Technical harvesting notes - English

search | Login Forums | New posts | Last 24h


	English

Forums » GENERAL » English » Technical harvesting notes

#1 Added by Forgotenland 9 years ago Report | Quote [en]

{translate}

I was really bored.

So I delved into the RyzomCore source code and tried to figure out how harvesting really works (unmarked spoilers everywhere!).

A lot of things are hidden in server-side config files. But we do have, in the RyzomCore repository, what looks like a copy of the egs config (as it was in early 2014), and light empirical testing suggests those numbers are accurate.

Feel free to suggest anything I've missed, or point out anything that doesn't match your harvesting experience.

Fun fact: at one point there was supposed to be a third "health" meter on sources that spawned creature(s?) when depleted. This is no longer a thing, but remnants of this idea exist in the code.

Edited 2 times | Last edited by Forgotenland (9 years ago)

---

#2 Added by Bitttymacod 9 years ago Report | Quote [en]

{translate}

Well, since my experience in digging hasn't changed in the last 5 years, I'd say the data was current.

However there are things that aren't in your analysis (at least to the limited extent that I can understand the writing). The main thing is that weather granularity must be modified by something in the code because I have seen exe mats that are bi-weathered, mats that have cutoffs that are not on the boundaries of the best/good/poor/worst, etc. There is a preponderance of weather change on the 1/6 1/3 .... changes in the weather number, but that is not universal.

There is also a definite and substantial random factor on the impact of any action on the stability and life of a particular source. If you included that, I'm sure I didn't recognize it.

The most important thing to a harvester (at least to me and various I know) is how does the location where you stand affect popping of nodes in various combinations of angle and range, and that isn't addressed by your extraction of knowledge at all.

All in all, some of the numbers reflect my experience (except for the random factors) and some of them might, but I'm unable to figure out what they mean.

---

Remembering Tyneetryk
Phaedreas Tears - 15 years old and first(*) of true neutral guilds in Atys.
(*) This statement is contested, but we are certainly the longest lasting.
<clowns | me & you | jokers>

#3 Added by Forgotenland 9 years ago Report | Quote [en]

{translate}

Thanks for the feedback! These are of course notes, and not yet a guide.

Bitttymacod (atys)

weather granularity must be modified by something in the code because I have seen exe mats that are bi-weathered, mats that have cutoffs that are not on the boundaries of the best/good/poor/worst, etc. There is a preponderance of weather change on the 1/6 1/3 .... changes in the weather number, but that is not universal.

Deposit filter setting is not exclusive; they can of course be flagged as showing up in a range of conditions. I've added a note to reflect that. There was also a bug in (recent?) clients that causes server weather and client weather to be completely different...

Bitttymacod (atys)

There is also a definite and substantial random factor on the impact of any action on the stability and life of a particular source. If you included that, I'm sure I didn't recognize it.

I do note that the actual impact of any action to D or E can be anywhere (uniformly drawn) from 0x to 2x.

Bitttymacod (atys)

The most important thing to a harvester (at least to me and various I know) is how does the location where you stand affect popping of nodes in various combinations of angle and range, and that isn't addressed by your extraction of knowledge at all.

Added a section on that. The mechanics aren't too interesting; basically it's a pick-and-draw until you get something or give up. Because of the way points are sampled (uniformly draw from angle range, then uniformly draw from distance range), you end up sampling close points more densely than far points.

Deposits themselves can be any shape allowed by a NLLIGO::CPrimZone, i.e. any valid polygon (?), and their exact locations are locked up in the server-side leveldesign data.

---

#4 Added by Mjollren 9 years ago Report | Quote [en]

{translate}

Apparently there are two types of "bad" and "good" weather: BMSite Atys Calendar , the page also displays the server value for weather.

Question regarding CRecentForageSite : does this contain all mat types? Regular deposits have basic/fine/choice spread all over the entire deposit area. If I prospect choice only with no angle, I usually manage to pop several nodes in a circle around me. If I dig <ForageSiteStock> mats, does that automatically deplete all 3 grades within 9m?

---

#5 Added by Forgotenland 9 years ago Report | Quote [en]

{translate}

A CRecentForageSite is attached to a single parent CDeposit and covers all mat types valid for that deposit. If you have a *single* CDeposit that's generating basic/fine/choice then performing ForageSiteStock harvest actions will sitelock all three classes. If on the other hand you have 3 overlapping CDeposits, one for each quality, then sitelocking e.g. the basics shouldn't affect e.g. the fines, since causing sitelock on one CRecentForageSite should only affect its parent CDeposit and not the two others.

