(cal...@lindenlab.com <mailto:cal...@lindenlab.com> ) wrote:

 
Yep - I saw a lot of memory related texture references too - I don't know if 
cards these days have more than 4GB of video memory - definitely a possibility 
soon if not already.

On Thu, Dec 15, 2016 at 5:01 PM, Niran <desmoulins.u...@googlemail.com 
<mailto:desmoulins.u...@googlemail.com> > wrote:
Funny, i just so happened to stumble across this a few days ago when i had this 
mindblowing realization that this might be the cause for the Viewer not 
properly reporting VRAM over 4gb but i don't happen to have a 4+gb VRAM GPU so 
i wouldn't be able to test anything i do and ultimately dropped the idea of 
touching it for now.

2016-12-15 20:13 GMT+01:00 Callum Prentice (Callum) <cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> >:
https://jira.secondlife.com/browse/BUG-41029

I'm taking a look at https://jira.secondlife.com/browse/BUG-41029and whilst it 
seems straightforward, it seems to be unraveling into something that touches 
dozens of files and I wondered if someone had done this work already.

There is a lot usage of 32 bit types (U32Bytes, U32Megabytes etc.) defined 
indirectly here: 

https://bitbucket.org/lindenlab/viewer64/src/9270caf3d4324f9c1c33aa158f80e0fe84036a48/indra/llcommon/llunittype.h?at=default=file-view-default#llunittype.h-824

that are used to count memory sizes/usage/difference etc.  I think we can 
convert them to their U64 equivalents for all viewers. 

Nat points out, rewriting this code using size_t as a return type would make 
more sense but that seems like it would involve more invasive code changes 
including changes in fundamental LL headers.

What does the collective wisdom say?

-- 
CALLUM PRENTICE | Software Engineer

LINDEN LAB | Create Virtual Experiences <http://www.lindenlab.com/> 


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev 
<http://wiki.secondlife.com/wiki/OpenSource-Dev> 
Please read the policies before posting to keep unmoderated posting privileges


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev 
<http://wiki.secondlife.com/wiki/OpenSource-Dev> 
Please read the policies before posting to keep unmoderated posting privileges



--
CALLUM PRENTICE | Software Engineer

LINDEN LAB | Create Virtual Experiences <http://www.lindenlab.com/> 


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Windows 64 bit build issue

[Cinder Roxley]

It’s not needed. We dropped it in Alchemy when adding 64-bit support, and it 
continues to run fine. /FIXED:NO just ensures that code can be position 
independent. Maybe at some point this was needed a long time ago for apr 
hijinks or Windows98 compatibility or something of that nature, but you can 
safely remove it.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On November 30, 2016 at 6:55:37 PM, Callum Prentice (Callum) 
(cal...@lindenlab.com <mailto:cal...@lindenlab.com> ) wrote:

 
I'm working with Nat Linden on the 64 bit viewer build and we've been 
encountering an odd error - A number of projects in 64 bit only configurations 
have an entry for "Force Includes" files set to XED:NO.  Nothing on Google so 
we were stumped for a while but eventually tracked it down to a number of lines 
in CMakeLists.txt files of the form: add_definitions(/FIXED:NO). Evidently, /FI 
is the compiler command to include a forced header file - hence the XED:NO 
entry. You can see one here: 
https://bitbucket.org/lindenlab/viewer64/src/3f7ba2a06e5cea596e3a4006d57e3fbc4703d90f/indra/llcommon/CMakeLists.txt?at=default=file-view-default#CMakeLists.txt-248

The FIXED command is evidently for the linker and not the compiler but I'm not 
sure (a) if it's needed or (b) if it is, how to direct it at the right place.

Has anyone else encountered this and already and figured it out?

--

​Callum Prentice
| Software Engineer

LINDEN LAB | Create Virtual Experiences <http://www.lindenlab.com/> 


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Windows 64 bit build issue

[Cinder Roxley]

We nuked that whole project and set volume directly on Windows using 
waveOutSetVolume(HWAVEOUT, DWORD).
https://bitbucket.org/alchemyviewer/alchemy/commits/b3e54a075de4440390bb1be8c9f73bfd94053a01

The shim is only needed up to XP and is completely unnecessary with the 
availability of WASAPI in Vista and above.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On November 30, 2016 at 7:36:24 PM, Callum Linden (cal...@lindenlab.com 
<mailto:cal...@lindenlab.com> ) wrote:

 
Thanks so much for the speedy reply Cinder. I hoped that was the case and the 
build seems okay without it. 

The other perplexing one is for the winmm_shim project. For 64 bit builds it 
fails with a bunch of unresolved externals which look to be related to MMX 
intrinsic maybe. Not at my computer right now and haven't really started to 
investigate but if that rings s bell after your work, please do let me know. 

Cheers! 

On Nov 30, 2016, at 5:16 PM, Cinder Roxley <cin...@alchemyviewer.org 
<mailto:cin...@alchemyviewer.org> > wrote:

It’s not needed. We dropped it in Alchemy when adding 64-bit support, and it 
continues to run fine. /FIXED:NO just ensures that code can be position 
independent. Maybe at some point this was needed a long time ago for apr 
hijinks or Windows98 compatibility or something of that nature, but you can 
safely remove it.

