[opensource-dev] Any hope of STORM-727 getting a little love?

[Dave Booth]

It's graded "Minor" but I really think it falls into the "Major" 
category - Functionality is seriously broken (landmark creation) and the 
only workaround is to relog. I also think its importance is increased by 
the fact that v2 (and this bug has been around since 2.0) is supposed to 
provide a "better new user experience" and this is a bug that will most 
seriously impact new users. It's minor for longtimers who already have a 
well populated and pretty static folder of landmarks but for new users 
having the places tab cease functioning every time they create a 
landmark is a much more significant impact.

Dave.
___
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] Any hope of STORM-727 getting a little love?

[Dave Booth]

On 12/9/2010 23:54, Vadim Savchuk wrote:
> Thanks Nicky, but STORM-727 is already fixed. :-)`

Confirm fix works for me in latest dev build.
___
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] saving textures

[Dave Booth]

Anyone else having issues SAVING textures with the dev viewer?

With the 2.4.0 release viewer I can pull up a texture (that I created, 
so no perms issues) out of my inventory and hit "save as" and pull a 
fresh copy down to my hdd with no problems..

With the current dev build it hourglasses for a long time and then pops 
a notification of "unable to download texture"

Related to the kdu upgrade perhaps Merov?

Dave.
___
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] saving textures

[Dave Booth]

On 12/30/2010 08:28, Oz Linden (Scott Lawrence) wrote:
>
> 'current' is not a version identifier - please be specific.
>

Sorry Oz - by "current" I meant an up to date daily build - in that case 
build 217881, although I've just repro'd it on 217920 too

___
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] Convexdecomposition for open source devs

[Dave Booth]

On 12/30/2010 08:03, WolfPup Lowenhar wrote:
>
> I have recently had a conversation with someone @ LL that has said 
> they would be willing to help develop and OS version of 
> llconvexdecompisition. I even I even discussed which ones they thought 
> was good but they have not had the time to look at the open source 
> version in a technical way. So what I thought of is start a thread 
> here that would allow the OS community both vote and comment on the OS 
> versions with the one most generally liked being used and converted to 
> be compatible to the LL mesh viewer.
>
> Versions that I have found(to place your vote on this put an x next it.):
>
> Bullet Physics Library(just the convexdecomp section):
>
> Web site : http://code.google.com/p/bullet/
>
> Votes>
>
>
>   John Ratcliff's :
>
>
>   Web Site :
>   
> http://codesuppository.blogspot.com/2009/11/convex-decomposition-library-now.html
>
>
>   Votes>
>
>
>   Comments for either system:
>
>

I cant exactly speak to the quality of either but the fact that Bullet 
is the physics implementation used by opensim should IMHO give that 
choice a bit of a leg up. The fact that the library is already in use as 
the phys engine on the most prevalent opensource server makes its use 
for decomp in an opensource viewer rational, just as the use of havok 
decomp in the LL-compiled viewer made sense since that physics library 
was already present in-house.
___
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] Review Request: VWR-24420: PNG images which specify "background color" lose alpha layer when imported.

[Dave Booth]

Thats a pretty good bit of thinking out loud there, Ponzu

Personally I'd say thats a "clean" fix for this. However it all depends 
on how things are stored server-side - if the asset server code 
automatically assumes that every texture has an alpha channel and 
supplies a "solid" one for those where the upload doesnt provide one, 
any benefit of specifying at upload time is moot.

On 1/10/2011 10:34, Ponzu wrote:
> A common cause of z-buffer problems in SL is textures that do not 
> *need* an alpha channel that contain one anyway.  Lots of content in 
> SL is created by rank amateurs (me for example).
>
> What if the upload dialog had a check box?
>
> [ ] This texture should have an alpha channel.
>
> If the box is not checked, the alpha channel is removed.  If it is 
> checked, it is retained, unless it is not there, in which case the 
> user gets an error message.
>
> Just thinking out loud.
>
> ponzu
>
> On Sun, Jan 9, 2011 at 9:59 AM, Thickbrick Sleaford 
> mailto:thickbrick.sleaf...@gmail.com>> 
> wrote:
>
> This is an automatically generated e-mail. To reply, visit:
> http://codereview.secondlife.com/r/74/
>
>
> Review request for Viewer.
> By Thickbrick Sleaford.
>
>
>   Description
>
> Current code composites RGBA PNG images that contain a bKGD chunk down to 
> RGB, discarding the alpha channel. This patch removes that code, since it 
> contradicts purpose of the bKGD chunk as described in the PNG spec and as 
> commonly used.
>
>
>   Testing
>
> Tested uploading the 2 images attached to VWR-24420 with and without the 
> patch. Before patch, "bad alpha.png" was uploaded as RGB, after patch, both 
> images were uploaded as RGBA.
>
> *Bugs: * VWR-24420 
>
>
>   Diffs
>
> * doc/contributions.txt (UNKNOWN)
> * indra/llimage/llpngwrapper.h (UNKNOWN)
> * indra/llimage/llpngwrapper.cpp (UNKNOWN)
>
> View 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
>
>
>
> ___
> 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] "Nearby" people tab

[Dave Booth]

Whats the expected behavior of this tab of the people sidebar? I'm 
currently sitting in a room with 14 other folks, half of whom are on my 
flist and all within chat range - but this tab tells me "no one nearby". 
Currently on Second Life 2.6.0 (21) Jan 29 2011 12:57:40 (Second 
Life Development) but to be honest I dont think I've EVER seen that tab 
populated, no matter what build or how crowded the venue.
___
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] "Nearby" people tab

[Dave Booth]

On 1/31/2011 07:52, Oz Linden (Scott Lawrence) wrote:
> On 2011-01-30 7:47, Hitomi Tiponi wrote:
>> I have been having this trouble on some sims for the last few days 
>> (using 2.5.1 Beta 2 and later).  I have been trying to figure out the 
>> cause or something that may tie in the various sims it seems to 
>> happen on - but so far without luck.  On other sims this works as 
>> expected.
>> And yes - this happens on sims at ground level as well.
>>
>> ---
>>
>> >Date: Sat, 29 Jan 2011 17:50:23 -0600
>> >From: Dave Booth > <mailto:d...@meadowlakearts.com>>
>> >Subject: [opensource-dev] "Nearby" people tab
>> >To: OpenSource Mailing List > <mailto:opensource-dev@lists.secondlife.com>>
>> >Message-ID: <4d44a7bf.9090...@meadowlakearts.com 
>> <mailto:4d44a7bf.9090...@meadowlakearts.com>>
>> >Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>> >
>> >Whats the expected behavior of this tab of the people sidebar? I'm
>> >currently sitting in a room with 14 other folks, half of whom are on my
>> >flist and all within chat range - but this tab tells me "no one 
>> nearby".
>> >Currently on Second Life 2.6.0 (21) Jan 29 2011 12:57:40 (Second
>> >Life Development) but to be honest I dont think I've EVER seen that tab
>> >populated, no matter what build or how crowded the venue.
>
> It would be interesting to see what was in the log file for the period 
> of time that the window was not properly filled in.

I intend to capture some more specific data in the next few days - will 
involve laying out precisely positioned prims in 3 dimensions for an av 
to sit on and probing what "NearMeRange" distances they appear and 
disappear from the nearby tab to an alt sat on the zero-ref prim. If 
properly designed, that study will establish any issues with height AGL, 
"flat" (2d) vector magnitude and "true" (3d) vector magnitude. Will 
include logfile capture and will advise.

