RHS 5 & 6.758 comprehensive update uploaded/frozen/final?

[el cid again]

I don't see it yet. Will be out of the loop for 9 hours.

[el cid again]

pwhex set 5 is released and in the upload process

for the first time this includes a regular pwhex and a panama pwhex file

the major change is to river systems in Burma and NE India - to make them more correct and
match some new art by Cobra

minor changes of great potential significance are connecting the Soviet Rail system to Korea
and Manchukuo at two points - where they were and still are connected.

Much work has been done in different places on identifying eratta in the 5.05 release -
and I hope to incorporate most or all of this in 5.10 - set for Tuesday release. This is supposed to
include the revised plane data - but I don't have the Japanese numbers - Mifune seems to be down.
I may have to do them myself.

Cobra has released art for the revised entry path proposal - but I have NOT done the pwhex file for it
NOR the changed location and ship files - which will be affected by that change. [Reinforcements change;
SLOC ships partly return to player control as well]. I am not working on either right now - and will not until
we have released 5.10 and started human testing based on it. This will be a while. It will be called 6.x

AFTER that we plan to do a more radical revision of the Panama area - opening up more battle space and
creating a Caribbean Entry Hex - probably linked to the South Atlantic Entry Hex and maybe to US East Coast.
That would be version 7.x

AFTER that we may tackle Madagascar - a dream of Cobra's so strong he long ago did the art!

[el cid again]

My above post remains valid except I have not got the Japanese plane data -
I have worked in a number of corrections of several kinds -
and 5.10 should release as planned Tuesday.

At that point I am going to "surrender" the files for still more data washing -
initiate human tests - and work on the unrelated pwhex file versions six.
When I get the files back - we will release 6.00 - which is just 5.10 with
any eratta fixed and the new entry scheme, and related impacts
on orders of battle/entry. Unless I get the new entry pwhex file done sooner.

[CobraAus]

from WITP patch 1.8.0.2 change notes

very helpfull

-Item 42 - Allow limited refueling from Soviet Bases once Soviets are active.

Fueling is not allowed from Soviet Bases until soviets are active. Found a few places where the "until soviets are active" portion of the check was not implemented. Corrected all checks to include soviet activation. Also limited refueling once soviets are active to Ships in Soviet Task Forces unless base has been captured by the Japanese, in which case only Japanese ships may refuel there.

Cobra Aus

[el cid again]

This change should impact ONLY RHS RPO and RHS CVO scenarios - where the Russians are not active and bases are defined in the normal way. In the Russian Active Scenarios Russian bases and ships are technically defined as French - so they will refuel ALL Allied ships, unless captured. I see no reason to change that. An ally is an ally. Further - vast amounts of shipping refueled in Soviet ports during the war. I am trying to figure out WHY we cannot send lend lease to Vladivostok - not how to make the ships carry it not refuel!

[witpqs]

Sid,

The item below might be relevant for the very large USN AO's in RHS. As I recall you had to reduce their capacity becuase the code converted it to a negative number. Now (with 1.8.0.2) they can be switched back.


From the 1.8.0.2 release notes posted by Don Bowen:

Item 37 - Fix overflow in fuel calculation

Ship classes with very large fuel amounts can lead to a numeric overflow in the refueling routine. This generates a very large negative number and actually adds fuel to the refueling base instead of subtracting it. Increased size of local variables to maximum unsigned integer to reduce chances of overflow. Also bulletproof with a check to ensure the result is not negative nor greater than class fuel.

[el cid again]

It affects ships on both sides - and it is good to hear. I wonder if it is fixed for AKs? The largest AK in the world (probably) was Brizialian Yamatama Maru - the size of Queen Mary. I have her running at about 1/3 her capacity!

[witpqs]

It's specifically a fuel fix - are you taking about 1/3 of her fuel capacity or cargo capacity? At any rate a query to Joe should get clarification for you.

[TulliusDetritus]

El Cid Again or Cobra Aus, when will the new maps be available? Weeks, months? I mean the next 2 steps:

1) "a more radical revision of the Panama area" and
2) the one with Madagascar

An estimate (even if it is not very accurate) will be highly appreciated [8D]

[el cid again]

The maps for the new entry path are done and even available on the RHS site.
Only the pwhex files and new location/ship files (changing reinforcements)
to match are required: ETA Friday.

The radical revision for Panama ALSO has art done - but probably it needs revision.
It ALSO needs a pwhex file. ETA 7 to 10 days.

