FAQs about Kokua

Howdy, gang! In the past few days, we’ve been receiving a lot of questions about Kokua, and we’ve noticed a lot of confusion and misconceptions about it. So, I’d like to take some time to answer some common questions, and help clear up a few things:

Read on for all the answers, or click any of the questions above to jump directly to it.

Will the Kokua Viewer UI be the same as Viewer 2?

No. Some people assume that because we will use Viewer 2 as the starting point for Kokua Viewer, that it will therefore be very similar to Viewer 2. That’s just not the case. Our goal is to make you go, “Wow! This is what Viewer 2 should have been like in the first place!”

You might be familiar with the television show “Extreme Makeover: Home Edition”. In each episode, they send a team of construction workers, designers, and decorators to completely renovate a crummy old house that no one wants to live in. Usually they will keep some aspects of the original house that worked well, but they rethink and remodel the bad parts to transform the house into an amazing place to live. And they aren’t afraid to knock down a wall or two if need be!

So let’s just say that we have our sledgehammers polished, and we’re ready to play “Extreme Makeover: Viewer Edition.” ;)

Does that mean the UI will be like the current Imprudence UI?

In many ways, yes, but not totally. Believe it or not, the current 1.x-style UI has many flaws, and the Viewer 2 UI actually has some good aspects. It would be silly to recreate those old flaws, and to throw out the parts that Viewer 2 got right, just for the sake of keeping the UI exactly the same.

Instead, our focus will be on transforming the Viewer 2 UI into something that is usable and comfortable for people who are used to the old 1.x-style viewers like Imprudence. So, Kokua Viewer will have many of the good parts from Viewer 2, but a significant amount of the UI will be changed to be more familiar, and to support a smoother workflow.

For example, you will be able to have more than one avatar profile open at a time, and the sidebar will be totally optional. We’re hoping to offer a choice between old-style pie menus and the new-style rectangular menus (but the timeline for that depends on how much effort it will be to recreate the pie menus).

And of course, we’ll be adding our own improvements, both ported from Imprudence and new ones created from scratch. :)

What do you mean when you say Kokua Viewer will “use Viewer 2 as the starting point”?

Allow me to explain by returning to the house metaphor. These days, the SL 1.x viewer is a dusty, rickety old shack. Granted, it’s a shack that we all know and are familiar with. We know which boards creak when you step on them, and we know not to flush the toilet while someone is in the shower. We can get along fine with it, for now. But that doesn’t change the fact that there are cracks in all the walls, and its plumbing and electrical wiring are a few decades out of date.

Imprudence is based on that rickety old shack. Sure, we can sweep the floor, pick out some new curtains, maybe throw a new coat of paint on the walls. But it’s becoming increasingly difficult to make any significant additions without the whole house collapsing on our heads. Whenever we want to add a new room or remodel an existing one, we end up having to fix all the plumbing, and re-wire the electricity, and replace all the wooden beams that termites have eaten through.

Viewer 2, on the other hand, has a more solid foundation, modern wiring and plumbing, a sturdier structure, and a hot tub. The only trouble is, the interior designers were told to make it into a nursery, then had to go through a bunch of revisions, and then the owners put the house up for sale when they were only half finished.

Fortunately, the problems are mostly on the surface, and the underlying structure of the building is intact and useful. Sure, it will take some hard work to remodel it into something most people will want to use, but the end result will be sturdier and easier to expand than Imprudence could realistically ever be.

When will Kokua Viewer be available?

Kokua Viewer is still a long way off. We announced it now to let you guys know in advance what will be happening in the months to come. We will be starting on initial development sometime this week or next. We are hoping to begin weekly preview releases in mid-December, and have a full release sometime in February.

Our focus right now is still on making Imprudence 1.4 the best possible 1.x-based viewer we can make it. We know it will take a while to hammer Viewer 2′s UI into shape, so we want everyone to have a reliable, featureful viewer to use in the meantime. We are aiming to release Imprudence 1.4 by the end of this month, but we may push that back a bit if we need more time to make it really solid.

But! We will be including changes in Imprudence 1.4 (starting with this weekend’s Experimental) so it can take advantage of the new features in Kokua Sim. So, you will be able to enjoy the benefits right away!

Will we have to switch from Imprudence to Kokua when it is released?

No. We will never force anyone to switch to Kokua. Imprudence will continue to work into the future, we just won’t be updating it anymore after 1.4. We prefer to spend our energy on creating a modern viewer that people will really want to use. If we do our job right, most people will want to switch, so they can take advantage of the new features and improved UI.

