Thalatta

Movement speed: slinger > javelineer ~ archer > crossbow, given their equipment and armor, which would be ordered the other way around (Chinese crossbowmen would use large shields, at least sometimes). As mentioned, depends. Certainly not for the Rhodians (historically I mean).

No clue, never read about their archers :P. Would need to find sources... They should. That's why I also gave them a large dodge defence.

I misread this, since before you said more damage and shorter range. You can't separate things like that. Some Asiatic recurved composite bows were way shorter than longbows, yet had longer maximum range (although what matters is effective range, but they didn't differ much on that). It's about the materials and complexity. They were indeed harder to build, but were necessary when going on horseback. Anyway, a fast ranged weapon with short range and less damage could be a repeating crossbow (the famous chu-ko-nu), supposedly invented at the time of the game, in the Warring States period, with the Han being involved. A slow ranged weapon with more range and damage would be the gastraphetes, but the game gives it a range of just 45m, which seems wrong to me. Then there's the Chinese crossbow, which maybe in the late Warring States period outranged their bows*, even when for the Han the game has the range for the Infantry Crossbowman as 45m, and for the Archer as 60m. I think this is wrong if we consider effective range (bolts lose less kinetic energy in flight than arrows). I haven't checked about other crossbows around that time, which weren't many, and probably weren't superior to the Chinese. Another interesting disadvantage of the crossbows not usually talked about is that ammunition was reused by the enemy, although you can't use arrows on crossbows (too much power) and bolts on bows (too short), but some cultures had a guide to use baby-arrows on bows that would allow them to also reuse bolts from the enemy. *Source: https://dn720004.ca.archive.org/0/items/science-and-civilisation-in-china-volume-5-chemistry-and-chemical-technology-par_20210927_1445/Science and Civilisation in China Volume 5%2C Chemistry and Chemical Technology Part 6%2C Military Technology Missiles and Sieges by Joseph Needham%2C Robin D. S. Yates (z-lib.org)_text.pdf, page 137. In other parts it's mentioned that the bows of the Huns and the Xiongnu couldn't outmatch the Chinese crossbows, although I haven't read enough to check if this means outrange (I'd guess effective range).

Indeed, I was talking for a given bow. And for example, slings lose less kinetic energy in flight. Same with crossbows, given that darts are heavier. But for short distance and high damage, it's basically javelineers (in all its variants like pilum, etc). And the Kestrophendone if you want to get fancy. I thought people were always complaining that archers were OP. They should be a support unit, melee should reign supreme (at least until gunpowder :P). But I'm all in for variety, and as I mentioned before, giving all those archers the same range doesn't seem the right decision to me (not even counting that things are the other way around). You should have, regarding range, Persian Archers > Cretan Archers > Other Greek Archers. Regarding damage, I've seen discussions if Cretan archers used anti-armor arrow points, but have to look into it again. I do care about history because there's basically one of it, while one could achieve in many ways a balanced gameplay.

Well, it's not to completely negate, but acts as a modifier. Thus, you have values of pierce (p), hack (h), crush (c), block (b), parry (a) and dodge (d), both for attack (A) and defense (D): pA, hA, cA, bA, aA, dA, pD, hD, cD, bD, aD, dD. For example, aA is how hard is to parry its attack, aD is how easily it parries attacks. The better for the unit the bigger the values are. As a proof of concept, a naive formula for the damage the attacker deals to a defender could be (pA/pD+hA/hD+cA/cD)(bA/bD)(aA/aD)(dA/dD), meaning that all damage, after being divided by each corresponding defender's resistance, is added up, and then multiplied by factors related to the probability (it's NOT directly a probability) of the attack being either blocked, parried or dodged (just adding them up is problematic). Here I'm showing the (rounded) results, with 10 taken as an average value, and other parameters like rate of fire, movement speed and range not yet taken into account: This means that the damage ratio for spearmen:cavalry is 3:1 (as wanted), for cavalry:archers is 6:4 (which makes sense, cavalry would get destroyed by archers if they don't close in, like in Agincourt), and for archers:spearmen is 2:5 (which makes sense, the advantage of archers being not this but keeping their distance). Remember that rate of fire, movement speed and range not yet taken into account, which would incline more the scales to what is wanted. Also, archers:archers is 4 times more destructive than cavalry:cavalry, which is twice as destructive as spearmen:spearmen, which makes sense considering how long these kind of engagements last. Would be nice to keep adding units.

