joelpurra (31/41)

Category index page 31 of 41 pages, listing 100 out of 4058 entries in the category joelpurra.

Prevent infinite loops caused by input data being close to chunk size, fallback to merging two and two

More notes

Add private-prefixes (as opposed to public-suffixes), aggregated as a single primary-domain

Rename to public-suffixes

Count TLDs

Count all domains in all files

Use bash execution instead of backticks

Updated URL splitting from jq-hopkok

Accidentally referenced an array by name twice

Stricter string type checking

Change ignored scheme blacklist to whitelist

Map more mime-types

Prefilter data before aggregation, to automate grouping

Add requests status to analysis

Don't limit mime-types

Don't limit the public suffix list

Add isUnsuccessful and isFailed request classifications

Also create inverted domain lists, for domain to group lookups

Add .se health status report 2013

Add a generic, simple reduce/merge deep addition of object with numbers

Expand URLs much more, fix expanded URL references, add isInternalDomain, isExternalDomain, isSuccessful classification

Use correct project name and tagline

Use github repository URL instead of github page URL

Split an URL into components, join them together

Readme, licenses

Initial commit

Rename boolean json name checking function to the opposite

Try to be more strict with type checking checking for failed requests

HTTPS for Google's font api

Flash files, URL components

Write about data analysis, add placeholder headings

Write about software considerations downloading domains

Fix GTM question parallelization problem by splitting up work to preparation and aggregation

Add Google Tag Manager question

Look at second level domains in domain lists

Retrieving websites and resources, spell check

Domain list preparation script for the .se health check domains

Write about .se health status and the public suffix list

Move code headings, data set headings

Write about domains lists, for example

Write more about scripts

Ignore temporary lyx files

Dump lyx pdf generation console output to a log file.

Don't crash on null mime parameter value (charset) strings

Use empty objects instead of ugly null values

Fix hacky malformed mime fix not returning mime parts

Parallelize parts of the data transformations

Try to protect against paths with unusual characters

Create disconnect dataset analysis as part of preparations

Run aggregate preparations in parallel

Explictly arithmetic comparison

Write about tools and code

Notes

Add official GNU parallel citation

Experimental parallelized map/reduce code for the final aggregation steps

Delete duplicated functions

Hack fixes for malformed/empty urls and mime types

Filter HAR data on failure/success of the initial request

Add neighbors to neighbors that were added just before

Add gonifs in two dimensions; fails partially for randomized start positions close to the corners

Draw the neighbor graph with the hexagons

Create a network of gonifs instead of hexagons

Create a new scope, pass fewer arguments recursively

Lazily cache line.center()

Break down hegaxon's getCacheKey into parts

Properties instead of functions, for a probably neglible effect

Comment on hexagons

Use for-loops instead of .forEach to reduce stack clutter while debugging

Better logic for undefined variables and properties

Name some anonymous functions

Use objects instead of arrays for cache based cache

Timeout phantomjs after 90 seconds

Use base64 --decode, which is supported across more platforms

Add check for GNU parallel

Use GNU parallel for improved CPU load handling, output domain as a progress indicator

Add the current domain as a progress indicator

Only look for .har files

Add sleep after opening a webpage, to allow additional async resources to load and render