Adaptive bucket splitting for distributed rendering.

[Pixel_Domain]

Any chances to implement Adaptive Bucket Splitting for distributed rendering? I found this feature as life-saving in many cases, but unfortunately it works only for local rendering. It could be one of the best feature in Vray since ever if it was available also for DR.

[dimitar_panayotov]

Yes, it is something we are working on, so stay tuned.

[Pixel_Domain]

Yes, it is something we are working on, so stay tuned.

Best news ever ))))

[Vizioen]

Best news ever ))))

Don't get your hopes up yet, it's been 3 years in the works. And seeing as the update cycle is slowing down, you might have to wait a while longer.

[Pixel_Domain]

Don't get your hopes up yet, it's been 3 years in the works. And seeing as the update cycle is slowing down, you might have to wait a while longer.

...emotional damage...

[^Lele^]

Don't get your hopes up yet, it's been 3 years in the works. And seeing as the update cycle is slowing down, you might have to wait a while longer.

Where do you get such rumors from?

[Vizioen]

Where do you get such rumors from?

I'm just being realistic, maybe a little too blunt but first mention of DR dynamic bucket splitting being worked on: https://forums.chaos.com/forum/v-ray...69#post1137269
Apart from me being around since Vray 1.49 and looking at the dates of the changelogs and releases lately, I have the impression it's slowing down with resources being pushed to the newer acquired products. Granted I was also disappointed with the release of the latest hotfix as well. No harm intended all in good faith .

[^Lele^]

Right, but that was hardly a promise of any sort, now, was it?
It turns out neither of the methods was truly viable, and nothing else was available to solve the issue..
Things like this then have to stay parked until a solution is found, it's not just a case of working harder (or not!) at it, and in this specific case the DR part has to play ball too, which isn't a given.
It's RnD, but the D can only come if the R was successful.

On the slowing down of feature releases, I don't think it's debatable.
Otoh, it's not 2005 anymore, the low-hanging fruits have all been picked, and these days finding improvements is absolutely not trivial.
Add to this that supporting open standards (OCIO/Aces, USD, MtlX, etc.) is exceptionally taxing (tons of simple and tedious, and not so simple code.), and a constantly moving target, and you get the gist of why things aren't as snappy as they used to be.
Yet, to be fair, I feel we have added a wee bit of general performance in the past years precisely "solving" long-standing issues (bucket splitting, adaptivity clamping, firefly suppression, and so on), so while not offended at all by your remark, I don't think it's entirely reflective of the quality of our efforts either.

It's definitely not going to be Chaos' vMax team to code for Enscape, not to worry.