John D Salt

No, not at all. I distinguished the two things I was talking about (hard-coding and use accessibility), in an attempt to avoid exactly the confusion you seem to have fallen into. I'm not the one who has mixed the two things up. And your evidence is...? I am not talking about table lookups for penetration. You seem to be becoming more confused by the minute. This is what I'm talking about (and not only armour, but all the other attributes of a kind of tank, vehicle, gun, personnel unit). And you're "not sure". So on your own admission you do not really know that these are hard-coded. Thank you. We got there in the end. As for "a string variable for each tank hard-coded wouldn't be that hard to do", it seems that my comments about the principle of locality were not understood. I'm afraid I can't think of a simpler way to explain it right at the moment. I'm mildly surprised that someone with "computer analyst" in his profile seems to be having such a hard time understanding such a trivial point. All the best, John.

Why so? I should have thought diversionary tactics were an entirely reasonable thing to engage in. All the best, John.

Uh huh... thus providing further proof that, where software, computing and consumer electronic are concerned, the exchange rate is roughly £1 = $1. Still, what you lose on the swings you also lose on the roundabouts; the Americans pay a lot less than we do for their food and petrol. All the best, John.

I fully understand the decision to prevent user access to the unit data. That is why I distinguished this aspect from the question of "hard-coding", which is quite separate. Rune seems to have failed to understand this. It is a precept of the most elementary kind of software engineering that you don't do things multiple times in different places if you can do them once in only one place (the principle of locality). If you actually type the values of unit category attributes into the code -- what "hard-coding" means -- you might, it could be argued, only have as much typing to do as if you type the same numbers into data files. Maybe true for the first time. However, as even a first-year (well, okay, maybe second-year) computing undergrad could tell you, this leaves you up the proverbial creek without a paddle when you want to add new attributes; you have to change a dozen, or a hundred, or several hundred statements in your code, instead of merely adding a column of data to your data file (however implemented) and modifying the single "read_AFV_attibs" (or whatever it's called) routine. Now, I don't know exactly how CM does this stuff, but I can't imagine that hard-coding is the way, as apart from anything else it makes life harder for the developer. All the best, John.

I'd certainly like to see such a thing, not because I ever intend to do any mods, but against the day I finally get round to assembling and painting the 30 1/76th T-34s currently lurking in my wardrobe. In particular, I would find it valuable to have a note of the vehicle and time period it was seen, for the historical examples. All the best, John.

Agreed. And let me add to the list of examples the case of a Matila I crew in the Arras counter-attack who, finding that it was disconcerting to have daylight showing through the armour where an AP round had penetrated their tank, bunged up the hole with a pair of socks and carried on. All the best, John.

I'm not sure that a first effort at it will produce the same times getting in and out as for a tank crew who do so every day, and have been trained in bailing out as a drill. Even so, are you honestly claiming that it would take you more than, say, fifteen seconds to climb out of a crew position and out of a hatch? Fifteen seconds is a very long time to sit in a dead tank under fire. I would be interested in finding any hard data on this subject, but it's not something I have ever seen dealt with numerically. Presumably there are "bogey times" to be made when training for bail-out drill somewhere, but I;ve never seen them. The only historical evidence that falls immediately to hand is John Foley's account of bailing out of his Churchill on p.82 of his excellent "Mailed Fist" (Panther, 1957; Mayflower, 1975): "McGinty had just got off the third round when the Tiger gunner recovered from his surprise. I was peering forward through the gloom when suddenly, and without any noise that I can remember, a sharp spike of yellow flame stabbed out of the 88mm gun in front of us. "Sparks flew from the front of Avenger, and she reared back on her hind sprockets, the nose lifting slightly off the ground. "A sudden heat singed the back of my neck and a rapid glance over my shoulder showed flames and smoke pouring from the engine hatches. ""Bale out -- round the back of the tank! I hollered, and snatching off my headset I dived from the turret straight to the ground. "Crosby, Pickford, and McGinty joined me as a second shell crashed through the length of the tank and into the engine compartment." That seem to put the time taken for all surviving members of the crew (Trooper Hunter, the driver, was killed by the first hit) at below the repetition time for an 88mm tank gun. So there's one data point. Any others? All the best, John.

This makes no sense to me at all. Presumably tanks (say) in CM all have the same attributes. It needs less coding to read the values of those attributes from a file than to set those attributes with a spearate statement for each kind of tank. Hard-coding is not a "compromise", it's a complete waste of time; so I find it hard to believe that it's been done. They haven't hard-coded the visual representation of tanks; why should they hard-code the logical representation? All the best, John.