Well, I just wanted to buff them a bit, maybe I went too far Regarding "short-range ranged unit with higher damage", wouldn't/shouldn't that be the javelineers? Longer (effective) range should be some slingers. I see that the Macedonians already have quite some specific units, but an interesting one is the Kestrophendone, which would enter late in the game, it's a heavy dart thrown with a sling, and it was really devastating (but apparently hard to make). Ranges are debated, but tests indicate shorter than a bow. Also, I just noticed that Rhodian Slingers have 45m range, while Cretan Archers have 60m range, I guess this comes from myths put forward by other games regarding Cretan archers... For the Persians, I see that Sogdian Archers and Immortals have also 60m range. Xenophon's Anabasis states: "For at present the enemy can shoot arrows and sling stones so far that neither our Cretan bowmen nor our javelin-men can reach them in reply" and "the barbarians were no longer able to do any harm by their skirmishing at long range; for the Rhodian slingers carried farther with their missiles than the Persians, farther even than the Persian bowmen", thus ranges were: Cretan archers < Persian slingers < Persian bowmen < Rhodian slingers. Cretan archers were good for the Greek world, but were outranged by other specialised range units. Also: "Now I am told that there are Rhodians in our army, that most of them understand the use of the sling, and that their missile carries no less than twice as far as those from the Persian slings. For the latter have only a short range because the stones that are used in them are as large as the hand can hold; the Rhodians, however, are versed also in the art of slinging leaden bullets. If, therefore, we should ascertain who among them possess slings, and should not only pay these people for their slings, but likewise pay anyone who is willing to plait new ones, and if, furthermore, we should devise some sort of exemption for the man who will volunteer to serve as a slinger at his appointed post, it may be that men will come forward who will be capable of helping us". This is interesting because Rhodians weren't even hired as slingers, they were hoplites, who happened to have slings and were extremely good at it.

I have no idea how formulas work in this game (would like to know though), but along the “simplified realism argument”, I think bonuses should be avoided, they seem to me a sign that the parameters used to reproduce reality are either not enough or are being inefficiently used. I rather have a couple more parameters than who knows how many random bonuses here and there. This keeping in mind that it would be nice to eventually show the relevant info not as a bunch of text like it’s now, but as a given set of icons and values that is as simple and standard as possible. It seems to me that with hack, pierce, crush, block (with a shield), parry (with a weapon) and dodge values, both for attack and defense, most things should be achieved (I don’t see any advantage to organic/metallic armor types, it’s just setting hack/pierce/crush defence values). Also, causing low damage should do very little, as a measure of actual penetration of defenses, to enhance differences, and penalising massing up the wrong kind of counter. Going through some of what has been said, and just as a first approximation: Spearmen: high pierce attack, hard to block, low dodging. Archers: pierce attack, easy to block, basically impossible to parry, hard to dodge. Cavalry: low blocking, high dodging. Maces: high crush attack, hard to block. Axemen: high hack attack, hard to parry. Swordsmen: high hack attack, hard to dodge, high parrying, high blocking if they have a shield. Javelineers: high pierce attack, low crush defense, high dodging. Elephants: very high crush attack, high hack and crush defense, can’t be dodged, can’t dodge. Rams: extremely high crush attack, very high pierce defense, high crush defense, extremely easily dodged, can’t dodge. Buildings: extremely high pierce and hack defense, very high crush defense, can’t dodge. Slingers: high crush attack, low defenses. Slingers should be good against troops that don’t have crush defense, cheap, sling faster and farther than bows (bow damage should decrease more with distance), but have no armor, be unable to form close formations when slinging, should have a straight line of sight (slinging over friendly troops was too risky, iirc), and, most importantly, take a lot of time to train. Regarding their slow demise (they would be around up to the Middle Ages though), certain late agricultural technologies should drastically increase how much their training takes, since they came from a more agrarian background, and besides some early “biconical lead projectiles” tech to increase their pierce damage, they should not be improved much by other techs (around the middle to late timeframe of the game sling technology was historically at its peak, while others kept improving). Going back to unit’s values: archers’ arrows should be somewhat easy to block with shields, they would lose against cavalry, who should lose against spearmen, who should lose against swordsmen (here I wasn’t sure if they should win or lose against archers, I guess it should depend on how much their shield can block). I like the hard to block/parry/dodge dynamic of the mace/axe/sword, going along with the reality that crush damage transfers more through shield and armor, axes are hard to parry, and swords have faster attacks. I have not emulated all possibilities in my head though, all remains to be tested. Javeliners are good against elephants (from a distance), which along rams are good against buildings. It’s a matter of fine tuning to get what one wants, as Enrico Fermi told Freeman Dyson: “with four parameters I can fit an elephant, and with five I can make him wiggle his trunk". No bonuses should be needed!

