trollson

"Magic Immunity" seems over-powerful for its level. I'd suggest it should only protect against magic affecting the individual, and not from physical effects of magic which takes place elsewhere. For example, a "fireball" spell would take place at the caster, but what affects the target is a ball of fire, which "magic immunity" doesn't protect against; heat/fire protection is needed instead. On the same principle, "magic immunity" should stop any magic affecting the individual, not just the ones they want. So if you protect yourself with "magic immunity", you cannot receive "heal" or "remote heal" spells.

JPEG vs PNG The (lossy) encoding used by JPEG (discrete cosine transformation) is great for natural images (photo-like). It is very good at encoding gradients and smooth transitions. In these cases it can be preferable to PNG, producing smaller files (depending on your encoding engines). However, if your image contains significant detail in lines or edges (eg, text, diagrams, screenshots), then JPEG with any compress will rapidly make that information illegible. Before PNG we'd use a lossless encoding such as GIF or TIFF for these1. Maps contain significant detail in lines and edges, so PNG is more appropriate than JPEG. If EL's in-game maps are intended to contain details of this type as well, then use PNG rather than JPEG2. Illustration from Wikipedia entry on JPEG: Illustration from Wikipedia entry on PNG:

The bushes are reflections, the trees are casting shadows; two different type of effects. Reflections are very sensitive to the surface-normal, and ripples will break up the pattern as seen in the illustration. Shadows are cast on to the water surface, and are not sensitive to surface-normal. The most you'd expect would be distortion at the edges in the presence of large waves.

Suggested in "Auction Forum (elBay)" back in May 2005, and a few times since then.

This was one of the problems I had in mind when I suggested "Virtual Items" some time ago -- providing some separation between inventory-slots and emu-capacity, so that you don't have to be a hulk to be an effective mage. But, of course, the application to spells was just one possible application for the metaphore.

So for non-PK deaths the stone always gets consumed, or never? Assuming always... This needs a definition of a death due to "PK" which the server is aware of, which may not be as straight forward -- if you lose all but your last health point in PvP, which is then taken by some other means, then is that a PK death? Though there could be interesting strategy in taking that last bit of health from an opponent by other means, just to peeve the stone-carrier. Death on a PK area from *any* means would be feasible and possibly interesting implementation -- mixers scoffing mushrooms, or mining in radioactive PK caves. Still I think this is targeted at a symptom rather than the actual underlying problem.

When I replied to this thread, I assumed that the suggestion was that the Rostogol Stone only *worked* on the fifth death; and that the previous four deaths would yield a death-bag (and the possible drop of the stone). From the subsequent discussion, do I take it that the intention is that a Rostogol Stone *works* for the all five deaths (no death-bag), but is only lost/consumed on the fifth death? In which case how is this different from just increasing the number of Rostogol Stones in the game fivefold (and buying in five-packs etc)? Increasing supply would decrease cost, making them accessible to more characters. Increasing their ability without increasing supply wouldn't make them more accessible. I do think that Rostogol Stones are treated as the scapegoat for PK-woes (along with BoD etc); even if I had a free supply of Rostogol Stones, I'd still find PK unappealing in this game.

...which is why PK seems inaccessible to new players, and I don't see how changing the Rostogol Stone affects this. The combat system needs to provide more opportunity or uncertainty in asymmetric engagements, or somehow separate characters of greatly unbalanced ability (search suggestions for "Law Level" as one possible approach to this). If there was a change to the Rostogol stone, I'd also change the mechanism for determining death bags to accommodate it: 1. Drop 40% (or whatever the percentage currently is) of each stack, not 40% of the stacks. Use a RNG with a binomial distribution to determine the quantity lost in each stack - this gives each item in a stack a 40% chance of being dropped, but at a cost of only ~2 RNG calls per stack. 2. Use the Rostogol stone to reduce the percentage dropped, say 10% rarther than 40%, rather than giving certainty of avoiding it. Similar effect to reducing the certainly of a RS working. 3. Add that say 10% of drops are destroyed (using the same mechanism). This was suggested by someone a while ago, and its rather a nice way to add uncertainly to the situation ("sorry your COL must have been destroyed "). For the OS, is the effect that RS only work: (i) Every fifth death, or (ii) every fifth death while carrying a RS, or (iii) a random 20% chance of working? If not (iii), then it would be too certain (die four times naked, then tool up for PK). If (ii) then can RS be dropped in DBs? Whats to stop players from arranging four "safe" deaths before tooling up for PK?