-- 
Cinder Roxley
Sent with Airmail

On November 30, 2016 at 6:55:37 PM, Callum Prentice (Callum) 
(cal...@lindenlab.com <mailto:cal...@lindenlab.com> ) wrote:

I'm working with Nat Linden on the 64 bit viewer build and we've been 
encountering an odd error - A number of projects in 64 bit only configurations 
have an entry for "Force Includes" files set to XED:NO.  Nothing on Google so 
we were stumped for a while but eventually tracked it down to a number of lines 
in CMakeLists.txt files of the form: add_definitions(/FIXED:NO). Evidently, /FI 
is the compiler command to include a forced header file - hence the XED:NO 
entry. You can see one here: 
https://bitbucket.org/lindenlab/viewer64/src/3f7ba2a06e5cea596e3a4006d57e3fbc4703d90f/indra/llcommon/CMakeLists.txt?at=default=file-view-default#CMakeLists.txt-248

The FIXED command is evidently for the linker and not the compiler but I'm not 
sure (a) if it's needed or (b) if it is, how to direct it at the right place.

Has anyone else encountered this and already and figured it out?

--

​Callum Prentice
| Software Engineer

LINDEN LAB | Create Virtual Experiences <http://www.lindenlab.com/> 


___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev 
<http://wiki.secondlife.com/wiki/OpenSource-Dev> 
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] post-configuration errors

[Cinder Roxley]

It’s probably better that they be fixed.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On January 4, 2017 at 11:04:22 AM, Jonathan Welch (jhwe...@gmail.com 
<mailto:jhwe...@gmail.com> ) wrote:

 
At the end of the configuration process I now am getting a screen full
of these errors. Should something be adjusted to suppress them or
what?

-- Configuring done
CMake Warning (dev) at media_plugins/libvlc/CMakeLists.txt:61 (add_dependencies)
:
 Policy CMP0046 is not set: Error on non-existent dependency in
 add_dependencies. Run "cmake --help-policy CMP0046" for policy details.

 Use the cmake_policy command to set the policy and suppress this warning.

 The dependency target "debug" of target "media_plugin_libvlc" does not
 exist.
This warning is for project developers. Use -Wno-dev to suppress it.

CMake Warning (dev) at media_plugins/libvlc/CMakeLists.txt:61 (add_dependencies)
:
 Policy CMP0046 is not set: Error on non-existent dependency in
 add_dependencies. Run "cmake --help-policy CMP0046" for policy details.

 Use the cmake_policy command to set the policy and suppress this warning.

 The dependency target "libboost_context-mt" of target "media_plugin_libvlc"
 does not exist.
This warning is for project developers. Use -Wno-dev to suppress it.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] windows build issues

[Cinder Roxley]

Looks like your python setup is fubar’d. Remove all traces of it and reinstall. 
pip install llbase and autobuild Do NOT use cygwin’s python and make sure that 
python comes earlier in your PATH than cygwin.
 
 
-- 
Cinder Roxley
Sent with Airmail
 

On January 4, 2017 at 3:20:53 PM, Lance Corrimal (lance.corri...@eregion.de 
<mailto:lance.corri...@eregion.de> ) wrote:

 
and again: autobuild itself seems to work, at least it starts to work. it dies 
at the point where it tries to download the prebuild packages.

here is the complete output of my problem:

http://susepaste.org/33664691




Am 04.01.2017 um 19:58 schrieb Nicky Perian:
cd into local repo  firestorm, viewer-release are any other.
>From there enter autobuild --version  an absolute path should not be needed to 
>reach autobuild.
It should give the version, If not, your path is not set or you may need to set 
AUTOBUILD='pathtoautobuild/bin/autobuild' environment variable.





On Wed, Jan 4, 2017 at 12:20 PM, Lance Corrimal <lance.corri...@eregion.de> 
wrote:
that's what i'm using. Like I said already, I had to reinstall my computer, 
before that I had been building viewers all the time, 32bit and 64bit, no 
problems whatsoever.

Now, it doesn't work with some error that looks like autobuild isn't 
downloading the prebuilts.



Any ideas?




Am 04.01.2017 um 19:15 schrieb Whirly Fizzle:
If you are building Firestorm 64bit, you must use Nicky Dasmijn's autobuild: 
https://bitbucket.org/NickyD/autobuild-1.0

NickyD / autobuild-1.0
bitbucket.org
Hg repository hosted by Bitbucket.




From: 
opensource-dev-boun...@lists.secondlife.com<opensource-dev-boun...@lists.secondlife.com>on
 behalf of Lance Corrimal <lance.corri...@eregion.de>
Sent: 04 January 2017 14:12
To: Oz Linden (Scott Lawrence); opensource-dev@lists.secondlife.com
Subject: Re: [opensource-dev] windows build issues
 actually I'm building the firestorm developer version, and that used to work 
just fine until I reinstalled my PC yesterday ;_;


Am 04.01.2017 um 15:01 schrieb Oz Linden (Scott Lawrence):
On 2017-01-04 08:15 , Lance Corrimal wrote:

d:\Users\lemmy\Build\autobuild-1.0>autobuild --version
autobuild 1.0


d:\Users\lemmy\Build\autobuild-1.0>which autobuild
/cygdrive/d/Users/lemmy/Build/autobuild-1.0/bin/autobuild
64bit here.. using your autobuild, which used to work just fine before I  
reinstalled my PC...