The Madagascar project is more complicated - it may not work at all.
If it does - it will be about a month.

[el cid again]

RHS 5.10 is in the upload process

It has revised plane maneuverabilty as posted on the maneuverability thread.

It has about 2 dozen more supply sinks separated from units - mainly commands -
and weakened. Many commands are now mobile.

It has a lot of eratta and revised fields for technical reasons.

Some Japanese bombers and flying boats lost their dual torpedos in favor of bombs -
because the code will not use more than one torpedo. And I found IJ specification
information indicating we had the wrong load for G8N and probably Ki-91.

I am turning this over for another session of debugging for some days -
and I am going to focus on human testing - and issuing the pwhex for the new entry
path. 6.00 may be 5.10 with the new pwhex and appropriate revisions to reinforcements.
The Allies get some of their SLOC ships back. We hope that by next week a 6.x version
will be out with the new entry system and (impossible though it may be) perfect files in
the sense of no eratta. We think this version of 5.10 is cleaner than any version of anything
EXCEPT with respect to leaders - who remain an issue. 6.x will clear that up. A heroic
effort to clean leaders for 5.10 failed - but we understand the problem now.

Note that we do things that generate error reports which are not errors - including creating some
generic leaders (generic allied, generic axis and generic axis engineer are associated with supply
sinks and some construction battalions). We also get leader errors when we say a Mongol Cavalry
unit is Chinese - but give it an IJA commander - and put it in a Japanese slot. It is not an error - it
is deliberate - but the error software won't understand this sort of thing. And so on. What we need
to resolve is mainly duplicate leaders - and also bad date leaders. The first is actually a problem when
a unit is destroyed: the other units then lose their leader and it is not replaced properly.

[witpqs]

ORIGINAL: el cid again
We also get leader errors when we say a Mongol Cavalry
unit is Chinese - but give it an IJA commander - and put it in a Japanese slot.

Does the code treat this as desired?

[Nemo121]

Well the code applies a negative combat modifier to Chinese troops so this is a means of ensuring these units have really negligible combat potentials. So, insofar as this is what Sid wanted to achieve, yeah, it works as desired.
 
Whether or not I agree that this choice is valid, especially when American,british etc engineer units don't get treated in the same way ( they get labelled as American, Indian, british etc and thus get superior combat modifiers), is another question entirely.

[witpqs]

What I mean is that the unit is said to be Chinese but put in a Japanese slot. Which side is the unit intended to be on - Allied or Japanese? Does the code treat it as desired by the modder in all ways (without some undesirable side-effect)?

[el cid again]

ORIGINAL: witpqs

ORIGINAL: el cid again
We also get leader errors when we say a Mongol Cavalry
unit is Chinese - but give it an IJA commander - and put it in a Japanese slot.

Does the code treat this as desired?

Well - no. The effects are substantially as desired - and the undesired side effects are a compromise. [Modding is, above all, almost art in its requirement for a fine sense of compromise - a point on which Andrew and I have commented many times]. The bad effects include that casualties are probably reported as Allied in this case. Now that may or may not be wrong - since they really are NOT Japanese casualties. But it isn't exactly right in the sense we would keep score that way. Of course, casualties are a bit wierd to begin with: the "weight" of a unit is used as its "manpower"! I just noted - in an AI vs AI run - a battle on Bataan on May 10, 1942 - in which the Allies are stated to have 78,000 men and change: remarkably close to the 80,000 men that were there at that time - and the first time I have seen such uninflated manpower numbers (in this case, the sum of 38 land units). [I am used to reports of 230,000 men at Port Moresby - on one side or the other - so this is a pleasant change for me]. When we discovered that "nation" does not mean 'nation" or even "service" - and that side is a function of slot rather than of this assignment - we realized we could play with the values assigned these nationalities for combat purposes. Thus, many of the really big supply sinks are Chinese, and others are Indian or Filipino -
and similarly for colonial/allied troops on both sides (where there was no secondary nationality we substituted the closese of the ones available) - and presumably we get weaker units (at any given size and equipment level) because we did. It will take human testing to be sure.

[el cid again]

ORIGINAL: Nemo121

Well the code applies a negative combat modifier to Chinese troops so this is a means of ensuring these units have really negligible combat potentials. So, insofar as this is what Sid wanted to achieve, yeah, it works as desired.