That said, there have been vague hints that Linden Lab might block access to Second Life for “old” viewers at some point in the future. So if you’re a Second Life user, you may end up having to switch away from Imprudence and other 1.x-based viewers anyway. Hopefully by starting now, we will be able to have Kokua Viewer ready with a clean, usable UI before anything like that happens.

Will Kokua Viewer work with non-Kokua versions of OpenSim? Will Kokua Sim work with non-Kokua viewers?

Yes, absolutely. We are still very much committed to the broader metaverse, with all its various viewer and sim versions. Kokua Viewer will be able to connect to Second Life, vanilla OpenSim, and custom OpenSim setups like those used by InWorldz, 3rd Rock Grid, and others.

Likewise, Kokua Sim will work with other viewers, including Imprudence and other 1.x-style viewers. Of course, some viewers have better OpenSim support than others, so we can’t vouch for their stability, but we won’t be doing anything to intentionally “lock out” non-Kokua viewers.

Of course, the whole point of our project is to create a viewer and sim that work especially well together. So, there will be some viewer features that other sim software might not support, and some sim features that other viewers might not support. But, both the Kokua Viewer and Kokua Sim will be open source, so other viewers or sim software will have the opportunity to integrate those features if they wish to do so.

Will Kokua Sim be open source? Are you making a Kokua grid?

Yes, Kokua Sim is and will remain open source under the BSD license, the same license used by OpenSim. Like OpenSim, Kokua Sim will have the ability to run an entire grid, but we have no plans to set up a new grid service of our own.

Instead, we will be offering Kokua Sim as an alternative version of OpenSim to be used by other grid operators, as well as average users who want a “standalone” sim, or one connected to a grid like OSGrid, or one available via Hypergrid. And, because it is open source, existing grids will be able to borrow features and enhancements from Kokua Sim, and integrate them into their current infrastructure.

What is “Aurora”, and how is it related to Kokua Sim? (Updated)

There has been some confusion about the name, but Aurora and Kokua Sim are one and the same. It has been quietly developed since summer of 2009, and some people have known it as Aurora since that time. It will be developed as Aurora for a little while longer, while some wrinkles are ironed out and we set up the new Kokua website, then it will be rebranded as Kokua Sim.

We decided to rebrand it to emphasize the fact that both the viewer and sim are being developed together as part of the same project, and built to work especially well with each other. We hope that everyone who have known it as Aurora will come to like the new name and what it stands for.

Update: Kokua Sim will be keeping its original name, Aurora.

Won’t developing the sim software distract you guys from viewer development?

Not really. Although we will be working together as part of the same team, each developer will have their own focus. In other words, the team members who have been working on the Imprudence Viewer so far, will still be focusing mostly on the viewer in the future. (But, from time to time a viewer developer might work on the sim, and vice versa. Some developers might even split their time half-and-half. It’s up to each developer to choose what to work on.)

How is Kokua pronounced, anyway?

“Koh-koo-ah”. To help you remember, you can think about the sentence “Coconuts are cool and awesome!” :D

Technically, the word is spelled kōkua, with a macron (aka a kahakō), the line above the “o” which indicates that it is a “long O” sound (as in “go”). But like many Hawai’ian words, kōkua is often written without the macron, to make it easier to type on standard English keyboards. We will tend to spell it as Kokua, to make it easier to write and search for.

Any other questions?

If there is anything else you’re confused about or would like to know, just leave a comment and we’ll try to answer it!

20 Responses to “FAQs about Kokua”

  1. Justin Clark-Casey

    I love your analogies, Jacek :). Best of luck with the new projects!

  2. Nalates Urriah

    Very well written… thanks.

    Where can we find info on setting up a Kokua/Aurora sim?

  3. DilSpi

    ohoh – Kokua

    Kokua spells ‘LIKEaBIKE’,that sounds to me a tiny bit conflicting http://www.likeabike.co.uk/about_likeabike_uk.php

    also Kokua is found sooo many times

    maybe an other name could be much more relaxing.

  4. EthanG

    Just a note about the right-click menu: A conventional menu (as Viewer 2 has) will be just unusable for some. Computer menus in general are difficult for many people; I thought it was just me but every time I mention it in-game someone agrees. (Less so when I’m talking to programmers. *shrug*)

    That said, you won’t necessarily need to make a pie menu to make it usable. A pie menu would be somewhat better because it’s somewhat easier to remember entries by position, but it’s not a huge help. (On that point changing the sit and touch text in a regular menu is a terrible idea — please at least keep “Sit Here” before the object-set sit text so users can find the menu entry.)

    The main problems with conventional menus are finding and clicking on the desired entry. Picking out a word or two in isolation does not come naturally to many people, including some prolific readers, primarily because you have to break concentration to do it. It’s also unrealistic to focus enough to aim the mouse at such tiny text.

    The answer for both problems is to make the text larger, but just for the right-click menu. The text in the viewer is tiny by default, and you can’t make it larger unless you want the UI to cover most of your view of the world. This is bearable in most areas but very much not in the right-click menu.

  5. Kakurady

    Will 1.4 still receive bug fixes, security fixes and protocol updates?
    If enough people write patches for 1.4, will you collect and publish them?

    Will we be able to see some cool mockups of the new viewer?

  6. Oz Spade

    This sounds really cool. I’m looking forward to see what comes out of the project. Good job on handling the questions.

    @EthanG: What about using icons instead of text? I’m thinking a pie menu-esque with images. Would that work for faster recognition by users? However then there’s the problem of “how do you communicate an action like ‘sit’ via an image?” Then you also get cross-cultural differences in imagery which is why text is so appealing, since it’s more-or-less translatable. But for how graphic/icon oriented many programs are I’ve always found it a little strange how much a 3D world like SL uses text in its UI.

    Note, I’m not involved with the Kokua or any viewer project, I just like to discuss UI design. :)

  7. Jacek Antonelli

    @Nalates: Setting up Kokua Sim/Aurora is very similar to setting up OpenSim, except that the organization of config files has been reorganized. At the moment, I think the best way to get info/help about setting it up is to join the #aurora-dev channel on the freenode.net IRC network. We’ll have some more tangible docs available in the future, of course.

    @Kakurady: If there are serious bug fixes, security fixes, or protocol changes in the 6 months or so after 1.4 is released, we would likely apply them and release patch-level versions (1.4.1, 1.4.2, etc.). But we don’t want to get stuck maintaining old software long after it is obsolete.

    I expect the first several rounds of changes to be mostly smoothing out the workflow and making things more familiar, which aren’t the sort of changes that are usually mocked up in advance. But, I’m sure we will have mockups to show off once we’ve gotten our bearings and start tackling bigger changes.

    @EthanG: Interesting thoughts about the menus. Each menu type has its trade-offs. Radial/pie menus do tend to be quicker to use once learned, and easier for people with limited motor control. But more people are familiar with how to use the conventional menus, and it’s easier to add new items to them (which I know was one of the Lindens’ big frustrations about the pie menu, and probably one of the main reasons they switched).

    You make a good point about the font sizes, too. We’ll have to check out whether the Viewer 2 UI toolkit is better at dealing with adjustable font sizes than the 1.x toolkit has been.

    @Oz Spade: It seems logical that good icons would improve recognition speed, but it doesn’t always work out that way (even setting aside the challenges of communication-via-imagery and cultural differences). I seem to recall reading about studies using eye-tracking that found users tended to sit and stare at an icon longer (vs. text) before deciding whether it was the action they wanted. The really curious thing is it still happened (but to a lesser degree) even after they were already familiar with the icons and knew exactly what they did.

    But, this sort of thing also depends a lot on other circumstances, so it may turn out that a radial/pie menu with icons would perform well. We’d have to test it out and see. We probably wouldn’t have the resources to do eye-tracking studies, but the “How does it feel?” test is feasible. I can’t make any promises at this point, though.

    Our first priority with recreating the pie menus would be to make the viewer more comfortable for people who have the pie menus ingrained in their muscle memories from years of use. After that, we can play around with concepts for improving them.

  8. christy lock

    The pie menu is an absolute must have for me. It’s faster to work with it and the position of the selections makes it where you do not have to read or move your eyes off your work. It becomes part of the flow and if you listen to loud rock music and dance when you build it really helps a lot.

  9. Dave Bell

    As you know, there’s an OpenSim on USB-stick implementation floating around out there.

    It’s something I’d love to see for Kokua.

  10. Ener Hax

    don’t you like OpenSim? why not develop with them to make it really good? are there personality issues?

    why do you think your sim platform will be better than OpenSim?

    i am more curious than anything and wonder if you can sustain and grow your sim software to be much larger than OpenSim (i know the answer is yes, but years of work has gone into OpenSim and they have a track record – i donate to them and believe in supporting what you use – but would likely not support two diff platforms)

  11. Tampa

    So then you gonna open a new site just for the new viewer or switch both to a new domain? ;) Just so my Bookmarks dont get lost

  12. Gavin Hird

    I hope you will take this opportunity to structure the Mac version of the viewer like a true Mac app with the menu at the top bar and not anchored to the floating window. Also supporting full screen behavior should be much easier if this is done right in the code (you basically get it for free.)
    That will in itself release a good deal of screen real estate on 16:9 aspect displays. In combination with true Mac OS X widgets (and the underlying frameworks), it would make these users feel much more at home.

  13. Andre

    and also mesh support would be nice. .dae or other types of meshes…

  14. Nica Pennell

    Ener, it sounds like Kokua is going to be based directly off of OpenSim’s codebase rather than being a parallel independent development. Which is great, because it means that any improvements the OpenSim folks make should be pretty easy to incorporate into Kokua going forward (and possibly vice-versa, once the OpenSim developers ease off on their restrictions about developers working on both viewer and server software). As long as some care is taken not to fragment the protocols being used I see this as nothing but a positive development for the open metaverse.

    BTW, now that the possibility of adding new simulator features significantly beyond what Linden Labs supports exists, I might as well mention my own personal dream item; the ability to paint ground texture explicitly rather than simply defining elevation ranges and hoping the resulting pattern looks okay. I long to be able to transfer ground texture patterns from World Machine along with the heightfields. Shouldn’t take all that much bandwidth to send the info. :)

  15. Ron Overdrive

    @Ener Hax: There has been a lot of politics with getting patches submitted in OpenSIM where its almost as bad as getting patches submitted to LL. One example is OpenSIM Core policies dictate if you’ve even looked at the code within 6 months you’re forbidden from submitting patches. Likewise any patches you might have made during that 6 month period will not be accepted. This makes cooperation and co-development impossible with OpenSIM.

    Aurora doesn’t have that limitation. This is why it makes sense to partner with the Aurora team instead of OpenSIM Core team. Bug fixes and new features are easily accepted and troubleshot making Aurora’s evolution much more rapid then OpenSIM’s. And unfortunately due to the policy differences Aurora’s patches will most likely not be accepted by the OpenSIM Core devs unless they change their policies.

  16. Aabye

    Wish you best of success with your new project.

  17. Ener Hax

    thank you Nica and Roin for the thoughtful answers and i am glad i did not come across as negative (i think in french and type in english and it sometimes comes out harsh – but us frogs have a chip on our sholder anyway) =)

    ah, a parallel track, that sounds like a good deal all around. resources are limited and i would hate for two groups of people to get burned out, but this type of work often brings out the best in each other

    and add me as a second to your idea on ground textures! that would be wonderful and so handy, especially for roadfs, paths and sidewalks!

    i had no idea about the politics with OpenSim. but that makes sense, it is like that in the gigher education system – the smaller the school the more ferocious the politics become (Henry Kissinger said something like “the politics are so fierce because the stakes are so small”)so i must be misunderstanding you Ron, if you have looked at OpenSim code you have to wait 6 months before looking at it again?

    if that is the case, they have made an elite core group that no one can move into. so if i come up with some great thing (no worries, i won’t) and they stick it in, and then because i am stoked and passionate, come up with another great thing in a month, that work is not looked at?

    if that is the case, holy cow, that is stoopid!

  18. Nica Pennell

    Just this morning I discovered that Naali has had a prototype of ground texture painting for several weeks now – someone posted to the OpenSim mailing list investigating how to go about integrating it into OpenSim’s core. So I imagine we’ll be seeing it in Aurora at some point soon as well. Woo!

    The 6 month delay thing is in regards to the viewer’s code since it’s released under a slightly more restrictive license than OpenSim’s BSD license (the viewer is LGPL, and used to be GPL). The concern was that OpenSim might get unintentionally “contaminated” with code taken directly from the viewer, resulting in licensing problems. A reasonable enough concern I think but a 6 month quarantine is perhaps too strict a solution to it.

  19. Lodragan

    I hate that Open Sim is based on Microsoft Windows .Net – it limits the platforms you can deploy on – and when you can deploy on non-native platforms you have to have an added layer (e.g. Mono on Linux) that negates the advantages of the host system (e.g. Unix IPC vs Windows RPC) as well as performance. I also would like to gain the benefits of distributed and P2P technology in the architecture.

    Additionally Open Sim is a pain in the butt to manage – I would like to see configuration management either very straightforward in the configuration files, or hide that altogether from the user and allow the configurations to be managed from the GUI interface (via the viewer).

    What are your views on this as you enter the ‘world hosting’ system arena?

    Thanks, Lod.

  20. Soto Hax


    Any rough estimate on a preview release date yet? >->