Dave.
___
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] Any possibility of playing animation b y uuid?

[Dave Booth]

On 4/17/2011 2:02 PM, Brian McGroarty wrote:
> On Sun, Apr 17, 2011 at 9:39 AM, Brandon Husbands  wrote:
>> We cal already play sounds, do textures why cant we play a animation by uuid
>> via lsl / viewer.
>
> Repeating what was done with textures and sounds is a non-starter;

agreed - thats a matter of basic protection for the hardworking 
animation creators.



> If reference by asset ID is important enough that you'd want to work
> on it, lay out a proposal detailing what permissions could be baked
> into an asset at upload time, and how the permissions could be honored
> by all viewers.



> Any proposal should maintain current restrictions against raw use of
> existing animation assets, which don't have extra permissions metadata
> attached.



Trying to think outside the box here, and not a fully thought out 
proposal, but are there possibilities for a lsl function something like 
llStartAnimByUUID(key agent, key animKey) that can only access an anim 
by uuid if the creator of the script and the anim are the same, and the 
script is nomod and one of nocopy or notransfer? If not the function 
call would just silently fail. I can see holes in that protection right 
off the bat, but most of them are avoidable by a creator taking care. 
Just a thought.

Dave.
___
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] OpenID based logins?

[Dave Booth]

DO NOT WANT.

My OpenID is linked to several RL sites that I want to have nothing to 
do with my SL. OpenID has already been pre-empted by 1L or RL  links. If 
it were 100% opt-in and guaranteed to never ever link to SL without 
explicit consent I'd support it because I know theres SLers out there 
that merge their identities that way and why shouldnt they? However, I 
have zero confidence that any SL implementation of openID  would meet 
those criteria.


On 4/21/2011 2:52 PM, Brandon Husbands wrote:
> Any thoughts to being able to link your SL account to an open id
> identity server?
>
> --
> ---
> This email is a private and confidential communication. Any use of email
> may be subject to the laws and regulations of the United States. You may
> not Repost, Distribute nor reproduce any content of this message.
> ---
> ---
>
>
>
> ___
> 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] Wierd experience last four builds.. just me?

[Dave Booth]

I've been experiencing a crash on initial login with the last four 
builds. The FIRST time I connect with a new build download it 
bluescreens my PC, subsequent logins work just fine..

Anyone else experiencing anything similar?

Config details below:

Second Life 2.7.5 (233708) Jun 23 2011 10:46:18 (Second Life Development)
Release Notes

You are at 250,670.0, 245,092.0, 4,011.0 in Mextli located at 
sim9935.agni.lindenlab.com (216.82.46.77:13004)
Second Life Server 11.06.14.232746
Release Notes

CPU: Intel(R) Core(TM) i7 CPU   Q 740  @ 1.73GHz (1729.02 MHz)
Memory: 6125 MB
OS Version: Microsoft Windows 7 64-bit Service Pack 1 (Build 7601)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: GeForce GT 425M/PCI/SSE2

Windows Graphics Driver Version: 8.17.0012.7533
OpenGL Version: 4.1.0

libcurl Version: libcurl/7.21.1 OpenSSL/0.9.8q zlib/1.2.5 c-ares/1.7.1
J2C Decoder Version: KDU v6.4.1
Audio Driver Version: FMOD version 3.75
Qt Webkit Version: 4.7.1 (version number hard-coded)
Voice Server Version: Not Connected
Built with MSVC version 1600
Packets Lost: 1/6,812 (0.0%)
___
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] Blocked

[Dave Booth]

Socksified works but yeah, takes more "liberal" corp IT attitudes than 
most will allow. I sent offlist the solution I used..  (openvpn  to home 
networks static IP and route all SL traffic down that using local routes 
on client and openvpn server)

On 7/15/2011 12:53 AM, Tateru Nino wrote:
> That's assuming they have the right kind of proxies there. HTTP proxies
> are commonplace, but SOCKS proxies are far less common and more tightly
> controlled on the corporate networks. He almost certainly has access to
> the former, but maybe not to the latter.
>
> On 15/07/2011 3:45 PM, Lance Corrimal wrote:
>> Isn't it a good thing that there's a proxified viewer now... can
>> someone point him/her at the downloads for a build with the proxy
>> stuff?
>>
>>
>> bye,
>> LC
>>
>>
>> Am Freitag, 15. Juli 2011 schrieb a...@skyhighway.com:
>>> Robert,
>>>
>>> thx for the info, but i think the page is telling me the answer is
>>> actually, "No," because i don't have any access to the company's
>>> network stuff.  They look at you funny if you even ask about it.
>>> i can configure a test machine any way i want to, but i can't do
>>> anything at all at all to the network.  How depressing :-(
>>>
>>> Since it is for a real work project, i think i'll show the page to
>>> the IT guy, but i think this is just going to be another one of
>>> those funny looks experiences, and a "I'll get back to you later"
>>> that ends up being a not overly polite "helpful suggestion" from
>>> my boss.  But i think it will be worth it just for the opportunity
>>> of getting to explain what i'm doing.
>>>
>>> - AK
>>>
>>> On Thu, Jul 14, 2011 at 10:17 PM,  wrote:
> Hey y'all, i had something i wanted to do at work a coupla days
> ago
>>> that i
>>>
> coulda used SL for.  It installed, but i couldn't get it to
> connect.  Is there some kinda proxy setting that i need to fix
> to get SL to start from where i work?
>>> short answer is yes
>>> long answer is
>>> http://wiki.secondlife.com/wiki/Configuring_your_firewall
>>>
>>> -- Robert L Martin
>>>
>>> ___
>>> 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

Re: [opensource-dev] Blocked

[Dave Booth]

On 7/15/2011 2:44 AM, Lance Corrimal wrote:
>
> how does that works in a http-proxy-only network?
>

if it's locked down that tight, with only http connections allowed 
outbound and all those forced to go via proxy then of course it doesnt. 
But all you need is ONE outbound service that can avoid the proxy - I've 
had best results setting the vpn endpoint to use the https port, but if 
necessary, all you need is outbound ssh whether its proxied or not - you 
can "add a layer" and tunnel a ssl-based vpn over that. I know 
"tunneling a tunnel" isnt exactly best practice, and theoretically you 
could just tunnel the SL ports with ssh alone but the advantage of using 
a vpn is that it appears as a separate network interface to the local 
machine and it's therefore easier to route the LL public subnets traffic 
through it reliably. It's always easier to poke holes in a firewall or 
proxy from the inside out, they are designed to prevent unpleasantness 
coming the other way :)
___
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] [JIRA] Proposal - Default Search settings on jira.secondlife.com

[Dave Booth]

Opinions inline...

On 8/26/2011 12:34 PM, opensourceobscure wrote:

> PROPOSAL:
>
> 1) exclude "Comments" from fields where keywords are searched for

Agree

> 2) exclude "Description" too (this may be excessive but usually works for me)

Disagree on the grounds that excluding this one would negatively impact 
one category of benefits you hope to achieve, namely the reduction of 
duplicate issues. Users tend to express the same issue in different 
words when dealing with two similar fields on the same form. Searching 
both makes the target of the search "wider" in that the original 
reporter has expressed it two ways and the search (reflecting the way 
the SEARCHER thinks of the issue) is more likely to hit one.