Aha. So under some circumstances the message "bailing out" would really mean "bailed out already but the bad guys haven't noticed yet"? That would certainly stop me worrying about people sitting about trimming their nails in a tank they should be getting out of sharpish. All the best, John.

[Peers over demi-lune glasses and sips port reflectively] Could someone a little more in touch with this new-fangled terminology tell me what, exactly, "Gold" means in this context? Is it, perhaps, a synonym for what people in computing might call "finished"? Oh, and an etymology to accompany the definition would be most welcome. [slouches back on sofa in dingy recesses of 1831 Club] All the best, John.

You baffle me strangely. Why on earth would the unit data be hard-coded? Madness. I refuse to believe it. It may not be accessible to the user, but that's a different thing. I can't believe that a development team talented enough to produce something like Combat Mission would hard-code data like that unless they were crazy on acid. All the best, John.

With the new "death clock", you can see the label "Bailing out" on friendly tanks that have been irremediably pranged, unbeknownst to the enemy. I have a couple of times had this "bailing out" label showing for most of a turn. This seems to me to be rather a long time for the process of bailing out, which I understand is usually done in haste. I can think of absolutely no source of information on the time tank crews typically took to bail out, but the chapter "Brewing up" in Don Featherstone's original "Tank Battles in Miniature" gives a good impression of the process, and suggests to me typical times of the order of a few seconds. What sorts of bail-out times are other people seeing? And do they think them reasonable? I'm wondering if the "long" bail-out times can be explained by trying to help a wounded crew-mate or un-jam a stuck hatch. Alternatively, the "bailing out" label might include some time before the order to bail out is given while the crew are trying to re-start the engine/get the turret moving/fight an intenal fire before the actual order to bail out is given. If any of the design team aren't too busy (HAH!) I'd be interested in hearing the thinking behind the spread of bail-out times that was chosen. All the best, John.

Is this definitely a violation of causality, or is it perhaps the "death clock" manifesting itself? Might you not have killed the Pz IV with a previous shot, but not know it until the next shot was on the way? All the best, John.

"The quality of the workmanship varies considerably. Whilst the highly-stressed parts have a finish comparable with that of British aero-engines of moderate output, the sand castings by contrast are exceptionally rough. In spite of this, however, the latter appear to be sound, there being no sign of porosity or blow-holes on the manufactured surfaces. Most of the important bolts and studs are stress-relieved and ground, and on a few components the standard of finish is very high. The large number of inspection stamps on certain components is very noticeable... The design shows a clear appreciation of the essentials of an effective tank and the requirements of war... When it is considered how recently Russia has become industrialised, and how great a proportion of the industrial regions have been overrun by the enemy... the design and production of such useful tanks in such great numbers stands out as an engineering achievement of the first magnitude." -- from a report on the T-34 by the British School of Tank Technology, quoted in Douglas Orgill's "T-34: Russian Armour", Purnell, 1970. All the best, John.

It's now called Volgograd, but it used to be called Tsaritsyn. I s'pose we all know the old joke about a Soviet citizen being asked questions by a government census-taker: In what city were you born? -- St. Petersburg. In what city were you married? -- Petrograd. In what city do you live now? -- Leningrad. In what city would you like to live? -- St.Petersburg. At least the chap in the joke got his wish in the end. All the best, John.

Yup. My first go as the Sovs (having previously played and won narrowly as the Germans) against the AI went something like this: [brief spolier space] * * * * * * I tried a free deployment for the Sovs. Maxim dets were evenly-spaced along the front, and the three rifle companies concentrated on the left flank in column of blunt wedges ("Left deviationism, a juvenile disorder", but concentrating on one flank worked for Epaminondas). I had the least-experienced company leading, so the new comrades could get the experience of being shot over, and the most experienced bringing up the rear. The tanks I trundled forward into close contact on their own to deal death to the fascists in accordance with the resolutions of the 21st all-Union deputies' congress. The God Of War placed smoke so as to try to isolate the left-flank third of the battlefield from fire from the central and right-flank positions. The smoke did not seem to do me a great deal of good. The leading company went to ground almost immediately, and the second wave made little more progress. Isolated sections filtered forward into cover progressively closer to the enemy, but strength was building only slowly in the forward locations. The whole feel of feeding infantry forward, rallying those gone to ground and feeding them forward again, and slowly accumulating firefighting strength closer to the enemy gave me a very strong impression of Chinese tactics in Korea as described by S L A Marshall. Not, I suspect, an inappropriate similarity. The tanks were the real battle-winners; one was lost to a door-knocker and one to a tank-killing team, but they took protected firepower forward, and as well as applying their own firepower effectively they scared up targets that could be malleted by Maxim and 50mm mortar fire. Even so, what appeared to me as the controlling player to be the most effective way of using my armour could, I can easily see, be interpreted as poor combined-arms co-ordination when seen from the other side. The picture of Soviet tanks operating on their own and remaining halted for long periods called to mind German descriptions of their early-war battlefield behaviour. I had the third company up and human-waving forward after the smoke stopped, an intrepid junior lieutentant and a rifle section had got into the left-flank German position and taken the first prisoner, and forward strength of effective sections was building nicely, although command arrangements were all to cock, platoons from the first two companies were scattered all over the place and people were following the nearest leader as company and battalion commanders moved forward to sweep up the stragglers. Then the Germans surrendered. I don't recall the exact score, but apart from the two tanks lost I had IIRC lost 61 casualties, a good deal less than the defenders' loss. I have to wonder whether at least as convincing a victory could not have been achieved without the participation of the rifle companies at all, except for their mortars (half-a-dozen mortars hiding in the same patch of bush with the battalion commander proved to be an effective means of dealing with individual foxholes as they were unmasked). "L'artillerie detruit, l'infanterie occupe", but with tanks and Maxims fulfilling part of the artillery role. I might give that a go next. All the best, John.

