* Move lastAugReset and lastNodeReset back to main API under getResetInfo, also included currentNode.
* The associated properties are deprecated on getPlayer()
* parseCommands no longer removes excess whitespace (it was unneeded and also had a bug in it relating to commands that ended in a quote mark)
* more documentation and some variable renaming
* Fix script editor focus bug on navigating to the editor from sidebar
* Fix initialization for lastNodeReset and lastAugReset
* Added new types for various file paths, all in the Paths folder.
* TypeSafety and other helper functions related to these types
* Added basic globbing support with * and ?. Currently only implemented for Script/Text, on nano and download terminal commands
* Enforcing the new types throughout the codebase, plus whatever rewrites happened along the way
* Server.textFiles is now a map
* TextFile no longer uses a fn property, now it is filename
* Added a shared ContentFile interface for shared functionality between TextFile and Script.
* related to ContentFile change above, the player is now allowed to move a text file to a script file and vice versa.
* File paths no longer conditionally start with slashes, and all directory names other than root have ending slashes. The player is still able to provide paths starting with / but this now indicates that the player is specifying an absolute path instead of one relative to root.
* Singularized the MessageFilename and LiteratureName enums
* Because they now only accept correct types, server.writeToXFile functions now always succeed (the only reasons they could fail before were invalid filepath).
* Fix several issues with tab completion, which included pretty much a complete rewrite
* Changed the autocomplete display options so there's less chance it clips outside the display area.
* Turned CompletedProgramName into an enum.
* Got rid of programsMetadata, and programs and DarkWebItems are now initialized immediately instead of relying on initializers called from the engine.
* For any executable (program, cct, or script file) pathing can be used directly to execute without using the run command (previously the command had to start with ./ and it wasn't actually using pathing).
* Update doc error, regenerate docs
* Update a bad error message in renamePurchasedServer@serverPurchases.ts
* restore true/false return (removed in #450 due to miscommunication)
This option adds additional time to the sleep in hack/grow/weaken before
the command takes effect. The critical difference between this and doing
your own sleep is that it creates a single, uninterruptible sleep: This
opens up multiple new avenues of gameplay for batching.
Note that use of this new feature is theoretically always suboptimal,
since extending the sleep time wastes RAM.
* Added new positive integer ns validation helper
* `run`, `exec`, and `spawn` verify threads as a positive integer.
* `run` terminal command also fails if the provided threadcount is not a positive integer.
* Removed some references to .script files in various documentation, and removed some of the NS1 example blocks
This is analagous to tprintRaw (enabled by ns.iKnowWhatImDoing()), but
for logs instead of the terminal. This provides a supported* method of
creating complicated UIs for scripts.
*No actual support, expressed or implied, is provided for use of this
function.
* handle enums differently
* Enums are frozen and fed directly to the proxy
* Enums are not included in the NSFull definition, allowing samekeys for RamCostTree<API>, InternalAPI<API>, and ExternalAPI<API>
* Rewrote a lot of the ramcalc test, with better typing thanks to the samekeys above
* Fix ns1 for proxy (args, pid, and enums after above changes were not being added to ns1 scripts.)
* Fixed an overview issue where the bars could display inaccurately.
Update changelog and bump version to 2.2.1
* ns API is wrapped once
* when a new workerscript is created, each layer of ns is stamped with a private workerscript field that allows the functions to work.
* Test has been refactored to account for new method of wrapping
* BREAKING: ns functions need access to `this` value of their parent ns layer (or any ns layer)
* Enums are passed directly to player (no cloning) but are frozen.
* Add specific types for sleeve tasks
* Moved LocationName enum to overall enums file
* Add LocationName enum to those provided to player
* remove trailing s on CompanyPosNames enum (now CompanyPosName, in line with LocationName)
* Also regenerated all ns documentation. This leads to an expected error when checking for generated files, because PRs should not normally touch markdown files. Also leads to a lint failure in generated file bitburner.d.ts
* also removed some exporting from NetscriptDefinitions.d.ts for anything that wasn't imported somewhere else.
* Some game enums moved to utils/enums. Others can eventually be moved there as well.
* findEnumMember function for performing fuzzy matching of player input with enum members, without needing separate fuzzy functions for every enum.
* Also used findEnumMember for safely loading save games (allows case changes in enum values)
* Changed capitalization on some enums.
* BREAKING: removed classGains work formulas function
* Split ClassType enum into UniversityClassType and GymType.
* Added universityGains and gymGains work formulas functions
* Provided the new split enums to the player on ns.enums.
* Add support for enums at other ns layers
* APIWrapper: simplified wrapping algorithm and modified to just use cloneDeep to copy enums instead of recursively wrapping enums as if they were new API layers
* Improve APIWrapper typing
* Changed some typings at RamCostGenerator to allow for enums at different levels without enums needing a ram cost
* Added enums to ns.corporation, removed getter functions that were being used instead.
* Add FactionWorkType for player use
* Add ClassType and CompanyWorkPos enums
* Change netscriptDefinitions to expect members of these new enums where appropriate.
* BREAKING CHANGE: Removed getSleeveStats and getSleeveInformation because this info is provided by getSleeve in a more usable form.
* BREAKING CHANGE: Removed tor, inBladeburner, and hasCorporation fields from ns.getPlayer. Functionality still exists via added functions ns.hasTorRouter, ns.corporation.hasCorporation, and ns.bladeburner.inBladeburner.
* Separated ns definitions for Person, Sleeve, and Player interfaces with both Player and Sleeve just extending Person.
Added getSleeve, which provides a Sleeve object similar to getPlayer.
* Renamed the sleeve ns layer's interface as sleeve lowercase because of name conflict. todo: May move all the ns layers interface names to lowercase for consistency
* Added ns.formulas.work.crimeSuccessChance and reworked to allow both sleeve and player calculations.
* Removed internal Person.getIntelligenceBonus function which was just a wrapper for calculateIntelligenceBonus. Any use of the former in formulas creates a conflict where ns-provided Person objects throw an error.
* Renamed helpers.player to helpers.person for netscript person validation. Reduced number of fields validated due to Person being a smaller interface.
* Fixed bug in bladeburner where Player multipliers and int were being used no matter which person was performing the task
* Fixed leak of Player.jobs at ns.getPlayer
* Person / Player / Sleeve classes now implement the netscript equivalent interfaces. Netscript helper for person no longer asserts that it's a real Person class member, only that it's a Person interface. Functions that use netscript persons have been changed to expect just a person interface to prevent needing this incorrect type assertion.
Snarling
David Walker
G4mingJon4s
Zelow79
lucebac
T.J. Eckman
Mughur