On Saturday October 26th 2013, the Firestorm team hosted another informal question-and-answer session. While the meeting was recorded, the Firestorm team are aware that many of their users have hearing difficulties, and / or prefer to read text. It is because of this that this transcript has been provided.
When reading it, please remember:
- This is not a word-for-word transcript of the entire meeting. While all quotes given are as they are spoken in the video, to assist in readability and maintain the flow of conversation, not all asides, jokes, interruptions, etc., have been included in the text presented here
- If there are any sizeable gaps in comments from a speaker which resulted from asides, questions to other etc,, these are indicated by the use of “…”
- Timestamps are provided as guidance should anyone wish to hear the comments in full from any speaker on the video
- Questions /comments were made in chat while speakers were talking. This inevitably meant that replies to questions would lag well behind when they were originally asked. To provide context between questions and answers, questions in the transcript are given (in italics) at the point at which each is addressed by a member of the Firestorm team, either in voice or via chat.
Please note: This transcript is provided for informational purposes only. As such, questions on technical issues relating to Firestorm and / or project-specific questions cannot be answered here unless one of the Firestorm team drops by.
The TL;DR Summary
The numbers in braces are timestamps which refer to the section of this transcript where more details can be read, and to the section of the video recording where the relevant comments can be heard.
- The upcoming release will be a public beta. It will have a number of known bugs, some of which will be annoying [0:01:42-0:02:28]
- There is a lot more coming from the Lab, so it is important the Firestorm team get a beta version out, so they can prepare for bug fixing and integrating the new code LL will be pushing out (and which is currently pending release by the Lab) including:
- SSA updates (AISv3)
- Group ban list
- Interest list
- HTTP updates
- It will be a beta release, but it will be fully supported [0:05:10-0:05:33]
- The Firestorm team have been accused of “holding things back”, they are not. There are valid reasons for any apparent delays (such as the amount of work some code merges have required and code for newer projects has yet to appear from the Lab [0:05:33-0:07:43]
- A full release is unlikely before February 2014 [0:07:43]
- The beta will be accompanied by a Windows 64-bit version of the viewer [0:11:01-0:13:41]
- Why feedback for the 64-bit version is being requested, what kind of feedback is sought, why feedback should be delayed for a week or more following the release, the placebo effect [0:11:37; 0:14:30-0:16:09; 0:32:52-0:34:13; 1:01:04-1:04:08]
- 64-bit versus 32-bit: what to expect and what the differences are [0:16:09-0:18:22]
- Linux 64-bit [0:28:50-0:29:34]
- Why it is unlikely any viewer will have multi-screen support (tear-off panels and menus which can be moved outside the viewer window & between monitors) in the near future [30:09-32:52; 34:13-34:41]
- Mac 64-bit [0:36:09-0:41:49]
- The issues in building 64-bit versions of Firestorm [0:41:49-0:44:47]
- Mac Maverick support [46:05-48:08]
- Both Windows 32-bit and 64-bit can co-exist on the same PC [49:47; 57:29]
- 64-bit and lack of Havok support [58:29]
- A discussion on Firestorm on mobile devices & LL’s coming beta with OnLive [50:39-54-15]
- Why clean installs are really, really necessary at times [1:14:47-1:18:39; 1:24:58-1:25:47; 1:32:50-1:35:24]
- Important notes on the 64-bit installer / installation [1:19:20-1:21:29]
- Why Firestorm should only ever be downloaded from the Firestorm website [1:22:14]
- OTR is coming, but not a priority [1:23:48-1:23:54]
- Questions on 32-bit and 64-bit throughout.
With thanks to North for the video.
Jessica Lyon (JL): All right, folks, we’re going to get started on this question and answer session. Thank you all for coming, we’ve got a full region … so welcome. We’ve got some devs up here … I’m sure you’ve all got a lot of questions about the 64-bit version that’s coming; there’s been a lot of questions on the blog and I’ve seen in other blog posts as well, from people asking all sorts of questions from people not fully understanding what 64-bit means to them.
00:41 JL: Here’s the disclaimer. I’m not going to pretend like I know what I’m talking about this time, because in a lot of this stuff, I’m probably as clueless as some of the most clueless people. I know a little bit, but I’ve seen people ask questions – and I’m going to get to this at some point – in our blog comments about multi-threading and stuff like that. What?! (laughs). I don’t know! So I’m hoping that Cinder and some of our devs up here will be able to answer some of these questions as well for us.
01:16 JL: Before we start talking about the 64-bit stuff we’ll get the 32-bit release out of the way. So we are releasing a public beta, as you probably all know at the end of this month. When? If I give you the actual date, it’ll jinx it, so I’m not going to give you the date. I’m just going to say “end of the month, roughly, give-or-take” – somewhere around there.
01:42 JL: It is a public beta because it has well-known bugs, and some of them are annoying and some of them you’ll probably be able to live with. Lassie is our QA lead …. Lassie is this, or is this not, ready for public release.
02:14 Lassie (L): It is as a release, not as a beta.
02:16 JL: OK. There you have it. It’s on him, so if it explodes, it’s his fault and not mine! Thanks for coming, Lassie! (laughs).
02:28 JL: Anyway, the fact is it does have bugs, but we want to get what we have out because Linden Lab has a whole bunch of new stuff coming, like server-side appearance improvements, that’s the polish-up code; inventory changes, which is also part of that. And that’s quite significant. There’s quite a bit of code in that. So it’s not something we can just, “let’s just *plink*, it’s done.” You can’t do that; it’s going to take some time.
02:58 JL: We’ve got the group ban code that’s coming, that we haven’t even seen yet; we’ve got interest list changes, which is a lot of stuff. You guys remember us complaining about CHUI before, being a lot of code? Well interest list changes are also a lot of code, and a lot of changes because what they’re doing now with that is allowing the viewer to handle a lot of the mathematics and decision-making. So that’s quite a bit of change in viewer code for us. A there’s a whole lot of stuff we have to merge. In that department we’ve got HTTP, which is like network stuff which is coming from Linden Lab. That’s a lot of stuff.
03:38 JL: We were hoping we could get that stuff sooner; most of it is not even public, or even what is public is by Linden Lab’s own admission not in any way stable or production ready, which means we shouldn’t even try to think about merging it right now. And by “us” I mean any third-party viewer; it’s just not ready, there’s lots more yet to come to it.
04:00 JL: So the decision was do we wait another couple of month, to 2014, before we get materials and whatever else we have out. Or maybe what we have is good enough to go out just as a beta; it’s not fully stable, it does have some bugs, we still have some interface bugs here and there because of the CHUI merge. So the idea is really to get it out to you now, as it is. Although we’re calling it a public beta, and I’m sure support would rather I said we’re not providing support for a public beta, but I think the reality is that most of you in Second Life who use Firestorm will be using this public beta, because it’s got materials and all the latest stable code, you could say. I don’t expect it to be any less stable or crashier than any of our other viewer releases’ it’s just that it’s got little quirks and things going on.
05:10 JL: So that’s going out at the end of the month. It’ll be the same as a basic release, it’s just going to be beta at the top…
05:29: Ed Merryman (EM): It will be supported and we will be teaching it in the classes.
05:33 JL: So basically, although we’re calling it a beta, it is going to be fully supported, we have to support it … Questions on the 32-bit?
How soon for another release after this beta? next year-ish?
06:13 Tonya Souther (TS): I will cite the Ernst and Julio Gallo commercials at you: we will release no code before its time.
06:20 JL: It’s going to be a little while, because like I say, we’ve got a lot of new stuff coming downstream from Linden Lab and … I don’t think we were expecting it to be as much of a difficult merge as it’s turning out it’s going to be, again. But we were expecting it. We were expecting it a lot sooner than we’re getting it right now. You guys probably remember in the last couple of Q&As we were saying we were hoping our next release would have all of this stuff, but the truth it, it’s not even available to us.
06:59 JL: I’ve been noticing this thing in blog posts from people, saying that we’re holding things back. OK, materials is not out in Firestorm yet, it’s also not out in most of the third-party viewers. Linden Lab and, I think, Exodus are, I believe, probably to only two viewer that have it, and Catznip1, I think. The major viewers don’t have materials out. So we’re not holding things back; we’re certainly not holding back the server-side appearance updates and things like that. Singularity doesn’t have materials yet, from what I understand. I was told that they’re planning to release it next week [Singularity had materials in an alpha version rather than a release at the time of the meeting].
07:43 JL: We’re not holding things back and we are moving along as quickly as we possibly can … I resent people that think we’re holding things back. I will grant to those people that mesh took us a little bit longer to actually get out released than it was anticipated, but we had other issues that held that back. But anyway, hopefully we can get this new code merged by the time Linden Lab has it stable and hopefully we can get something out by late February.
08:26 JL: So one of the things I want to do with the team right now is, once we get this public beta out, Thanksgiving is coming for the Americans, Christmas is coming, holidays, all that stuff. I think it would be nice for our team to have a little bit of a break without panicking about trying to cram-in all this new stuff.
08:57 EM: Hey, wait a minute, did you just say that we all get a break?!
09:00 JL: No, no break. Broken bones, maybe!
Any chance the Friends list is going to be customisable, same as contacts list?
09:21 JL: Not likely. I think contact sets is about what you’re going to get in that regard, otherwise it’s going to require some Linden Lab server intervention.
You going to fix the media in Linux?
09:45 TS: The media on Linux is a 64-bit thing, so we’ll have to address that when we get into that side of things.
10:08 JL: I think that’s about all the questions we’re seeing for the 32-bit public beta – which, by the way, should be the highlight on your minds, not the 64-bit. The fact that we’re putting out a public beta, a new release with new stuff, with .DAE export, for example, [thanks to] Cinder for that; and import / export of all kinds of things, and cool stuff coming. And that’s what you really should be excited about, because the 64-bit that’s coming is an experiment. A lot of people are, “Oh! 64-bit, can’t wait!” Let me tell you something, it is a complete experiment and it’s alpha. So let’s talk about that really quickly a little bit.
11:01 JL: If you are one of those explorers and brave people who want to try something new, and decide to download and install the 64-bit Windows version, and you find a bug, and they you say, “OK, I’m going to go to the support group.” And you’re going to say, “How come this isn’t working?” Or, “This is doing this, and that’s doing that?” And they’re going to say, “Thank you for your feedback. If you could just e-mail that to Jessica-dot-Lyon-at-phoenixviewer-dot-com, it would be much appreciated.”
11:37 JL: And the reason they’re going to tell you that is because we want your feedback on it. We’ll be able to help you with questions like, “Where is this setting?” Or, “Where’s that setting?” Or “How do I change this?” … But the fact is, we’re not convinced – I’m not convinced – that a 64-bit Windows version actually has that big a difference. I’ve been running it for the past week or so, and switching back to the 32-bit brother of it, same version and everything, and I can’t really notice a significant improvement. Maybe you will. I’ve crashed on it, and I’ve also been in situations like this right now, where I’ve got a significant draw distance where everybody’s rezzed, there’s seventy people here, and I’m actually doing pretty good performance-wise, but I would be with a 32-bit as well. So, we’re not convinced that it’s all that great, although everybody has been asking for it for years.
12:55 JL: Some of you may have read Inara Pey’s blog post recently. Inara interviewed us about the 64-bit. The question my be what brought us to doing it. Well, we didn’t do it actually. Nicky Dasmijn did, who is our lead Firestorm developer, and she did it because she just had an itch, which is the nice thing about open-source; you can just have an itch and decide to scratch it. and she was playing with it for a while and she felt it was ready and she dropped it into our code respository, and she opened, in that process, a very big can of worms.
13:41 JL: We were actually all a little bit surprised, although some of us knew she was working on it, I don’t think any of us expected her to drop it in the code just like that (chuckles), “Surprise!” But there it is, and suddenly we’ve got … there are other third-party viewers, Singularity, I spoke to Latif [Khalifa] a couple of days ago and they actually have their own version of 64-bit, which they’ve released in an alpha form … I don’t know how theirs is doing, I haven’t talked to them since they’ve released it, so I’d be interested to hear how that’s going.
