Search results

Oops, I made an assumption that the only intrinsic effects that really matter are the ones granted by equipment - looks like we finally have some exceptions to that.

You mean the "inv" command? There's no regex involved, it just looks for what you typed as a substring of the item name. If you're talking about name matching in general, "fuzzy matching" as it's generally called, that's in ItemFinder.java - but again, there's no regex involved.

The update of cached prices only occurs when you use the CLI or ASH 'buy' command. It can't reliably learn anything from Purchase tab activity, since you aren't necessarily viewing all of the lowest-priced sources of an item - you might be looking at the contents of a particular player's store...

Yes, and in fact that was done with the "Spell Damage" modifier to distinguish it from "Hot Spell Damage", etc., but there would be performance implications for doing this in general. In Java at least, a regexp starting with at least four characters of literal text is vastly more efficient...

The modifier names in the data file generally can't be a suffix of any other name: "Combat Mana Cost: <expr>" would match the regexp that looks for "Mana Cost: <expr>", for example.

Does 37 sound right as the number of logins you managed before running out of memory? There's that many instances of KoLDesktop in your heap dump, each retaining a bunch of GUI-related objects. CLI mode seems like a reasonable thing to try; any actual fix for this problem appears to be deep in...

This is going to depend on the number of items in the list, and which one happens to match; if 'a' is almost always item1, then the first form is likely to win on practical grounds, even though it's theoretically slower (O(n) vs. O(log n)). The second form is a definite win assuming no such...

The floating-point results from monster expressions are ALREADY being explicitly converted to an int; the program wouldn't even compile otherwise. Anyway, adding a conversion couldn't possibly solve the problem - after all, it was only affecting one of the basement monsters, but they've all got...

Should be fixed in r9235, but not tested extensively so there may be side-effects.

If you had somehow managed to miss the current semi-rare (by spending the actual turn in an inappropriate place), that 162 could have been the next semi-rare... but 30 is too early for that (and too late for the current one).

This describes mafia's old behavior (and current behavior when macrofication is skipped for whatever reason); it's impossible to implement in a macro due to lack of a predicate for testing whether you still have two of a combat item available for use.

Current maximum number of ingredients is 7, for the Staff of the Well-Tempered Cauldron. It would be over 100 if we actually represented the Misshapen Animal Skeleton properly. Having a separate field per ingredient is insane, anyway - and you can ALREADY get the ingredients as an array for...

That's already available as string_modifier(whatever, "Modifiers").

You've still got the original version of htmlform.ash, I suspect.

This script has nothing to do with libram summons. All the relevant settings are built-in: enable the desired libram(s) for breakfast summoning, enable summoning during mana burning, set your start-burning/burn-down-to thresholds as desired (for example, I have mine set to 90%/40%, which means...

"use item spooky putty sheet" is not a valid macro; the word 'item' doesn't belong in there. There is absolutely no point in submitting a single action as a macro, anyway - there's no benefit over the existing consult script functions.

How are you even imagining that to be possible, without having to do an individual mall search on each item to be repriced? The ONLY information available to the auto-repricer is the lowest price currently in the mall (taken from the in-game repricing page, which gives no hint as to the...

Preliminary results: hola did it... Your heap dump included 217 MB of data from 12199 cached files with names like "aa/aa-19418.txt". As of r8962, unbounded use of file_to_map() / map_to_file() is basically illegal, since the data file cache that was added back then has absolutely no provision...

Will take a look at it this evening.

Look for the instructions I've previously posted for using the jps and jstack tools to do a heap dump. Do so when you're in an out-of-memory state, and post the dump - I can analyze it to figure out exactly what's taking all the memory.