> 3) replace the "OR" logic operator (which is implicit in searches)
> with the "AND" logic operator

Oh heck, YES. Include the "quoted multiword string as single search 
token" functionality too.

>
> I think that this would lead to a better use of JIRA for
> SL users willing to give feedback. More specifically,
> the number of duplicate issues may be reduced because
> the search for existing issues would be more efficient.
>

___
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] Review Request: storm-1578: change default Away timeout to never

[Dave Booth]

With Soft's suggestion implemented, I'd set an away timeout..  as things 
stand its always "never"

On 8/31/2011 10:53 AM, Brian McGroarty wrote:
> What about removing the automatic disconnect instead?
>
> Avatars are disconnected after some amount of time in the away state,
> which is silly given how few resources a minimized viewer window uses.
> The reason many people turn off the away timer is that they don't want
> to be disconnected, not that they don't want others to know they will be
> unresponsive.
>
> Without automatic disconnect, we'd get a useful social cue back.
>
> On Wed, Aug 31, 2011 at 7:29 AM, Oz Linden  > wrote:
>
> This is an automatically generated e-mail. To reply, visit:
> http://codereview.secondlife.com/r/459/
>
>
> Review request for Viewer.
> By Oz Linden.
>
>
>   Description
>
> Simple change to the Away timeout default value (to never).
>
> *Bugs: * storm-1578 
>
>
>   Diffs
>
>   * indra/newview/app_settings/settings.xml (04642a178228)
>
> View 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
>
>
>
>
> --
> Brian McGroarty | Linden Lab
> Sent from my Newton MP2100 via acoustic coupler
>
>
> ___
> 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] Current status of Mesh??

[Dave Booth]

On 8/31/2011 12:16 PM, Robert Martin wrote:

> Questions i have
>
> 1 what is the conversion of Vertexes to Prims?? (how many vertexes do
> you get for a "prim")
Already answered by many that it aint that simple, not going there.

> 2 are there any simple tools to make models (Blender does not qualify).
There are many simple tools to make BAD meshes, there are no simple 
tools to make GOOD ones.

> 3 are non-biped avatars supported in mesh?? (quads are the biggest concern)
We have it on record that custom bones are one of the features the folks 
coding mesh would love to implement, we just dont have a timescale.
___
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] ok wtf is this?

[Dave Booth]

With the latest build I keep getting an alert box for "memory pool low, 
some sl functions disabled to avoid crash" and sure enough I crash a 
couple minutes later..   but memory pool low?  I've usually got over 2g 
of physical memory free when it hits, and the machine isnt even breaking 
a sweat. I dont see a ramp up in memory use that might indicate a leak, 
I look at the logs and I see the latest ERROR level entry is a "bad 
memory allocation" for a texture. Somebody who's more current with the 
code tell me where I should look to narrow this thing down enough to 
file a JIRA...
___
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] ok wtf is this?

[Dave Booth]

On 9/5/2011 12:53 AM, Tateru Nino wrote:

> In order of likelyhood (most to least), I'd say: corrupted texture
> header data, corrupted heap, fragmented heap.
>

After a few moments of reflection of my own, I agree with your 
assessment..  The question that springs to mind is why does it happen in 
an area where I know all the textures have been loaded successfully in 
previous builds without any of the three issues showing up?
___
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] ok wtf is this?

[Dave Booth]

On 9/5/2011 2:32 AM, Nicky D. wrote:

> There is a memory pool that got enabled 4 days ago. Which could cause
> this problem aswell.
>
> The change is here:
> https://bitbucket.org/lindenlab/viewer-development/changeset/298722ecdb2e
>

oh yeah, that's an "aha!" moment for sure.. Thanks, Nicky.

Checking the "changes since last good" on the download page there it is 
on build 239990, the most recent, which correlates exactly with when I 
started to see this failure mode. If switching it off means I dont crash 
at all tonight then I'll leave a little gift on JIRA for the kindly 
folks at the lab to discover after their long weekend :)

Dave.
___
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] ok wtf is this?

[Dave Booth]

On 9/5/2011 11:14 AM, Tateru Nino wrote:
>
> Indeedy. An excellent catch. If it _is_ the memory pool, though, it is
> hard to figure out why unit-testing didn't catch the underlying fault.

At a guess, and this is total speculation because I havent done any side 
by side test runs with the setting on and off yet followed by spelunking 
through the logs, I suspect that the easiest way to escape notice in the 
unit tests is if it's a flaw elsewhere that was masked by not using the 
memory pool - unit tests could easily not catch that, things like that 
usually show up in integration testing and UAT and well, that's what 
folks like me are for isn't it - pulling a new build every day and 
spending time inworld with it to see what breaks :)
___
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] ok wtf is this?

[Dave Booth]

On 9/5/2011 10:13 PM, Erin Mallory wrote:
> maybe their machines have more memory? maybe they didnt have the proper
> test? I dunno but i see it too...

Even if the size of the memory pool allocated was impacted by the total 
available RAM on the machine, which I'm not sure it is, I can see them 
having more memory on their test machines than I do on my laptop, which 
has only 6G but I've twice that on my gaming rig and I saw the issue on 
both. Too soon to be absolutely certain yet, but so far I've been 
inworld for hours on my laptop tonight with the private memory pool 
switched off in the debug settings and havent had a problem, so at least 
there are indications of a workaround. I'm thinking Bao still has a 
little debugging to do. Once I get a chance to run a test or three on 
the gaming rig I'll file the JIRA with my results.

___
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] ok wtf is this?

[Dave Booth]

Having experienced several hours without the issue recurring on both my 
systems  since turning off the use of the private memory pool, have 
filed https://jira.secondlife.com/browse/VWR-26864

Have fun finding out what crawled out from under the rock you flipped 
over, Bao..  Sooner you than me in tracking this one down ;)

___
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] ok wtf is this?

[Dave Booth]

On 9/6/2011 1:05 PM, Robert Martin wrote:
> On Tue, Sep 6, 2011 at 3:25 AM, Nicky D.  wrote:
>
>> As strange as this might sound, 6/12 GB might be too much.
>>
>> There are a lot of casts down to U32. There might be a hidden overflow there,
>> turning some counter into a very small number.
>> ___
> it may then be something in the pool causing a C5* type error where
> the code does not handle that large of a total memory and wraps around
> the memory space.
>
>
> * a C5 Galaxy is a very large cargo plane and if you are not careful
> things can get very messy. Even some airports that handle Jumbo Jets
> can't handle a C5.
>
> Empty weight: 380,000 lb (172,370 kg)
> Loaded weight: 769,000 lb (348,800 kg)
>
> anybody wanna bet that some counter in the code should be the next size 
> bigger??
>
>
>

The only ones I ever shipped on were your good old reliable herky-birds 
:) I'm betting the level of discomfort when they were rigged with seats 
probably scaled up in proportion to the size with the galaxy? Sooner you 
than me dude.
___
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] ok wtf is this?

[Dave Booth]

Gotta give kudos to Bao on this one..

After Grumpity assigned it, 1 minute from status "Acknowledged" to "Fix 
Pending" - If you can zero in on it that fast, Bao, you are really in 
tune with your code :)

Looking forward to seeing your fix hit the trunk :)
___
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] Weird issue with mouselook flight in newer viewers

[Dave Booth]

Where did the limits on mouselook angle come from in the newer builds? 
In current development a mouselook-steered aircraft cant pitch up beyond 
80 degrees and cant pitch down beyond 40. It's got to be in the viewer 
code somewhere because using TPVs like the current Firestorm beta I can 
pitch freely between +90 and -90 in the same vehicle. This is breaking a 
LOT of existing content, every aircraft that uses ML steering. Any 
ideas, folks? It's hard to notice without fully functional flight 
instruments but its a real problem - when a flight envelope can change 
just based on the version of viewer code aircraft combat is blown into a 
cocked hat...
___
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] Weird issue with mouselook flight in newer viewers

[Dave Booth]

On 10/20/2011 11:06 AM, Ricky wrote:
> Those angle limits have been in place for many, many years.  Though I
> agree that having them has little purpose, they have, nevertheless
> been there the whole time.
>

So in other words I've been living with it without really noticing until 
I actually started building my own aircraft and made myself an 
instrument panel with a fully functional artificial horizon.

I guess it's JIRA time, although its hardly going to be a high priority 
as it will be considered an enhancement rather than a fix. In the 
meantime I guess I'm switching over to Firestorm whenever I'm going to 
be flying through any area I might get shot at, since having a wider 
flight envelope than anyone using a LL viewer could well be useful :)
___
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] Weird issue with mouselook flight in newer viewers

[Dave Booth]

On 10/20/2011 1:51 PM, Dave Booth wrote:
>
> I guess it's JIRA time

VWR-27241 filed, requesting removal of the mouselook limits imposed by 
code in /indra/newview/llagent.cpp - There is no reason to limit this so 
tightly and nearly all TPVs out there allow pitch between -90 and +90 
without restriction.
___
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] Missing details in Snowstorm development builds

[Dave Booth]

On 10/19/2011 10:56 AM, Oz Linden (Scott Lawrence) wrote:
> I've asked our tools team to look at this...

Have they made any progress?
___
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-development-builds] Successful Build Nr. 245237 for snowstorm_viewer-development (cfa78777453c) on CYGWIN

[Dave Booth]

On 11/18/2011 11:20 AM, Tillie Ariantho wrote:
>
> That build crashes the nVidia driver pretty often. Right now I can't
> even log in with the dev client, crashes the driver right after
> clicking Login.

Confirmed here too, same driver. Previous workaround does not work,
viewer crashes as soon as basic shaders are enabled no matter what point
it's done - on my system this is a regression from previous behavior
when the GL bug could be avoided by setting gfx prefs to mid or lower 
for login and setting them back up to desired settings after everything 
rezzed. Only workaraound for current behavior is a registry tweak:
Create or set the following key in
HKLM\System\CurrentControlSet\Control\GraphicsDrivers

TdrDelay:REG_DWORD set to 60 decimal to allow a 60s graphics driver
timeout rather than the 2s default.

Note that changing this value from default is considered a serious
failure by Microsoft and absolutely precludes an app from gaining
windows logo certification if it is required for normal operation.
___
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] Is the build machine stuck?

[Dave Booth]

On 11/20/2011 12:41 AM, Hitomi Tiponi wrote:
> It does appear to be stuck still - it's not the first time this has
> happened. Also the problem with not showing changes in the builds has
> still not been solved when it does complete.
>

Ricky found the issue with the latter of those.. posted it here

>  Turns out that the reason why it isn't displaying
> changesets or JIRA entries is because the JSON format was changed
> without correcting the Javascript.  What's more, the JSON information
> doesn't have the needed data anymore.

He even mocked up a fix...
___
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-development-builds] Successful Build Nr. 245237 for snowstorm_viewer-development (cfa78777453c) on CYGWIN

[Dave Booth]

On 11/18/2011 5:58 PM, Dave Booth wrote:
> On 11/18/2011 11:20 AM, Tillie Ariantho wrote:
>>
>> That build crashes the nVidia driver pretty often. Right now I can't
>> even log in with the dev client, crashes the driver right after
>> clicking Login.
>
> Confirmed here too, same driver. Previous workaround does not work,
> viewer crashes as soon as basic shaders are enabled no matter what point
> it's done - on my system this is a regression from previous behavior
> when the GL bug could be avoided by setting gfx prefs to mid or lower
> for login and setting them back up to desired settings after everything
> rezzed. Only workaraound for current behavior is a registry tweak:

245610 is even worse..  even the registry hack doesnt work. As soon as 
you enable basic shaders you're looking at a hard crash. Guys, this isnt 
progress, this is progressively worse regressions. We need to get the 
shining fixes into the trunk ASAP - it's progressing from an 
inconvenience to a showstopper real fast. I hope the GL fixes in shining 
solve 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] [viewer-development-builds] Successful Build Nr. 245237 for snowstorm_viewer-development (cfa78777453c) on CYGWIN

[Dave Booth]

On 11/21/2011 9:09 PM, Dave Booth wrote:
> On 11/18/2011 5:58 PM, Dave Booth wrote:
>> On 11/18/2011 11:20 AM, Tillie Ariantho wrote:
>>>
>>> That build crashes the nVidia driver pretty often. Right now I can't
>>> even log in with the dev client, crashes the driver right after
>>> clicking Login.
>>
>> Confirmed here too,
...
> 245610 is even worse..  even the registry hack doesnt work.

Do we have an eta for the shining branch GL fixes hitting the trunk yet? 
Even a back-of-the-envelope guesstimate? I'm tired of crashing every 
time I set my graphics higher than I'd use on a machine half a dozen 
years older than the one I am actually on, or even up to the settings I 
used to use on my previous box..
___
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] Is the build machine stuck?

[Dave Booth]

On 11/19/2011 9:26 PM, Cincia Singh wrote:
> I have seen it saying "in progress" for 2 days now. Is it not finishing?
> Both Snowstorm and the shining fixes web page for the OpenGL builds.
>

It's doing it again. 245787 has been showing in progress for  over 24h.
___
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] Is the build machine stuck?

[Dave Booth]

On 11/29/2011 6:30 PM, Dave Booth wrote:
> On 11/19/2011 9:26 PM, Cincia Singh wrote:
>> I have seen it saying "in progress" for 2 days now. Is it not finishing?
>> Both Snowstorm and the shining fixes web page for the OpenGL builds.
>>
>
> It's doing it again. 245787 has been showing in progress for  over 24h.

And this breaks the quicklink from the wiki to the latest build too. 
That still points to Wednesdays 245711.


___
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] mesh upload crash in recent builds...

[Dave Booth]

Somewhere between the branch for the current beta and the tip is a 
mistake leading to an instant crash when selecting "upload -> model" 
from the inventory floaters "+" menu. It's a 0xc5 exception, the 
windows equivalent of your garden-variety segfault, pulling up the dmp 
file in vs2010 looks like the viewer is passing a null pointer into a 
call to strstr somewhere.
___
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] mesh upload crash in recent builds...

[Dave Booth]

On 1/24/2012 4:20 PM, Jonathan Welch wrote:
> Do you get the same result/crash when trying to upload via a different
> menu entry point?
>
Yep, Build > Upload > Model produces the same result as Inventory 
Floater > + > Upload > Model. Immediate crash with the same exception 
before the dialog to select a .dae is displayed.

Its mesh-specific, no crash trying to upload any other kind of asset. 
It's relatively recent, whichever changeset caused it, a little while 
ago I was able to upload meshes no problem and the problem is not 
present in the current beta download.