14:30 JL: So I want your feedback, we want your feedback, and probably what we’re going to do … basically what I want to know if you find it’s fast; send me an e-mail, I want to know. If you find it’s slow, send me an e-mail, if you find it is stable or unstable, things like that, that’s what we want to know. What I’m going to do is make a filter in my e-mail and have all your e-mails go to a folder, and then I’m going to copy / paste everything you send me – providing it’s PG – and without your names onto a public Google doc that I’ll link to from somewhere on the website and then we’ll try to sift through all the results.
15:14 JL: I expect there’s going to be a lot of placebo. and I’ll tell you why I think that … Ed how many times have we seen placebo with viewer stuff?
15:26 EM: far too many. Every release?
15:27 JL: We’ll go from one release where nothing has changed, and you’re telling us that it’s ten times better than it was yesterday, but nothing’s changed. It’s placebo, I’m telling you. So there’s going to be a lot of placebo; there’s going to be a lot of people saying, I expect, that the 64-bit is amazing! It’s so much better! And it might be. I know there’s some truth to that; some computer systems will probably find some difference to performance gains than other computer systems. I’ve not noticed it; maybe I’m not lucky enough to have one of those systems.
16:09 JL: so we’ll talk a little about 64-bit versus 32-bit … I’ve seen this question a lot: I’m on Windows 7 64-bit, but I’ve only got 4 gig of RAM. Am I going to have an improvement?
16:23 JL: In theory, you should have some improvement. Let me try telling you the difference between Large Address Aware and 64-bit. And this is something I learned just recently from Ansariel and Cinder, because I knew that this question was going to be asked, and I needed to fully try to understand it.
16:59 JL: Let’s say you’ve got Windows 64-bit operating system, 8 gigs of RAM and you’re running Firestorm with Large Address Aware, and you notice that it never really goes past 1.5 gig – I think that’s the most the viewer uses anyway. So without Large Address Aware, the amount of memory the operating system allocates to the viewer is somewhat dependent upon the other applications you have running as well. When you have Large Address Aware, it will allow the viewer to have a little bit more memory than the other applications that you also have running in the background or wherever.
17:51 JL: The problem is a lot of people think Large Address Aware means it can use all 8 gig. It can’t. There’s still a 3.5 gig cap; it can’t go beyond that, because it’s a 32-bit viewer. When you have a 64-bit viewer, suddenly it can allocate a lot more memory. It can use-up even all 8 gig if it needs to, although I can assure you it won’t. In fact I would be surprised if it went past two gig.
18:22 JL: So is there a real advantage to it? I don’t know. I’m not convinced; but I’m a sceptic.
18:53 JL: We asked in the blog post for people who couldn’t make it, if they wanted to ask some questions and we’d try to answer them. There weren’t all that many questions, so I’m just going to skim through these really quickly.
I was wondering if you would be able to add an inventory filter by permissions. I realize this might be complicated and there are at least two ways this could be approached
19:20 JL: Before I go any further, he’s absolutely right, this is going to be complicated, and it would very likely require server-side support, and inventory is very complicated anyway. And inventory is a whole lot of convoluted code and any time we make changes to something that has a lot of code, it causes a lot of merge nightmares with us and Linden Lab every single time we merge with them. And so when it comes to changes to inventory, it better be a feature that has a lot of value, because it’s going to be a lot of work maintaining it going down the road.
20:03 JL: Somebody today messaged me as well, had a great idea for inventory improvements. Inventory is one of those areas where there’s tonnes of room for improvement, but it mostly requires server-side support. And I think really, we’re all a little bit scared of inventory because you break inventory and you’re going to pay the price … That’s like the Bermuda Triangle of Second Life viewer code.
True multi-thread from the Linden Lab source, or from additional input, to take advantage of 8-integer cores from AMD or eight hyperthreads from Intel, although 4 cores and 4 threads are quite common occurrence.
20:56 JL: First of all – What?! I didn’t even understand the question.
21:09 TS: All right, basically what they’re asking is, can Firestorm be made to use eight cores or eight threads; a fantasy processor with lots and lots of processing power that can do multiple things at one time. And the answer to that is, not without a whole bunch of re-writing.
21:34 TS: Taking an existing programs and turning them into something that can handle doing more than one thing at a time is a notoriously difficult problem. Firestorm and the Linden Lab viewer, I can’t speak to the others, they have some things that have been moved off into other threads. That mean that if you have a 4-core processor, it can do something else at the same time as running the main viewer code. But breaking up the viewer into lots of different threads of execution is a problem, in part because you have to have to understand what you can “parallelise”, what can you break-out and have another CPU do at the same time as the rest of Firestorm.
22:30 TS: Doing that would require a lot of intensive measurement and effort and understanding of the code. And at this point, I’m not sure anybody understand the code to that level of detail. We probably don’t; I’d be surprised if Linden Lab really had that level of understanding at the deep level required. That doesn’t mean that they don’t, it means that there’s maybe one or two people there, and they’re up to their ears now as it is.
23:07 TS: That’s something Linden Lab would probably not see as a high return effort for them.
Audience comment: Firestorm is doing multiple things; one can if I recall with one command; make it that things don’t use the same CPU; for example mesh no need be in same as chat / networking.
23:23 TS: That’s true, it’s doing multiple things, but the question is, can it parallelise them? Can it do those things at the same time and not have one piece depend on the results of another piece? That problem is much harder than you think. You can scatter interlocks all over every place, and tell one piece of the code to wait until this other piece is done. But if you’re going to do that, you might as well leave it in the same thread. This is a very, very, difficult problem, and probably one that’s beyond our ability to tackle.
24:04 JL: I hope that answers that! Not going to happen!
I’ve read materials are coming in this beta. Are ribbon/glow particles coming with it too?
24:27 JL: Yes they are. Although particle glow is broken, and that is not our fault.
I think a useful feature for those of us who have massive friends lists, and who want to show as if we are offline, would be an invisible or show offline option, at the top of the contacts list, or in the communications menu option.
25:01 JL: As a matter-of-fact there is that feature
25:06 EM: It’s there, but I don’t want people using it. And the reason I tell you not to use it, is you can change them manually quicker using that eye icon. There’s an eye icon on the People panel, if I remember correctly, on the Friends tab of the People panel.
25:30 Kadah Coba (KC): What it does is it goes sequentially through your contacts and toggles off Show My Online Status to Friends … It takes a long time as it does each one sequentially. So it does one, waits for confirmation, then does the second one. If you do them all, one after another, there could be failures, but it would just be that one just not synchronising properly.