Something like this would be reasonable, but I think it would be more interesting as a permanent condition confined to within the temples of the gods. Lets use the gods, followers, and temples a bit more. There needs to be a "pull" for followers of the opposing gods to put themelves into danger (remembering that the game is more than just PK). Why would followers of Zarin (crafters) risk entering the domain of Unolas (magic)? Just placing a skill-relevant resource there would not be sufficient, since you could just send someone else to get it. Could it be tied to do with the sponsored skill itself? Increase the god bonus when in the domain of the enemy (taunting the opposition)? However, that would be too contrived really. Why would making medallions under the alter of Unolas be such a big deal? So maybe something else needs to be added to the game, even if its just place-the-funny-nose on the statue Meanwhile, "Godless" characters wouldn't be welcome in the shadow of any deity, and conversely followers of the god could be non-PK'able in their temple (if they were PK'able for other reasons; punishment or guild war).

Half-life Model of Storage If you wanted to represent "decay-in-storage", this could be represented as a half-life. This can be applied as a check at the start of a transaction with storage. The implementation is (briefly) as follow: 1) Save a time-stamp of the last access. 2) On access, compute a decay quantity based on time-stamp and half-life, use a binomial random number generator to smooth things out [1]. This is done before the transaction. 3) Reset time stamp after the transaction. Of course, if the time stamp is very recent the decay step could be skipped. The resolution of the time-stamp is up to the implementation, and could as much as 1 hour resolution in 1 byte (10 RL days) (we have this freedom without making the model clunky thanks to the binomial RNG). To be fair on infrequent players, the time-stamp should be based against playing-time rather than real/game-time. This would add complication, unless decay computation is applied on log-off and time-stamps reset on log-in (meaning time-stamps only need to be part of the runtime structure, not the data file). ...this would also discourage players from sitting AFK for long periods. Time stamps could be held per slot or for the whole inventory -- lower runtime storage on the latter, but you have to run the half-life calculation on *all* slots whenever any slot is accessed. Probably best to take the storage penalty and save runtime. But... One interpretation of the *world-spanning* storage industry is that it is more of a bank than a warehouse; the items you deposit are not the same instances you withdraw. So your special potion has actually been used by somebody else, and you get a different, fresh one back when you withdraw it. Personally, I'd prefer to see high-level tasks which are more than collect-n-mix; such as having to be in the right place at the right time to create something. [1] The half-life and time-period will give you an expected percentage loss (eg, 10%). This should not mean that you lose 10% of the stack, but that each item in the stack has a 10% chance of being lost -- you could lose none or all, if you are very unlucky. While its not practical to check each of 100,000 flowers individually, a binomial RNG will take the probability (0.1) and quantity (100,000), and give you a quantity back, as-if you'd rolled for each item. A good implementation of the binomial RNG will cost ~2 random numbers. This could (should) be used for death bag drops as well.

The current crowns provide extra reserves of health or mana which are useful in a crisis situation, so the time taken to build up to the increased levels (either naturally or through boosters) is worthwhile. Food is not usually consumed in a crisis situation by articifers; it would give you a minute or so grace before you are back to being limited by food-cooldown. On the other hand, having a high food level would be of great benefit to AFK fighters, combined with power saving and regeneration. ...on reason why this wouldn't be introduced? A crown of half-food-cost would be too easy; with other crowns you have to "invest" in wearing them before you get the benefit (build up to the new levels). A half-food crown would be too easy to put on/take off as needed (unless an equipment cooldown is introduced). This seems too different to the other crowns to fit into the same category of item. Actually, search for "Crown of Hunger".