Is this work around because of some coding limitations? Naively, I would think of normal bridges (the Persian was a particular case) as a structure that behaves like the port and the wall: start on one shore, finish on another (under certain depth/lenght limits), and the construction animation doesn't have to be that accurate (none is)... maybe there's some problem about walking on top of it if done in this way?

Ok, I just installed this and have a few questions: -Regarding "new quotes and tips" in the game load screen, it goes away too fast for me, I can't read it. As suggested at some point, "press any key to continue" could be implemented. -I have no Scenarios under Campaigns anymore, like I had with R27. And as suggested at some point, I'd remove the redundant "Continue Campaign" option from the main menu and leave only the one under "Single-player". -I like that civilians are both men are women now, but has there been any discussion regarding gendered roles according to each civilisation? I'd like to see more female priestesses for the Greeks for example.

These could be considered: -Artemisia I for Persians (or Carians, if part of the game at some point): naval commander. The one from the battles of Artemisium and Salamis. -Artemisia II for Persians (or Carians): naval strategist and medical researcher. Sister/wife of Mausolus, ordered the construction of his tomb: the Mausoleum at Halicarnassus. -Lǚ Mǔ for Han: rebel leader that played a role in the restoration of the Han dynasty. -Teuta for Illyrians: not a playable faction still (eventually?), but since Gaia's units being more challenging has been mentioned, they could even have Heroes, and Teuta would be compatible with Gaia's piracy units (although I'd name them differently than Gaia, and reserve that for basically abandoned things).

Probably not, but I think better not to drift towards the unrealistic direction. The big ships are one of the things I liked about the game the first time (and many have commented the same). I think it would be the wrong call to make things look more and more like any other RTS. I hope this is done at some point, would also solve the issue of strategising against ships with modified stats because of the troops being carried. Maybe. Personally I prefer battles that look closer to the Shogun 2 Total War ones, not so much on the side of numbers, but tactics. Although numbers is also good. On a side note, maybe different water depths could be considered, differentiated by color, which would be important the more relevant the bodies of water become: -Shallow: for the transit of most land units (except siege for example) and small boats. -Moderate: for small and normal boats, but not the largest ones. -Deep: for all boats. Maybe eventual bridges should not be built over it (except for the Persians, who could have also bridges made of ships).

Instead of making the ships unrealistically smaller, why not make the bodies of water realistically bigger? And then one would zoom out a bit for a proper naval battle.

