投稿時間:2022-02-09 08:46:39 RSSフィード2022-02-09 08:00 分まとめ(60件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
IT	気になる、記になる…	【セール】スマホ版｢ドラゴンクエストIV 導かれし者たち｣が約33%オフに（2月13日まで）	https://taisy0.com/2022/02/09/151789.html	天空シリーズ	2022-02-08 22:48:15
IT	気になる、記になる…	CalDigit、Thunderbolt 4を含む18種類のポートを搭載した新型ドック｢CalDigit TS4｣を発表	https://taisy0.com/2022/02/09/151785.html	caldigitts	2022-02-08 22:43:55
IT	気になる、記になる…	スリープ中でもMacのバッテリー残量が減る問題は｢macOS Monterey 12.3｣で修正か	https://taisy0.com/2022/02/09/151783.html	macosmonterey	2022-02-08 22:20:23
IT	気になる、記になる…	Apple、開発者に対し｢iOS 15.4 beta 2｣や｢iPadOS 15.4 beta 2｣などを配信開始	https://taisy0.com/2022/02/09/151781.html	xcodebeta	2022-02-08 22:07:47
IT	気になる、記になる…	Apple、開発者に対し｢macOS 12.3 beta 2｣をリリース	https://taisy0.com/2022/02/09/151779.html	apple	2022-02-08 22:06:59
TECH	Engadget Japanese	日本古来の釣法「テンカラ」を愛する米国人が作った、アウトドアに最適な伸縮式の竿「Tiny Ten」	https://japanese.engadget.com/tiny-ten-225044356.html	日本古来の釣法「テンカラ」を愛する米国人が作った、アウトドアに最適な伸縮式の竿「TinyTen」日本古来の釣法「テンカラ」を、リデザインして新しい釣り竿として、米国から逆輸入渓流に生息するイワナなどを毛バリだけで釣るテンカラが海外でも人気です。	2022-02-08 22:50:44
TECH	Engadget Japanese	アップルオリジナル映画として初、アカデミー作品賞に『CODA あいのうた』がノミネート	https://japanese.engadget.com/apple-nominated-for-its-first-academy-award-for-best-picture-222046202.html	appletv	2022-02-08 22:20:46
IT	ITmedia 総合記事一覧	[ITmedia エグゼクティブ] ヤマハ発動機社長　日高祥博さん　新中計画、成長領域にシフト、世界生産最適化	https://mag.executive.itmedia.co.jp/executive/articles/2202/09/news069.html	itmedia	2022-02-09 07:41:00
IT	ITmedia 総合記事一覧	[ITmedia News] Stripe、Tap to Pay対応開発キットのクローズドβ申込受付を開始	https://www.itmedia.co.jp/news/articles/2202/09/news068.html	iphone	2022-02-09 07:38:00
IT	ITmedia 総合記事一覧	[ITmedia News] 米司法省、2018年にBitfinexから盗まれたビットコイン4160億円相当を押収、2人を逮捕	https://www.itmedia.co.jp/news/articles/2202/09/news066.html	bitfinex	2022-02-09 07:22:00
IT	ITmedia 総合記事一覧	[ITmedia News] Apple、iPhoneが決済端末になる「Tap to Pay」を発表　2022年後半に米国で展開	https://www.itmedia.co.jp/news/articles/2202/09/news067.html	iphone	2022-02-09 07:19:00
TECH	Techable（テッカブル）	もうかさばらない！ 象印、マイボトルを店舗保管・ロッカーでドリンク受け取りの新サービス実証	https://techable.jp/archives/173101	受け取り	2022-02-08 22:00:13
python	Pythonタグが付けられた新着投稿 - Qiita	[AWS] Chaliceのサーバレスアプリにドメイン駆動設計（DDD）を適用してみました	https://qiita.com/hiro-tech1192/items/d0c4f2567fa4e44d7881	リポジトリでも同じようなものを定義していますが、層を跨いで使用するとメンテナンス性が悪くなる気もするのでアプリケーションサービス専用のクラスを用意してみました。	2022-02-09 07:42:28
AWS	AWSタグが付けられた新着投稿 - Qiita	[AWS] Chaliceのサーバレスアプリにドメイン駆動設計（DDD）を適用してみました	https://qiita.com/hiro-tech1192/items/d0c4f2567fa4e44d7881	リポジトリでも同じようなものを定義していますが、層を跨いで使用するとメンテナンス性が悪くなる気もするのでアプリケーションサービス専用のクラスを用意してみました。	2022-02-09 07:42:28
golang	Goタグが付けられた新着投稿 - Qiita	Go言語におけるエラー型の生成方法 | erros.Newとfmt.Errorf	https://qiita.com/hayabusa_3288/items/0b8fab4a508d13543e0b	Go言語におけるエラー型の生成方法errosNewとfmtErrorfGo言語のコードを書いている時に意図的にエラー型を生成したい時がある例えば、osOpenやosCreateの関数などはerrorを返す関数になっています。	2022-02-09 07:10:02
海外TECH	Ars Technica	Kansas lawmakers attack medical board for probing ivermectin cases	https://arstechnica.com/?p=1832608	casesone	2022-02-08 22:37:49
海外TECH	MakeUseOf	How to Import Safari Bookmarks to Brave	https://www.makeuseof.com/brave-import-safari-bookmarks/	browser	2022-02-08 22:02:06
海外TECH	DEV Community	Book notes: Designing Data-Intensive Applications	https://dev.to/danlebrero/book-notes-designing-data-intensive-applications-2bp2	Book notes Designing Data Intensive ApplicationsThese are my notes on Designing Data Intensive Applications by Martin Kleppmann A very data intense book It made me smile that there is one chapter dedicated to the perils of distributed programming when the fact is that the whole book is a warning after another of all the possible things that can go wrong We are doomed Key InsightsReliability Systems should work correctly even in the face of adversity including human error Every legacy system is unpleasant on its own way Data models affect how we think about the problem that we are solving Graph model Good for evolvability ease to add new relations and properties Datalog declarative query language Better for complex data Less convenient for simple one off queries Column oriented storage and bitmap Base encoding increases data size by RPC location transparency there is no point to make a remote service look too much like a local object because it is a fundamentally different thing Two operations are concurrent if neither happens before the other Replication Single leader replication Scalability of read only replicas requires async replication Multi leader replication Multi datacenter offline clients collaborative editing Conflict resolution Leaderless replication Quorum writes and reads High availability low latency occasional stale read ACID Consistency is a property of the application not the database SQL standard definition of isolation levels is flawed In a system with thousands of nodes something is always broken If you send a request to another node and don t receive a response it is impossible to tell why When a timeout occurs you still don t know whether the remote node got your request or not or if is still queued Human error is the major cause of network outages Phi Accrual failure detectorGoogle assumes ms drift for clock synchronized with NTP every secs secs if synchronized once a day Clock reading should return a range of time confidence level instead of point in time Fencing token Monotonically increasing id Server can check if the client still holds a lock lease by remembering the last writer fencing token Linearizability Make a system appear as if there were only one copy of the data and all operations on it are atomic Due to network delays quorums do not guarantee linearizability Linearizability is slow and this is true all the time Consistency and consensus Need to reread this chapter more times Causal consistency is the strongest possible consistency model that does not slow down due to network delays and remains available in the face of network failures Two Phase Commit PC blocks if coordinator crashes XA transactions “Just a C API for interfacing with the PC coordinator In practice making data available quickly even in a quirky difficult to use format is more valuable than trying to decide on the ideal data model up front Messaging systems Key design questions What happens if producer is faster than consumer What happens if nodes crash or temporarily go offline Are messages lost Turning the DB inside out Transactions are not enough TOCFoundations of Data SystemsChapter Reliable Scalable and Maintainable ApplicationsChapter Data Models and Query LanguagesChapter Storage and RetrievalChapter Encoding and EvolutionDistributed DataChapter ReplicationChapter PartitioningChapter TransactionsChapter The Trouble with Distributed SystemsChapter Consistency and ConsensusDerived DataChapter Batch ProcessingChapter Stream ProcessingChapter The Future of Data Systems Foundations of Data Systems Chapter Reliable Scalable and Maintainable ApplicationsReliability Systems should work correctly even in the face of adversity including human error Fault component deviating from its spec Failure System as a whole stops providing service Scalability As a system grows including complexity there should be reasonable ways of dealing with that growth Latency duration that a request is waiting to be handled during which is latent Response time what the client sees Head of line blocking Typically fast request being slow because they are queued due to concurrent request being slow and using all resources This is the reason to measure response time from client side Tail latency amplification in a fan out service response time is the slowest of the called services hence high percentiles become very important Maintainability Over time people should be able to work productively Every legacy system is unpleasant on its own way Operability simplicity evolvability Chapter Data Models and Query LanguagesData models affect how we think about the problem that we are solving Models Relational Document weak join support many to many great hierarchical one to many Graph DocumentRelationalSchema flexibilityBetter joinsBetter performance due to localityBetter many to one and many to many relationshipsSchema on readSchema on writeUpdates require rewrite of whole documentRead always the whole docRelational and document databases are becoming more similar Relational DB support JSON XML RethinkDB support joins Graph model Property graphs Neoj Triple store Datomic RDF equivalent to vertex gt prop gt vertex Query languages Declarative Cypher SPARQL Datalog Rules can be reused and combined Better for complex data Less convenient for simple one off queries Imperative Gremlin Graph processing framework Pregel Good for evolvability ease to add new relations and properties SQL recursive common table expressions WITH RECURSIVE syntax can express graph queries Chapter Storage and RetrievalStorage engines OLTP or OLAP Transactional optimized OLTP Log structured Hash indexes Like HashMaps All keys must fit in memory Bitcask storage in Riak Write only file segments compaction tombstones like Kafka Range queries are not efficient SSTables Stored String Table As Hash indexes but segment files sorted by key Merging segments simple and efficient In memory index sparse Less memory Find value by scanning between two other keys Block read compressed to save IO disk space LSM Tree Log Structured Merge Tree parts Memtable for current segment Read back tree or AVL tree Periodic compaction leveled or size tiered Unordered log for recovery LevelDB RocksDB Cassandra InfluxDB HBase ScyllaDB BigTable Lucene for the term dictionary Bloom filters to speed up reading unknown keys Page oriented B Trees Most common type of index Key value pairs sorted by key Break DB in fixed size blocks or pages k typically Branching factor Depth O log n levels typically Update in place of pages Write ahead log for resilience Multiple optimizations Clustered index Store the value with the key MySQL InnoDB primary key SQL Server can specify one clustered index per table Covering index Store some columns with the index Multidimensional index PostGIS R trees Lucene for similar words Levenshtein automation Similar to trie In memory DBs Non durable Memcached Durable Either append only log or replication or periodic snapshot Relational VoltDB MemSQL Oracle TimesTen Key value RAMCloud Redis Couchbase weak durability due to async writes to disk Faster May support more data structures like sets or queues Anti caching To support bigger than memory datasets Evict to disk based on LRU Future Non volatile memory Analytics optimized OLAP Star schema aka dimensional modeling Fact tables events that reference to dimension tables Dimension tables the who what when how why of the event Snowflake schema when dimensions are broken down in subdimensions More normalized but harder to work with Very wide tables over columns typically Column oriented storage Each column file contains the rows in the same order Less work Better compression Bitmap encoding Indices except for primary require an entire copy of the data LSM tree vs B Trees All the following are typical and depend a lot on the workload LSM trees are faster for write LSM trees compress better B Trees faster for reads B Trees have higher write amplification LSM trees compaction process can cause operational issues Chapter Encoding and EvolutionBase encoding increases data size by Avro More compact than Thrift Protocol buffers but reader needs the writer schema Friendlier to dynamic generated schemas Friendlier to dynamic languages In schema evolution fields are matched by name weaker connascense than position RPC location transparency there is no point to make a remote service look too much like a local object because it is a fundamentally different thing RPC REST assuming servers are updated before clients Backward compatible on request Forward compatible on responses Distributed Data Chapter ReplicationReasons Increase read throughput Increase availability Reduce latency Algorithms for replicating changes Single leader Writes always go to the leader Replication Synchronous Asynchronous Semi synchronous follower sync others async Chain replication Microsoft Azure Storage Failover issues In async replication latest writes maybe lost Further issues if other storage systems have seen the lost write Split brain What is the right timeout before a leader is declared dead Replication implementations Statement based Ship the insert update delete statements Write ahead log shipping Issue Tightly coupled to storage format PostgreSQL Oracle Local row based log Specific log format for replication MySQL binlog Allow easier change data capture Trigger based replication Custom logic flexible Scalability of read only replicas requires async replication Replication lag issues Read your own writes Monotonic reads When second requests goes to a replica with more replication lag than the first request Fix always read from the same replica Consistent prefix reads In partitioned DBs event happens before event but if events are in different partitions a client can see event before event Multi leader replication Use cases Multi datacenter GoldenGate for Oracle BDR for PostgreSQL In general considered dangerous Offline clients CouchDB Collaborative editing Google Docs Conflict resolution Somewhat fix All writes to a key always go to the same datacenter Convergent conflict resolution All replicas arrive to the same final result Last write wins data loss Higher numbered replica wins data loss Concatenate values Preserve all values and let the user resolve Custom conflict resolution logic On write Burcardo On read CouchDB Usually at the row level not the transaction level Automatic conflict resolution CRDTs Riak Mergeable persistent data structures Leaderless replication Dynamo Riak Cassandra Voldemort Quorum writes and reads Read repair anti entropy process Strict quorum write replicas read replicas gt replicas Hard to monitor staleness High availability low latency occasional stale read Sloppy quorums Accept writes in nodes that are not the owners of the key Hinted handoff Issues Sloppy quorums can return old data Concurrent writes Concurrent write and read Partial write failure in quorum Node failure can bring writers down Two operations are concurrent if neither happens before the other Version Vectors To keep track of happens before Version number of key for each replica Client must send the version vector when writing Chapter PartitioningAKA shard in MongoDB ElasticSearch SolrCloud region in HBase Tablet in BigTable vnode in Cassandra Riak vBucket in Couchbase Skewed partitions and hot spots Partition by Key range Hash of key Range queries not efficient MongoDB or not possible Riak CouchBase Voldemort Secondary index Partition by Document aka local index Query requires scatter gather all partitions Tail latency amplification MongoDB Riak Cassandra ElasticSearch SolrCloud VoltDB Partition by Term aka global index Writes require talking with multiple partitions Usually updated asynchronously Rebalancing Fixed number of partitions partitions way bigger than nodes Riak ElasticSearch Couchbase Voldemort Dynamic partitioning Split partition when becomes too big merge when too small HBase MongoDB RethinkDB Proportional to nodes Fixed partitions per node Cassandra Chapter TransactionsACID Atomic All or nothing Consistency Invariants are always true Property of the application not the DB Isolation Pretend that only one transaction is running at a time Serializability Oracle does not implement serializable but snapshot isolation Durability Weak transaction isolation levels Read uncommited Avoid dirty writes mixing writes from several transactions Read commited Avoid dirty reads read uncommited data Snapshot isolation Aka serializable in Oracle repeateable read in MySQL PostgreSQL Avoids nonrepeable reads aka read skew Reading twice in a transaction and getting different results Transaction can only see values commited before it started Multi version concurrency control MVCC SQL standard definition of isolation levels is flawed Write conflicts Dirty writes Read modify write aka lost update Atomic writes set value value Cursor stability Explicit locking Automatic detection by DB Snapshot isolation Compare and set Write skew and phantoms Constraint depends on object A and B and one transaction update A but no B and the other B but not A Phantom a write in one transaction changes the result of a search query in another transaction Fix Serializability Materializing conflicts create a table with rows to be able to lock the rows Serializability implementation options Actual serial execution VoltDB Redis Datomic Through stored procs send action to data Single CPU throughput Scalability through partitioning Two Phase Locking PL PL phase commit Readers block writes and writers block readers Implemented with shared lock exclusive lock Predicate locks Phantoms avoided Queries are stored as predicates and any row changes are matched against them Index range locking Serializable Snapshot Isolation SSI Optimistic keep track of read write rows by a transaction and check no concurrency issues on commit Used also in distributed DB FoundationDB Chapter The Trouble with Distributed SystemsPessimistic and depressing overview of things that may go wrong in distributed systems In a system with thousands of nodes something is always broken If you send a request to another node and don t receive a response it is impossible to tell why When a timeout occurs you still don t know whether the remote node got your request or not or if is still queued Human error is the major cause of network outages Delays and queues everywhere Phi Accrual failure detectorUnreliable networks Telephone networks guarantee a fixed bandwidth for the call hence there is no queueing and a maximum end to end latency TCP network are designed for bursty traffic Variable delays are a consequence of dynamic resource partitioning Unreliable clocks Google assumes ms drift for clock synchronized with NTP every secs secs if synchronized once a day Time of day clock System currentTimeMillis Can go backwards Monotonic clock System nanoTime Always move forward Useless across computers Leap seconds and smearing Precision Time Protocol Clock reading should return a range of time confidence level instead of point in time Google TrueTime API is Spanner returns earliest latest Process pauses Example GC A node in a distributed system must assume that its execution can be paused for a significant length of time at any point even in the middle of a function Treat GC pauses as outages Notify others that a GC is about to happen Restart process when full GC is required A node in the network cannot know anything for sure Fencing token Monotonically increasing id Server can check if the client still holds a lock lease by remembering the last writer fencing token Byzantine fault a node is maliciously behaving Untrusted networks like Bitcoin Chapter Consistency and ConsensusLinearizability Aka atomic consistency strong consistency immediate consistency or external consistency Make a system appear as if there were only one copy of the data and all operations on it are atomic Once a new value has been written or read all subsequent reads see the value that was written Two Phase locking and actual serial execution are typically linearizable SSI is not Usages Distributed locks and leader election Constraints and uniqueness guarantees Cross channel timing dependencies Example store a image in a DFS and then send a message to a queue for the image to be rescaled No linearizability could mean that the rescaling process does not find the image Due to network delays quorums do not guarantee linearizability Linearizability is slow and this is true all the time Ordering helps preserve causality Linearizable systems have total order of operations there are no concurrent operations Causal consistency is the strongest possible consistency model that does not slow down due to network delays and remains available in the face of network failures Lamport timestamps Causal consistency Tuple counter nodeID Timestamp is sent to by clients and server always returns a counter Timestamp ordering is not enough for uniqueness constraints as checks is done after the fact Total order only emerges after collecting all operations When do you know you have collected all operations Total order broadcast aka atomic broadcast Requires Reliable delivery Totally ordered delivery Used by Consensus Zookeeper and etcd DB replication state machine replication Serializable transactions Log Lock services Linearizable compare and set registry and total order broadcast are both equivalent to consensus Two Phase Commit PC blocks if coordinator crashes If coordinator cannot recover manual intervention is required Three Phase commit assumes a network with bounded delays and nodes with bounded response times In MySQL distributed transactions are ten times slower than single node transactions XA transactions Just a C API for interfacing with the PC coordinator PC coordinator usually implemented as a library in the app issuing the transaction Fault tolerant consensus Leader is unique within an epoch On leader dead an election is hold with a higher epoch number Before leader decides anything a quorum of nodes approve the proposed leader This way the leader checks that there has not been an election an it is still the leader Most consensus algorithms assume a fixed number of nodes Uses Linearizable compare and set registers Atomic transaction commit Total order broadcast Locks and leases Membership coordinator service Uniqueness constraints Derived Data Chapter Batch ProcessingThe importance of MapReduce is now declining Map Reduce Putting the computation near the data Mappers send messages to the reducers the key being the address Actor model Joins Reduce side joins Mappers emit records to join with the same key Reducer merges the records Map side joins Faster than reduce side joins but input must oblige some conditions Broadcast hash join When joining small table with big one all mappers read the small table in memory Partitioned hash joins Join input have same number of partitions with same key and same hash function Mapper has all the data Map side merge joins Same as partitioned hash joins but also sorted by same key In practice making data available quickly even in a quirky difficult to use format is more valuable than trying to decide on the ideal data model up front Alternatives to MapReduce Dataflow engines Spark Tez Flink Explicitly model the flow of data between processing steps Advantages Sorting and other expensive operations only when necessary Not unnecessary map tasks More locality optimizations possible as the scheduler knows about all steps Immediate state stored in memory or local disk A processing step can start as soon as some input is available Reuse of JVM Fault tolerance Snapshotting Recompute Spark RDD Pregel processing mode For graph data Iterative processing Calculate one step Check completion condition Yes done No Go back to calculate one step One vertex sends a message to other vertex along the edges in the graph Vertex contains state are fault tolerant and durable Fault tolerance Snapshotting vertex states after iteration Chapter Stream ProcessingMessaging systems Key design questions What happens if producer is faster than consumer Drop messages Buffer max size durable Backpressure What happens if nodes crash or temporarily go offline Are messages lost Types Direct messaging from producers to consumers UDP multicast ZeroMQ UDP messaging StatsD for example WebHooks Message brokers JMS AMQP Load balancing redelivery messages processed out of order Log based message brokers Kafka Kinesis Keeping systems in sync with dual write has issues Race condition if two clients write at the same time and the second writer is faster writing to the seconds system Fault tolerance if second write fails Change Data Capture CDC Starting point is an initial snapshot offset of that snapshot Kafka log compaction make snapshot not required Some CDC tools integrate the snapshot creation Event sourcing Easier to evolve apps Easier to debug Guard against app bugs Dealing with delayed straggler events Ignore them Publish a correction Event timestamp for client events when offline specially Event timestamp using device clock Timestamp when event is sent to server using device clock Timestamp when received by server according to server clock is offset between server and client Apply offset to all events Types of windows Tumbling fixed length no overlap Hopping fixed length fixed overlap Sliding fixed length continuous overlap Session no fixed duration triggered by inactivity Time dependant joins Ordering of events is not deterministic across partitions Fault tolerances Micro batching second batches using processing time Spark Streaming Checkpointing Triggered by barriers in message streams Apache Flink Atomic commit Efficient enough if restricted just to the internal event stream Google Dataflow VoltDB Kafka Idempotence Store message offset in DB Rebuild state Flink gt store in HDFS Samza KafkaStreams gt store in Kafka Or just rebuild from scratch Chapter The Future of Data SystemsAuthor opinions XA has poor fault tolerance and performance characteristics which severely limit its usefulness Better protocol is possible Unification of batch and streaming to simplify lambda architecture Unix esque approach of low level abstractions to the domain of distributed OLTP data storage Dataflow across an entire organization looks like one huge DB Instead of implementing all features in a single integrated DB implement them in different services administered by different teams Two possible routes Federated DBs Unify reads Route of single integrated DB Example PostgreSQL foreign data wrapper Unbundled DB Unify writes Follow Unix philosophy Async event log with idempotent writes is more robust and practical the distributed transactions across heterogeneous systems Desired equivalent of mysql elasticsearch Differential dataflow Dataflow spreadsheet like Dataflow to the web browser or mobile app Desired fault tolerant abstractions that make it easy to provide application specific end to end correctness properties Transactions are not enough Desired more self validating or self auditing systems that continually check their own integrity rather than relying on blind trust on DBs HW or app code Event based systems provide better auditability Maybe certificate transparency or distributed ledgers Ethics	2022-02-08 22:09:04
海外TECH	DEV Community	API trends to discover in 2022	https://dev.to/mehaknarula/api-trends-to-discover-in-2022-23gi	API trends to discover in The world of APIs is evolving at a higher pace and there are some important trends to pay attention to this year Here are the top ten API trends Economy growthThe API economy is not just about selling access to your APIs and data it is far more than this Changing landscape of how enterprises are organizing their teams resources and budgets The entire supply chain being power driven by APIsOrdering food to having it delivered to your doorstepRequesting a cab and tracking it from beginning to endStreaming channels and shows to televisions computers and mobile phones around the worldOrdering groceries and having them delivered from the store to your door APIs are just the enablers and the gateway to the digital or physical products that you are creating They are what enables your business to engage with consumers in the digital world APIs are how your products are viewed and gives great consumer insights into their buying behavior In order to be very competitive in the API market organizations are increasingly adopting “API first strategies They are realizing that their APIs are a product and not just an add on to their existing products Such decisions to put budget and resources behind their API teams will allow them to succeed in API led modernizationAPIs are driving the modernization of businesses across all verticals Mainframe and legacy systems can be a big dependency for many companies We are seeing an increase in the number of businesses using APIs and API platforms to replace or extend the capabilities of their legacy systems Integration experienceA lot can be said about the user developer experience in terms of APIs API documentation developer portals and artifacts The expectations that users and developers have when they go by to integrate with an API product has had a corresponding growth API products and portfolios require to Be intuitiveBe well documentedHave a self service developer portalOffer a great onboarding experience Have consistent error messagingHave great UX and aestheticsProvide a great analyticsInclude code samples and Postman scriptsTo get an edge on your competition it is mandatory to have a great developer and user experience for your APIs and integrations Open API standardsOpen integration and API standards are being adopted at an increasing rate It is recommended to get involved with the groups that support these standards and participate in the process of growing the API community If you are consuming and using these standards make sure to keep up to date with the changes to ensure you are producing and integrating with APIs in the best possible way Take a look below at some of the open standard communities in the past year OpenAPI Specification v GraphQL Specification October edition JSON Schema Specification AsyncAPI Specification v API Best PracticesOften best practices might have conflicting messages or be vague as to lead to branches of differing micro best practices Best practices include many API topics including API designAPI securityAPI integrationsAPI transformationsLogging and tracingCode and design lintingError handlingMonetizationOnboardingObservationAnalyticsDocumentationTesting auditsBulk data handlingEventing ScalabilityAPI and Integration automationThe goal of automation is not to replace workers but to allow workers to focus on more important tasks Keep a look out for an increase in Automation for API design with various no code solutionsAutomation with API IntegrationsAutomation in API creationAutomation for API Management and GovernanceAdvancements in automation for API SecurityAdaptive API managementManaging the entire API lifecycle can be a challenge Every organization is different and often the requirements can be complex You need to adapt quickly to changes in the market in government regulations in technology and even how or where people work Whether you are adapting new API policy requirements API security threats API integration requirements or API consumer or partner opportunities the need for an adaptive API management system is increasing tremendously Hence will see a continued growth of adaptive API management systems that provide a robust and holistic approach to API management Some organizations require on premises solutions others want cloud based SaaS solutions But many will need something in between with a mix of both Seamless integration solutions will see the need for seamless API integrations that scale even further with solid and proven integration tools standing out from the rest because of their smooth and seamless handling of integration processes both on premises and in the cloud This increased demand for secure integrations with other API systems organizations require and need products that can remove any hassles from their integration experience API cybersecurityAPI security and cybersecurity is critical to every organization Organizations are working on their understanding and proper implementation of encryption to protect their data Everything from storage and transmission of data needs a critical security eye when planning and implementing DevOps security Privacy is also directly related as API cybersecurity is essential to securing the privacy of users and employees Not only does data need to be secured but at times the best decision is not to store the data at all There are many government regulations surrounding privacy in certain parts of the world but security and privacy requirements can also have a strategy that reduces unnecessary data to protect both consumers and enterprises API cybersecurity as it is important to understand what is at risk and how to protect and defend your APIs or other systems from various attacks Composable APIsYou will start to see more enterprises that may forego the building part or all of an API and instead composing a new API together from existing to add features instead of internally building new and using existing SaaS services Composable APIs have the potential to reap the benefits from both the flexibility of a custom built API and the seamlessness of a SaaS API subscription or service Consumable APIs are heavily driven by the no code SaaS and on premises API integration solutions that exist today	2022-02-08 22:06:29
海外TECH	DEV Community	What CTOs Say vs. What Their Developers Hear w/ DataStax's Shankar Ramaswamy	https://dev.to/linearb/what-ctos-say-vs-what-their-developers-hear-w-datastaxs-shankar-ramaswamy-31i0	What CTOs Say vs What Their Developers Hear w DataStax x s Shankar RamaswamyAnyone who s been in a rapidly scaling company with an ever expanding engineering team knows that communication is never as simple as it seems That s why we were so excited when Shankar Ramaswamy decided to sit down with Dev Interrupted Shankar is a veteran leader in the engineering space having helped recruit build and guide developer teams at companies like Amazon eBay Paypal and Google where before leaving he managed a team of people Now the Head of Engineering at Datastax Shankar took the time to talk to us about some of the practical lessons he s learned managing teams of several hundred people and why something he calls the “intent perception gap is so tricky for managers to get right We also took the time to talk about Shankar s views on the future of cloud computing what healthy conflict looks like at unicorn companies and why you might want to rethink your single vendor strategy Episode Highlights Include Shankar s career from product management at eBay to eng at PayPal Why Amazon is the only place where junior devs argue about customer experience What is the intent perception gap Future of cloud computing Why single vendor strategies are a bad idea Devs have no formal training to handle people problems You re Invited to INTERACT on April thJoin engineering leaders from Netflix Slack Stack Overflow American Express amp more at LinearB s virtual engineering leadership conference INTERACT on April th day speakers s of engineering leaders all driven by the Dev Interrupted community If you are a team lead engineering manager VP or CTO looking to improve your team this is the conference for you Learn more here	2022-02-08 22:04:27
海外TECH	Engadget	Sony's next PS5 system update will add voice commands	https://www.engadget.com/sony-playstation-4-5-system-update-223001341.html?src=rss	Sony x s next PS system update will add voice commandsThe next PlayStation and PS system update will add a handful of new accessibility and quality of life features to Sony s consoles Among the additions is support for voice commands on PlayStation Sony s previous generation console has had that feature since launching in but it s now making its way to the company s latest console too Starting with a beta Sony will make available to English speaking users in the US and UK first the company is adding a system option that will make the console respond to “Hey PlayStation It s a feature you can turn off but leaving it on will allow you to use your voice to launch games and other apps as well as open the system menu and control media playback The update will bring other new accessibility features including one that makes headphones output mono sound Sony notes that s something that should be particularly helpful to players with unilateral hearing loss As part of the same update Sony is also tweaking how group chats work Moving forward they ll be known as parties and you ll have the option to decide whether they re private or open to the public Should you leave your party open not only can your friends join without an invite but so can their friends as well Another new PS feature will allow you to filter your games by genre as well as keep up to five of them to your console s home screen for quick access Sony has also updated the design of trophy cards and added support for more screen reader languages among other changes You can help Sony beta test the update by signing up to do so on the company s website Look for an email in your inbox on Wednesday to find out if you ve been selected to take part Sony will release both PlayStation and PS system updates later this year	2022-02-08 22:30:01
海外TECH	Network World	Major security vulnerability found in top servers	https://www.networkworld.com/article/3649365/major-security-vulnerability-found-in-top-servers.html#tk.rss_all	Major security vulnerability found in top servers Security firm Binarly has discovered more than vulnerabilities hiding in BIOS UEFI software from a wide range of system vendors including Intel Microsoft Lenovo Dell Fujitsu HP HPE Siemens and Bull Atos Binarly found the issues were associated with the use of InsydeH a framework code used to build motherboard unified extensible firmware interfaces UEFI the interface between a computer s operating system and firmware Get regularly scheduled insights by signing up for Network World newsletters All of the aforementioned vendors used Insyde s firmware SDK for motherboard development It is expected that similar types of vulnerabilities exist in other in house and third party BIOS vendor products as well To read this article in full please click here	2022-02-08 22:20:00
海外科学	NYT > Science	With Mask Restrictions Set to Lift, a Haze of Uncertainty Lingers	https://www.nytimes.com/2022/02/08/health/covid-mask-restrictions.html	mandates	2022-02-08 22:32:59
金融	金融総合：経済レポート一覧	FX Daily（2月7日）～ドル円、115円台前半で推移	http://www3.keizaireport.com/report.php/RID/484073/?rss	fxdaily	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	上昇するプライム市場選択企業の英文開示実施率	http://www3.keizaireport.com/report.php/RID/484074/?rss	大和総研	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	貸出・預金動向（2022年1月）	http://www3.keizaireport.com/report.php/RID/484076/?rss	日本銀行	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	高止まりしそうな米国インフレ圧力、問われる量的緩和の功罪：国際金融トピックスNo.2	http://www3.keizaireport.com/report.php/RID/484077/?rss	国際通貨研究所	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	日本銀行の金融緩和解除で長期金利はどの程度上昇するか～日銀の金融緩和政策による長期金利の下押し効果の測定：基礎研レポート	http://www3.keizaireport.com/report.php/RID/484081/?rss	日本銀行	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	懐かしきイールド・ハンティング 強まる選別の目 消去法の日本株も：Market Flash	http://www3.keizaireport.com/report.php/RID/484085/?rss	marketflash	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	MMT派の信用創造理解：その貢献と限界：マクロ経済・経済政策	http://www3.keizaireport.com/report.php/RID/484099/?rss	信用創造	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	セイファート（東証JASDAQ）～美容室向けに美容師の求人広告、紹介・派遣、美容学校向けに教育事業を展開。Z世代向け求人情報アプリの投入と美容顧客向けプロモーション拡大で成長目指す：アナリストレポート	http://www3.keizaireport.com/report.php/RID/484106/?rss	jasdaq	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	企業年金ガバナンス高度化への取組み～ガバナンスコード対応・企業価値向上のための組織対応	http://www3.keizaireport.com/report.php/RID/484108/?rss	企業価値	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	日銀の出口、ECBは「炭鉱のカナリア」～ECBが出口に動けば日本の出口観測一段と強まる：高田レポート	http://www3.keizaireport.com/report.php/RID/484114/?rss	岡三証券	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	ESG四半期レポート：ESG規制競争始まる：フォーカス	http://www3.keizaireport.com/report.php/RID/484115/?rss	規制	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	Monthly Guide 2022.02 ～世界経済拡大と低金利環境が併存し株式等の成長資産選好続く	http://www3.keizaireport.com/report.php/RID/484118/?rss	monthlyguide	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	【石黒英之のMarket Navi】QT（量的引き締め）と株価の行方を考える	http://www3.keizaireport.com/report.php/RID/484119/?rss	marketnavi	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	グローバルリート市場レポート（2022年2月号）～2022年1月のグローバルリート市場（除く日本）(配当込み)（Ｓ＆Ｐ指数ベース）(前月末比)は6.4％下落	http://www3.keizaireport.com/report.php/RID/484120/?rss	配当	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	保険代理店のInsurTechに関する日本・中国での比較研究を実施（2021年）【概要】～大手の差異は軽微も、「InsurTech企業　兼　保険代理店」において大きな差異	http://www3.keizaireport.com/report.php/RID/484121/?rss	insurtech	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	2021年4-12月期決算の途中経過と株価の反応：市川レポート	http://www3.keizaireport.com/report.php/RID/484130/?rss	三井住友	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	主要な資産の利回り比較（2022年1月）～FRBのタカ派化懸念等から主要3資産の利回りが上昇	http://www3.keizaireport.com/report.php/RID/484131/?rss	三井住友	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	金融政策正常化で問われる暗号資産の投資価値～本格的な投資対象資産として定着するかの試金石に：リサーチ・アイ No.2021-064	http://www3.keizaireport.com/report.php/RID/484133/?rss	日本総合研究所	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	日本市場における株式PTSの現状：大崎貞和のPoint of グローバル金融市場	http://www3.keizaireport.com/report.php/RID/484137/?rss	pointof	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	グローバルREITウィークリー 2022年2月第2週号～先週のグローバルREIT市場は、前週末比では▲0.1％	http://www3.keizaireport.com/report.php/RID/484140/?rss	日興アセットマネジメント	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	【注目検索キーワード】サイバー安全保障	http://search.keizaireport.com/search.php/-/keyword=サイバー安全保障/?rss	安全保障	2022-02-09 00:00:00
金融	金融総合：経済レポート一覧	【お薦め書籍】5秒でチェック、すぐに使える! 2行でわかるサクサク仕事ノート	https://www.amazon.co.jp/exec/obidos/ASIN/4046053631/keizaireport-22/	結集	2022-02-09 00:00:00
金融	ニュース - 保険市場TIMES	損保ジャパンら、レベル4自動運転サービス向けの保険を開発	https://www.hokende.com/news/blog/entry/2022/02/09/080000	損保ジャパンら、レベル自動運転サービス向けの保険を開発月日金発表損害保険ジャパン株式会社以下、損保ジャパンは、株式会社ティアフォー、アイサンテクノロジー株式会社、国立大学法人東京大学大学院情報理工学系研究科の加藤真平准教授の研究室とともに、レベル自動運転サービス向けの「自動運転システム提供者専用保険」を開発したと年月日金に発表した。	2022-02-09 08:00:00
海外ニュース	Japan Times latest articles	Day 4 recap: Yuzuru Hanyu makes shaky start at Beijing Olympics as Eileen Gu wins thrilling gold	https://www.japantimes.co.jp/sports/2022/02/09/olympics/winter-olympics/beijing-olympics-day-4-recap/	Day recap Yuzuru Hanyu makes shaky start at Beijing Olympics as Eileen Gu wins thrilling goldJapan s top skater has a fight on his hands to win a third gold medal after Nathan Chen smashes his rival s world record in the	2022-02-09 07:02:47
ニュース	BBC News - Home	Brit Awards 2022: Ballads over bangers as Adele wins big	https://www.bbc.co.uk/news/entertainment-arts-60312788?at_medium=RSS&at_campaign=KARANGA	arena	2022-02-08 22:45:44
ニュース	BBC News - Home	Rodriguez strike denies Man Utd as bottom club Burnley claim battling draw	https://www.bbc.co.uk/sport/football/60210237?at_medium=RSS&at_campaign=KARANGA	Rodriguez strike denies Man Utd as bottom club Burnley claim battling drawManchester United are left frustrated as Jay Rodriguez s second half goal sees bottom club Burnley claim a battling draw at Turf Moor	2022-02-08 22:20:10
ニュース	BBC News - Home	Newcastle out of drop zone with win over fellow strugglers Everton	https://www.bbc.co.uk/sport/football/60210236?at_medium=RSS&at_campaign=KARANGA	Newcastle out of drop zone with win over fellow strugglers EvertonNewcastle move out of the bottom three with a huge win over fellow strugglers Everton in Frank Lampard s first Premier League game in charge	2022-02-08 22:24:40
ニュース	BBC News - Home	Zouma starts in West Ham win after being condemned for hitting pet cat	https://www.bbc.co.uk/sport/football/60210235?at_medium=RSS&at_campaign=KARANGA	Zouma starts in West Ham win after being condemned for hitting pet catWest Ham reignite their top four push with a narrow victory over Roy Hodgson s Watford as Kurt Zouma controversially starts for the home side	2022-02-08 22:14:05
ビジネス	ダイヤモンド・オンライン - 新着記事	フォードとGMが取り締まり、ディーラーの価格上乗せ販売 - WSJ発	https://diamond.jp/articles/-/295786	販売	2022-02-09 07:23:00
ビジネス	ダイヤモンド・オンライン - 新着記事	現代自動車、インドで不買運動　カシミール巡る販売店の投稿で - WSJ発	https://diamond.jp/articles/-/295787	不買運動	2022-02-09 07:08:00
北海道	北海道新聞	復帰の大迫傑、２８年五輪視野　マラソン前日本記録保持者	https://www.hokkaido-np.co.jp/article/643567/	日本記録	2022-02-09 07:20:11
北海道	北海道新聞	パレスチナ人３人を射殺　西岸でイスラエル治安部隊	https://www.hokkaido-np.co.jp/article/643573/	治安部隊	2022-02-09 07:20:07
北海道	北海道新聞	茨城・袋田の滝が「氷瀑」　完全ならずも凍結９割の日も	https://www.hokkaido-np.co.jp/article/643574/	日本三名瀑	2022-02-09 07:20:02
北海道	北海道新聞	ＮＹ株続伸、３７１ドル高　好決算企業に買い	https://www.hokkaido-np.co.jp/article/643582/	続伸	2022-02-09 07:10:00
北海道	北海道新聞	ＪＲ札幌―新千歳が再開　大幅減便し運行	https://www.hokkaido-np.co.jp/article/643581/	除雪作業	2022-02-09 07:11:07
ビジネス	東洋経済オンライン	地方銀行｢経費率｣ランキングが示す効率性の違い 経費率90％台から40％台の銀行まで差は大きい | 金融業界 | 東洋経済オンライン	https://toyokeizai.net/articles/-/510088?utm_source=rss&utm_medium=http&utm_campaign=link_back	地方銀行	2022-02-09 07:30:00
仮想通貨	BITPRESS（ビットプレス）	[日経] 暗号資産5000億円を洗浄か　米で夫婦逮捕、押収額最高	https://bitpress.jp/count2/3_9_13051	逮捕	2022-02-09 07:47:12