Notes from my presentation at Word Camp Kansai 2014. Be sure to visit http://codebyjeff.com/blog/2014/03/make-your-own-wordpress-mvc-plugin as well for a tutorial on creating this yourself
The document discusses configuring a minimum WordPress configuration including using Nginx instead of Apache as the web server, adding CDN integration, making the web and database servers redundant, and tuning the MySQL configuration using mysqltuner.pl to check security, storage engines, and performance metrics and recommendations.
Notes from my presentation at Word Camp Kansai 2014. Be sure to visit http://codebyjeff.com/blog/2014/03/make-your-own-wordpress-mvc-plugin as well for a tutorial on creating this yourself
The document discusses configuring a minimum WordPress configuration including using Nginx instead of Apache as the web server, adding CDN integration, making the web and database servers redundant, and tuning the MySQL configuration using mysqltuner.pl to check security, storage engines, and performance metrics and recommendations.
This document proposes using a BehaviorSubject and custom scoped ViewModel to manage progress dialogs in a way that is independent of activity lifecycles and avoids common issues like dialogs dismissing on orientation changes or remaining on screen. The BehaviorSubject would emit true when an API call starts and false when it ends, while the custom scoped ViewModel ensures the loading state can be revived even if the activity is recreated.
GAE can be adapted to work within China's Great Firewall through various techniques:
1. Custom domains can be used instead of *.appspot.com domains to avoid image and mail API restrictions.
2. Edge caching can be implemented for blobstore downloads to improve performance behind the Great Firewall.
3. Alternative services like local account authentication, custom mail servers, and domestic analytics tools can replace restricted Google services.
4. While some Google APIs may time out, others still function normally, so GAE is not impossible to use in China with some workarounds.
My client wanted their apps synced, and I made it with GoToru Furukawa
The document discusses using Go to run simulation apps in parallel while keeping them synchronized. It describes using shared memory and message passing between processes to delegate simulation work to DLLs and coordinate their execution. Key points include spawning goroutines to handle requests concurrently, suspending tasks if not ready, and flushing suspended tasks when conditions are ready to proceed in parallel while staying synchronized. The goal is to leverage Go's concurrency features to efficiently run multiple simulation objects in parallel instead of serially while maintaining sync between processes.
Google Cloud Dataflow can be used to build TensorFlow pipelines. Dataflow allows training multiple TensorFlow models in parallel and writing results to Cloud Datastore. A sample pipeline shows generating training parameters, mapping over them to train models, and writing accuracy results to Cloud Storage. Dataflow provides autoscaling and machine types can be configured. The new DatastoreIO allows reading from and writing to Cloud Datastore from Dataflow pipelines using Protobuf and entity conversion helpers.
Go 1.8 introduced new configuration options for net/http timeouts including ReadHeaderTimeout and IdleTimeout. These help address issues with request timeouts that were not configurable in previous versions. The document discusses various timeout settings for sockets, HTTP servers, and HTTP clients in Go and how the context package can be used to implement request cancellation. It also summarizes that Go 1.8 has made the http.Server more stable in regards to timeouts.
This document summarizes a talk about using context.Context for concurrency in Go programming. It recommends using context.Context to pass cancellation and deadline information between goroutines instead of manually synchronizing access to shared resources with mutexes. Context makes code cleaner by making cancellation explicit and avoiding ambiguous APIs. It provides examples of running multiple goroutines with a common timeout by passing them the same context.
This document introduce the literature 'Connecting Generative Adversarial Networks and Actor-Critic Methods' written by D. Pfau, O. Vinyals. This is used in the event named 'The meeting where we discuss DRL model or else'.
This document proposes using a BehaviorSubject and custom scoped ViewModel to manage progress dialogs in a way that is independent of activity lifecycles and avoids common issues like dialogs dismissing on orientation changes or remaining on screen. The BehaviorSubject would emit true when an API call starts and false when it ends, while the custom scoped ViewModel ensures the loading state can be revived even if the activity is recreated.
GAE can be adapted to work within China's Great Firewall through various techniques:
1. Custom domains can be used instead of *.appspot.com domains to avoid image and mail API restrictions.
2. Edge caching can be implemented for blobstore downloads to improve performance behind the Great Firewall.
3. Alternative services like local account authentication, custom mail servers, and domestic analytics tools can replace restricted Google services.
4. While some Google APIs may time out, others still function normally, so GAE is not impossible to use in China with some workarounds.
My client wanted their apps synced, and I made it with GoToru Furukawa
The document discusses using Go to run simulation apps in parallel while keeping them synchronized. It describes using shared memory and message passing between processes to delegate simulation work to DLLs and coordinate their execution. Key points include spawning goroutines to handle requests concurrently, suspending tasks if not ready, and flushing suspended tasks when conditions are ready to proceed in parallel while staying synchronized. The goal is to leverage Go's concurrency features to efficiently run multiple simulation objects in parallel instead of serially while maintaining sync between processes.
Google Cloud Dataflow can be used to build TensorFlow pipelines. Dataflow allows training multiple TensorFlow models in parallel and writing results to Cloud Datastore. A sample pipeline shows generating training parameters, mapping over them to train models, and writing accuracy results to Cloud Storage. Dataflow provides autoscaling and machine types can be configured. The new DatastoreIO allows reading from and writing to Cloud Datastore from Dataflow pipelines using Protobuf and entity conversion helpers.
Go 1.8 introduced new configuration options for net/http timeouts including ReadHeaderTimeout and IdleTimeout. These help address issues with request timeouts that were not configurable in previous versions. The document discusses various timeout settings for sockets, HTTP servers, and HTTP clients in Go and how the context package can be used to implement request cancellation. It also summarizes that Go 1.8 has made the http.Server more stable in regards to timeouts.
This document summarizes a talk about using context.Context for concurrency in Go programming. It recommends using context.Context to pass cancellation and deadline information between goroutines instead of manually synchronizing access to shared resources with mutexes. Context makes code cleaner by making cancellation explicit and avoiding ambiguous APIs. It provides examples of running multiple goroutines with a common timeout by passing them the same context.
This document introduce the literature 'Connecting Generative Adversarial Networks and Actor-Critic Methods' written by D. Pfau, O. Vinyals. This is used in the event named 'The meeting where we discuss DRL model or else'.