___
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] mesh upload crash in recent builds...

[Dave Booth]

Clarifying "relatively recent" - my last successful mesh uploads on the 
dev snapshot builds were on Jan 2nd, using 3.2.6.246889 from Dec 20th (I 
only update my download when the commits list gets a note) so whatever 
introduced this was in the 31 changesets on 1/6, the 41 on 1/8, the 6 
(in two merges) on 1/9, the 15 on 1/17 or the 11 on 1/23. Obviously any 
of those merges predating the branch for the current beta can be 
excluded, which makes me look exceptionally suspiciously at the two 
builds since the version was updated to 3.2.8, namely 247796 on 1/22 and 
248202 on 1/23 - but that dont stop me suspecting every build since I 
last successfully used it to upload a mesh :P

Spending 30 years as a sysadmin makes me VERY willing to believe that a 
developer can break stuff in ways that defy logic :)

On 1/24/2012 4:32 PM, Dave Booth wrote:
> On 1/24/2012 4:20 PM, Jonathan Welch wrote:
>> Do you get the same result/crash when trying to upload via a different
>> menu entry point?
>>
> Yep, Build>  Upload>  Model produces the same result as Inventory
> Floater>  +>  Upload>  Model. Immediate crash with the same exception
> before the dialog to select a .dae is displayed.
>
> Its mesh-specific, no crash trying to upload any other kind of asset.
> It's relatively recent, whichever changeset caused it, a little while
> ago I was able to upload meshes no problem and the problem is not
> present in the current beta download.


___
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] mesh upload crash in recent builds...

[Dave Booth]

reverified with todays build, VWR-28207 created.

On 1/24/2012 4:02 PM, Dave Booth wrote:
> Somewhere between the branch for the current beta and the tip is a
> mistake leading to an instant crash when selecting "upload ->  model"
> from the inventory floaters "+" menu. It's a 0xc5 exception, the
> windows equivalent of your garden-variety segfault, pulling up the dmp
> file in vs2010 looks like the viewer is passing a null pointer into a
> call to strstr somewhere.
> ___
> 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] mesh upload crash in recent builds...

[Dave Booth]

Dan took it into SH as SH-2918. Guess its gonna be fixed :)


On 1/25/2012 8:46 PM, Dave Booth wrote:
> reverified with todays build, VWR-28207 created.
>
> On 1/24/2012 4:02 PM, Dave Booth wrote:
>> Somewhere between the branch for the current beta and the tip is a
>> mistake leading to an instant crash when selecting "upload ->   model"
>> from the inventory floaters "+" menu. It's a 0xc5 exception, the
>> windows equivalent of your garden-variety segfault, pulling up the dmp
>> file in vs2010 looks like the viewer is passing a null pointer into a
>> call to strstr somewhere.
>> ___
>> 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] Removal of the "MultipleAttachments" debug settings ?

[Dave Booth]

  On 8/26/2010 15:54, Marine Kelley wrote:
> I understand, it makes sense, thanks for your reply Nyx. I did see the 
> Add option on 2.1.0 along with MultipleAttachments there too so I 
> assumed both were needed for some reason (this debug setting was used 
> in a few places in the code).
>

So, Marine, I'm guessing theres going to be a new UI element in the next 
RLV release  - a checkbox that toggles the behaviour of the force-attach 
option between "wear" and "add". Either that or you've got the pain of 
revising the API yet again and having @attach:folder=force do a "wear" 
and @attach:folder=forceadd do an "add" - along with a note that for 
backwards compatibility, any viewer that doesnt implement add must treat 
forceadd as equivalent to force, just to spread the pain down to the 
other folks that implement 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

Re: [opensource-dev] Removal of the "MultipleAttachments" debug settings ?

[Dave Booth]

  On 8/30/2010 19:38, David Couchenour wrote
> Someone can edit and add the additional attachment points to the
> avatar_lad.xml file for the other viewer(s). I've done this for the
> 2.1.2 viewer and it seems to work. Note that you can't simply copy the
> avatar_lad.xml from Emerald to 2.x.
>
> This way a 2.x viewer can see the Emerald attachments correctly, and
> still have use of the newer multi-attachment feature too. Essentially
> making it backwards compatible with Emerald, but Emerald still cannot
> see the newer attachments.

I disagree. The proper way to implement multiattach is to recognize a 
list and render them all..  everyone that was using Emeralds ugly hack 
needs a backwards compatible kludge that isnt ever going to be forwards 
compatible. So to equip viewer 2 with the kludgy emerald attachment 
points would be to bake the kludge into the mainstream and we dont want 
that. I would suggest instead the following user story... "As a user who 
has become familiar with the TPV extensions to avatar_lad.xml I want the 
official viewer to recognize those "hacked" additions and silently 
convert them to multiattach-compatible additions to the attachment list 
on the appropriate point."
___
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] Blocking viewers.

[Dave Booth]

  On 9/8/2010 16:39, Tom Grimshaw wrote:
>Dear Linden Lab,
>
> It's absolutely none of your business what software I choose to run on
> my PC.
>
> Blocking emerald is a step of pure arrogance - and ignorance - on Linden
> Lab's behalf - it's not having an adverse effect on your servers, in
> fact THE ONLY WAY you can tell i'm runing Emerald is by the channel and
> version provided in login, and this has been proven by the number of
> clones that have popped up with their channel renamed (and the ID
> texture changed of course). You cannot censor Open Source software, and
> the fact that you're trying to makes you a despicable organisation.
>
> Stop policing my computer. I will decide what viewer I use, thank you.
>
> ~T
> ___
> 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
>
ye gods, enough with the drama. We expect it on forums and blogs but on 
THIS list??? just grow a pair. Bottom line, Emerald devs screwed up and 
violated TPV, they are gone and so is their malware-infected viewer. LL 
did exactly what any responsible org would do and nixed 'em. Done. Over. 
END.

Now either get over it and pick an alternative viewer or dont let the 
door hit your posterior on the way out, ok?
___
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] The Plan for Snowglobe

[Dave Booth]

  On 9/10/2010 17:30, Henri Beauchamp wrote:
> ROFLOL 
>
> You are kidding, aren't you ?... If not, then please try using a good
> TPV and see how many times it crashes in a week... 0 for the Cool VL
> Viewer (and I'm using it every day). Fact is that TPVs got fixes that
> v1.23.5 doesn't have and that makes them MUCH stabler than 1.23.5, and
> v2 (which is even worst, stability-wise, than v1.23.5 !). The reason
> is simple: should I crash, I trace the crash down and fix the code.
> Crash gone !
Well, Henri, I've been using the daily build of viewer-development for 
two weeks and have only crashed once and I can definitively blame that 
on a video card overheat not the v2 code.

I appreciate the wish to backport rather than refactor everything to 
match the v2 codebase but as an inworld creator for a niche market I 
have to switch to v2 since the primary coder in that market is also 
migrating. I would regret being unable to use Cool for my primary 
testing alt. If LL is unwilling to wrap a 1.x UI around 2.x code, then 
surely that is an opportunity for skilled devs like you rather than the 
slap in the face that you seem to be taking it as.

Dave
___
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] where is the source

[Dave Booth]

  On 9/13/2010 19:56, Yoz Grahame wrote:
>
> On 13 September 2010 16:48, Joshua Bell  > wrote:
>
>
> The server side changes to deliver a static URL to an LLSD
> resource from login are minimal; if someone wants to take a stab
> at the client side changes and defining a forward-looking LLSD
> format, I'm sure you'll find a server-side champion.
>
>
> Hello!
> (Ghengis and I recently made a long-overdue login.cgi fix that makes 
> this kind of change only require login.xml changes, though we still 
> need to merge it.)
>
> -- Yoz
>
>
> ___
> 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
Get it QA'd and merged dude :)
___
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] Bug report or feature request?

[Dave Booth]

  In pretty much all v2 viewers, including development, I've noticed a 
weird behavior with group notices... When they arrive you get a little 
notification in the bottom right of the screen and that persists for a 
short time and then goes away exactly as one would expect... However if 
you subsequently go to the icon in the bottom bar and pull up the 
notification history then click on one of the notifications for a group 
notice because you want to read it, the floater that comes up to display 
the notice ALSO fades out after a few seconds, even when its focused and 
you're half way through reading the notice. Intended behavior or not? 
Either way I need to file a Jira, but if its intended behavior I'll be 
making it a feature request to allow me to stop this rather than a bug 
report. Seems to me that when a notice has been explicitly pulled up to 
read it then it shouldnt go away until explicitly dismissed.

Or am I just missing something in the settings somewhere?

Dave
___
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] VWR-20094 - Maybe still a hope for scripters with viewer 2?

[Dave Booth]

  On 9/24/2010 04:57, Satomi Ahn wrote:
> Hello all,
>
> I just wanted to mention that I found how to fix the bug that was
> keeping all scripters from switching to viewer 2 (the one that made
> the viewer freeze on loading large scripts).
Oh trust me, its a good one to fix but it sure isnt "the" bug :) 
Personally, the one that gets furthest up my nose is the cursor drift - 
I typically have my UI sized way down and the incorrect cursor tracking 
makes editing anything text-based, script or notecard, a royal pain in 
the fundament. It doubly annoys me because one of my RL hobbies - 
letterpress printing. Sometimes I'd swear I could get the same text into 
my composing stick more efficiently than into a notecard and theres 
really no excuse for that. I appreciate that font handling and cursor 
tracking has almost as many quirks as 3d rendering but still.
___
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] VWR-20094 - Maybe still a hope for scripters with viewer 2?

[Dave Booth]

  On 9/24/2010 08:49, Opensource Obscure wrote:
> On Fri, 24 Sep 2010 08:38:38 -0500, Dave Booth
> wrote:
>> On 9/24/2010 04:57, Satomi Ahn wrote:
>>> Hello all,
>>>
>>> I just wanted to mention that I found how to fix the bug that was
>>> keeping all scripters from switching to viewer 2 (the one that made
>>> the viewer freeze on loading large scripts).
>> Oh trust me, its a good one to fix but it sure isnt "the" bug :)
>> Personally, the one that gets furthest up my nose is the cursor drift
> what is the PJIRA entry for that?
> does it affect all viewer releases?
>
> opensource obscure
>
Theres plenty of them - most refer to notecards, because the problem 
shows up easier in paragraph-based stuff like notecards. The more text 
you have in a "line" (which a nc will wrap as a paragraph) the worse it 
gets. An explicit \n resets it but in a long code line it happens too, 
particularly when you have a UI scaling to anything other than 1.0 (some 
folks report it when scale>1.0, some when scale<1.0) You can click 
somewhere in a line of code and start typing to insert characters and 
then see them inserting several places to the left of where the click 
and the blinking cursor indicate they should, or see the cursor getting 
further and further ahead of the characters you're typing, if you're 
typing a new line.

https://jira.secondlife.com/browse/VWR-20178 and its dupes are the 
clearest description of it for notecards,
https://jira.secondlife.com/browse/VWR-18875 and 
https://jira.secondlife.com/browse/VWR-21388 point it out in the script 
editor

As for versions, I screamed and fled from the 2.0.x viewers... only came 
back to the 2.1 beta so I can tell you its definitely in 2.1 beta and later

 From a printers perspective, it looks to me very much like the kerning 
and tracking data in the fonts used are a little off. This would have 
least effect with UI scale = 1, but as soon as you apply a scaling 
factor other than unity then it show much more rapidly - my digital 
typography knowledge is a little rusty but I believe that particularly a 
tracking error would grow as a square function with line length when 
scaled that way. If the data in the font itself is good, then its in the 
editor 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] My latest build, objects on self are back in Mouselook

[Dave Booth]

  On 9/28/2010 20:41, Ponzu wrote:
> Just reporting.  Not a big deal.  Wasn't this fixed?
>
> Actually, I sort of like it.  My hair hangs over my eyes, and I am 
> looking out thru my glasses, just like real life 8-)
>

I've not seen a fixed daily build - makes using aimed weaponry or 
piloting a mouselook-steered vehicle nigh impossible. As far as I can 
tell this one has persisted in all viewers since 2.1
___
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] Fix for "Attachments displayed in mouselook" bug

[Dave Booth]

  On 10/9/2010 10:07, Ponzu wrote:
> Photo attached, from behind my eyes 8-)
>
>
>
>
> ___
> 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
looks like you wear the same glasses I do :)
___
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] O.O Display name code DROP!

[Dave Booth]

  On 10/15/2010 01:46, Stickman wrote:
>> The new chat/IM logs are saved in LLSD of all things?
> NOT MY CHAT LOGS! That's where I keep all my text!

Me too - thats a change needs a backout immediately.

Whiskey Tango Foxtrot?


___
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] O.O Display name code DROP!

[Dave Booth]

  On 10/15/2010 06:53, Stickman wrote:
> If it's not, the Lindens should be waking up soon and see a huge
> thread of people screaming that the sky is falling and put some
> authority down on the situation. Looking forward to it, those clouds
> are looking awfully close.

The sky may not be falling and its not a question of unreasoning panic 
or hyperbole, but the whole point of logs is to be able to go back to 
them after the fact and see who said what. I'd even grudgingly accept a 
format change provided theres more in there as an identifier than just a 
bare uuid - its GOT to have at a minimum both the real name and the 
display name they are using at the time. Without both those the utility 
of any kind of logging becomes questionable. Logs will still be of value 
to resolve disputes since they record the uuids of the participants 
(although that process becomes a LOT less straightforward since one has 
to look up the uuid to find any name) but what about situations where 
inworld activities are being logged for another purpose, like 
documenting RP story arcs? Somebody changes "roles" in the RP a few 
months down the line and switches their display name to match and 
suddenly all the logs of their previous character seem to refer to their 
current one since theres no contemporaneous logging of the display name.

Certainly one of the RP groups I spend time with is going to have a hard 
time, since theres a lot of stuff they use logs for - copying them over 
to group websites to provide records and narrative so that we keep 
decent continuity.

To provide a more silly example of why its a bad idea, to know that uuid 
----0 was yelling "Fear my sparkly 
colors!" just isnt funny if you cant remember that user Joe.Lame was 
using the display name of "Phils Codpiece" at the time..

___
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] O.O Display name code DROP!

[Dave Booth]

  On 10/15/2010 17:38, Marc Adored wrote:



Bollocks.
___
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] Reflections in SL. Re: Project-MESH viewer

[Dave Booth]

  On 10/16/2010 18:00, Argent Stonecutter wrote:
> On 2010-10-16, at 17:31, SuezanneC Baskerville wrote:
>> I'd be quite content with mirror surfaces that only, for example, reflected 
>> avatars.
> They had a lot better than that...
>
> http://www.sluniverse.com/pics/pic.aspx?id=139865
>
> ___
> 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
>
Yeah, windlight added so much but we lost a lot at the same time - Thats 
one of the biggies we lost. That single image of Argents has more 
"reality" to it than we'll ever see in a current viewer no matter how 
far we tweak our windlight settings or how carefully we create and 
texture our meshes Just what was it about introducing windlight that 
required the nerfing of that kind of awesome anyway?
___
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] Stuff I'd like to see in the coming sprint

[Dave Booth]

On 10/19/2010 13:58, malachi wrote:
> I would also like to see (since there are so many new changes to the
> server.)
>
> issues like
>
> VWR-11683
>
>
> or any other new scripting functions that will reduce lag and speed up
> efficiency for developers. i dont think focusing only on functions and
> features that are going to ADD to lag should be going on here. i think as
> a community we should be working on ways to reduce the lag as well.
>
> any chance that someone will drop the few lines of code into the new
> server version and spit out the keys to scripts or do we all have to keep
> using tpv clients and 65535^3 prims to scan regions?

I love that idea but since its a scripting enhancement then apart from 
the script editor floaters syntax highlighting thats purely a 
server-side change. It's not something the snowstorm team or any of us 
can work on since its not in the opensource code. Sad to say that jira 
is likely going to end up closed as misfiled, because if its a scripting 
thing it aint a VWR project issue. You can count on my voice adding to 
you advocating for it as a server enhancement, but in its current jira 
location its likely gonna sink without trace.
___
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] Enhanced Script Editor Request

[Dave Booth]

On 10/20/2010 13:20, malachi wrote:
> Of all the developers who are working on the client someone has to be
> smart enough to implement this.
>
> For you windows developers who are using Visual Studio, When you type a
> function name and get to the ( point of the function it pops a tip up
> telling you what is needed to complete this function. I believe it is
> called intellisense. I could be mistaken. Why can we not have the same
> type of script editor? An auto complete type of compiler for the editor.
> As we type in our scripts popups would help us complete what it is we are
> typing. this would increase speed of script writing, help beginners learn
> LSL programming faster, and simplify the process. this is just my
> thoughts. it is already implemented in LSLEditor just figured it could be
> done in the client as well.

hover the mouse over the highlighted function name and you get a tip 
popup showing the functions prototype. Personally I prefer this to 
environments where the popup happens automatically - the popup would be 
a distraction for the functions I use really frequently and never need 
reminding about their parameters but I'll use the tip to remind myself 
of the correct parameter order for those I use infrequently. Perhaps 
this should be a selectable behaviour - Theres already a hook to spot 
when a function name is typed and highlight it, possibly leverage that 
and have function tip popups in the editor be on-hover, on-highlight or off?

Dave.
___
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] Enhanced Script Editor Request

[Dave Booth]

On 10/20/2010 18:40, Ponzu wrote:
> Alas, so many generations of coder who learned vi from someone who 
> already didn't know vi.
>
> You will note that Google's new keyboard commands are vi commands. 
>  There is a reason for this.
>

same reason that most chat apps commands are IRC commands :) /me does 
something in SL for example ...
___
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] Mesh Source Code ETA

[Dave Booth]

Theres even precedent for that kind of approach, where image decode is 
handled by kdu in the linden builds, but it falls back to jpeg2k if kdu 
isnt available. Indeed, the absence of a functionally equivalent 
fallback wherever code that is encumbered by proprietary licenses that 
restrict its opensource release is referenced is a severe blow to the 
entire snowstorm project, as it guarantees that any viewer built from 
source will never be feature-complete. Getting a working decomposer 
thats releasable ported in there should be a pretty high priority for 
the mesh project. I'd even go so far as to suggest that if proprietary 
code is to be used in a viewer feature, the availability of a 
functionally equivalent (doesnt have to perform as well, just has to 
work) source-releasable fallback should be a requirement for the 
introduction of the proprietary code to pass review.

On 10/22/2010 04:12, Zabb65 wrote:
> I think this would be the most obvious way to add proper decomposition
> support back into the client, and allow for a full third party
> implementation. Bullet has a fairly basic decomposition engine, but it
> should work for what is being done, even if the results are not 100%
> idea, and possibly can be worked into snowstorm given sufficient time
> and testing.
>
> On Fri, Oct 22, 2010 at 05:03, Aidan Thornton  wrote:
>> On 10/22/10, Zabb65  wrote:
>>> Looks like this does not need an answer now. Code is up.
>>> http://hg.secondlife.com/mesh-development/
>>> \o/
>> Looks like convex decomposition support has been pulled.
>> "LLConvexDecomposition is a proprietary library based on Havok (TM)
>> physics libraries. In its place, Linden Lab shares the code for
>> LLConvexDecompositionStub, a stub version of the same library with no
>> proprietary components."
>>
>> I'd suggest porting the convex decomposition code from something like
>> (for example) Bullet instead might be a good idea.
>>
> ___
> 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] LGPL violation

[Dave Booth]

On 10/28/2010 06:29, Carlo Wood wrote:
libmedia_plugin_webkit.{sp,dll,dylib}

Make sure you quote examples of static linking when you're talking about 
static linking :)

Dynamically loaded libraries (that is, after all, what "dll" is an 
abbreviation for) are by definition not statically linked. if they were 
statically linked you wouldnt need the .dll, the .so or the .dylib file 
because the object code would have been incorporated into the binary at 
link time.  Statically linked progs dont require external library files, 
dynamically linked ones do. Thats the reason that on Solaris systems 
roots default shell is /sbin/sh rather than /bin/sh - its statically 
linked so that even if the lib loader is hosed the sysadmin can still 
get to a working shell to fix it. (historically, statically linked progs 
went in /sbin not /bin - not always a convention honored these days 
since so few progs are ever completely statically linked any more but in 
this case it was)

Now get off my lawn ;)

Dave

___
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] Down tools and play the "game" -- was Re: What was the point of display names?

[Dave Booth]

On 11/7/2010 19:06, Scruffy wrote:
> Can we not have the whining about Display Names on here please?  It's 
> happening so get over it. 

of course its happening, but yes its relevant to this list - because it 
impacts several areas that are VERY significant to the folks that read 
it and try to contribute.
1: it impacts the viewer-side appearance of any scripted content that 
relies on knowing nearby names,
2: it impacts any inworld experience where a residents name has become 
synonymous with a brand
3: it impacts display of friends lists and logging behavior

I dont intend to argue against its implementation, that is LLs privilege 
and they are welcome to do it - I'd even welcome it for some of the RP 
scenarios I am a member of.  However, its implementation has holes I 
could drive a Mack through without touching the sides. The vast majority 
of these problems would go away with one simple change to its 
implementation - disallow the use of an existing username as a display 
name unless the avatar in question is logged in on that username. if 
that were implemented it WOULD take display names out of any viewer-side 
context, but the word we have right now is that that isnt going to 
happen - even though it would only take a simple username lookup for the 
selected display name at any change, real trivial. So we have to discuss 
and deal with the viewer-side issues, given that  as you remind us, it 
IS going to happen.
___
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] Touch Broken, Stuttering Issues

[Dave Booth]

Oz, I cant add this to the jira because I can view it as a guest but get 
perm violation when logged in...

Repro on my system too with a build I just downloaded. help/about below.