Separate the concept of centre-of-vision from actor-position; the server tells the client that the character is at position (x,y) (cf. teleport), and reports other actors visible about that point, except for the character's avatar which remains the previous location for others to see (and interact with, the character would be rather helpless). Does the client link avatar and pov positions, or does it treat the avatar as just another actor? Still, while it would be an interesting item, does the effort justified the benefit? Footnotes: 1. Characters attacked (locally) while peering through binoculars would be at a great disadvantage of course. 2. Could cast a spell through binoculars on a visible target (are they in some way magical?), though any action should pull the user back to their own position. 3. I don't think that telescopic sights benefit archery. Maybe since EL has very limited ranges it lets you target further, but there should still be a range penalty (longer shots have more arched flight path). Update 2008-08-27 Another use for "remote viewing" is to support peeking through spy-holes or key holes; were the "other side" is some distance away on the map grid (or on another map). This is illustrated in the first panel of the episode of "The Noob" below: The Noob, episode 319 "The Patrol" Similar use would be using devices such as crystal balls, linked to either a fixed, random, or controllable location.

Since the server knows who is a bot, it would be possible to throttle bot messages to public channels. Hold bot messages in a queue, with only one message per bot in the queue at any time (replace on update). Messages are then dequeued at a set rate -- such as 1 per 10 non-bot messages or 1 minute; bot message rate is then independent of the size of the bot population.

Not true. The game just have to give players' choices during their characters development; where you choose to take one path the the exclusion of another.

On changing the damage done by RoP: It doesn't have to be a fixed amount of hit points, it could be defined in other ways. For example, inflicting damage equal to 50% of the targets remaining health. So, you couldn't kill somebody with RoP alone; it would be ineffective against weak targets, but a potent leveller against high-health opponents. This may place the item in a different niche tactically, rather than just being a bigger version of the RoD. Alternatively, there are still other forms of "damage" which are not exploited enough in the game -- damage to attributes, nexus, or skill levels.

I suggested something similar to this, but more flexible, in reply to the thread on "A field for low lvlpkers". It would be simple enough to code as a test applied when a character tries to initiate combat with a target; but would have to be server-side. Update -- Outline and update of the original suggestion in the aforementioned thread: Basically, to initiate combat against a target, the following must be true: your attack is less than the target's defence plus modifier Furthermore, a "dummy" attack roll is then made between you and the target, which must succeed for combat to be initiated. This protects newbies from accidentally initiating combat with stupidly superior opponents. modifier can be positive or negative, and may be determined by a number of components and aspects of the target: law level of the area. Current PK & non-PK areas would have strong values, say -100 and +100 respectively. Otherwise, closer to civilisation, the higher the law level, and the more wild the lower. law level should, for the most, vary gradually across a map to reflect the nature of the areas -- stay on the path and you will be safe! equipment modifiers reflect how "up for a fight" you look. If you wander around the streets in full dragon plate and carrying a huge flaming sword, then you are more of a legitimate target than a monk in a robe. This also adds a role playing incentive to put away your weapons and dress appropriately when in civilisation. outlaw level as mentioned above appears as another character property or attribute, but is not increased (or decreased) by experience or pick points, but by other means (to be determined, but possibly linked to karma). Like an attribute, outlaw level has a base value, and a current (applicable) value which drifts back towards the base at 1 point/minute. outlaw level can be temporarily adjusted by a number of events: Initiating a combat adds something (say +10). Rogue- or thief-like activities, if they become available. ... Note that your current outlaw level shouldn't be detectable by other players. area effects if they become available. For example, if you are in range of a Guard then this effectively increases the law level. There could also be area effect spells which cause calm or rage, and an area undergoing an invasion must have a decreased law level. politics can give you casus belli, which can be guild or deity based; intra-guild relationship could add a modifier, as could ranks in opposing temples. ... Note that all these modifiers apply with respect to the target. This is deliberate; it would be to easy to exploit a system in which the attacker could increase their opportunity to start a fight! Similar mechanisms could be applied to other PvP activities; attacking with summons, hostile magic, and now ranged attacks. I think that the modifiers above are generic enough to still apply to these other forms of interaction, though with different sets of skill & attributes being contested.

Nothing wrong with the idea of a medallion of missile protection; just that the details need to be worked out so as to fit within the game. I seem to recall a stat for missile protection (?), which would be enhanced by the medallion. Then we can work out appropriate negatives to maintain game balance -- ones which can't easily be avoided if you want to use the medallion.