25:59 EM: Takes about ten minutes to do fifty people. If you have a Friends list of fifty people, it takes about ten minutes for it to happen. I don’t know about you, but I can click 50 times quicker than ten minutes.
26:14 KC: Another thing to keep in mind is if you do use it, it will toggle all of them off, and if you use it again, it will toggle them all back on. So if there is someone on your contacts list that you had Show My Status To turned off, it will turn it back on if you use that to turn it back on; it won’t keep that remembered, as it is toggling.
26:39 JL: And of course there’s the well-known fact that scripts can totally bypass any of those settings that determine whether you’re online or not.
[Discussion on whether scripted on-line checking tools still work, with the consensus being they still do]
28:50 JL: So, Linux 64-bit. Who wants to know about Linux 64-bit? Linux 64-bit is not new, as you all know. At least if you’re a Linux user, as you know. And unfortunately, it is highly unlikely that we’re going to have Linux 64-bit on time for the release. I tasked Techwolf [Lupindo] with doing that, and we even have instructions somewhere on how to do it; it’s not that complicated to do. The trouble is, Techwolf has been really busy in real life, and he just hasn’t had time to do it.
29:34 JL: That’s the bad news. The good news is that it is coming, it’s just that it probably won’t be available when we release the 64-bit Windows alpha. The instructions are on the wiki somewhere. Compiling Firestorm, I believe it was under. I have no doubt that Whirly or somebody is going to post a link to it at some point.
Can we drag chat out the window now? (multi screens)
30:09 EM: No.
30:13 TS: Major rewrite.
30:25 JL: It is true, I think, That in order to have tear-off windows, the viewer would need to be 64-bit.
30:44 TS: The limitation is not that the viewer is 64-bit. The limitation for tear-off windows and multi-screen support and all that is that the user interface code has the assumption that everything is drawn in one window baked into it very, very deeply. You’re talking about a total rewrite of the user interface code to enable that; and that’s just not going to happen.
You need fork it through a plug-in
31:13 KC: [No] the inclusion of a third-party interface library that is OS-dependent and then putting different interfaces outside. That’s just another mess all together.
31:26 JL: Having tear-off windows is what I consider the Holy Grail … Apparently it would be easier, at least, if it were 64-bit Windows, from what I understand. It’s possible; I mean, people have worked on tear-off windows … Let’s just say tho, that tear-off windows is the Holy Grail, and it might happen; it could happen; possibly, hypothetically, in the future.
Right now a lot of large programs like 32 bit Mozilla Firefox are running into a memory wall. I haven’t run into the problem yet with Firestorm; I’m able to build in 32 bit and cross compile on 64-bit. Do you see this as a problem with the viewer down the road? Firestorm does use more memory when compiling. And yes, the 64-bit system build works well up to the 4..6.0 GCC.
32:52 JL: We do have some crashes where the viewer goes beyond the amount of memory that’s allocated for it, and it crashes. We call them memory crashes. And we do have some of those, and we’re thinking the 64-bit might actually fix that, simply because there’s more memory available. Whether it’s an improvement? That’s where we’re going to need your feedback. We’re going to need your feedback on what is your experience.
33:18 JL: And I don’t want your feedback like, “I just downloaded it and installed it and just launched it. I’ve been on it for five minutes, and I’m so happy with it!” I don’t want that. I want you to send us an e-mail after a week or maybe two weeks, or even a month, because using it for five minutes or an hour or a day is not really enough to give good feedback. And what’ll end up happening is you’ll give feedback after an hour, and then a week later your feedback will be completely different. And they you’ll have to update me with that, and then I’ll be confused because I’ve got conflicting things coming from you … So take your time with it, play with it, enjoy it, put together your thoughts before you send us your feedback.
Question about tear-off windows: You mean like dragging my IM window outside the viewer?
34:13 JL: That is having the ability to rip off your conversations panel and move it outside your viewer. That would be really, really beneficial to people with multiple monitors.
Tear off windows would be the ultimate nightmare for support.”Where’s my XXX window gone ?” “Dunno .. where did you put it?”
34:41 JL: Yes, that is true. But I think two years ago somebody would have said, “It’s really unlikely that there would ever be a 64-bit windows version of the viewer.” And probably, I may have even said that myself, and I’m sure it has been said back then. and here we are coming out with a 64-bit version of the viewer, Singularity’s got one and I think in a very, very short time all the viewers out there are going to have one. I suspect that if the crash rate is reduced enough on the 64-bit, Linden Lab will adopt it. I understand Oz is interested; in fact, Oz wanted us to have a different channel for the 64-bit, which we’re already planning anyway, because he wants to know if it’s going to reduce the crash rate and be more stable. So if the 64-bit turns out to be stable, I bet you Linden Lab will do it as well.
35:34 JL: I think the reality is though, the can of worms is open, the cat’s out of the bag, and 64-bit viewers are just going to be part of the landscape moving forward now.
35:48 JL: Tear-off windows is probably the next big thing.
36:09 JL: Who wants to know about Mac 64-bit? … Cinder, you’re up, my dear … Windows is going to have a download on the website that’s 64-bit and one that’s for 32.-bit. I understand that Mac can do both 64-bit and 32-bit in one application. Please explain.
38:16 Cinder Roxley (CR): Yes, the mach-o architecture in Mac allows for bundling binaries for more than one platform. so you can have a 32-bit and a 64 in one application.
38:25 JL: Which means if you have a Mac 64-bit available, and when we do, it’ll be in one build instead of two? Do people know how to switch that? How does that work?
38:42 CR: No. The OS knows how to do that by itself.
39:00 JL: so does that happen during the install?
39:10 TS: It Just Works, like everything else on OS X.
39:21 JL: So the next logical question is, are we going to have a Mac 64-bit capable release with this public beta? Yes or no?
39:42 CR: No, but by release.
39:43 JL: So there you go, you’re all disappointed. sorry. Thanks for coming (laughs). The reality is we’re not going to have Linux in time and we won’t have Mac on time. but I think there’s a good chance that within the next week or two after the public beta a new Mac might appear that’s 64-bit compatible and a Linux may appear that’s 64-bit compatible.
What’s the minimum processor and OS for Mac 64?