If you're trying to build viewer64 you should be aware that it's still a Work 
In Progress and has some limitations... Also, building that repository requires 
using autobuild-1.1 (https://bitbucket.org/lindenlab/autobuild-1.1). All this 
will be updated on the wiki once this project has reached the Release Candidate 
stage.
lindenlab / autobuild-1.1
bitbucket.org
Hg repository hosted by Bitbucket.
--
OZ LINDEN | Engineering Director, Second Life email or hangouts: 
o...@lindenlab.com | Real Life: Scott Lawrence LINDEN LAB | Create Virtual 
Experiences

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___ Policies and (un)subscribe 
information available here: 
http://wiki.secondlife.com/wiki/OpenSource-DevPlease read the policies before 
posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

[opensource-dev] SL Mini-Meetup at Libreplanet?

[Cinder Roxley]

I’m going to be at LibrePlanet (https://libreplanet.org/) this weekend in 
Cambridge, MA, USA if any Second Life open source developers want to get 
together for drinks or a fistfight or anything like that.

-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] macOS -DVIEWER-CHANNEL

[Cinder Roxley]

GIGO


On August 11, 2017 at 2:47:04 PM, Nicky Perian (nickyper...@gmail.com)
wrote:

When -DVIEWER_CHANNEL="\"SecondLife Test Channel Assignment\"" is passed on
configure a malformed filename is constructed for dmg. This produces
Second_Life_SecondLife\\_Test\\_Channel\\_Assignment_5.1.0_x86_64.dmg

and the dmg is not useable.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Update to mercurial and conflict with "hg policy" local extension.

[Cinder Roxley]

Mq is bordering on deprecation (likely going to happen when hg evolve is
finished). Consider switching to modern tools like hg rebase, hg histedit,
hg graft, hg strip, and hg commit --amend or try out hg evolve which needs
more testers and has backend support on bitbucket already.


On July 22, 2017 at 7:59:36 AM, Nicky Perian (nickyper...@gmail.com) wrote:

Mq is bordering on deprecation (likely going to happen when hg evolve is
finished). Consider switching to modern tools like hg rebase, hg histedit,
hg graft, hg strip, and hg commit --amend or try out hg evolve which needs
more testers and has backend support on bitbucket already.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] viewer64 SL_Launcher

[Cinder Roxley]

I think it’s likely that the initial launcher isn’t going to switch
discrete graphics mode in to detect it. (Isn’t it just a python script?)

You’re either going to have to go off of the chipset detected or maybe add
some GL intense splash screen to the launcher that causes the system to
kick over to discrete mode in order to detect the discrete GPU.


On June 30, 2017 at 11:02:28 AM, Nat Goodspeed (n...@lindenlab.com) wrote:

On Fri, Jun 30, 2017 at 11:52 AM, Nicky Perian 
wrote:

On my alienware it is a Nvidia setting to use its driver as default and can
> also be set for a specific *.exe program name.
>
> On Fri, Jun 30, 2017 at 10:26 AM, Jonathan Welch 
> wrote:
>
>> Some systems have two built in graphics chips such as Lenovo laptops with
>> a system called Optimus. One graphics chip is very weak and the other much
>> stronger, bit also more power hungry. As I recall it took a while to
>> register the .EXE name with Nvidia so the stronger chip runs the viewer.
>>
>> Is this registration underway already?
>>
>
I don't know about that registration, but will ask.

What I think you're saying, though, is that detection of the issue may be
complicated. In fact it might be impossible for SL_Launcher.exe to detect
which potential graphics driver is configured to handle our viewer. Is that
a fair statement?
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Please accept my invitation to Nextdoor

[Cinder Roxley]

No, I don’t think so.
 
 
-- 
Cinder Roxley
Sent with Airmail
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer Manager (Launcher) and AntiVirus

[Cinder Roxley]

Submit a sample to Avira so they’ll whitelist it.
https://analysis.avira.com/en/submit Bitdefender and Kaspersky don’t detect
it as a virus (and Bitdefender thinks git is malicious, so there you go.)

Further, code signing is your friend. Signed bins are much less likely to
be flagged as malicious by crappy free antivirus suites.

On September 28, 2017 at 8:47:06 AM, Niran (desmoulins.u...@googlemail.com)
wrote:

Avira (and presumably all others) goes nuts on vmp_util.exe whenever it
appears anywhere, worse, when i configure the viewer and it downloads and
extracts the viewer manager package it immediately flags it as potentially
dangerous in such a way that said file becomes completely phased, i cannot
remove it, move it or otherwise interact with it and the configuration
fails with "permission denied" until i restart the PC. I wonder what i can
do about it, the moment i'm going to pack this thing into the Viewer
package and offer a download people will chop my head off that its infected
with Viruses.

I know AntiVirus applications are shit but i'm worried about this happening
not just to me but to basically every single user, this would be absolutely
horrible.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Windows 64bit compiling error.

[Cinder Roxley]

Would need more context about what you changed in your viewer. If I have
time, I’ll pull later and see if I can’t figure out what went awry in CMake.