I was thinking about this wonder issue some weeks ago. I think all wonders should have different effects, to enhance how each faction plays differently, while keeping it historically accurate. I agree that the "wonder tech" should be merged with the wonder, if that's the only thing they'll do. Also, I don't think the 7 Wonders of the Ancient World have the prominent role they should have, considering the time span of the game: 3 were built before 500 BC, the other 4 being built before 280 BC. The Persians have the Hanging Gardens of Babylon, when they should build the Mausoleum at Halicarnassus. I think they should start with the Gardens, thus working as a civ. bonus that could be captured or destroyed. Worse is the case of the Ptolemies, they have the Pharos (I would add “of Alexandria”, for everyone to get it just in case), yet it’s not even their wonder, but the temple of Horus (a bigger version of the temple of Isis). I would make the Pharos of Alexandria the wonder, and put the Great Pyramid of Giza close to them for capture, kind of another civ. bonus for the player to decide if and when to get it. This would make the wonder play of these 2 civs quite particular (historical ideas first, balance later). Their mentioned civ. bonuses could be called Echoes of Babylon (a known phrase, used in a book title, and a shoot out to the Voices of Babylon AoE campaign) and Sands of Time (another known phrase, the Pyramid is by far the oldest of the 7, the only one remaining, and hints at the race against time for capturing it). The remaining 3 of the 7 Wonders (not belonging to any playable faction) could be included in different game modes: the Temple of Artemis at Ephesus to be “rebuilt” at a specific location to win, while the Statue of Zeus at Olympia, if the map is continental, or the Colossus of Rhodes, if it’s on an island, could be captured to win (the scenario could start with Gaia building the wonder, preferably finishing it before a reasonable time to capture it after it’s finished).

Indeed, I always want to do the opposite, fill fields one by one. And if I send them to a filled farm, they already go to a nearby empty farm (not sure how far away the farms have to be for this to work). What!? How was I supposed to know this... there has to be some indicator for this kind of stuff (maybe there is?). I haven't played this game for too long. I agree that it would be nice for this kind of stuff to be added to the base game. I like this, although if the point filled first are the ones closer to the dropsite, not to add micro. They should. Actually, I'd like for unit abilities to be added, and them making a sound could have a temporal aura effect (maybe increased damage stats for now, but more realistically enemy morale decrease if this is added). I'd like to see day and night, seasons, weather, and other natural phenomena, eventually. In line with what was said before, I think they meant that they just become inefficient, since storehouse and farmstead techs would apply only to civilians. I think the idea is interesting (considering specialisation of trades), but faction considerations should be taken into account, since not everyone had the same dynamics between civilian and soldier. This could be solved by fine tuning which civilian, citizen-soldier and soldier units each faction has access to (and at which cost). For example, around the timespan of the game, Greeks and Romans would indeed mostly progress from civil militias to professional soldiers (that would find mercenary service more profitable than farming), but other factions like the early Germanics wouldn't make this shift in the same scale, and their armies (and male civilians) would mostly consist of farmer warriors even towards the end of the period (only later being more influenced by the Roman ways). All this would influence how the different factions should be played.

Well I thought pressing them in somewhat fast sequence (just like a dobule-click can be distinct form two clicks), but I see that when you press a given number in fast sequence it centers the camera on that group. Maybe it can't be done in a clean way (not disrupting what's already established). One possibility could be that repeated numbers can't be allowed (no 00, 11, 22... groups). Another that two different click speeds could be implemented: faster clicking for double digit numbers, a bit slower for centering on groups, the fast clicking working as the slower one if there are no double digit groups, making things work as they do now when you only have groups form 0 to 9.