40:34 TS: Snow Leopard, and at least a Core 2 processor.
A link is provided for people wishing to check the Intel Mac processors.
41:42 JL: So no Mac 64-bit on this release, at least not at the time of this release. Probably not Linux at the time of this release.
41:49 TS: The big thing about building 64-bit versions of Firestorm on all three OSs is that the distributed Firestorm includes a bunch of independently-produced libraries, upwards of thirty. And all of those libraries as distributed by Linden Lab have all been built for 32-bit. They have to be rebuilt for 64-bit in order to be distributed. So that’s what Nicky did to made 64-bit work. Firestorm itself has been 64-bit clean for some time. And if you’re building a standalone version of Firestorm on Linux for your 64-bit Linux OS, like a lot of people are doing, that’s what you’re doing. But instead of the 30+ libraries, you’re using libraries installed on your operating system.
42:52 TS: The problem is that if you build that against your operating system, then it has to be distributed only to people that use your operating system because there are version level dependencies that exist there, and Firestorm will simply refuse to run. So the major work is in rebuilding all those libraries. And that’s what Nicky did, that’s what Cinder is doing for OS X … and that’s what we will need to do for Linux. Once we do that, the rest is easy … Once the libraries are in place, then building and distributing Firestorm is comparatively simple.
43:42 JL: So one of the drags of being a developer who does compiling is that they have to be on old operating systems in order to be able to support you guys with old operating systems. Which is why it’s useful to have a build farm, because if you’re compiling a release on your computer that you’re using, you can’t really take those updates like Windows updates and graphics card and whatever updates; you have to have an old system to be compatible with other people. We had that problem with Phoenix for ever. In fact, we had to have LGG [LordGregGreg Back] compile Phoenix for some people because he had a very, very old system, and there were a lot of people with very old systems that his compile would work on. And so that’s a problem, and being a developer means you have to have a crappy computer … well, not up-to-date.
44:47 JL: We had a problem with Snow Leopard … we’ve had it for a little bit now in our internal work, and cinder had a fix for it. but when she compiled it, the fix wouldn’t work because her computer is just too up-to-date. So she had to end-up giving it to Nicky [Dasmijn], whose build Mac – the build she uses for releases that you guys use, happens to be older, so that worked.
Did you say that the new version will be fully compatible with the new Mac Mavericks?
46:05 JL: It should be, yes. We’re not going to say it’s fully compatible because Mavericks just came out.
46:10 CR: Yes, I’ve been testing for months on Mavericks.
MacOS 10/9 (Mavericks) and audio streaming issues fixed?
46:38 CR: Yes. all the current version problems are fixed for the next release.
Does the Maverick build require particular or different settings in any of the preferences?
47:35 CR: No, it should work. Just don’t try full screen mode.
My question was more about what aspects of Mavericks & Firestorm is causing audio streaming to break, specifically?
48:08 CR: FMOD … the FMOD sound system is incompatible with Mavericks, but the next release will have FMODex.
Do you all see it staying feasible for all three builds stay in the now maintained source. or split to different operating systems.
48:40 TS: Yes, all three builds will stay in the same source … we’re not going to break it up unless and until Linden Lab do.
48:52 JL: There was actually some question in the blogsphere and I’ve had people ask me as well: does this mean that we’re moving away from 32-bit? Absolutely not. The nice thing about 32-bit is pretty much everyone can run 32-bit. So no, we’re not moving away from 32-bit. We’re just adding a new 64-bit into the family. But no, definitely not, and we can’t, really move away from 32-bit because there are a lot of people on 32-bit operating systems.
On the Windows 64 and 32-bit versions. Can we have both installed without interference from each other for our own personal testing?
49:47 JL: Yes you can, because one will install to your Program Files (x86) folder, which is the Windows 32-bit version, and one will install to your Program Files folder, which is the 64-bit version. But they can share the same settings.
Will there be a Firestorm viewer for the iPad?
50:39 JL: Cinder and I got an e-mail from a developer … “I’m planning on converting the Firestorm code to Object C for an eventual release on a mobile platform for iPad Air and iPhone 5s.”
51:14 JL: It could happen.
51:19 KC: (Laughing) Well, somebody’s got an infinite amount of time.
51:21: TS: Sorry, I have to agree with Kadah there … It would be easier to re-write the code in x86 assembler.
51:34 KC: That would be easier.
51:36 JL: Well, the answer to the question is, somebody is thinking about it. They’ve got their hands full; I wish them the best of luck, because I don’t want to discourage them.
51:50 KC: As far as SL on mobile, Linden Lab is actually trying to do something there, but it’s just OnLive.
52:05 TS: Remember I’ve been saying major re-write all this entire Q&A about different things? That is well beyond major re-write; that is, might as well start over from ground zero and work your way back up. I don’t know how much time Alina has invested in Lumiya, but it would be a similar commitment to get Firestorm on iOS, if not more.
52:35 JL: Part of the problem, though, is the processing capabilities of things like the iPads and the iPhones and, granted, they are pretty impressive nowadays, but a proper viewer like Firestorm, which is already resource hungry would be pretty heavy.
52:53 KC: Well, to put it in perspective, what Linden Lab is doing for their mobile offering … what they did instead of trying to figure out how to make it render on mobile devices, they’re using OnLive which is just a video game streaming service.
53:15 JL: So they’re basically streaming video to you?
53:19 KC: That’s what OnLive is. The game runs on cloud-based servers and then the game is streamed to you as video. So that offering is going to be pretty bandwidth intensive, I would imagine.
53:36 JL: that might be the easiest way to do it with today’s technology.
53:42 KC: It’s the easiest way to do it without lots of investment in making a new application. Which, for Linden Lab, I don’t see them ever making that consideration, because you’d have to get like 90% equivalency to the desktop application just to get into beta.
I’ll take that as a solid no.
54:15 JL: Well I wouldn’t take it as a solid no. Kadah and Tonya are not believers … I’m of the mind that two years ago it was pretty much thought that a 64-bit viewer in Second Life would never happen, and suddenly, ta-da! Here it is; it’s about to become reality. And who’s to say where we’re going to be in two years? Advancements are happening very quickly; not perhaps as quickly as they could be in Second Life, but advancements are happening pretty quickly. Certainly, mobile technology and hardware is moving really, really quickly. I mean they’ve got now, I think it’s Nokia, a 40 magapixel camera on a phone … That’s insane! So I would never say it’s never going to happen.
[Discussion of the 40 megapixel camera and a conversation about Australia and spiders]
If we can install 32 and 64bit at the same time, will they also have separate cache paths? Because I’ve had both betas and release versions installed, and if I don’t manually separate the cache textures go wacko!
57:29 JL: No. Not unless you give them separate cache paths … actually, I don’t think you’re going to have a problem with textures … I don’t think you’ll have a corrupt cache by sharing the cache between the 32-bit and the 64-bit. don’t hold me to that, though, because I’ve not actually tried it. The 64-bit I’m running does actually have it’s own separate cache location that I’ve given it … I all honesty, we’re expecting that the 32-bit and the 64-bit viewers are going to have the same bugs, and they should be no different. The only thing really that we’re wondering what’s going to happen is, is the 64-bit going to be faster, is it going to be more stable? Because otherwise it’s pretty much the same.
58:29 JL: One other thing I should point out, though, is the 64-bit does not have Havok. It is essentially an OpenSim build, which means it works in Second Life and in OpenSim. And, don’t get confused with what I’m going to say; it can’t properly upload mesh with physics. That doesn’t mean you can’t see physics. It fully supports physics and rendering physics and seeing physics and it supports even uploading physics, but you can’t upload mesh with physics, but it will work fine with mesh. And if you use pathfinding in your region, and need to view navmesh with pathfinding, it’s not capable of doing that. Otherwise, it’s identical.
Can we expect bigger draw distances with 64-bit?
59:23 TS: Draw distance is a function of your graphics hardware, not the amount of memory you have available.
59:41 KC: If you want your draw distance to go over four kilometres, you need to go back to viewer 1.
59:52 JL: A lot of people don’t realise that when you raise your draw distance to 512, that’s not 512 metres in front of you; that’s 512 metres in front of you, behind you, to your right, to your left, above and below you. That’s a lot.
1:00:07 EM: Doubling your draw distance actually increases the amount of area you potentially render eight-fold. So you’re not doubling your draw distance. And if you’re at 1,024 metres and you put yourself in the right place, you can see the better part of 64 different regions.
1:01:04 JL: To clarify the original question, can we expect bigger draw distances with 64-bit, if your finding after three or four weeks of using it that 64-bit lets you have a higher draw distance without crashing than the 32-bit does, send me an e-mail because I’d like to know about it … don’t it now, because you don’t even have it yet. And I’d really like to not start getting e-mails for at least a week, because if I start getting e-mails the say we release it … probably the first three days that anybody sends me, I just going to maybe ignore, because you’ve not given it enough time to really formulate a proper opinion. And the other thing is, of course the placebos. There’s going to be a lot of people with placebo effect, and it happens to the best of us, where you think it’s better.
1:02:24 JL: Try to go back and forth between the 32-bit [and the 64-bit] and actually make real comparisons if you can. I know that’s difficult to do properly in Second Life; in fact I don’t know if there is a proper way to do it in Second Life, but the more facts, the better. But I also want to know your opinion as well. How does it feel to you? Does it feel better?
1:02:49 EM: The placebo effect is very, very strong in some of you. It truly is. When the 4.4.1 release went out, there were people complaining about this and that and the other thing. When we had to withdraw that and put out the 4.2.2, which was essentially the same viewer, a debug setting changed. All of a sudden, people were saying, “Oh, this is so much better than the 4.4.1! Thank you very much!” And it was the same viewer! Same code!”
1:03:23 JL: It was the same thing. It was one commit which had nothing to do with anything that you would have noticed, and suddenly it was way better! That’s the best example, actually of placebo … We had people say it was worse, as well.
So interest list fixes, SSA revision and group ban updates are all scheduled for the next official/non-beta release?
1:04:08 EM: that’s correct.
You should just increment the version number and release a “new” viewer to keep people happy.
1:04:16 JL: We have considered doing that. Because people also have this idea that a higher number is better. We’ve noticed that with preferences a lot. People say, “Oh my God, my bandwidth can slide all the way to 10 trillion! I should do that, because it must be better! And draw distance goes really high too! Let’s crank that up!”
Will the 64-bit alpha be released concurrently with the public beta?
1:05:22 JL: That is the hope. In truth, Lassie, how many beta testers have the 64-bit right now?
1:05:36 L: None.
1:05:37 JL: None. There you have it. None. So when I tell you the 64-bit is alpha, I mean it’s not even beta tested … the hope is that it’s going to go out at the same time. But we still have yet to test it.
1:06:23 JL: How many of you are in our Preview group? For those of you not in the Preview group, don’t expect to see the 64-bit before anyone else. That’s all I’m going to say.
[Discussion about the Firestorm Halloween party, held on Sunday October 27th.]
What was the recent issue with Snow Leopard?
1:12:02 CR: Apache portable runtime. Library incompatibility.
Preview is getting it before the Beta Testers group?
1:13:32 JL: It’s worth repeating. If you don’t want 64-bit before everybody else, then don’t join the Preview group … Beta will get it before the Preview group, I just can’t promise how much sooner you’ll get it, but you will get it before the Preview group.
1:14:47 EM: I have two little things that I’d like to talk about … I asked the support team in our meeting last week for topics to bring up. Things that we should talk to the users about. And it was almost unanimous: tell people not to clear their cache, is what they said to me. So I’d like to ask all of you to help us out and pass the information on to people that clearing your cache is not always the right thing to do … Far too many people cache their cache at the drop of a hat, and it really is counter-productive. So I’m asking you all to spread the word.
1:16:17 EM: when this beta release comes out, please, please do a clean install … A clean install’s important, you can use the backups, but please do a clean install and save yourself some headaches as well. [see also clean reinstall.]
Sometimes clearing the cache fixes problems like items disappearing from your inventory.
1:16:54 EM: In that case, you should only be clearing your inventory cache. So that’s the two things I wanted to bring up.
1:17:22 JL: Lying to support? People lie to support?