Whether or not I agree that this choice is valid, especially when American,british etc engineer units don't get treated in the same way ( they get labelled as American, Indian, british etc and thus get superior combat modifiers), is another question entirely.

Actually - Japanese engineer units are usually treated the same way Allied ones are (but see notes below) insofar as they are rated as IJA or IJN. ONLY one formation is treated otherwise: the independent construction engineer battalion. [Note this formation is used by BOTH IJA and JAAF]. These units were almost worthless as combat units IRL - and they were often the ONLY IJA units to surrender in the battles for Pacific islands. They were NOT ethnic Japanese - but mainly Korean - and this is the reason I have classified them as "chinese" - which I do for almost all Manchukuo, Mongolian, Jehol and Korean units. [There is one "Russian" Manchukuo unit - a wierd one really composed 100% of ethnic Russians - including its commander] Technical notes on Japanese engineers:

1) These are the worst units in the game for support. This is based on real organizatoin. They are composed of various combinations of skilled people, pioneers and equipment - in the general order (higher getting better equipment and more skilled ratings):

a) Independent Engineer 'regiments"
b) Road Construction "regiments"
c) Independent Construction battalions [All three are battalions]

Japanese engineers were intended to be supported by other folks - and if you put support heavy units in the same hex they will build a LOT faster.

2) Japanese engineers are poor in motorized support and elements - in the same order as above - relative to the Allies.

3) Japanese engineers are poor in heavy weapons - in the same order as above - but the higher up the list the better they are in combat. The third (c) category construction units are supposed to be almost worthless in combat - but quite worth having for construction - and this is why they are classified as "chinese"

4) There is one real Japanese combat engineer unit in most scenarios - the 7th "Electric" Engineer - to which EOS adds a second - the 27th.
These units have a unique form of remote controlled armor - an actual device - which is similar to the German Goliath. [The Japanese invented this in the early 1930s and probably had more of this sort of thing than any other nation; otherwise they are second to Germany]. Electric refers to radio or wire control systems. This (or in EOS these) units also have some heavy mortars - just a section - because the Japanese believed in these for combat support.

5) Japanese "assault engineer squads" are the only units considered to have flamethrowers. Japanese soldiers did not like these - only educated engineers would use them.

[el cid again]

ORIGINAL: witpqs

What I mean is that the unit is said to be Chinese but put in a Japanese slot. Which side is the unit intended to be on - Allied or Japanese? Does the code treat it as desired by the modder in all ways (without some undesirable side-effect)?

It appears to work. Similarly, ships can be assigned to any nation - and they appear to work. Only air units are a problem - they must be classified as IJA or IJN to work properly. If they have Allied planes - these work fine in Axis units BUT they appear NEVER to replace. These units may upgrade to Japanese planes which then replace. Since this is done only for Thailand - and since during the war Thailand could not get US aircraft (with one exception) - it isn't a big problem. RTAF DID get 9 B-10s from captured Dutch equipment - but I cannot get the code to let them be Thai - no matter what I try.

[m10bob]

With all you have been able to "remedy" or work-around ,is it possible to give certain air units a "night bonus" for either search or attack? I refer specifically to the "Black Cats", (of which were used by USN and Australia)..?
Yes, I realize you have already burnt your hands and elbows with so many irons in the fire, just offering an idea for the back of your mind..[&o]

[el cid again]

Don't know yet. But maybe radar helps? It is in RHS in many cases - first of all on the Swordfish. We need to play some human games to find out how it is working?

As for errors, our policy of addressing possible errors - not all "errors" are wrong even if software thinks they are -
is paying off. Some indications are that processing time is going down (by more than half) - and tools are showing much lower error counts (some remain - because tools and I disagree about what is an error). If we get the leader issues whipped there will be very few errors in the RHS data set - except to the extent you disagree with some value given on purpose. I long ignored leaders - except to set a very few to known leaders - but have learned that duplicate leaders are a problem for code when one of the units dies. So we will clean them up - and by the grace of God we have people interested in that - so I don't have to - for once. Even so - at RHS 5.10 level - you probably have fewer actual errors than in any variant of WITP ever issued - if tools can be believed.

Our biggest problem now is seeing about longer range issues. The supply sinks may appear too early for example.
Stuff like that.

[m10bob]

Since the real "glory days" of the Black Cats were when utilized in the Solomons slot and near shorelines,maybe the bonus could be tied into night port attack or something like that?..Just an idea..