Systems selections: Difference between revisions
(→Software: expanding on the role of each sofftware) |
(→Software: stating some infra software in use) |
||
| Line 53: | Line 53: | ||
=== Current software === | === Current software === | ||
* System: [[Debian]] [[GNU]]/[[Linux]] | |||
* Hypervisors (by hosting guys) Xen and KVM | |||
* Webservers: Nginx and Apache | |||
* [[MediaWiki]] ([[PHP]], [[MariaDB]] and more) | * [[MediaWiki]] ([[PHP]], [[MariaDB]] and more) | ||
Revision as of 10:32, 1 September 2016
This article is for choosing good systems (technical) for The Consumerium Effort - Enhancing Consumer Informedness
Factors affecting systems selections
- Copyleft 'free to modify' and 'free of cost' software is strongly preferred over other solutions.
- Minimal modifications made (MMM) is an preference as well. This will ensure that we don't get swamped down on patching ad-hoccy, unstandard aged solutions to keep 'em working with the new stuff.
Encodings
- Integers w:BASE10 for integers, size depends on available solutions. Escape integer ceiling by bluntly encoding the integer as string of BASE-10 chars, dirty hack but could do the trick.
- Condense URL encoding w:BASE64 and 64 possible characters, exactly 6 bits from w:ISO/IEC 646 (hard compatible with 7-bit ASCII)
- w:UTF-8 encoded w:Unicode for text. UTF-8 is efficient and allows 2^32 different characters.. unless escaped to 40-bit in future revision
Copyright and copyleft protections system selections
- Copyleft licensing for content, code and Consumerium Commons media repository
- Automatic plagiarism testing which usually probably also translates to copyvio
- Disqualify and rapidly remove all copyvio
- Discourage piratism. Piratism on copyleft much stronger than on copyright.
Languages
Programming languages
Currently used languages
Considered languages / languages of interest
Databases
Current databases
Main article: Databases
Currently using all major free software RDBMS:
Considered databases
- Graph databases and Subject-predicate-object databases seem to perform the same thing just with different architecture and suitability for certain searches. See Talk:Databases for more on this.
Candidates for testing
- All copyleft licensed triplestore and more advanced i.e. graph
Frameworks
Frameworks for glueing into various systems
- Semantic MediaWiki compatibilities: PHP, Semantic Forms, enter data in wiki code -> RDF <-> Graph -> Outputs wikirenders and public SPARQL endpoints
Software
Current software
- MediaWiki (PHP, MariaDB and more)
- w:diaspora (software) (Ruby on rails ,Ruby, MariaDB, Node.JS and more) - role: consumer protection oriented free social media since 2010
- w:Draft:Hubzilla (PHP, MariaDB) - Excellent channel-based distributed free social media where subscription is to the whole network allowing nomadic and double citizenship. A model the others should follow suit.
- w:Friendica (PHP, MariaDB, fully contained in database w/ images stored as blobs. The least learning curve free social media for facebook refugees. Initiated by the same guy as Hubzilla.
- w:GNU social MariaDB - Microblogging platform
- w:GNU MediaGoblin (Python2 (Installing Python3 was impossible last time I looked at it, theoretically compatible with Py3 but broken because outside code supplier.) and PostgreSQL) Role: Copyleft code, content and codecs based media repository.
...
APIs
APIs to communicate with:
- Free social media ones
Data sources
Data sources: own
- Semantic MediaWiki well thought out and set up on implementation wiki (c.consumerium.org)
- Development Wiki - User database, userspace sharing or importing and Unified Login are being looked at as options.
Data sources: Semantic web
- Other copyleft Semantic MediaWikis
- DBpedia Public SPARQL endpoint of DBpedia.org and Projects and efforts using DBpedia are very relevant.
- Wikidata Public SPARQL endpoint of Wikidata.org (Wikidata Query Service)
Lists of Sparql endpoints
Media sources
- w:Wikimedia Commons
- Consumerium Commons
- Upcoming federated w:GNU MediaGoblin