changes.md

August 20, 2026 · View on GitHub

7.0.0

Breaking changes

  • Coverage and tracker history do not survive this upgrade. Stored documents now carry the metadata the new merge protocol needs, and keys are scoped by a generation token, so the storage format version changed for both RedisStore and the cache adapters. Old keys are ignored rather than migrated, and coverage accumulates fresh from the upgrade. Delete the old coverband_3_2.* keys once you are happy with the new data.
  • Removed configuration settings that no longer had production consumers: s3_region, s3_bucket, s3_access_key_id, s3_secret_access_key, track_gems, gem_details, and simulate_oneshot_lines_coverage. The S3 and gem-tracking settings were already deprecated no-ops.
  • Coverband::Utils::SourceFile and Coverband::Utils::Result now accept only enriched file records from stored Coverband reports. Passing the legacy bare line-coverage Array is no longer supported.
  • Coverband::Adapters::MemcachedStore is deprecated and is now a thin subclass of the new ActiveSupportCacheStore. Its memcached_namespace option and reader keep working; its stored data is not migrated.
  • background_reporting_sleep_seconds now defaults to 600 for the new cache adapter (HashRedisStore still defaults to 300, everything else to 60), because each report rewrites a whole document.
  • Adapters::Base#size now returns bytes as an Integer or nil; size_in_mib owns the "N/A" presentation. A store returning the string "N/A" previously reported 0.00 MiB.
  • Adapters::Base#get_coverage_report now reports the receiver's coverage instead of routing through Coverband.configuration.store, which quietly returned another store's data when the receiver was not the configured one.
  • AbstractTracker#redis_store is deprecated in favor of #storage, and trackers no longer reach through the store to raw Redis commands.
  • Removed Adapters::Base#raw_store and the Redis adapters' raw-client escape hatch. Redis maintenance now uses the narrow store.cleanup capability (clear_legacy! and clear_orphans!); adapters that cannot enumerate keys return nil.
  • Work is retained when generation resolution fails, so a backend blip during a reporting cycle no longer discards that cycle's coverage. Previously only deltas that had reached the pending queue were protected, and the generation is resolved before the payload is enqueued. The retention caps apply to that held work too, so a sustained outage cannot grow the queue without limit
  • A cache store that is not resolvable yet (ActiveSupportCacheStore.new { Rails.cache } running before Rails assigns Rails.cache) is no longer a NoMethodError on a nil target. It holds the cycle's work, logs, and resolves on a later cycle. A resolver returning something that is not a cache store now names the problem instead of failing deep in the storage layer
  • Rake tasks that reach the store (coverband:clear, clear_coverage, clear_tracker, clear_legacy, clear_orphans, coverage, coverage_json, coverage_html) load the Rails environment first. With a lazily configured store they previously failed with an unavailable cache unless rake environment was named explicitly
  • Coverage data loss (eviction, dropped pending deltas) is now surfaced on the web report index. It was logged and shown on the tracker tabs, but the coverage index reported partial numbers as though they were complete
  • Adapters::SessionCoverage#save_report returns the protocol state from Session#record rather than nil, so a caller can tell a write that landed from one that was deferred or refused
  • Coverage now absorbs the two-cycle outage the design commits to. The cap bounds the queue, and the cycle that recovers takes a slot in it, so a cap of 2 only ever absorbed one; it is now 3
  • Data loss from dropping a delta that was already written and merely awaiting confirmation is reported as unconfirmed_dropped rather than pending_dropped, and the web report says the numbers may be undercounted instead of claiming earlier results are unavailable. Only the second kind is lost coverage
  • Trackers keep their own keys whenever a report is not stored (:failed, :unavailable, or an error), so an unreachable backend is no longer lossier for them than one that raises. TrackerStorage::Base#retains_pending? is removed; the decision comes from what record returns
  • A write refused or errored after the delta was enqueued now reports :retained rather than :failed, since the work is in the storage queue and a caller holding a second copy would replay it. This is the documented memcached >1MB path, where it double counted QueryBurstTracker
  • QueryBurstTracker reports on a quiet cycle when storage is holding work it could not write, instead of waiting for the next burst to flush it
  • A document that cannot be written is now reported while it is happening, rather than an hour later. Its retention caps never fire on a quiet tracker -- nothing new arrives for them to drop -- so only the absolute age cap eventually turned a permanently refused write (a tracker document past memcached's 1MB limit, for instance) into a loss event, and until then the tab read "0 used" exactly like an app that used none. The web report now says how many reports could not be stored and since when
  • Presence trackers (views, routes, translations) re-supply their known keys when storage reports work dropped at its retention caps, so a backend outage longer than the queue no longer leaves a used view reading as unused. Their merge is idempotent, so this cannot double count; QueryBurstTracker sums, so its loss stays irreducible and reported
  • save_report no longer mutates the report it is given. The merge used to sum stored counts back into the caller's line arrays, so a caller that reused its report object re-submitted counts that were already recorded.