@Genava55, I was on vacation, and then had things to do. I had a partial answer written down, which I’ll post now considering your new posts. Chronologically: A. Starting with Proto-Indo-European, according to Ringe, haimaz, “settlement”, comes from ḱóymos, “resting place”, while þurpa, “farmstead, village”, comes from treb-, “building” (this is your own source), which fits with the first Proto-Germanic dictionary entries for haimaz, “home”, and þurpa, “village, rural settlement”. Even in PIE my main concern is reflected: haimaz appears mainly as an ambiguous term, not necessarily structural, while þurpa appears mainly as a physical place, and not preferentially reconstructed as a single unit. B. Regarding your “6.3. The Household” screenshot of my source (Lehmann), you fell for a mistake there. It states that in Old Norse þorp means “farm, estate”, and that only in later West Germanic texts it means village. But ON texts are not earlier than OE, and you can check that ON þorp translates as village (this is something I’ve mentioned before, that basically only Gothic and Old High German support your view). The whole point of the discussion is, taking (the cognates of) haimaz, þurpa and wīhsą, how would they be urbanistically ordered? Both PGmc sources I cited (Lehmann and the dictionary) state that þurpą and wīhsą are more similar to each other than to haimaz (as can be read in your screenshot of my source, and from others). The only outlying source I find is Kroonen, who gives haima- as “village, home” and þurpa as “crowd”, which anyway blatantly contradicts your “singular first” interpretation (making all 4 sources mentioned against it), and what is a crowd of either houses or people anyway? Maybe he went for something figurative (I then found that the ON dictionary alings with this, more later on). C. You complain that, centuries later, Gothic þaurp translates as “farmland” (only once, which is not very useful), but disregard that the same can happen for weihs (at least twice). In fact, it also even happens with haimom. Now I see you addressed that “it seems Weihs was equivalent to Haims”, but sometimes looks even worse, in Mark 6:56 one has haimos as “villages” and weihsa as “country”, exactly the other way around of your proposal. Again, luckily Gothic is not the only thing used to reconstruct PGmc, if we rely too much on it no decision could be taken. D. The next reference seems to be the Old English one from 725 AD, which treats þrop closer to a village than a single farm. Basically all languages (except Gothic and OHG, as far as I can remember) follow this position, which is the position of the PGmc sources already cited. E. You dismiss the PGmc entry ordering, but insist on the OHG one, from a millenia after the fact, thinking a semantic shift right there can be traced back unchanged, independent from all context. You choose 2 datapoints there and just draw a linear function going all that way back, ignoring points A, B and D, instead of trying of harmonising all sources, since it could have been that what was a collection in PGmc times (as Ringe, Kroonen, Lehmann and the dictionary state) became singular for Gothic and OHG, and then OHG collected it again (a known behaviour in evolutive systems like linguistics and biology). Niemeyer might have referred to the previous singular stage when he stated “originally”, otherwise he would be contradicting all PGmc sources mentioned. You are only relying on a couple of languages from many centuries later, and that’s not that chronologically close, much less with something so subtle and apparently oscillating as what we are discussing here. That’s why “‑heim names are usually assigned to the oldest naming layer” is not only not old enough, but in a previous post of yours it is even clear the many meanings “heim” had. You even say the “only reason” wīhsą was your “second choice is that it seems to have been used much later, in the early Middle Ages, to refer in some cases to small Roman towns”, and this is nonsense to me, that’s way too late. I find this “critical approach” flawed for this reason, combined with that you give zero credence to what the experts have already reconstructed for reasons and with knowledge well beyond what has and will be discussed in this thread (this attitude doesn’t seem quite “mainstream”), and as if they have not already considered all that you have been saying. As if they don’t know what the Gothic Bible says! To finish this post (I dislike writing so much since Brandolini's law is BS, one can refute the Flat Earth theory with a couple of sentences, no matter how many books are written on it, so, it’s never a matter of quantity), I see you commented on the git commit https://gitea.wildfiregames.com/0ad/0ad/pulls/8722, and the problem with your points are: 1) We are not discussing the exact word reconstructions forms, but their meanings. For this, many disciplines were indeed used, it’s just wrong to state that it's “based solely on phonological rules and sound laws”. 