Which of these happens in practice depends on how the world was built (leveldesign files) and must otherwise be determined empirically.

---

#6 Added by Karu 9 years ago Report | Quote [en]

{translate}

Mjollren (atys)

Apparently there are two types of "bad" and "good" weather

There is no two of them. There is single 'good' that spans two sections.

---

Hello!

#7 Added by Bitttymacod 9 years ago Report | Quote [en]

{translate}

Mjollren --

There is only one "weather number" (WN). The labels "good" and "bad" are irrelevant. Exe moon resin in GoC is present in "worst" weather (or from WN=0.85 to 1.0). It is not types of "good" or types of "bad" it is only the WN. The exe yana node in void is active when WN < 0.166 ("good") or when 0.5 < WN < 0.833 ("bad" -- but not for the digger). Karu's graph is spoiling the heck out of me in digging exe's and drastically increasing the length of the labels on my flags.

Forgotten --

The bug in the client weather report has been there since I've been playing, which is over eight years now, so it isn't new.

Forgotenland

I do note that the actual impact of any action to D or E can be anywhere (uniformly drawn) from 0x to 2x.

So you do. However, there is also an interaction between D (source life) and E (stability) ((and the number of mats and time remaining- but that seems to be a second order effect))

For instance in mode 3, the more that stability is greater than life (E>D) the more effect there is (on average) on life vs stability.

In mode 5, as long as E>D both E and D tend to be affected together. Once E < D E is effected much more by each action.

There are numerous other interactions that I have observed, and of course they depend on which stanzas you are using. ((The above observations are for q250 gentle speed 6 harmful rate 6, gentle extraction, material spec, land spec.))

---

#8 Added by Forgotenland 9 years ago Report | Quote [en]

{translate}

Bitttymacod (atys)

there is also an interaction between D (source life) and E (stability) ((and the number of mats and time remaining- but that seems to be a second order effect))

For instance in mode 3, the more that stability is greater than life (E>D) the more effect there is (on average) on life vs stability.

In mode 5, as long as E>D both E and D tend to be affected together. Once E < D E is effected much more by each action.

It's an interesting hypothesis, but not one borne out by the RyzomCore source. In all the relevant code paths, D and E are independent and do not couple to each other.

Some of your observations may have alternative explanations that don't contradict the available information. Take mode 5 for instance: D couples to speed, which will end up somewhat more strongly damped (21 coefficient and 4x absorption from matspec) than E, which is coupled to quality (39!! coefficient and 3x absorption from ecospec). But with all the absorption being slung around the impact ranges may end up similar enough to each other that you'll frequently observe the bars going down together at first, until E takes (and statistically almost certainly keeps) the natural lead.

---

#9 Added by Bitttymacod 9 years ago Report | Quote [en]

{translate}

I understand what you are saying but after several years of near-daily digging, I can say with some certainty that the correlations that I have seen are real, predictable, and in some cases, useful.

If, in mode 5, I heal stability to the point where E>D I will have many actions where E remains greater than D. If I do not, then E will rapidly degrade to 0 and blow me up. I think I have enough trials with mode 5 to make a statistically valid sample.

I'm not saying that you haven't looked at the code. I just think there may be other bits of code.

---

#10 Added by Gidget 4 years ago Report | Quote [en]

{translate}

Based on the fourth section (Sources), I've figured the following;

SM 0 - Use high/harmful Rate to degrade Life; use gentle Q to preserve Stability; Terrain spec preserves both; Mat spec is irrelevant

SM 1 - Use high/harmful Rate to degrade Life; use gentle/lower Speed to preserve Stability; Terrain spec preserves Life; Mat spec preserves Stability

SM 2 - Use Harmful Q to degrade life; use Gentle/lower Rate to preserve Stability; Terrain spec preserves both; Mat spec is irrelevant

SM 3 - Use Harmful Q to degrade life; use Gentle/lower Speed to preserve Stability; Terrain spec preserves Life; Mat spec preserves Stability

SM 4 - Use Harmful Speed to degrade Life; use Gentle/lower Rate to preserve Stability; Terrain spec preserves Stability; Mat spec preserves Life

SM 5 - Use Harmful Speed to degrade Life; use Gentle Q to preserve Stability; Terrain spec preserves Stability; Mat spec preserves Life

Preservation - Gets a +20% boost from Mat spec

Stability - Gets a +10% boost from Terrain spec

---

Do not assume that you speak for all just because you are the loudest voice; there are many who disagree that simply have no desire to waste words on you.

Last visit Thursday, 18 April 17:38:40 UTC

P_: