Wednesday, July 13, 2011

still, I can't bear the thought of you now

All right. We are where we are. Nothing changes that. But...

What if, when v.2 was announced, as the horrendous reconfiguring that it was, but one that needed to be done to get some form of movement back in the game...what if the user community had responded the way every other game on the planet has responded, and worked out ways that dulled the horror to a very mild eeping?

Think I'm kidding? exists to supply mods that gamers in WoW, Rift, Starcraft II, Runes of Magic, Warhammer Online, Aion, Age of Conan, and scattered smaller free-to-plays believe are vitally necessary to continue playing their games of choice.

Minecraft doesn't even hide it; their mods are posted right in the user forums.

Instead, what's happened? (And I'm as guilty of this as anyone else.) Hate. Hate hate hate. Hate loathe despise hate. All heaped onto v.2's frail shoulders.

According to those who've actually done the hard work and gutted out the interior of the game, it's not that hard to mod. So...why hasn't anyone done it?

I'm serious. Why isn't the Second Life community banding together and making mods to let v.2 look like v.1, or look like WoW, or look like Minecraft, or be all-over pink and chartreuse, whatever? Why aren't we doing this?

We can make entire alien worlds, but we can't fix a little pie menu difficulty? We can sculpt robot parts that whir and turn and rotate and light up, but we can't find better ways to deal with the sidebar?

Instead, what are we doing with our time, our energy, our life and our livelihoods? Bitching about it. Instead of doing anything about it.

And I'm as guilty as everyone else. We're all guilty of this, every single avatar who's ever shook her head at v.2 and said Yeah, well, I'M never using that thing...

All those months. All those blog entries. All those forum posts. It's all such a goddamn waste of space!

We need to do this differently. Now I'm tracking down code bits in the .xml wiki pages, seeing what I can do that others have tried, first.

This "morning" (my morning, at least), I found an email from Second Life marked Meet Residents -- Your Connection Recommendations:

(from the bizarre album)

Now, disregarding the fact that I know three of those people (in fact, Ameshin Yossarian was one of the first people I met in SL, when I was a young and unvarnished newb), I'm typically the type to become mildly resentful when I get these messages from other companies.

I will find my own friends, thank you, is generally always my initial reaction. Especially that if it's by matching people via interests alone, at this point, I'm virtually guaranteed to get some set of recommendations for people who dislike me, have actively muted me, or whom I otherwise don't resonate with for various reasons, because we all seem to circle in the same communities!

But maybe--I said maybe--it's a step in the right direction, and could give people who don't know many people on the grid that feeling of common ground. Why look at that, they could think. Those people also like spam and arson! I should friend them right away!

There's a new JIRA on the horizon, and this one's been amusing to sit back and watch. It's all on how much the look of items changed with the application of lighting effects and shadows. Now, that particular JIRA isn't bound to go anywhere, because Dessie Linden determined that it immediately linked to this JIRA, about essentially a similar problem.

There were a couple of really interesting visuals applied to the secondary issue, too. This one in particular:

(from the bizarre album)

This is a neatly graphed progression, split into two sections: shadows ON, and shadows OFF. The upper (grey) orbs are without bumpmapping applied; the lower (gridded) ones have bumpmapping.

If you can't make it out in the smaller size, I heartily encourage viewing the whole thing, or, if Picasa won't give you the full size, try the direct link to that particular image. This one demonstrates, clearly and completely, that is is not a bug or a malfunctioning procedure of any kind; this demonstrates that the workarounds we developed for getting through the initial Windlight issues need to be abandoned. Lighting effects and shadows are exactly like Windlight was, a radical viewer advancement--some computers can't run them, and content is going to break because of the new display stats.

Still unconvinced that this isn't a bug? How about this image:

(from the bizarre album)

This is another perfect example (and again, if you can't see McMahon's image clearly enough, look at it directly): when shadows are off, shine exists; when shadows are on, shine is reduced but depth is added.

Remember, when we first developed the shiny/bumpmapping tricks, while we could--and can--use shine with existing textures, for the most part, we don't. Most of us texture an object we intend to make shiny with a very simple texture, or no texture, so that the metallic concept "shines" through...aheh. My apologies.

In the above image, the pipes and tanks are clearly colored from blank stock, then have shine applied to them. Applying shadows kills this effect, but adds depth and distinction.

Translation: for all the content you think is broken, it's not. You just need to go out and buy (or make) better textures, now, and find the next workaround. (You also need to realize that lighting effects and shadows can't be seen on all systems, so you're going to have to pick your market. Do you make it pretty for those who can't use the new advances, or do you make it pretty for those who can? Or do you start putting out non-shadowed/shadows on versions of everything?) While these are all important questions for creators of content, it's not something the new viewer structure broke.