On September 1, 2017 at 8:50:12 AM, Niran (desmoulins.u...@googlemail.com)
wrote:

So after the painstaking process of going through the very lacking wiki
guide, getting everything set up, fixing all the issues and getting the
Alex Ivy repository successfully build, i'm now in the process of switching
my own Viewer (Black Dragon) to Alex Ivy base, after a few more problems i
could luckily solve (such as the 100% missing guide on how to build FmodEx)
i'm now at a successful configure, starting the viewer VS project file and
getting fatal errors basically as soon as the first file start to compile.

6>f:\blackdragon64\indra\llcommon\llapp.h(307): error C2061: syntax error :
identifier 'siginfo_t' (F:\BlackDragon64\indra\llcommon\llapp.cpp)
6>f:\blackdragon64\indra\llcommon\llmemory.h(93): fatal error C1189: #error
:  "LL_ALIGN_PREFIX and LL_ALIGN_POSTFIX undefined"
(F:\BlackDragon64\indra\llcommon\llapp.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(142): error C2146:
syntax error : missing ';' before identifier 'DERIVED_TYPE'
(F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>  f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(143) :
see reference to class template instantiation
'LLThreadLocalSingletonPointer' being compiled
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(142): error C4430:
missing type specifier - int assumed. Note: C++ does not support
default-int (F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(146): error C2146:
syntax error : missing ';' before identifier 'DERIVED_TYPE'
(F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(146): error C2143:
syntax error : missing ';' before '*'
(F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(146): error C2065:
'DERIVED_TYPE' : undeclared identifier
(F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(146): error C2923:
'LLThreadLocalSingletonPointer' : 'DERIVED_TYPE' is not a valid template
type argument for parameter 'DERIVED_TYPE'
(F:\BlackDragon64\indra\llcommon\llapr.cpp)
6>f:\blackdragon64\indra\llcommon\llthreadlocalstorage.h(146): fatal error
C1903: unable to recover from previous error(s); stopping compilation
(F:\BlackDragon64\indra\llcommon\llapr.cpp)

The error list goes on and on and on, every second file errors out, every
project fatally fails compiling from what i assume is a problem with the
compile not knowing i'm trying to compile for Windows. From the llmemory.h
file i can see that the error should only ever happen if neither Mac/Linux
nor Windows are specified to be used.

#if LL_LINUX || LL_DARWIN

#define LL_ALIGN_PREFIX(x)
#define LL_ALIGN_POSTFIX(x) __attribute__((aligned(x)))

#elif LL_WINDOWS

#define LL_ALIGN_PREFIX(x) __declspec(align(x))
#define LL_ALIGN_POSTFIX(x)

#else
#error "LL_ALIGN_PREFIX and LL_ALIGN_POSTFIX undefined"
#endif

This is the specific snipped i'm talking about. However, that brings up the
question, why does this happen, where and how can i fix it, i hg checked
against the successful Alex Ivy build, every single file of importance, all
cmake files and so on, they all seem fine and basically exactly the same
except the name of the Viewer.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] macOS 10.14 deprecation of OpenGL, what does this mean for SL?

[Cinder Roxley]

Apple deprecated AGL in 2009 and yet programs linking to it still run
nearly ten years later. The sky Isn’t falling any time soon.


On June 4, 2018 at 5:26:00 PM, Kadah Coba (kadah.c...@gmail.com) wrote:

So just heard about this, not sure if this was known about before this.
https://developer.apple.com/macos/whats-new/

"Apps built using OpenGL and OpenCL will continue to run in macOS 10.14,
but these legacy technologies are deprecated in macOS 10.14. Games and
graphics-intensive apps that use OpenGL should now adopt Metal."

There's doing the same on iOS too, not that it effects us.

Knowing Apple's way, could likely assume 10.15 or 10.16 may not support
OpenGL at all, or at worse, a later update of 10.14.x.

So what does this mean for us since its the thing we breath
? Blue alert or do we change
the bulb to red ?
May 31, 2018 at 11:22:38 AM ___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] macOS 10.14 deprecation of OpenGL, what does this mean for SL?

[Cinder Roxley]

SSLv3 was disabled because it’s exploitable!
https://www.us-cert.gov/ncas/alerts/TA14-290A

On June 5, 2018 at 4:05:36 PM, Kadah Coba (kadah.c...@gmail.com) wrote:

On 6/4/2018 10:41 PM, Dahlia Trimble wrote:
> I can't believe that Apple will remove an API that so many
> applications use.

I've seen Apple do worse in the past with less warning.

Apple disabled SSLv3 by default in iOS 9 and an OSX update (maybe around
10.11, can't remember) without warning. This would have been fine since
any updated server would also support TLS, except Apple screwed up
somewhere and many Apple apps, like Mail, would still attempt to connect
using SSLv3 but then fail with a typically Apple style unhelpful error
when the OS doesn't allow it to use SSLv3 because the app's plist config
doesn't allow for an exception.

I got around this by disabling SSLv3 and enforcing TLS only, which
flagged dayed email for Apple users. Before I did that, if they updated
to iOS9, they couldn't get email; after I made the changes, they
required iOS9, and likely an account delete and readd on their phone
because Apple.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

[opensource-dev] AIS3 docs

[Cinder Roxley]

I’m having a hell of a time writing AIS support for libremetaverse and 
Radegast. From an initial capture from wireshark, not everything matches up 
exactly with http://wiki.secondlife.com/wiki/Linden_Lab_Official:Inventory_API. 
Any chance we could get some up to date docs? Maybe a swagger spec? I’d rather 
not have to do this entirely via wireshark and don’t want GPL poisoning from 
the viewer code. Thanks!

signature.asc
Description: Message signed with OpenPGP using AMPGpg
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] compiling viewer-release on linux

[Cinder Roxley]

On August 6, 2018 at 5:44:54 AM, Alex (sl...@vlan1000.net) wrote:

Hi All,

I am being brave and giving this a go..