Then consider alternate state-less behaviour. For example: The item explodes on breaking, inflicting damage. To avoid a loophole, this requires equipment to be breakable from a missile strike, as from a hand-to-hand blow. Is this the case? I don't know. For that matter, other sources of physical damage should behave the same (eg, rockfall when mining); then an exploding medallion would count as _another_ trigger for further item breakage!

...or make storage areas no-mix-zones. Far less mass production. Player run "factory zones" using hbags & couriers, guards, cooperation & team work. Far more bitching from players who want to be spoon fed...

Okay, thanks! I was missing the 'skybox' and 'shaders' directories (new since initial checkout - must do 'cvs up -d'). I wasn't proposing fixing the cracks in the geography, just if the "undersky" could be kept black they may not be noticeable. Edit: Except that would bugger up places were the horizon dips into the lower hemisphere, such as where you might have a valley or gorge.

Can you make the lower hemisphere of the skysphere/box black? I've been wandering around with the new camera in Idaloran, with low camera angles to give a near first person perspective. The sky turns red at the horizon, and this red leaks through "cracks" in the ground, it is rather obvious. In this image you can see the "undersky" leaking out of the cracks either side of my character. It was far more pronounced earlier with a brighter red sky, and from other angles during my walk (particularly, the entrance to the mines).

Immediately after the server update, a newly compiled client from cvs, with an updated data directory, worked fine. I didn't have to update the ".elc" directory. Following some other system update that night, EL started to give SEGV at the end of the progress bar. This also happened with the downloaded client, a completely fresh data directory, and a completely fresh ".elc". Now I have re-installed OpenAL, fresh ".elc", rebuild the cvs client. The client will now log in, but gives a FPS of <1. Same with the downloaded client. I see that the "official" ebuild of EL is now on the Gentoo repository, so I'll try that next. ---- Okay, updated from the Gentoo ebuild still gave the <1 FPS, and EL was consuming 100% CPU when on 3D view. Un-selecting "Toggle Frame Buffer Support" and I'm back to my cap of 35FPS, even with all other eye candy active. The client built from CVS is also working with this option disabled. Not sure why this was set, as it is not set in previous el.ini or the one from the distribution. ...hmm Still some odd behaviour (pitch black on one restart, not on others) but not related to OP.

I'm seeing a SEGV on Gentoo/Linux (32bit) in the same place. Odd thing -- I d/l the new client and data when the server was updated, and had a nose around then, it all seemed okay. But, following an overnight update (spot the cause...) it started to SEGV. Both the download and cvs built client. GDB has problems with threads at the moment, so isn't much help. I view this as my problem, since it was working before the update, so I'll try to work out what is going on. Note: Other OpenGL/Cal3D applications are still working. Have to check OpenAL. Nvidia drivers could be a problem (latest != greatest). Note: I did get to the "New Character Creation" 3D interface (didn't go through with character creation though - just confirming that EL could display 3D). ((This may not be soon... shivering, hot forehead... time to go home to bed for a few days...))

All of which you can do with CVS.

How you calculate the "Ranking" for a Guild depends on what you want to be measuring. But I'd suggest that the formula should have a few simple properties: An increase in level within a population should result in a increase in score. Actually, just that one. I was going to suggest that increasing the population shouldn't decrease an existing score, but that depends on what you intend to measure. Certainly, guild-capability increases if you add a horde of newbies, if not your "cutting edge" capability. The current formula fails (1) since members improving from skill level 0 to 1 decrease the score. I was going to suggest a Lehmer Mean with p=2: L = sum ( a[i]^p ) / sum ( a[i]^(p-1) ) L = sum ( a[i]^2 ) / sum ( a[i] ) As this absorbs the level zero characters, gives a bit more weight to cutting edge, and is independant of guild size. However, it doesn't meet (1) either, since a character going from 0 to 1 suddenly decreases the score. Infact, a Lehmer Mean is an even worse model than I thought! Take a guild of all level-N's, add a newbie who starts levelling from zero. The overall guild score decreases for each level the newbie earns upto level N/2. Definitely not following (1). Ah, the joy of statistical modelling. I'll be stuck thinking about this now. The question I have to ask is: So what property is guild rank intended to represent? Update: Rephrased (1).