Adrian Colyer

@adriancolyer

Author: , Venture Partner with Accel, London after many years as CTO for SpringSource and then apps at VMware, Pivotal

Romsey, England
Vrijeme pridruživanja: ožujak 2009.

Medijski sadržaj

  1. 31. sij

    "Seamless offloading of web app computations from mobile device to edge clouds via HTML5 web worker migration" Jeong et al., Web workers let you offload computation to another thread... and now to the edge too!

  2. 27. stu 2019.

    Part II of our look at Mergeable Replicated Data Types (MRDTs): merge operation derivation, and the Quark implementation:

  3. 27. stu 2019.

    "Mergeable replicated data types" Kaki et al., OOPSLA'19 - Part II. MRDTs brought to life in the Quark implementation...

  4. 25. stu 2019.

    Mergeable Replicated Data Types with merge semantics defined in the relational domain, allowing automated derivation of merge operators given the mapping functions:

  5. 20. stu 2019.

    A fantastic vision (and some great sample apps!) for 'local-first' applications from et al., - I'd love to see more apps built this way... Check out for a demo!

  6. 15. stu 2019.

    Taiji is Facebook's routing infrastructure that maps user requests from the edge to a datacenter. By exploiting connections in the social graph, FB were able to significantly increase caching effects and reduce backend load by up to 17%

  7. 15. stu 2019.

    "Taiji: managing global traffic for large-scale Internet services at the edge" Xu et al., SOSP'19. How Facebook reduced backend query load by 17% with (social) connection-aware routing.

  8. 11. stu 2019.

    Snap is Google's user space networking microkernel, and the Pony Express data plane component eschews TCP/IP for a ground-up rewrite of the transport engine.

  9. 11. stu 2019.

    "Snap: a microkernel approach to host networking" Marty et al., SOSP'19 In which Google reveal they're steadily moving away from TCP/IP in the datacenter...

  10. 1. stu 2019.

    Risk-scoring models so easy anyone could use and understand them are not so easy to create! But RiskSLIM has the answer...

  11. 1. stu 2019.

    "Optimized risk scores" Ustun & Rudin, KDD'17 Creating optimal, human usable, risk-scoring models.

  12. 30. lis 2019.

    How to create the best possible simple, interpretable, models : Try this first before doing anything more complex...

  13. 30. lis 2019.

    "Learning certifiably optimal rule lists for categorical data" Angelino et al., JMLR 2018 Generating the simplest and best model possible...

  14. 25. lis 2019.

    How to match visualisation types to tasks for best performance. Bonus item: is it ok to use a pie-chart?

  15. 25. lis 2019.

    "Task-based effectiveness of basic visualizations" Saket et al., 2019 Choosing the right visualisation for the job...

  16. 23. lis 2019.

    Insights into the world of Exploratory Data Analysis from Alspaugh et al., .

  17. 23. lis 2019.

    "Futzing and moseying: interviews with professional data analysts on exploration practices" Alspaugh et al., VAST'18. What really goes on during Exploratory Data Analysis?

  18. 21. lis 2019.

    Way back when the "Grammar of Graphics" taught us a principled way to think about visualisations. Vega-Lite gives a grammar of _interactive_ graphics (and a pretty neat open source project too!):

  19. 21. lis 2019.

    "Vega-Lite: a grammar of interactive graphics" Satyanarayan et al., IEEE Transactions on Visualization and Computer Graphics, 2016 A powerful platform for declarative specification of _interactive_ visualisations.

  20. 18. lis 2019.

    When Facebook unshackled Hack from PHP it opened the door for 'sweeping language advancements.' One of those it turns out was support for probabilistic programming. Here's a quick look at Facebook's HackPPL extensions:

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·