Can someone explain this:

(SNIP)


Assuming its not needed, how can I drop the requirement for this during
the configure phase of the build?

You are building with the proprietary libs. You need to build an opensource
autobuild target:

autobuild configure -c ReleaseOS
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac Build - Where do I start?

[Cinder Roxley]

Hi Kittin!

I haven’t built the viewer in a little while, but you should be able to
build with the latest version of macOS 10.13. As far as Xcode, I think the
latest will work too. If it doesn’t, it’s probably clang being picky. All
the autobuild steps are pretty much the same. Feel free to reach out to be
directly if you get snagged, I should probably try building the viewer
again soon too.

On April 19, 2018 at 11:20:41 AM, Kittin Ninetails (
kittin.nineta...@gmail.com) wrote:

Wow, a lot has changed since I last pulled the Viewer source.  And with the
big warning on the wiki saying the instructions are out of date, I'm not
sure where to start.

If anyone has some quick & dirty notes for a command-line build, I'd
appreciate them.  My current questions are:

   - What version of OSX do I have to be on?  I'm probably going to have to
   downgrade something.
   - What Xcode is supported?  Apple keeps pushing the latest on me, so
   I'll probably have to downgrade that, too.
   - What is is the LL_BUILD env variable and how do I use it?
   - Is it still:  autobuild configure -c ReleaseOS  ?

I'll probably have a lot more questions, but if someone can help me get
started (again), I'd appreciate it.

I'm also not sure why autobuild.xml references "../indra".  Shouldn't that
be "./indra"?  Or am I supposed to now launch autobuild from a subdirectory?
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac Build - Where do I start?

[Cinder Roxley]

I was looking into this a bit over the weekend. I hadn’t really tried to
build SLViewer in a while. I wish I had better news, but I couldn’t figure
out what was going on with it now either. Seems to have become a massively
over engineered and undocumented process developed for CI not individual
development, no doubt. Doesn’t help that autobuiild.xml was removed and
recreated making hg blame useless.

Some docs would be nice, until then, I don’t think figuring it out is worth
the time. Sorry.

On April 20, 2018 at 1:23:40 PM, Kittin Ninetails (
kittin.nineta...@gmail.com) wrote:

Hiya,



I’m snagged.  Very snagged.   First, it complains about a LL_BUILD env var
I don’t have set.  So I set it to -Wall and I get “must contain -isysroot
or something like that.  So I add that, and still no luck.



I don’t know why automake.xml keeps calling for ../indra when indra is a
subdir of the current directory.



It’s all very confusing right now and none of the current documents seem to
mention any of this.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Mac Build - Where do I start?

[Cinder Roxley]

pip install llbase


On April 23, 2018 at 12:36:36 PM, Kittin Ninetails (
kittin.nineta...@gmail.com) wrote:

Thanks, Nicky!  That got me much further.  It stops with an error with
llbase and llsd?  I'm not familiar with those.


PhaseScriptExecution CMake\ PostBuild\ Rules
> /Users/kittin/Dev/Personal/SL/Viewer/build-darwin-x86_64/newview/SecondLife.build/Release/secondlife-bin.build/Script-8E4364E25B1E434A98F894BF.sh
> cd /Users/kittin/Dev/Personal/SL/Viewer/indra
> /bin/sh -c
> /Users/kittin/Dev/Personal/SL/Viewer/build-darwin-x86_64/newview/SecondLife.build/Release/secondlife-bin.build/Script-8E4364E25B1E434A98F894BF.sh
> /usr/bin/python
> /Users/kittin/Dev/Personal/SL/Viewer/indra/newview/viewer_manifest.py
> --actions=copy --arch=x86_64
> --artwork=/Users/kittin/Dev/Personal/SL/Viewer/indra/newview
> --build=/Users/kittin/Dev/Personal/SL/Viewer/build-darwin-x86_64/newview
> --buildtype=Release --configuration=Release
> --dest=/Users/kittin/Dev/Personal/SL/Viewer/build-darwin-x86_64/newview/Release/Second\
> Life.app --grid=agni --channel=KittinTest
> --versionfile=/Users/kittin/Dev/Personal/SL/Viewer/build-darwin-x86_64/newview/viewer_version.txt
> --bundleid=com.secondlife.indra.viewer
> --source=/Users/kittin/Dev/Personal/SL/Viewer/indra/newview
> Traceback (most recent call last):
>   File
> "/Users/kittin/Dev/Personal/SL/Viewer/indra/newview/viewer_manifest.py",
> line 49, in 
> from llbase import llsd
> ImportError: No module named llbase
> make: *** [secondlife-bin_buildpart_0] Error 1
> Command /bin/sh failed with exit code 2
> ** BUILD FAILED **



On Mon, Apr 23, 2018 at 9:09 AM, Nicky Perian <nickyper...@gmail.com> wrote:

> Step 1.
> hg clone ssh://h...@bitbucket.org/lindenlab/viewer-build-variables
>
> Or
> hg clone ssh://h...@bitbucket.org/kokua/viewer-build-variables  //setup for
> multiple Xcode installed in /Xcode  currently Xcode-9.3
>
> Set an environment variable AUTOBUILD_VARIABLES_
> FILE=/Users/bill/viewer-build-variables/variables
>
> Step 2.
> Pip install autobuild1.1
>
> Following is a bit customized. I switch between autobuild-1.0 and
> autobuild-1.1. I source .bash_profile for part of the switching.
>
> {code}
> # araxis
> export PATH=$PATH:/Applications/Araxis\ Merge.app/Contents/Utilities
> # virtualenv
> export WORKON_HOME=~/virtualenvs
> source /usr/local/bin/virtualenvwrapper.sh
> # autobuild
> #workon autobuild-1.1
> #alias autobuild="/Users/bill/virtualenvs/autobuild-1.1/bin/autobuild"
> #export AUTOBUILD="/Users/bill/virtualenvs/autobuild-1.1/bin/autobuild"
> export AUTOBUILD_VARIABLES_FILE="/Users/bill/viewer-build-
> variables/variables"
> #export PATH='/Library/Python/2.7/site-packages/autobuild:$PATHworkon
> export DEVELOPER_DIR='/Xcode/Xcode_9.3.app/Contents/Developer/'
> export CMAKE_OSX_ROOT='Xcode/Xcode_9.3.app/Contents/Developer/
> Platforms/MacOSX.platform/Devloper/SDKs/MacOSX10.13.sdk'
> {/code}
>
> I source another file to move to older Xcode and autobuild-1.0
>
>
> Step 3.
> Configure and build.
>
> From my build script. Full paths are not needed for CL entry
> {code}
> /Users/bill/virtualenvs/autobuild-1.1/bin/autobuild \
> configure --address-size=64 -q -c ReleaseOS --   -DLL_TESTS:BOOL=OFF
> -DFMODEX:BOOL=ON -DOPENAL:BOOL=OFF -DPACKAGE:BOOL=ON  \
> -DUSE_KDU:BOOL=OFF -DRELEASE_CRASH_REPORTING:BOOL=OFF
> -DVIEWER_CHANNEL="Kokua Release RLV”
> …
> /Users/bill/virtualenvs/autobuild-1.1/bin/autobuild build
> --address-size=64 --no-configure -q -c ReleaseOS 2>&1 |tee -a
> DarwinSLRLV.log
> date 2>&1 |tee -a DarwinSLRLV.log
> {/code}
>
>
>
>
>
>
>
> On Apr 23, 2018, at 9:56 AM, Cinder Roxley <cin...@alchemyviewer.org>
> wrote:
>
> I was looking into this a bit over the weekend. I hadn’t really tried to
> build SLViewer in a while. I wish I had better news, but I couldn’t figure
> out what was going on with it now either. Seems to have become a massively
> over engineered and undocumented process developed for CI not individual
> development, no doubt. Doesn’t help that autobuiild.xml was removed and
> recreated making hg blame useless.
>
> Some docs would be nice, until then, I don’t think figuring it out is
> worth the time. Sorry.
>
> On April 20, 2018 at 1:23:40 PM, Kittin Ninetails (
> kittin.nineta...@gmail.com) wrote:
>
> Hiya,
>
>
> I’m snagged.  Very snagged.   First, it complains about a LL_BUILD env var
> I don’t have set.  So I set it to -Wall and I get “must contain -isysroot
> or something like that.  So I add that, and still no luck.
>
>
> I don’t know why automake.xml keeps calling for ../indra 

Re: [opensource-dev] specifying wich python binary to use?

[Cinder Roxley]

I usually setup a virtualenv for each viewer since several tpvs use
different forks of linden’s autobuild.

virtualenv ./autobuild
source autobuild/bin/activate
pip install llbase
pip install 'https://bitbucket.org/lindenlab/autobuild-1.0'


On September 23, 2018 at 3:57:07 AM, Lance Corrimal (
lance.corri...@eregion.de) wrote:

Am Samstag, 22. September 2018, 16:53:27 CEST schrieb Lance Corrimal:
> Am Samstag, 22. September 2018, 16:00:31 CEST schrieb Henri Beauchamp:
> > On Sat, 22 Sep 2018 15:43:14 +0200, Lance Corrimal wrote:
> > > I have python 2 and python 3 installed on my system, is there a way
to
> > > specify which python version autobuild should use? by default it
tries
> > > python3, which fails.
> >
> > Depending on your distribution, you may specify (or not) which Python
> > version is to be used when invoking 'python', with: alternatives
--config
> > python
>
> looked into that already, didn't work.
> The weird thing is, /usr/bin/python already points at /usr/bin/python2,
but
> autobuild still picks up python 3...
>
> > If it does not work, you may trick the build system by creating a
"python"
> > link to the python2 binary and then changing the PATH so that your link
is
>
> > found first. E.g. from a terminal:
> gonna try that. Why didn't I think of that myself (no I do not want the
> obvious answer :P)

the trick with the link did not work either, cmake is evil enough to
actually
search for python binaries WITH the attached version number.