2) Gothic is chronologically closer, but that doesn’t mean it’s enough, and as I’ve shown, it has sentences that directly contradict your proposal. The PGmc reconstruction takes into account all available knowledge. 3) Basically what I said on 2), and þaurp cannot be given much importance since it’s only used once, it was rare for them. 4) Just confirms that haimaz can be (and has been) many things, which makes it not specific enough. 5) Again, centuries after the fact, and doesn’t add anything to the issue of multiple and ambiguous meanings of haimaz. 6) Yet again, it cannot be said that þurpą “does not seem to have been used immediately to designate villages in place names” by citing a few examples centuries after the fact. 7) It is not true that the “oldest writings in Germanic languages” (as if that goes beyond Gothic) support your point, as B and D prove. As mentioned in E, with OHG you are a millenia off already, any shift happening right there means nothing. Points 2, 3, 5, 6 and 7 are just milking the same cow over and over again with Gothic and OHG, while completely ignoring the rest, which as I’ve said an infinite amount of times, has been taken into account for the PGmc reconstructions we have. As I explained in B, I did not contradict myself, you just fell for a clear mistake in that source (about something I wasn't even pointing out). 8) Since those few places you posted would be maxed out “Phase III” settlements (given you didn’t come up with anything bigger), then what would “Phase I” be? Seems you realised that and tried to preemptively explain things away, but the whole thing is unclear. Either you are saying that there was not much differentiation between types of dwellings (kind of my point), or that they had both farms and non-farm hamlets (which doesn’t seem right). In “The Germanic People”, Francis Owen states that there were basically two forms of settlement: the individual farm and the farm village (https://babel.hathitrust.org/cgi/pt?id=mdp.39015008272570&seq=176), so, hamlets again fall back to (a collection of) farms, which is how things were kind of everywhere actually. It’s clear that this early all those concepts are not that far removed given the context, translations, and witnesses. That’s why Lehmann says that Tacitus says that they were basically farms, not just because they were a bunch of isolated buildings, but surely because of all the crops and animals around. Caesar also talks about their annual land redistribution for farming, so, again, I don’t understand where you are trying to get with this paragraph. All this explains why the reconstructions are what they are (sometimes overlapping), and why one cannot just focus on particular choices one or two languages took centuries later. Many things I've said can be nicely wrapped up with what the (largest) ON dictionary says regarding þorp (https://cleasby-vigfusson-dictionary.vercel.app/word/thorp): “turba is taken to be the same word, this word, we think, was originally applied to the cottages of the poorer peasantry crowded together in a hamlet, instead of each house standing in its own enclosure”, “the etymological sense being a crowd”, “a hamlet, village, rarely of an isolated farm”. Although this proves Lehmann’s mistake, I’m not making the mistake of claiming one can extrapolate this all the way back to PGmc (even when I could point out the "originally" and "etymological sense" phrases being used), although I think that’s the right conclusion given what the reconstructions using all available knowledge state (and that ON comes from a separate North Germanic branch). Thus, the question is the same: which term is closer to wīhsą: þurpą or haimaz? Just go back to points A and B, which relate directly to PGmc, given that one can’t rely only on Gothic, where haims and weihs appear in interchangeable urbanistic order, and þaurp appears only once. The following instance this word appears in sources is OE þrop, "village".