In what way? Israeli/German ubertanks smash invincibly through Egyptian/Russian anti-tank defences in the Sinai/Kursk salient, or Egyptian/Russian anti-tank teams armed with fearsome Saggers/PTRDs give the Israeli/German M-48s/Pz IVs a bloody nose? I'm having a bit of difficulty spotting the parallels here... All the best, John.

A crew who had been ordered to spike their gun would spike it, sure. If they were "running for their lives" I doubt that they'd have time to do so. Within the scope of a single battle, though, there is no reason at all that the detachment shouldn't take cover in their shelter trench when the fire becomes too heavy, and return to the gun when it needs to come into action again. Read an account of the Rifle Brigade's action at Outpost Snipe for a description of this. Likewise, if the original detachment are all shot down at their posts, there is no reason that the gun should not be manned by someone else. I'm afraid we are up against the design limitation of treating the gun-and-detachment as a single entity. I hope this will be changed at some future date; but coding for collaboration of distinct entities in the way that would be required is bound to take a deal of thinking. I imagine that enabling one entity to use another as an instrument in this fashion, or more generally, enabling entities to colaborate in specified ways, might also help with coding multi-turret tanks, dismountable cavalry, Goliaths, and perhaps even enable the AI to handle APCs and assault boats. All the best, John.

Quite right. The first time I saw the exagerrated shock-front effect in CM:BO I was reminded of a clip of bombs being dropped over the Mekong Delta. This suggested a more tropical climate than I am accustomed to in Calvados. I think the CM:BB style is much more satisfactory. Let's save the shockwave effects for the version of CM that happens in hot and sticky places (CM:BBBB -- "Combat Mission: Bataan, Buna, Betio and Burma"). All the best, John.

Ah, splendid, my suspicions confirmed. The issue of true canister in the light ATk calibres was news to me. You evidently have tapped into an excellent source on Soviet artillery ammunition -- could you tell us your sources, please? (Thinks: I hope this doesn't need me to buy a ticket to the St. Petersburg artillery museum.) I'm never sure what the letters in Soviet ammunition designations stand for, but might it be safe to guess that the "SH" in "USH" is for "Shrapnyel'"? All the best, John.

Check you haven't got some annoying thing like a Radeon card with an "out-of-date" (a few months old) driver. I was getting 13 secs waits for the Demo, which reduced to 3 or under with the new drivers. All the best, John.

As long as it's $100 or less I'll consider it a bargain. All the best, John.

I'd say the learning curve was very steep indeed (meaning you learn fast). Both CM:BO and CM:BB are models of what good interface design should be. The greater number of command options will make CM:BB a little bit more work to come to terms with for the complete tyro, if only because some little explanation of things like the difference between "advance", "move" and "assault" will be needed. I think one of the many aspects of CM's excellence is the very useable, quickly-learnt interface, which could serve as a model of good UI design. This hardly ever gets mentioned, of course; but you'd expect that. Good interfaces don;t get in the way, so people don't notice good interfaces. All the best, John.

I'd be very interested to know what evidence there is for Soviet use of canister in WW2. I do not recall ever hearing of such a thing, and it doesn't get a mention in the places I'd expect to see it (TM-30-430, Zaloga & Grandsen's "Soviet tanks and fighting vehicles", Zaloga & Ness' "Red Army Handbook", usque ad nauseam). They did, on the other hand, keep shrapnel in service long after it had become unfashionable with other countries, and that's what I assumed the new mystery nature to be when I first saw it. So, is this "canister" really shrapnel? All the best, John.