BUT there is a variable that you can pass to autobuild, to pass it on to
cmake, that contains a list of the version numbers to search for...

-DPython_ADDITIONAL_VERSIONS:STRING="2.7 2.6 2.5 2.4 2.3 2.2 2.1 2.0"

with this autobuild does not find python3 binaries, and all is well.

Cheers
LC






___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] LL_TESTS:BOOL=ON

[Cinder Roxley]

Short answer, in C++11, dynamic exception specifications have been
deprecated.
http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2010/n3051.html


On June 4, 2019 at 2:28:57 PM, Nicky Perian (nickyper...@gmail.com) wrote:

This is a Kokua viewer question but, should  apply to other third party
viewers and to LL's viewer-release once Linux is re-implemented.

Note: all tests pass on MacOs and VS2013.

Kokua's branch with no rlv is used as it is closest to viewer-release code.

My approach is to use skip(..) to bypass tests that do not work on a linux
build then once the viewer builds go back an attack each skipped test one
by one.

Test:
Building CXX object
llcorehttp/CMakeFiles/INTEGRATION_TEST_llcorehttp.dir/tests/test_allocator.cpp.o
Faiis with:

home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp: In
function ‘void* operator new(std::size_t)’:
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:51:47:
error: declaration of ‘void* operator new(std::size_t) throw
(std::bad_alloc)’ has a different exception specifier
 #define THROW_BAD_ALLOC() throw(std::bad_alloc)
   ^
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:155:39:
note: in expansion of macro ‘THROW_BAD_ALLOC’
 void * operator new(std::size_t size) THROW_BAD_ALLOC()
   ^
In file included from
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:27:0:
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.h:40:8:
error: from previous declaration ‘void* operator new(std::size_t)’
 void * operator new(std::size_t size) throw (std::bad_alloc);
^
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:51:47:
error: declaration of ‘void* operator new [](std::size_t) throw
(std::bad_alloc)’ has a different exception specifier
 #define THROW_BAD_ALLOC() throw(std::bad_alloc)
   ^
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:161:41:
note: in expansion of macro ‘THROW_BAD_ALLOC’
 void * operator new[](std::size_t size) THROW_BAD_ALLOC()
 ^
In file included from
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.cpp:27:0:
/home/bill/kokua-sl-np-sandbox/indra/llcorehttp/tests/test_allocator.h:41:8:
error: from previous declaration ‘void* operator new [](std::size_t)’
 void * operator new[](std::size_t size) throw (std::bad_alloc);
^
The code for this test is the same as what is in viewer-release.

Distribution: gcc (Ubuntu 5.5.0-12ubuntu1~16.04) 5.5.0 20171010

My thought was that MARCO expansion had changed with -c++11 or between
gcc-4.6 and 5.5.

Bottom line is, I need help.

Nicky








___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting
privileges
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

[opensource-dev] ViewerAsset cap

[Cinder Roxley]

Greetings on a pleasant Sunday morning!

I am updating Radegast and LibreMetaverse to support HTTP asset fetching
via ViewerAsset capability. (https://github.com/cinderblocks/libremetaverse)

I have scoured the wiki for API doc on ViewerAsset, and one does not exist
publicly. I have gotten most of it to work, but I need to know what exactly
is and is NOT supported. It appears textures, mesh, sound, animations,
landmarks, do indeed work, and this seems to cover the bulk of what was
obsoleted and removed.

However, the viewer code would suggest any asset type should be
downloadable, but when I attempt to use the scheme GET http:///?_id= notecard, lsltext, and other asset types return
“Invalid Syntax”.

So my question is, what asset types are supported? Is migration to HTTP
ongoing? eg. Will notecards and scripts be moved to the capability away
from UDP? If so, I will simply handle the error code and fallback. If
there’s no plan to move them, I won’t bother with that and leave fetching
on UDP for unsupported types.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer changes for Premium changes

[Cinder Roxley]

On February 27, 2020 at 10:10:28 AM, Oz Linden (Scott Lawrence) (
o...@lindenlab.com) wrote:

On 2020-02-27 08:21 , Cinder Roxley wrote:

On February 25, 2020 at 2:17:16 PM, Henri Beauchamp (sl...@hotmail.com)
wrote:

On Tue, 25 Feb 2020 10:48:03 -0500, Oz Linden (Scott Lawrence) wrote:

Sadly, you used an LLSD sub-array into the LLSD map, and "old" viewers
(i.e. all viewers not based on LL's v3 viewer code for the XML RPC part)
do not know what to do with such an array (they can only deal with simple
key/value pairs, not with key/arrays); this was the case of my viewer
(but I thankfully and by pure "luck" noticed the issue a few weeks before
LL did stealthily modify the login server on the main grid, because the
beta grid already had the changes which caused me to fail to login in it
at that time, and I could diagnose and fix the issue).

Lost a day out of my weekend diagnosing and resolving this in
LibreMetaverse/Radegast-ng. It really is a death blow to the unmaintained
OMV library. Heads up before this kind of deployment would be very
appreciated.

The definition of LLSD and our open source implementations of it have
always included the possibility of arbitrary nesting in arrays and maps
(and we use it extensively internally without problems). We're not able to
constrain our designs to maintain compatibility with incomplete
implementations we may not even know about, much less ones that are
unmaintained.

Understood. Naturally, the onus for compatibility is on the client
developer. In this case, the failure happened because the XmlRpc library
being utilized is poor quality… most C# XmlRpc libraries are. This has been
remedied in LibreMetaverse/Radegast. I had only wanted to mention advanced
warning is appreciated because these “rouge” implementations do exist and
people use them to connect.

Seconded. Always appreciated when API changes are documented on the wiki as
well.

I admit that I expected that an entry in a map that you didn't expect
wouldn't be a problem; I assumed that it would be ignored (as it is in
ours) and that it would be more useful to give you changes when there was
something you could test against. I suggest that in the future your
implementations be guided by Postel's Law.
<https://en.wikipedia.org/wiki/Robustness_principle>

Indeed. I have no commit access to OMV (which is why I forked
Libremetaverse) and a lot of it hr… questionable to say the least. :)

That having been said, we'll try to provide more advance warning for future
changes when possible. Note that the further in advance the notice comes,
the less specific and actionable it can be.

Thanks kindly, Oz. I do appreciate the notice and communication you and the
rest of the lab have and do provide via this list, open e-mail inboxes,
meetings, and wiki articles. It’s a lot more developer support than I’ve
had from Mojang and other companies writing third party addons and I don’t
take that for granted.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges

Re: [opensource-dev] Viewer changes for Premium changes

[Cinder Roxley]

On February 25, 2020 at 2:17:16 PM, Henri Beauchamp (sl...@hotmail.com)
wrote:

On Tue, 25 Feb 2020 10:48:03 -0500, Oz Linden (Scott Lawrence) wrote:

> *As we mentioned at the last Third Party Viewer meeting,

Oh yes, those meetings a non-English speaking developer cannot follow
because they are held in voice chat... How nice and practical a way
of communication !


Agreed. Even as a native English speaker, the meetings are not accessible
to those without access to voice. Summaries of the meetings on third party
blogs don’t really suffice for api breaking updates.

> Why are there changes?
>
> The viewer currently has some messages and costs hard coded

They are _*NOT*_ hard-coded. The cost is modified after login on
receiving the EconomyData message (the process_economy_data() callback
is implemented in indra/newview/llviewermessage.cpp).
By the way, that's how OpenSIM grid can (and always could) deal with
different costs than in SL.

Agreed on this point. When I implemented some of these same features in an
OpenSim grid, it was by extending the EconomyData message combined with
simulator cap.

Won’t go into implementation issues with the way it was done because it’s
already been done, outside of mentioning the rampant
singletons-calling-singletons-calling-singletons pattern is problematic. :)

Sadly, you used an LLSD sub-array into the LLSD map, and "old" viewers
(i.e. all viewers not based on LL's v3 viewer code for the XML RPC part)
do not know what to do with such an array (they can only deal with simple
key/value pairs, not with key/arrays); this was the case of my viewer
(but I thankfully and by pure "luck" noticed the issue a few weeks before
LL did stealthily modify the login server on the main grid, because the
beta grid already had the changes which caused me to fail to login in it
at that time, and I could diagnose and fix the issue).

Lost a day out of my weekend diagnosing and resolving this in
LibreMetaverse/Radegast-ng. It really is a death blow to the unmaintained
OMV library. Heads up before this kind of deployment would be very
appreciated.

> For example, one of the benefit tags is "texture_upload_cost"; its value
> is the number of L$ required for this user to upload a texture. The
> viewer displays that cost in the upload dialog so that dialog must be
> modified to use the value returned in the tag rather than the L$10 that
> is currently hard coded.

This particular usage is useless and redundant, since EconomyData/
process_economy_data() already can change that value in real time (and
in fact, it could even occur on a per-sim basis if needed, and not just
at login !).

Very true. It also handles the case where a customer changes premium levels
while logged in. They would only need a balance refresh not completely
exiting the viewer and logging back in (including an out of sync state
where an account is downgraded while the agent remains online.)

> Where are the changes?
>
> The changes are in the 'DRTVWR-481
> ' branch in
> the viewer git repository. A viewer built from that branch will be
> available as a Release Candidate soon.

It would have been nice to give a sufficient forewarning to avoid breaking
the login process on the main grid for many viewers (or viewer versions):
Radegast, my viewer (all versions before v1.26.24.2 are unable to login
in SL any more), OMV, etc, etc, etc.

Seconded. Always appreciated when API changes are documented on the wiki as
well.

I am (yet again) extremely disappointed with LL:
- No communication (a message in this list should have been posted even
before the change would have gone live on Aditi !),
- No forewarning (it's already too late for Agni as well !),
- No anticipation of the problems induced by the planed change,
- Not even a sane, simple, trivial precaution, such as respecting LL's
own viewer protocols design (here via the requested_options mechanism)
for something as essential as the login process !

Agreeing with this. There’s a wealth of knowledge that can be tapped into
via this mailing list that could have prevented some of these compatibility
issues. Remember that many of us have been blackboxing the API and
extending features without any access to the server side code for over a
decade now. Those of us with that breadth of experience rarely attend the
inworld meetings due to language, accessibility, and geographic constraints.
___
Policies and (un)subscribe information available here:
http://wiki.secondlife.com/wiki/OpenSource-Dev
Please read the policies before posting to keep unmoderated posting privileges