Please enable JavaScript.
Coggle requires JavaScript to display documents.
QCon New York 2018 for Pantheon (Joshua Bloch - Effective Java author.…
QCon New York 2018 for Pantheon
Camunda
Bernd Rucker - co-founder, developer advocate
Topic about some problems in microservices flow and how to avoid
Distributed systems - regular, using compensations, relaxed
Some recommendations for using event based systems
Notes
BPMN colored
BPMN with badges
BPMN with gradient background
Machine Learning
Apache Spark
Can be used for application optimization (
Zoran Sevarac
)
I.e., optimized options for billing app (no refactors/code) have found - as a result 15% faster
DoorDash - cathering software, was trained step-by-step
Simple rules - Learnt coeficients - First ML model - Many ML models - Full fledged ML model (now)
Blockchain
Where it useful, where not
Yes if immutable data, not trusted interaction, other
No if trusted interaction, banks, servers
Problems and solutions
Scalability
Small capacity
Discrimination, wallets ban, trust requirement
My opinion about
Need to introduce in Odyssey
Use for all transactions: sign audits with blockchain
Provide Pantheon's blockchain platform for clients
Customer can set its own blockchain repository
Developers as a malware distribution vehicle
Developers are powerful as never before
XcodeGhost - malware in China, undetected 4 months, affected to 300+ apps in AppStore
As lower level of code as harder to detect
Affected to libraries
Sometimes fixing of bug is parallel to introducing a vulnerability
How to decrease risk
Learn past
Automate security controls
Make as easy to secure as possible
Education
Manage access like tech giants even if you're small
Always control access
Automated test cases
Use open source
QCon summary
900+ attendees
Paypal, Google, Microsoft, Yandex, Amazon, Camunda, a number of other companies
15 themes totally (5 themes a day), in about 150 speakers, 100+ topics
Some sessions were useful some not
Hotel Marriott Marquis
London, San Francisco, Beijing, Shanghai, Sao Paolo, New York, Tokyo
Joshua Bloch - Effective Java author. Session about API evulotion, and session on crucial changes in Java 9
Lambdas
Changing Template Method to Strategy - generally
6 basic standard functional interfaces
43 standard functional interfaces totally
Streams - use generally but avoid in some cases; use parallel streams caution
Regarding to API evolution - now is very dangerous moment: reimplementing of API can not be free by law
photo :smiley:
IT evolution
UI evolving, stages
Architecture evolving
AI is a new UI
API from Google, Facebook, Microsoft, Amazon, IBM provided
Example: shopping - qrcode scanner, camera recognition, automatic charging
Messaging as a new platform
Starting at channels in Facebook, Whatsapp, Telegram
Bots in Telegram, Skype
Full business processes integration in Slack
Extras
IntelliJ Idea
Using automated test cases widely
Developer knows the implementation
Provides simple use cases as is