Back on the inventory keystroke search JIRA that I covered some days back, I was very, very wrong in how I described it. Namely, I was talking about what has now been explained to me as standard search. Which works in all viewers.

Where I had it wrong was in not comprehending exactly what was being described, because in five years on the grid (woo, I'm five, yay, blah), I have never done this. Not once.

If you have used it, imagine how wide my eyes got, never having known that form of inventory existed before! And if you haven't, here's why it's important:

In 1.x viewers, you can go to a specific folder in your inventory, select it, and start typing. This will find you the first item in that folder and bring you right there.

Looking for a piece of the Anais gown? Type it in after selecting the folder it's in, and whoomp--you're there. Want to track down the first stuffed bear you have? Go to Objects, type in 'bear'--and it finds the first one. This is the definition of "Fast, Easy, Fun".

And in 2.x viewers? IT NO LONGER WORKS.

Yeah, that's kind of a big deal.


Winter said...

Prior to v2 being announced, Linden Lab had spent something like a year, saying "We can't accept or commit any new patches, we can't change anything in the viewer until after the new viewer comes out, because the new viewer is changing the paradigm drastically."

That's just common sense. they outsourced a major overhaul to huge swaths of code, and they couldn't change that code without invalidating what the outside developers had in their hands.

Thing is, the volunteer Open Source developers, who were contributing patches and fixes to the 1.23.5 set.. got bored. I mean really it was ages and ages of LL basically ignoring their efforts, or dismissing them outright. A lot of developers started quitting trying to help, working solely to create or improve Third Party viewers.

Then LL, in a fit of legal gymnastics, released a "special" version of the viewer to opensource, so they could take all the "Second Life" labels off of it for legal reasons. In the process, fracturing the developer community. Now people were developing solely for this thing called "Snowglobe".

Then in the midst of playing Musical CEOs and preparing to fire 1/3 of it's employees, LL finally got v2 out into the world, they reaped the crops that they sowed in the opensource fields. A fractured opensource community, a huge scandal when Emerald did what Emerald did. And most of the Open Source developers had wandered off, washing their hands of the whole thing. Whether out of boredom, out of desire to see their fixes acknowledged, or out of plain ol hubris, who knows.

The Third Party viewers got dozens of dedicated developers, even a few ex Lindens, so rumour would suggest. And poor poor v2 had to struggle on life support.

A new development team was formed, and within a year, all but one member of that team would leave Linden Lab. Eventually, some developers did return to LL, once LL started accepting patches again.

But now the Jira is full of bug reports, and SnowStorm is a small team with only so many hours per day.

And mistakes, big mistakes, like those present in viewer 2.7.4, make it out to the public, and the viewer development team is no longer agile enough to simply retract the viewer deploy until it's ready.

So we have to live with issues (like 'show transparencies' and 'prim counts' being broken) for days, weeks, or months until the next batch of fixes and features runs through whatever QA process they've got.

It's sad that the Open Source project is in the state that it's in, with third party developers actually refusing to allow LL to implement their fixes and patches (Fuck you Qarl)... but that's the situation as it sits now.

The community once upon a time, did indeed try to help Linden Lab fix it's viewer. Linden Lab refused a lot of that help (like changing the chat interface back to pre-windlight functionality.. that's what started the first 3rd party viewers). Linden Lab has changed it's mind, and it's position several times on the issue. Suffering from a sort of corporate multiple personality disorder.

But how do you fix it? In a climate when everyone is so happy to just say "Viewer 2 sucks, use __________."

Emilly Orr said...

Exactly. Thankfully, I downloaded 2.8 for Snowstorm, so I never ran into the Highlight Transparencies bug, but for people still using the official viewer, it's still there, and it sucks.

And the bigger problem now--in fact, why I'm tacking v.2 again at all--is that it has finally sunk in that if I don't, when mesh arrives that will utterly kill the game for me.

And when mesh arrives, where are all these TPVs going to be? Getting patches in small dilute doses, taking months to update new content released from the Labs, plus, as you pointed out, refusing the help the Labs be better--I mean, how screwed up are things going to be then?

Anonymous said...

That assumes the 1.x TPVs don't get mesh. I'm still betting Henri will manage it.

Emilly Orr said...

Maybe, but the problem's larger than that.

For better or worse--actually, there's no better in this, but still--viewer 2 demanded a different implementation of code. The Labs are steadily, day by day, hard-coding that in to the servers. Right now, there are visual problems in how TPVs observe things. There are JIRA bugs being filed which the Lindens are never going to deal with, and openly STATE they are never going to deal with, because the problem is how that TPV analyzes the new code.

The world is changing. And even if viewer 1-structured analogues code in new patches, they'll STILL not be interpreting the code right.