1:17:38 EM: “Oh yeah, I’m pretty sure I did a clean install.” “So you got all three Firestorm folders, the program one and the two in [C:\Users\[username]\]AppData?” “What two in AppData?” … That’s the thing. When we give you advice, and we send you to the wiki to do something, please [do it] one at a time and in order and be careful and do what it says. It does make a big difference. And for any of you that don’t come out to our classes, come out to our classes. It’s not like high school or college, we tend to have some fun.
I have a silly question. What’s a clean install of Firestorm? Install SL whole together, or just remove Firestorm from the Control Panel?
1:18:39 EM: It’s getting rid of all the Firestorm related settings. Uninstalling the program, making sure that everything is gone form that folder, if not the folder itself; and then on top of that, the two folders in AppData. One for your cache and the other one for your settings.
1:19:02 JL: speaking of settings, I’ve got some good news for you. If you have made a settings backup from your 4.4.2, you can import that into the 64-bit.
1:19:20 JL; But I’ll tell you, installing the Windows 64-bit is quite different from installing the 32-bit. The installer is different … it’s very basic and needs a lot of work … This 64-bit that we’re giving, there will be as some point, no doubt, another update … our current installer is going to make you uninstall what you’ve installed, and in doing so you have to go through your Add/Remove programs, or in Windows 7 it’s Programs and Features in the Control Panel, which is annoying. but it’s just because the installer is primitive … it was difficult to get a 64-bit installer in a short time, so it’s not ideal. So there’s a heads-up to that. I’ve had to go through this uninstall process several times and it is a pain in the butt.
1:20:31 JL: Also, when you first launch it, it is not going to recognise your settings for the 32-bit, so you’re going to have to enter your name and your password, so hopefully you’ve not forgotten it [contact Linden Lab support if you have] … the Windows 64-bit is quite a bit bigger in file size to download.
What if I told you I’ve never done a clean install of Firestorm since I started using it?
1:21:07 EM: I’d say you were one of the lucky few.
Will the 64-bit version come in a version for other grids to?
1:21:13 JL: It will only come in the version which can be used to log-in to OpenSim and Second Life.
1:21:22 Whirly Fizzle (WF): Windows 64-bit installer doesn’t have language choice. But you can change language once logged in.
1:21:29 JL: There’s no language choice with the installer. Also … you know with our 32-bit installs it shows the publisher as being the Firestorm project? We actually pay for that every year so that we can assure you that when you’re installing something that says “unknown”, then it is not coming from us, and you should not be installing it if it is pretending to be an official Firestorm viewer. The installer for the 64-bit doesn’t have that; it’s not signed, so it will return “unknown”. Just make sure you only download it from our website.
1:22:14 JL: Unfortunately there are people who will take Firestorm, and they’ll do this with other viewers as well. They’ll take the viewer, they’ll compile it themselves, they’ll put into it a keylogger that sends them everything you type, including your password and all that kind of stuff. And then they’ll start-up a fake group that’s like the official Firestorm group, and they get people to join the group and then they send people the links … and people download it and install it, and then their accounts get hacked. and that’s because they didn’t download it from us. Don’t download it from a link, go to the website to click the download and take it from the website.
1:23:48 TS: We’re still going to do it, but it’s a really low priority, so don’t hold your breath.
1:23:54 JL: Actually that came up at the last Q&A … OTR, for those of you who don’t know, that’s encrypting your communications so that Linden Lab can’t spy on you, and neither can the NSA. Well, the NSA can spy on the leaders of other countries, tsk, tsk, even allied countries … they can read your chat if they want, but why would they bother? Anyway, we are going to do a form of OTR at some point in the future, but it isn’t a huge priority.
I’m curious, why am I one of the lucky ones[in not doing a clean install]?
1:24:58 EM: Have you ever been watching our support group right after a release? The number of people coming in and saying, “Oh this isn’t working, and this isn’t showing up right,” and they go off and do a clean install, and come back and all the problems are gone. We really don’t like telling people to do a clean install, we don’t like telling you that you have to clear your cache and clear your settings and all that. We really don’t. But it makes a huge difference to the average user if they do a clean install; then at least we know any problems are with the installation and not any leftovers.
1:25:47 JL: I think what doing a clean install helps with, is when you’re doing support, it’s a process of elimination, and asking the user to do a clean install usually eliminates everything, straight off the bat. If you’ve done a clean install and you’ve done it properly and you still have the problem, then that has eliminated a huge amount of problems and questions support would have to run through with you and try to work out. So doing a clean install, although it may take you ten minutes, is actually saving you and support like an hour of time and in most cases, it actually does fix the problem, in which case it saves you even more than that.
I was actually talking to someone about a week ago and the issue they were having, everyone else they could see fine, but they couldn’t see themselves fine; they were grey. I told them to turn off their HTTP [Preferences > Graphics > Rendering > Use HTTP], clear their cache and relog. So she clears her cache, relogs, turns off HTTP, she can see herself and everybody fine…. so as far as toggling HTTP on and off, should you also clear cache as well?
1:27:22 EM: No.
1:27:22 JL: Clearing cache probably wouldn’t have made a difference. It’s possible the HTTP might have. Linden Lab actually has a whole bunch of work coming to us that’s HTTP related, improvements in performance, and some people will say degrading performance by slowing things down, but I’m not convinced that’s going to happen …
1:23:43 JL: There are some people with some modems and some routers that do not handle Second Life well. They don’t handle the viewer very well. And Monty Linden, bless him, he’s an amazing guy, has been working for over a year now trying to work out for these few people how to fix this problem. And it’s kind-of like re-writing all the network architecture behind the viewer and server-side and it’s no easy task. The hope is in the coming months, and in the next release, possibly, we’ll have Linden Lab HTTP updates and they will all work and people won’t have problems with HTTP any more.
1:28:48 JL: It also depends on your service provider. The viewer uses more connections than almost any other game or program or anything, aside from p2p sharing programs, and some service providers actually incorrectly think that their customers are doing file sharing when they’re using Second Life, and I believe we’ve actually had users get shut down by their ISPs. Of course, it’s been resolved, but just because it comes across that way. There’s a lot of network stuff going back-and-forth with the viewer.
[Further discussion of the specific issue encountered (and resolved) in the question above, essentially requesting that people with issues be pointed towards Firestorm support]
Would it be possible to incorporate the clean install functions within the Firestorm install? In other words force the clean install at the Firestorm download/install.
1:32:50 JL: Yes, we have thought about that. The problem is that part of that clean install involves files in your AppData folder, which the Firestorm files themselves are not that sensitive, necessarily, aside from the chat logs which are also stored in that same location. so we have considered that we could have the installer just go in and wipe the settings, but it’s very difficult for the viewer to determine what those settings are, especially when you have people who have different settings files. So which settings files do you have the installer wipe? It would basically end-up being unpredictable or just dangerous. We could have it wipe everything, but then we’d have people just lose their chat logs, they lose their saved password files, they lose all kinds of stuff, and that’s really annoying. And it’s possible you might put something else in that folder … and if the installer wipes it, it’s gone. It’s just dangerous and it’s just better that we give you it step-by-step, you actually see what’s in the folder, the support wiki does say to back-up your chat logs.
1:34:18 JL: The correct fix is not to have the installer do the clean install. The correct fix is to figure out why clean installs fix things, and fix it. The fact is, clean installs shouldn’t be needed to diagnose problems and be a fix all. At the moment it is. We have spent a lot of time in trying to figure this out, why some people get corrupt settings while others don’t. I’ve never had the problem … I’m one of the worst ones for not doing a clean install. Our beta testers are told almost every single time they get a new version to test, they have to do a clean install; and our support people do it as well.
1:35:24 L: Beta testers are being told not to do clean installs any more … we’re looking for that settings thing. If they come across something that’s fixed with a clean install, it’s a lot easier for me to get a beta tester to give me their settings files than it is to get those files from a normal user.
What’s the status of the ATI crash on mesh? I’m still on Catalyst 12.1.
1:36:57 JL: You mean on selecting mesh, right-clicking on mesh?
1:37:35 (WF): Give the latest Catalyst a try. It might be fixed. That crash is a lot less common these days. There’s is another nasty mesh crash at the moment, though, actually on all viewers and all systems. It happens when you wear certain very high poly rigged meshes and you edit on something. The viewer will freeze in a loop.
As a point of interest I am now past the amount of memory used that I would have crashed on[with] 32-bit – and I haven’t crashed.
1:38:36 JL: That’s good to hear . I’m presuming you’re on the 64-bit self-compiled, right? … So that was one of our hopes, because we do have a memory crash. I think Linden Lab has it, I think all viewers have it, actually. Although letting the viewer just use more memory is not the proper fix for memory crashes. But hey, if it works!
Do we have a fix in for Interest list?
1:40:54 JL: No. Linden Lab is releasing soon(ish) possibly not-so-soon, new code with interest list fixes. For those of you who don’t know, that’s the bug that when you land somewhere and you know your house is right in front of you, yet it’s not there until you right-click on empty space, and it’s there. That’s an interest list bug. The fix for that is not a small fix; it is a whole crap-tonne of code. It’s going to be a nightmare to merge, and it’s not ready yet. So we expect to have it in the release after the public beta. So I’m afraid you’re going to have to live with that bug a little bit longer. But one is coming, yes.
1:42:41 WF: LL haven’t even released it yet. You can roll your own if you want though: https://bitbucket.org/lindenlab/viewer-interesting. But it’s still really buggy.
[Minor conversation, with re-iterations of release times for beta / 64-bit & about the Halloween party]
1:44:06 JL: OK, folks. I’m going to call an end to this … Thank you all for coming, hope you got your questions answered. If you guys could do us a favour, if you see somebody in some other group ask a question we may have answered tonight, please feel free to pass on the information. The more people know, the easier it is for them, the easier it is for us. And it’s just nice to help people out.
1:45:14 EM: I want to thank you all for coming out to. It’s very encouraging that you’re willing to come out and ask questions and get more information straight from the horse’s mouth, so to speak, rather than running around listening to rumours.
- Catznip 8.1 is without materials. There are a number of v3 and v1 viewers supporting materials, which are a mix of release and alpha / beta / experimental versions.
7 thoughts on “Firestorm Q and A, October 26th: video and transcript”
Jessica is so so wrong on this:
If she even bother to read your blog she would know that:
Black Dragon, kokua, UkanDo, Cool Sel viewer experimental branch, all those have materials support already!
Catznip does not have materials!
You may have noticed there was quite a bit of uncertainty in my statements. This is indicated by me saying things like “I believe, probably…” and “I think.” Meaning.. I wasn’t actually sure. I also stated “The major viewers don’t have materials out” Black dragon, UKanDO, Kokua, Cool Viewer, RLV are hardly major viewers. They barely have enough users to show up on the stats most weeks. The major viewers are Firestorm, Singularity, Catznip. Even Phoenix has a lot more users than the ones you mention. So even if my statement was false, which I don’t believe it is.. it wasn’t intentional or an effort to mislead anyone. Also, I do not consider ‘alpha’ as something having been released. Alpha is not even beta.. not even tested. That is not something that gets adopted by more than 1% of the viewers users in most cases.
Ignoring most of the tpv’s giving only credit and consideration to what you call the majors! Nice aproach jessica!
It has nothing to do with giving credit. The only credit that should go out in regards to materials is toward LL and the developers who made it. TPV’s who released it should get no credit as they didn’t make it. Nor did we.. nor was I taking any credit for FS releasing it. No, the comment was purely in regards to adoption of materials and how many can see it. It was about the accusation that Firestorm is holding back SL features like materials. Of the viewers you mention there may be about 4000 +/- users combined (roughly) (not including LL) compared to the well over 300,000 on the ‘major’ viewers. Without actually doing the math, that demonstrates that a very small number of users actually seeing materials without or without us. I’ll repeat to make sure it is clear. The comment was in regards to how widely adopted materials is… as in how many people can see it. Does that make sense? Perhaps this is a pointless argument discussion… I’ll leave you to your opinion.
I can agree and for sure, we wish all could use Alm at all times, be cause of materials or not, cause it really makes All look so much more amazing (and not only in SL)!
But i do think Firestorm should dare to do more, being the leader You could change things, like the default camera settings, that are horrible as we all agree and that for sure would make a lot more impact and would show for many how the world can be perceived and enjoyable much better!
I would like to take a step in here and say “The Default settings are the developer’s way of telling you to get off your ass and change it the way you like”
Comments are closed.