When to expect a roadmap?

[Ernest Burden III]

A Roadmap is a good idea. But I just fear that it might be a road to nowhere, and that NL are starting ( or maybe already know?) to realize this. Gosh, I hope I'm wrong.

 

You're onto the one point I've had a hard time selling--that it may already be too late, and we will not get a working product or a refund. But if that's where we are there isn't anything to do about it. So all there is left is to find a way to help NL succeed, assume its possible.

 

A roadmap is the best way to get an honest look at where we are, and then to prioritize what's left.

 

If NL refuses to play ball then that will be the best indicator that they are in trouble, and it may be time for everyone to pull the financial plug.

 

One detail of online creditcard acceptence is that the banks have a right (because they say so, that's why) to remove money from your bank account when there is a dispute. NextLimit can say 'no refunds' and VISA can say 'that's what you think' to them and hit them with what's known as a 'chargeback'. What starts the process is a customer claiming they haven't gotten what they paid for. If a company's chargeback rate hits a certain percentage of monthly transactions (it may be 1.5%) their bank will stop them from getting any new money. So where does this go if NL's monthly sales drop and their chargebacks climb? It ain't good. Oh, and the banks usually side with the cardholder.

 

So anything that raises customer patience is good for all.

 

Roadmap.

[Fran]

To paraphrase Clarence, I think the RC5 will be the best indicator as to where the "Roadmap" leads.

 

Hope it's not to this:

 

When I see RC5, I will decide whether or not to talk to my credit card company about my 2nd license purchase. I really don't regret my original purchase. It's the second one that I wish I hadn't made.

[_PopArt]

Well, as Vlado put it on the Vray forum (my paraphrase) - 'there's a reason why no one has implemented these mlt algorithms for the last 10 yrs'...Buffos mentioned here somewhere that he would give them more or less 3 months before we will see if its a dead end or not, and that sounds pretty reasonable. (I hope that's correct). The RC1-4 core engines had so many problems that it hard to know where to start...

[Devin Johnston]

Well, as Vlado put it on the Vray forum (my paraphrase) - 'there's a reason why no one has implemented these mlt algorithms for the last 10 yrs'.....

 

There must have been numerous tests done by NL before they decided to go in the MLT direction. Those tests however crude must have convinced them that MLT was a viable technology, or they devised some modifications for the algorithms that made them think it was probable for them to develop an application out of it.

[_PopArt]

For sure. Plus the algorithms, as we've seen, are capable of an astounding level of realism. I think it has'nt been considered up until now mainly because it hasnt been considered as being 'fit for production (meaning slow).' Surely after 2 initial years of development NL must have been confident that they could have overcome all of the obstacles (and they did make great progress), and sooner or later they'll bring it to completion. Its just a question of how long. And even when completed, speed issues will still prevent it from being used in cg production, at least for some time to come.

[Ernest Burden III]

When I see RC5, I will decide whether or not to talk to my credit card company

 

I love it--'slow children at play'. But is that us, or nextLimit?

 

Like I said, if its too late, its too late. But working on the idea that it is not, I think a somewhat unified demand from a group or 'customers' is the last, best hope. But it needs to be a group effort, just me isn't doing much so far, just any one of us, either.

 

From what I've read of RC5 so far it reads like we're being prepared to be under-whelmed. Its still 'you will be amazed---later'. If its 'RC5 and out' for Fran then its going to be that way for a lot of others, too. Fran isn't a hothead like some of us. So let's see if we can put together something here.

 

So does anyone have the feature list from days of old?

[Guest imanobody]

I think RC5 will be the final chapter in Maxwell's history (thank God, I need to get back to having a life). There may be a few other versions, or even a V1, but no one will care by then. History tells us that if NL doesn't state it, they don't have it. So there's no Sun/Sky, clipmaps, volumn fog, etc. The images they posted look pretty good, but so did the RC/Beta comparisons and we all know how that turned out to be a scam... I'm sorry, it wasn't a scam; they just forgot to tell everyone that it was rendered with a "hybrid" BETA engine.

 

I think they started out honestly believing that they could do this, but then found out too late that it was not feasible and now are just looking for a way out of this mess without giving refunds.

 

This technology will be feasible... in another 10 years (My lightscape screams now:D).

[AdamT]

All I can find is a feature list for Alpha 1.2.2, e-mail dated 3/10/05:

 

"Dear Customer,

 

We are writing to inform you that a new version of Maxwell render is available.

Maxwell Render Alpha (version 1.1.22)

 

"New Features

 

· Motion blur of camera ( movement, rotation, fstop and focal length )

· Motion blur of objects ( controlled by shutter )

· Clipmap support

· Region bitmap render

· Fog (volumetric render)

· New camera parameters: ISO

· Now z-layer, alpha-layer, cosAngle-layer have depth of field and motion blur.

· 4 processors per license.

 

Performance:

· Less noise with the same render time.

· Less memory requirements for voxelization

· New MXS format with size optimization"

 

Ironically, since that alpha we've had no clipmap support, no volumetrics, and in C4D we've *never* had motion blur or region render.