Well, Banana Lord's suggestion was actually relevant, because mafia had an updated daily build with the change to can_equip 10 hours before your post saying you were going to closet the shirt waiting for mafia to update...
Yay for unintentionally intentional relevance, obfuscational sentences, and creative spelling!
EDIT: My dictionary informs me that the correct term is "obfuscatory". As a grammar nazi I apologize for this transgression
There has to be a way to turn that into a thing. Let's coin that phrase right now.
Option to use Modifier maximizer
The maximizer is already used in the rollover section, and Harvest will save the most optimal outfit for you (using the name you specified for your rollover outfit in the relay script). As for the farming section, the best way to farm is not always to maximise +meat or +items, sometimes combat modifiers are optimal, and sometimes we need special items equipped to farm at all (i.e.: The Sea). In most situations like this Harvest leaves it up to the user to decide on the most optimal course of action. My goal with Harvest was to make it easy for the user to implement a particular farming strategy, not to design one for them. Just out of interest, has anyone tried dj_d's MakeMeatFast.ash? Does it/did it work well?
[Divison by zero problems]
Do you have a blank line at the bottom of the script file?
turns_per_cast($skill[none]) -> 0 is the only situation I can think of that would give that error. If it was a spelling mistake Mafia would throw a "bad skill value" error.
Quick consumption suggestion. In the fortune cookie consumption section:...
Yep, good idea. I wasn't originally going to be fussed about minor optimizations like that, but hey, who am I to deny an inch where I've already given a mile?
If you don't have ANY accordion, your turns per cast would be 0. And you can't divide num_turns by 0. Similar problem exists with any turtle tamer buffs if you don't have a turtle totem.
There's a function to make sure the player has a mace of the tortoise and a rock and roll legend (or better). The theory was that that would never be a problem, but in the current version of the script they're called in the wrong order. Wouldn't a simpler fix be to make sure that
get_buffing_aids() gets called before
request_buff(...) and
cheapest_at_buff()?
Would you mind giving me a quick tutorial on the "?" and the ":" characters? I've never used those before
.
For what it's worth the same is also true for saucespheres if you have no saucepan.
Hmm. My rationale for always getting the accordion and turtle casting items was that it would almost always be necessary for a character, and where it wasn't it would either become so as the character's farming ability was improved, or be beneficial for other reasons. That doesn't apply to the saucepan. Are saucespheres ever optimal/useful? I didn't write Harvest with characters below level 13 in mind, or those with combat difficulties. It might be worth revising that opinion though.
Another suggestion is to do the same with munchies pills -- if you're making 1300 MPA or more (which is slightly more involvement than mere castle farming), then this becomes optimal, as well.
See what I meant about that inch mile thing? Let that optimal crowd have their way once, just once, and you'll never get a moment's piece
. Added to to-do list.
I asked about this a while ago, finally wrote the patch and been using it for a few weeks now.
https://gist.github.com/4012381aa0b3c6b28e42 that adds the idea of a familiar_script for each mode, which if set overrides the normal familiar/familiar equipment logic. I can share my test script which cycles through item-producing familiars if people are interested, but it is far from polished.
I'm still hesitant to add this. In general I believe that "more settings = bad". The only advantage of adding a new setting for each section (bountyhunting, farming, etc.), each of which specifies a separate script, is so that you can cycle through item-producing familiars. Now that might be a worthwhile feature, but in order for it to be worth adding I need to be convinced that the irritation of additional complexity is outweighed by the benefits to
most users. So, convince me
. If I do add it, I don't think I'll do it like you have. I'd make it as invisible as possible by providing a checkbox to toggle automatic familiar switching on or off. Now comes the fun part of design, options and problem solving.
- Harvest could automatically look for item-producing familiars, or the user could specify them in a data file.
- Harvest could decide which familiars were optimal, or simply use all of the familiars from a predefined list that were available to the character.
- The script could decide the optimal number of items to collect from each familiar before switching to another based on the character's most recent MPA stats, or it could let the user specify this number in the afore-posited datafile.
- The script could work out which familiars to use in certain sections (modes), or let the user specify that somehow.
As you can see the main decision to make is to automate or not to automate, and taken together automating all of that would not be trivial. So perhaps I could automate some things and leave others up to the user, but again, this forces the user to work out what's optimal and what isn't, and strange though it sounds, I wonder if most people wouldn't be happier not making this decision at all. The argument that adding a new feature is at worst harmless because users who find it difficult to use can simply ignore it is demonstrably false (look at open source software). I'm interested to know what you all think.