Second Life 2.4.0 (214370) Nov  9 2010 18:24:07 (Second Life Development)
Release Notes

You are at 250,693.0, 245,078.0, 4,011.1 in Mextli located at 
sim8914.agni.lindenlab.com (216.82.41.90:13003)
Second Life Server 10.11.03.213781
Release Notes

CPU: Intel(R) Core(TM)2 Duo CPU T5750  @ 2.00GHz (1994.99 MHz)
Memory: 4094 MB
OS Version: Microsoft Windows Vista 64-bit Service Pack 2 (Build 6002)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: GeForce 8600M GS/PCI/SSE2

Windows Graphics Driver Version: 8.16.0011.8766
OpenGL Version: 3.1 NVIDIA 187.66

libcurl Version: libcurl/7.20.1 OpenSSL/0.9.8j zlib/1.2.3
J2C Decoder Version: KDU
Audio Driver Version: FMOD version 3.75
Qt Webkit Version: 4.6 (version number hard-coded)
Voice Server Version: Not Connected
Built with MSVC version 1400
Packets Lost: 56/101,996 (0.1%)

On 11/10/2010 09:33, Oz Linden (Scott Lawrence) wrote:
> On 2010-11-10 6:13, Trilo Byte wrote:
>> I was busy checking out some other items/banging on different parts of the 
>> viewer, and didn't realize until just now that "Touch" appears to not be 
>> working properly in the latest builds.  Clicking left/right on a panel 
>> vendor seems to work, but touching a notecard giver, landmark giver, dance 
>> ball, and vendors for purchase appears to be broken in the Mac client (I 
>> checked multiple regions/builds/shops).  Build 214319 appears to work.
> It looks like this was probably a problem with a change that was made
> yesterday.
>
> I've filed https://jira.secondlife.com/browse/VWR-23782 to track the fix.
>
>> While investigating this problem, I noticed that some 'stuttering' problems 
>> appear to be back.  Popping open the debug console, it appears I'm getting a 
>> string of messages timed to the exact moments things freeze (a little faster 
>> than one per second).  I'm not sure what it means, can anybody tell me if 
>> this is a problem in the Viewer, or something possibly something gone wrong 
>> in my settings?
> If you can better describe how to reproduce this, please?
>
> ___
> 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] Touch Broken, Stuttering Issues

[Dave Booth]

and fwiw, this was introduced for me with changesets that happened 
between build #214319 and #214370

On 11/10/2010 14:14, Dave Booth wrote:
> Oz, I cant add this to the jira because I can view it as a guest but get
> perm violation when logged in...
>
> Repro on my system too with a build I just downloaded. help/about below.
>
> Second Life 2.4.0 (214370) Nov  9 2010 18:24:07 (Second Life Development)
> Release Notes
>
> You are at 250,693.0, 245,078.0, 4,011.1 in Mextli located at
> sim8914.agni.lindenlab.com (216.82.41.90:13003)
> Second Life Server 10.11.03.213781
> Release Notes
>
> CPU: Intel(R) Core(TM)2 Duo CPU T5750  @ 2.00GHz (1994.99 MHz)
> Memory: 4094 MB
> OS Version: Microsoft Windows Vista 64-bit Service Pack 2 (Build 6002)
> Graphics Card Vendor: NVIDIA Corporation
> Graphics Card: GeForce 8600M GS/PCI/SSE2
>
> Windows Graphics Driver Version: 8.16.0011.8766
> OpenGL Version: 3.1 NVIDIA 187.66
>
> libcurl Version: libcurl/7.20.1 OpenSSL/0.9.8j zlib/1.2.3
> J2C Decoder Version: KDU
> Audio Driver Version: FMOD version 3.75
> Qt Webkit Version: 4.6 (version number hard-coded)
> Voice Server Version: Not Connected
> Built with MSVC version 1400
> Packets Lost: 56/101,996 (0.1%)
>
> On 11/10/2010 09:33, Oz Linden (Scott Lawrence) wrote:
>> On 2010-11-10 6:13, Trilo Byte wrote:
>>> I was busy checking out some other items/banging on different parts of the 
>>> viewer, and didn't realize until just now that "Touch" appears to not be 
>>> working properly in the latest builds.  Clicking left/right on a panel 
>>> vendor seems to work, but touching a notecard giver, landmark giver, dance 
>>> ball, and vendors for purchase appears to be broken in the Mac client (I 
>>> checked multiple regions/builds/shops).  Build 214319 appears to work.
>> It looks like this was probably a problem with a change that was made
>> yesterday.
>>
>> I've filed https://jira.secondlife.com/browse/VWR-23782 to track the fix.
>>
>>> While investigating this problem, I noticed that some 'stuttering' problems 
>>> appear to be back.  Popping open the debug console, it appears I'm getting 
>>> a string of messages timed to the exact moments things freeze (a little 
>>> faster than one per second).  I'm not sure what it means, can anybody tell 
>>> me if this is a problem in the Viewer, or something possibly something gone 
>>> wrong in my settings?
>> If you can better describe how to reproduce this, please?
>>
>> ___
>> 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] Dynamic shadows and ATI

[Dave Booth]

On 11/13/2010 06:04, Laurent Bechir wrote:
>
> Is it a hardware problem of ATI or just a software problem that can be
> solved by SL developers ?

ATIs drivers have bugs - ATI + OpenGL FBOs = crash or render artifacts 
or both.
___
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] Dynamic shadows and ATI

[Dave Booth]

On 11/22/2010 15:48, Laurent Bechir wrote:
>
> Perhaps Linden should put something like "You will have a better
> experience with Second Life if you choose NVidia graphic card". I know
> it's quite stupid, but from what I've read there is not a chance that we
> get good support for ATI cards and dynamic shadows which is quite
> disapointing.
> _

I think what would be even more appropriate would be "LL recommends 
NVidia graphics cards, because of bugs in ATIs implementation of OpenGL. 
ATI cards are supported but tickets related to known issues specific to 
ATI OpenGL will be closed, since these cannot be fixed by LL and must be 
addressed by ATI" - and pop it up every "first login" of a new or 
updated viewer install when the viewer detects an ATI card.  The more 
folks that do something like that, the more ATI will feel the pressure 
to make their OpenGL implementation conform to the standard. I actually 
LIKE ATI hardware designs and the performance of their cards, I switched 
from being an ATI fanboy over to using NVidia exclusively because with 
NVidia everything just works and their drivers are rock solid, When I 
formed my preference for ATI then both had issues with their high-end 
cards and drivers playing nicely with every 3d protocol. However NVidia 
fixed those bugs while ATI did not and so that preference was 
reluctantly discarded.

Dave
___
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] does anyone else think this would be a good thing?

[Dave Booth]

I've been thinking about the anims I've been making for my products 
inworld.. I know the mesh guys are screaming for custom bones but it 
strikes me that even in the standard animation setup theres a possible 
enhancement. More hand positions

I know its not a viewer side thing really.. but what do you guys think 
about these...

The "thumb" position - enables thumbs up, thumbs down and incidentally 
the Sicilian thumbnail flip..
The V position - enables the classic anime gesture, along with the 
"Winston Churchill" and unavoidably the "Harvey Smith" (which is the 
brit equivalent of...)
The bird..  Its gotta be there.. the flip off is part of human 
interaction and even if it isn't kind it is an essential part of 
immersion in the world
May as well add the "horns" too - useful for inworld pagans as well as 
folks wanting to ward off the evil eye...
___
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