@Genava55 As I said, literally, "proper villages [...] in the eyes of the Romans", you can just go and read Tacitus: "It is well known that the nations of Germany have no cities, and that they do not even tolerate closely contiguous dwellings. They live scattered and apart, just as a spring, a meadow, or a wood has attracted them. Their villages they do not arrange in our fashion, with the buildings connected and joined together, but every person surrounds his dwelling with an open space, either as a precaution against the disasters of fire, or because they do not know how to build", he's trashing their way to do it, for him they looked like big farms (and the ones you posted seem to be the biggest ones, a few extended families). But all this is, yet again, not even the point. The point is, according to A Grammar of Proto-Germanic, and the PGmc dictionary, and whatever other Proto-Germanic source we can cite, and if we agree that wīhsą can describe some kind of big farm or small village (which, again, could also be put into question according to yet another one of your cherry-picked criteria, since weihs has been used to translate agrós), then which term is closer to it, þurpą or haimaz? That is the ONLY question, and for whoever reads those sources the answer should be clear, as I already stated in my previous post.

I do check my info and make an effort I shouldn’t even do because I’m really not interested in arguing if the accepted Proto-Germanic reconstruction is wrong, which it isn’t, and what you have to check is your tone. Re-read my comments and realise that it’s not me stating you are “defending something weird that nobody talks about”, that’s you when for example called BS on the opinions I quoted from experts in the field. “Once again” it is you who acts like a condescending 12-year-old brat that feels the disgusting need to make pseudo-patronising “Brandolini's law” comments. It's not my fault if you can’t check info properly, just go to https://archive.org/details/diealthochdeutsc01steiuoft/page/n7/mode/2up, the column header meanings are on page 1. From there (and volume 3) you get the Sg. 911 and Sg. 242, which are dated from the very late 8th century (around 790 AD) and the 8th-11th centuries, respectively, as can be checked in https://www.e-codices.unifr.ch. It took me 20 minutes, no library needed (perfect place to make a condescending comment, I know, but lets have some class, shall we?). I only missed giving more span to the latter, because, again, I really don’t give a darn about spending time questioning the accepted reconstruction. I’m just giving evidence on why things are how they are, while you keep digressing towards centuries apart changes and ignoring any source that contradicts you regarding what matters: Proto-Germanic, from 500 BC to 1 BC. The only relevant thing is if haimaz is better or not for what is wanted. Nothing else. Luckily, I think I found exactly what’s needed: https://folksprak.org/common/material/pdf/A-Grammar-of-Proto-Germanic.pdf, which states that the Germans did not form villages but rather lived in isolated homesteads. Old Norse heimr, Old English hām and Old High German heim mean house or home, while Gothic uses haimos (only appearing in accusative plural) for village, and translates agrós 'land' to þaurp 'land, lived-on property', like Old Norse þorp 'farm, estate'. In West Germanic it means 'village', as in Old English þorp, Old High German dorf. In Gothic weihs 'village' also translates agrós. This is exactly what I meant with the demographic change, þurpą means what we need because there were no such things as proper villages, and both it and wīhsą seem to refer to whatever was there, call it land, property, farm, estate, with surely an extended family or more, and in the eyes of the Romans. I don’t see how any of them would be smaller or less appropriate than haimaz (taking from ON, OE and OHG). All this is exactly what the preferential reconstructions (are ordered entries important or not? Or is it just cherry-picking?) from the PGmc dictionary are telling us: haimaz is “house” first, “home” second, and “village” last, for þurpą the order is “village, settlement”, “gathering of people, crowd”, and “cleared land”, and for wīhsą it’s just “village, settlement”. All fits perfectly. If one travels in time it will look like a big farm or estate, conceptually it was the closest you could get to a village. They were not thinking in Phase I, II and III. And, as I said before, some branches kept it literal, while others kept the concept (which is what matters), and depending on each word. You are not going to convince me that the dictionary is wrong because, considering the source I cited and all the methodologies used that go way beyond your knowledge, it just isn't. If you have a problem with that, just take it to the ones that put it together, and use whatever word you want, I have better things to do honestly.

I agree, that's kind of my point 9 of the 20 points I suggested a couple of weeks ago

This sounds like just going with what the dictionary states, which is my position, mostly because in general it would be too much work to question what they say, and because things beyond our knowledge were also taken into account. In any case, I find using warją for the fortress a more pressing issue, I don’t think it is good to have repeated words if there are alternatives, all these situations deserve a second look given that @Vantha said “that's the case for some other civs too already”.

I’m just skeptical in ignoring what the Proto-Germanic dictionary says because of what Gothic says (which seems to have made choices opposite than the rest regarding this), when clearly this was already considered when the consensus was reached. It’s like reinventing the wheel. An interesting thing is that þurpą has been proposed to be related to the Latin turba, and for any Spanish speaker this is clearly related to a collective (meaning something like a mob). The Italic-Germanic split happened over 3000 years ago. Uncertain, but on the table, and shows that it is not generally assumed that þurpą was a singular unit early on. Only Gothic does that (as said, “in anderer Bed.”). With haimaz the Proto-Germanic reconstructions are home, house and village. I think there’s no way around that. It’s just like modern “home”, I could refer to my house, my hometown, or my homeland. That’s why I questioned it a bit, not that it can’t mean village (not from -heim though, but directly), but that it seemed too ambiguous. I don’t have a better alternative, I think we have exhausted what the PGmc dictionary has to offer. Alhs can mean settlement, but it seems too singular, and many other meanings are preferred https://kaikki.org/dictionary/Proto-Germanic/meaning/a/al/alhs.html.