Features

  • Replaced the generation lifecycle mixin and its implicit instance-variable/hook contract with a state-owning coordinator and typed generation-change events (#656)
  • Added structured storage health through Coverband.storage_health, normal JSON reports, and the MCP get_storage_health tool, so monitoring can distinguish healthy, stalled, data-loss, and unsupported diagnostic states (#658)
  • Added Coverband::Adapters::ActiveSupportCacheStore, which stores coverage in any ActiveSupport::Cache::Store — Redis, Memcached, files, or Solid Cache for Postgres/MySQL/SQLite. The cache target can be passed lazily (ActiveSupportCacheStore.new { Rails.cache }) so it works from config/coverband.rb, where Rails.cache does not exist yet (#533)
  • Trackers (views, routes, translations, query bursts) now work on non-Redis stores. They previously reached through the store to raw Redis commands, so file and memcached users got trackers that silently collected nothing and empty web UI tabs
  • Fixed the read-modify-write conflict Coverband has always had on RedisStore: concurrent reports could silently drop one process's contribution. Reports are now applied under a per-writer sequence, so a conflicting write is detected and repaired on the next cycle without double counting. QueryBurstTracker had the same conflict on its cumulative counters and is fixed the same way
  • Resets are now strong: clear! and reset_recordings retire a whole generation key, so a straggling write from before the reset cannot undo it
  • Coverband now detects when a backend has dropped its data (cache eviction, Rails.cache.clear) and reports it in the web UI rather than silently showing partial numbers
  • Trackers are skipped with an explanatory log line when the configured store cannot support them, instead of collecting nothing
  • Query bursts no longer count Coverband's own storage queries against the surrounding request, which would otherwise inflate query counts on a database backed cache
  • Added persistent_coverage?, supports_trackers?, and supports_paged_reports? capability predicates on adapters, plus file_count and cached_file_count defaults so non-Redis stores no longer raise from the web UI's datatables path
  • redis_ttl applies to stored coverage documents again; generation pointers deliberately never expire, since a pointer that outlived its document would leave live coverage unreachable
  • Storage reads degrade to empty and log when the backend is unavailable (an unreachable cache, or a Solid Cache table that has not been created yet) rather than raising into the request rendering the report; write failures still reach the reporting paths that log them
  • Added rake test:solid_cache, an ENV gated SQLite backed Solid Cache suite covering atomic pointer creation, batched pointer reads, ActiveRecord being unavailable, a missing Solid Cache schema, and Coverband's own SQL not being attributed to the request that triggered the report
  • Added rake coverband:clear_orphans to reclaim generation keys no longer referenced by a pointer, and scoped coverband:clear_legacy to Coverband's own namespaces and tracker names so it cannot delete unrelated application keys
  • CI now runs a memcached service, so the memcached backed adapter is actually exercised
  • Coverage adapters share one implementation of the merge protocol, and the generation lifecycle is shared between the session and the Redis hash repository, rather than each reimplementing the same bookkeeping

Unreleased

  • Feature: Added Coverband::Collectors::TrackerRegistry so applications and gems can register their own trackers, along with Coverband::Configuration.add_tracker_flag for their enablement flags and Coverband.configuration.tracker_for / Coverband.track_key support for trackers without a dedicated accessor; bundled trackers now self-register and are initialized generically at Rails boot (#651)

6.2.2

  • Feature: Added a dead-code review session to the web report — hide files/folders by hovering any path segment, mark files for deletion with an optional comment, and export marked entries to CSV to drive an out-of-band code-removal process; session state lives in browser localStorage; thanks @obie (#647, #649)
  • Feature: Added Coverband::Collectors::QueryBurstTracker to detect N+1/burst SQL patterns per request or job, with configurable query-count and SQL-time thresholds and a new Query Burst Tracker tab in the web UI (#644)
  • Fix: Scoped the ruby-head memory-leak benchmark check to only fail on Coverband's own retained objects, fixing CI false positives caused by stdlib/gem string-deduplication artifacts on newer Ruby (#648)

6.2.1

  • Fix: Search now includes full filenames hidden by truncation in the web UI; users can search for complete file paths even when displaying truncated versions (#569)
  • Feature: Added support for TLS Redis connections via rediss:// URL scheme for use with serverless ElastiCache and other TLS-required Redis services (#580)
  • Fix: MCP HTTP handler now properly uses StreamableHTTPTransport for full HTTP protocol support including GET, DELETE, and OPTIONS methods (#604)
  • Fix: Suppressed Rails logger stdout pollution in stdio MCP mode to ensure clean JSON-RPC stream for Claude Desktop and other AI clients (#625)
  • Fix: Tracker persistence now gracefully handles non-Redis stores (FileStore, NullStore) without raising errors at process exit (#637)
  • Fix: Resolved Rack 1.6 compatibility issue in web reporter by detecting Rack::Files vs Rack::File class (#639)
  • Docs: Added TLS Redis setup documentation to README for serverless ElastiCache configurations

6.2.0

  • Fix: Correct Rack::Static URL handling in the web reporter by using an empty string URL prefix; thanks @alpaca-tc (#635)
  • Fix: Skip missing files during dead method scans to avoid errors in dynamic or removed-file environments; thanks @doug-hall (#632)
  • Fix: Improve MCP dead method formatting for clearer tool output; thanks @doug-hall (#633)
  • Docs: README updates (including master → main wording updates); thanks @jjb (#627)

6.1.8

  • Feature: Added temporal data to GetFileCoverage tool for MCP by @fabienpiette (#626)
  • Optimization: Significant performance improvements to FileList avoiding array allocations and reducing redundant iterations (#619, #620, #622, #623)
  • Optimization: Optimize HTMLFormatter with template caching (#616, #618)
  • Optimization: Optimize memory usage by removing redundant dup calls across Coverband::Utils::Result (#617, #621)
  • Optimization: Optimize array_add logic with in-place modification avoiding overhead in newer Ruby (#615)
  • Optimization: Optimize ViewTracker unused keys and path normalization (#608, #609, #613)
  • Optimization: Optimize AbstractTracker to use bulk Redis hset and direct Hash lookups (#610, #612)
  • Optimization: Cache Regexp in short_name and optimize RelativeFileConverter with Regexp.union (#606, #607)
  • Optimization: Optimize delta collector ignore check order (#611)
  • Refactoring: Extract Line class to Coverband::Utils::SourceFile::Line (#605, #614)
  • Fix: Improve symlink handling in file converters (#603)

Coverband 6.1.7

  • add back a running demo site
  • Feature: Added MCP (Model Context Protocol) server support for AI assistants
    • Allows AI agents (like Claude) to securely query coverage data, find dead code, and analyze usage
    • Includes HTTP and Stdio transports
    • See docs/mcp_security.md and README for setup instructions
  • Feature: Added Coverband::MCP::HttpHandler for mounting MCP server in Rack apps
  • Improvement: Added routes_tracker alias for route_tracker in configuration
  • Improvement: Better error handling when Redis is not available (defaults to NullStore with a warning)
  • Improvement: Updates to lib/coverband/utils/lines_classifier.rb to better handle :nocov: markers
  • Docs: Added agents.md for AI contributor guidelines
  • Docs: Added docs/mcp_security.md
  • Fix: Logging to $stdout instead of STDOUT to avoid warnings
  • Fix: Added GitHub Issue templates and Dependabot config

Coverband 6.1.6

  • Check for Rails::Engine - Stephen Wetzel
  • Support views located in components directory, as well as sub-directories like packs/app/views - Derek Kniffin

Coverband 6.1.5

  • add testing for ruby-head
  • drop support for older Ruby and Rails version as matching our policy
  • remove ostruct as it is no longer in ruby-head
  • fix issue requiring base64 to be added to folks gem files
  • added support for track_key method allowing for code to specifically mark things tracked
  • improvements to CI thanks ydah
  • Update README to include functionality details for "Coverage first seen" and "Last activity recorded" thanks juanarevalo
  • fixes for Rack 3 - thanks Aaron Pfeifer
  • Do not mark single-line and empty methods as dead - fatkodima
  • fix on clear_trackers rake task
  • Consider eager loaded code too when detecting dead methods - fatkodima

Coverband 6.1.4

  • fix on ignoring to many view files thx @fatkodima
  • improve docs on clear

Coverband 6.1.3

  • memory optimizations thx @fatkodima
  • drop support for Ruby less than 2.7 thx @fatkodima
  • improved 0 missed styles thx @fatkodima
  • improved sidekiq swarm support thx @fatkodima
  • default include memcached adapter thx @colemanja91
  • improved long filename web admin views thx @fatkodima
  • reduce logging noise thx @jamiecobbett
  • improved readme thx @hotoolong
  • fix for ignoring views @thijsnado
  • improved documentation @jjb
  • add back static output summary @bessey

other cleanup, small fixes, and updated mostly basic maintenance, by me.

Coverband 6.1.2

  • Fix for paging that would pull empty pages after getting all the data.

Coverband 6.1.1

  • Performance fix making paged report loading 10X faster

Coverband 6.1.0

This release has a number of smaller fixes and improvements. It includes a sizable refactoring around the UI which should simplify improvements going forward. This release is mostly targeting large projects with 6K+ ruby files, use the new config.paged_reporting = true option with the HashRedisStore to enable paged reporting for large projects. The HashRedisStore now also includes the last time a line in a file was executed.

  • Thanks to @FeLvi-zzz for the last time accessed support for the Hash Redis Store
  • Thanks to @alpaca-tc for the improvements on the route tracker
  • Thanks to @ydah for typo fixes, doc updates, adding ruby 3.3 to build matrix, improvements on CI, standardrb fixes
  • Thanks to @trivett, @khaled-badenjki, @IsabelleLePivain for improved docs
  • Thanks to @prastamaha for the memcached adapter
  • Thanks to @ursm for a yaml fix
  • Thanks to @Drowze for a layered cache approach for perf improvements
  • Thanks to @vs37559 for a sinatra pandrino fix
  • This release addresses large projects and adds in paged reporting
    • to ensure even on projects with 10K+ files it can load on heroku under the 30s timeout
    • only supports HashRedis store
    • faster UI for web UI in general which should be noticeable on non paged reports
    • reduce redis calls

Coverband 6.0.2

  • thanks makicamel for improved deferred eager loading
  • thanks Drowze for two performance improvements to reporting coverage

Coverband 6.0.1

Coverband 6.0.0

  • The 6.0.0 release is all that was the 6 different RC releases of 5.2.6
  • Added Rails test matrix to github actions to test on all the supported versions
    • Rails: 6.0.x, 6.1.x, 7.0.x, 7.1.x

Coverband 5.2.6

  • add support for translation keys
  • refactor non Coverage.so based trackers
  • adds CSP report support (thanks @jwg2s)
  • add JSON reporter (thanks @mark-davenport-fountain)
  • README improvements (thanks @jfvanderwalt)
  • Fix on test mode configuration (thanks @rileyanderson)
  • Text Improvements (thanks @etagwerker)
  • readme update on configuration options (thanks @D-system)
  • fix for rack mount of class vs instance (thanks @vs37559)

Coverband 5.2.5

  • (colemanja91/danmayer) experimental support for route tracking, opt-in
  • remove deprecations for old coverband, rails, and ruby versions out of support
    • Note: official support is now:
      • Ruby: 2.7.6, 3.0.4, 3.1.2 (min versions)
      • Rails: 6.0.x, 6.1.x, 7.0.x, 7.1.x

Coverband 5.2.4

  • add install task with example configuration file

Coverband 5.2.3

  • fix for thread error bubbling up
  • additional documentation around Redis

Coverband 5.2.2

  • fix for deprecated redis pipeline usage by @casperisfine

Coverband 5.2.1

Coverband 5.2.0

  • (Fumiaki MATSUSHIMA) Fix CI setting for Ruby 3.0
  • (Fumiaki MATSUSHIMA) Avoid calling coverage with the same type twice on get_coverage_report
  • (kei-s) Fix filling zero for not loaded files
  • (wonda-tea-coffee) cleanup readme / fix typos
  • support deferred eager loading data
  • (Donatas Stundys) Fix disabling views tracking
  • (Simon) Improve View tracker to track render collection too

Coverband 5.1.1

  • Fix early report error
  • Improved view tracker perf Hosam Aly

Coverband 5.1.0

  • add support for dead method detection rake task
  • add support for tracking email view templates/partials

Coverband 5.0.3

  • fix for non standard root paths for view_tracker thx @markshawtoronto
  • support basic auth for Rack prior to 2.0 thx @kadru

Coverband 5.0.2

  • change default port of local server
  • update on readme about issue with scout app, thanks @mrbongiolo
  • fix on configuration page not loading when redis can't load

Coverband 5.0.1

  • fix for race condition on view tracker redis configuration setting ensuring it is set after it is configured... bug only impacted apps that have multiple redis connections for the same system and have Coverband not on the default REDIS_URL

Coverband 5.0.0

  • Full JRuby support
  • Reduced footprint
    • drops S3 support
    • drops static report support
    • drops gem support
    • only loaded web reporter files when required
  • configuration improvements - improved load order allowing more time for ENV vars (better dotenv, figaro, rails secrets support) - all config options can be set via coverband config, not requiring ENV var support - deprecation notices on soon to be removed config options - config exceptions on invalid configuration combinations - additional testing around configurations - improved defaults and reduced configuration options
  • improved resque patching pattern
  • improved default ignores
  • additional adapters
    • supports web-service adapter for http coverage collection
    • support log/file adapter
    • extendable pattern to support any additional backends
  • reduce logs / errors / alerts on bad startup configurations
  • fix: #301 runtime vs eagerload should always remain correct

Coverband 4.2.7

  • Ignore patterns too aggressive #382, thanks @thijsnado
  • Stack level too deep error when running certain activejob jobs #367, thanks @kejordan and @hanslauwers

Coverband 4.2.6

  • Address Redis exists deprecation warning by baffers

Coverband 4.2.5

  • alpha support of jRuby
  • fix for rails 4.0 by rswaminathan
  • do not error on branch coverage / simplecov compatibility by desertcart

Coverband 4.2.4

  • fixes related to startup without Redis, skipping Coverband on common rake tasks (assets:precompile), etc
  • avoid calls to redis if coverage not updated
  • split out lua work and added testing around lua scripts use for the hash redis store
  • improvements to hash redis store resulting in 25% faster perf
  • avoid duplicate hash calculations
  • debug data now includes md5_hashes
  • added support to download coverage and view data in JSON format
  • documentation about working with environment variables
  • add cache wiggle to avoid Redis CPU spikes (cache stampede on Redis server)
  • make the nocov consistent on the data download and html view
  • small performance improvements

Coverband 4.2.3

  • readme fixes
  • improved defaults for the next redis hash adapter
  • add tooltips
  • improved messaging around non-loaded files
  • fix on last updated nil issue
  • view tracker improvements
    • clear all
    • reset individual file
    • timestamps on last seen activity

Coverband 4.2.2

  • new experimental hash redis store for high volume collection (hundreds of clients), thanks @kbaum
  • view_tracker supports tracking view layer files like .html.erb
  • documentation improvements, thanks @brossetti1, @jjb, @kbaum
  • we now have discordapp for discussions, thanks @kbaum, https://discordapp.com/channels/609509533999562753/609509533999562756
  • perf fix on Rails initialization, thanks @skangg
  • simplified logging

Coverband 4.2.1

  • larger changes
    • reduce memory usage
    • fix issue where reports didn't include files with 0 activity
    • updated runtime relevant lines and runtime percentages
    • add Oneshot coverage support for Ruby 2.6.0 thanks @zwalker
    • I would consider this our test oneshot release, please report any issues
    • improved control over memory vs functionality
      • oneshot support is the best memory and speed if you are on Ruby 2.6.*
      • simulated_oneshot works for Ruby prior to 2.6.*, keeps lower runtime memory, at a trade of only being able to detect eager_load or runtime hits not both
      • full hit tracking (The previous and current default), this uses more memory than other options, but provides full robust data
  • small changes fixes
    • further improvements on eager_loading detection, thanks @kbaum
    • fix on ignore configuration options
    • fix broken static server
    • fix issue where clear and coverage trigger coverage reports
    • improved logging
    • fix on gem runtime code coverage support, thanks @kbaum
    • fix sorting of runtime coverage
    • add runtime coverage to gem summary pages
    • documented redis TTL, thanks @jjb
    • resolve regression on coverband rake tasks recording coverage
    • improved runtime / eager loading tracking
    • fix on small memory leaks
    • added warnings that gem tracking isn't production ready
    • built in support for basic auth on the web rack app, thanks @jjb
    • don't clobber rake environment method definition, thanks @shioyama
    • readme improvements, thanks @yuriyalekseyev
    • fix duplicate requires thanks @yuriyalekseyev
    • various cleanups and improvements, thanks @kbaum
    • additional benchmarks to ensure proper memory handling, thanks @kbaum
    • default redis TTL, thanks @jjb

Coverband 4.2.0

NOTE: This release introduces load time and runtime Coverage. To get the full benefit of this feature you need to reset you coverage data (rake coverband:clear or clear via the web UI). Until you clear the combined result is still correct, but your runtime data will be mixed with previous data. This is due to an additive change in how we store coverage. We didn't reset by default as the coverage history for some folks may be more important than runtime breakdown.

  • loadtime vs runtime coverage
    • This fixes the coverage last seen date to exclude just load time data
    • now you can see boot time coverage vs code hit during production execution
    • list view shows runtime percentage and runtime hits
    • file view shows line hits load, runtime, and combined
  • perf speedup on gem details views
  • ajax load code views for more responsive pages on large projects
  • fix for issues with COVERBAND_DISABLE_AUTO_START
  • silence warnings
  • move from thread based reporter instance to process based instance
  • thanks Karl Baum for much of the above mentioned work
  • clear coverage on individual files
  • we have a logo, thanks Dave Woodall, http://davewoodall.com

Coverband 4.1.1

  • fix bug on info page when using namespaces
  • fix bug related to forking Ruby processes (Resque for example)
    • bug caused negative line hits
    • bug seemed to cause memory leaks (not entirely sure how)
  • various test improvements
  • fix bad requests on coverband web admin to /coverage/favicon.png
  • improved documentation

Coverband 4.1.0

  • default disabled web clear, add config option to allow it
  • out of the box support for resque
  • readme improvements
  • fix on regression of merging directory changing deployments
    • fixes for duplicate root paths
  • pilot release of Gems tracking (disabled by default) - todos - support multiple gem paths (various version managers setup multiple gem paths) - speed up page load by allowing multiple pages
  • added web settings and debug views
    • added support for seeing coverage data size consumed in redis
  • support coverage data migrations from 4.0.x to 4.1.0
  • fixes for heroku /tmp asset building

Coverband 4.0.1

  • drops Simplecov runtime dependency
    • still used to measure our own code coverage ;)
  • thanks SimpleCov for all the years of solid HTML reporting, and support!
  • reduced the S3 dependencies to minimal set (not loading all of aws-sdk, only aws-sdk-s3), ensured they are optional
  • Improved Coverband web admin
  • Coverage reports include timestamps of Coverage collection
  • Added Coveralls to the dev process thanks @dondonz
  • now tested under Ruby 2.6.0 thanks @Kbaum
  • improved full stack testing for Rails 5 & 4
  • warning before clear coverage on coverband web

Coverband 4.0.0

  • Add support for Railties integration
  • Reduce configuration options
  • Default to background reporting vs middleware reporting
  • Resolves issue requiring submitting initial coverage data pre-fork
  • Simplified setup with just works sensible defaults for configuration out of the box
  • Fixes on the pilot release of background reporting in 3.0.1
  • Rake tasks automatically configured
  • Updated and simplified documentation
  • Thanks to Kbaum for all the extensive feedback on the PR

Coverband 3.X

Will be a stable and fast release that drops maintenance legacy support in favor of increased performance and maintainability.

  • expects to drop Tracepoint collection engine
  • drop anything below Ruby 2.3
  • release begins to simplify ease of use
    • drop collectors adapter
    • reduced configuration options
  • add memory benchmarks showing memory overhead of coverband
  • use full stack tests to prove no memory leaks when used in Rails

Coverband 3.0.1

Coverband 3.0.0

  • drops Tracepoint
  • drops Ruby <= 2.3.0
  • drops JSON Gem dependency
  • drops various other features not needed without Tracepoint
    • memory cache, sampling, restricted to app folders, etc
  • standardizes on Coverage array format vs sparse hash
  • rewrites store methods, for 60X perf
    • implemented for Redis and File store
  • improved mountable web interface

2.0.3

  • don''t include docs in the gemfile thanks @bquorning
  • pipeline_redis to reduce network overhead thanks @Kallin
  • various additional benchmarks @danmayer
  • Filter out files with no coverage thanks @kbaum

2.0.2

  • fix possible nil error on files that changed since initial recording @viktor-silakov
  • add improve error logging in verbose mode (stacktrace) @viktor-silakov
  • improved logging level support @viktor-silakov
  • launch Coverband demo and integrate into Readme / Documentation
  • fix on baseline to show an issue by @viktor-silakov
  • remove all coverband:baseline related features and documentation
  • dropped Sinatra requirement for web mountable page
  • fix on filestore by @danrabinowitz
  • fixes to the MemoryCacheStore by @kbaum

2.0.1

  • add support for fine grained S3 configuration via Coverband config, thanks @a0s
  • Using the file argument to self.configure in lib/coverband.rb, thanks @ThomasOwens
  • added redis improvements allowing namespace and TTL thx @oded-zahavi
  • fix warnings about duplicate method definition
  • Add support for safe_reload_files based on full file path
  • Add support for Sinatra admin control endpoints
  • improved documentation

2.0.0

Major release with various backwards compatibility breaking changes (generally related to the configuration). The 2.0 lifecycle will act as a mostly easy upgrade that supports past users looking to move to the much faster new Coverage Adapter.

  • Continues to support Ruby 2.0 and up
  • supports multiple collect engines, introducing the concept of multiple collector adapters
  • extends the concepts of multiple storage adapters, enabling additional authors to help support Kafka, graphite, other adapters
  • old require based loading, but working towards deprecating the entire baseline concept
  • Introduces massive performance enhancements by moving to Ruby Coverage based collection
    • Opposed to sampling this is now a reporting frequency, when using Coverage collector
  • Reduced configuration complexity
  • Refactoring the code preparing for more varied storage and reporting options
  • Drop Redis as a gem runtime_dependency

1.5.0

This is a significant release with significant refactoring a stepping stone for a 2.0 release.

  • staging a changes.md document!
  • refactored out full abstraction for stores
  • supports hit counts vs binary covered / not covered for lines
    • this will let you find density of code usage just not if it was used
    • this is a slight performance hit, so you can fall back to the old system if you want redisstore.new(@redis, array: true)
    • this is the primary new feature in 1.5.0
  • Redis has configurable base name, so I can safely change storage formats between releases
  • improved documentation
  • supports SimpleCov.root
  • show files that were never touched
  • apply coverband filters to ignore files in report not just collection
  • improved test coverage
  • improved benchmarks including support for multiple stores ;)

1.3.1

  • This was a small fix release addressing some issues
  • mostly readme updates
  • last release prior to having a changes document!