@Genava55 Definitions 1 and 2 refer to Gl 1 and 3, both from the 8th century, so the meanings seem more simultaneous than sequential (one would need to believe the shift happened exactly there). The second line states "got. thaúrp (in anderer Bed.)", showing that in Gothic (the black sheep in all this) the meaning is different. Looking at the Old English lexicon: https://wehd.com/94/Thorp.html, the earliest reference, from 725 AD, treats conpetum, tuun, and þrop as equivalent, closer to a village than a single farm. The ending -heim doesn’t originally refer to a settlement, it’s the genitive ending of a singular personal name, as in Mannheim and Ingenheim meaning "Manne’s home" and "Ingo’s home" (or homeland), with the settlement meaning derived later.

@Genava55 The Codex Abrogans is from the 8th century. I shortened “villa worker” to “villager”, maybe changing the exact meaning, but the point remains that these things happened, even when they are later examples. I’m just trying to make sense of what the dictionaries actually say. Another issue could be how big these settlements are in context. You say the original meaning is farmstead, but what if a very small collection of farmsteads is also valid? At Proto-Germanic times this would have been the equivalent of a village (again, there has to be a reason why this has been reconstructed as such, besides commonality), and what meaning exactly had on the different branching languages could have turned out to be quite arbitrary and relative, Gothic could have retained a more farmstead position, while all the others considered the increased demographics. And related to this, coming back to the very beginning, what does say that þurpą has to be smaller than haimaz, which is primarily reconstructed as home? Maybe you can quote the relevant passages from the Codex Argenteus and Abrogans.

@Genava55 But Euler’s work comes after Ringe’s. Downgrading happens particularly with things perceived as "less than": ‘villain‘ comes from Latin ‘villanus‘, meaning ‘villager‘ (which would give credence to ‘village‘ being downgraded to ‘farmstead‘ and not the other way around), or ‘sinister‘, coming from Latin ‘left-handed‘, obvious for Italian speakers. So, what happened? A downgrading for Gothic or a collection for West and North Germanic languages? Even if Gothic was the more conservative, a consensus considering many other variables was reached and written down in Proto-Germanic dictionaries. If we grab a time machine we might find out that this was the wrong call, but it wouldn’t be an error on our part, but the academic consensus, and what can one really do about that but guess.

@Genava55 those books are over 20 years old (almost 40 in the first case), studies now support the notion that they were mutually unintelligible, for example linguist Wolfram Euler’s (curiously, yes, related to the famous mathematician) Das Westgermanische summary, https://www.verlag-inspiration.de/euler-das-westgermanische, already states “by the time of the Gothic translation of the Bible, Western and Eastern Germanic were already so dissimilar that Gothic and for example Frankish people could not have held a fluent conversation”. He got his PhD in 1979 but maybe you can send him an email stating that his work is BS anyway :P. Philologist Friedrich Maurer states that Old English, Old Dutch, Old Saxon, Old Frisian and Old High German (which has the word dorf) were already quite different early on, instead of just branching off from a common Proto-West Germanic, and for all those languages the meaning is village. In any case, even if we are dealing with scholars not agreeing among themselves (and we can cite their books ad infinitum), the elephant in the room is still the same: why not to think that Gothic was the one that suffered the semantic shift? And, more importantly, why has þurpą been reconstructed in PGmc dictionaries as village? (https://kaikki.org/dictionary/Proto-Germanic/meaning/%C3%BE/%C3%BEu/%C3%BEurp%C4%85.html).