Ubiquitous Cloud Media · Reaching audiences

Online video and streaming platforms

Online video and streaming platforms make audiovisual media available through networked, software-mediated delivery rather than fixed broadcast schedules, physical carriers or complete local downloads alone. They combine upload or ingest, compression, transcoding, storage, manifests, segmented delivery, content-distribution networks, players, identity, rights management, moderation, advertising, analytics and.

When it emerged
Web video-sharing during the mid-2000s; adaptive and subscription streaming expansion during the 2010s
What changed
Reduces scheduled-channel, physical-copy and specialised-distribution barriers to global audiovisual publication and on-demand playback
Reading time
20 minutes
The essential questions

Online video and streaming platforms, clearly explained

Online video and streaming platforms make audiovisual media available through networked, software-mediated delivery rather than fixed broadcast schedules, physical carriers or complete local downloads alone. They combine upload or ingest, compression, transcoding, storage, manifests, segmented delivery, content-distribution networks, players, identity, rights management, moderation, advertising, analytics and recommendation into one operational system.

What is it?

Online video and streaming platforms are defined here as networked systems that ingest live or recorded audiovisual content, prepare one or more encoded representations, store or relay the media, expose it through software players, and govern discovery, access, monetisation, rights and measurement at scale.

What problem did it solve?

Online video platforms reduce the audiovisual reproduction, scheduling and distribution constraint.

How did it work?

They combine upload or ingest, compression, transcoding, storage, manifests, segmented delivery, content-distribution networks, players, identity, rights management, moderation, advertising, analytics and recommendation into one operational system. Streaming is not a single transport method. A service may use progressive download, adaptive segmented delivery over HTTP, real-time low-latency protocols or live contribution links.

What came before?

It built on Television broadcasting, Motion-picture film and cinema, Fibre-Optic Communication, World Wide Web and Social Networking and Microblogging Platforms.

What did it make possible?

It helped make possible Generative Image, Audio and Video Models and Digital Provenance and Authenticity Systems.

What survived?

Television seasons, channels and scheduled premieres survive inside on-demand catalogues.

Why does it still matter?

Individuals and organisations can publish moving images without owning a broadcast transmitter or physical distribution network. Recipients choose when to start, pause, resume and revisit on-demand media. Players can switch among representations according to network and device conditions.

Deep dive

The deeper story

Online video and streaming platforms make audiovisual media available through networked, software-mediated delivery rather than fixed broadcast schedules, physical carriers or complete local downloads alone. They combine upload or ingest, compression, transcoding, storage, manifests, segmented delivery, content-distribution networks, players, identity, rights management, moderation, advertising, analytics and recommendation into one operational system. [S01-S08]

Streaming is not a single transport method. A service may use progressive download, adaptive segmented delivery over HTTP, real-time low-latency protocols or live contribution links. Modern on-demand platforms commonly encode one source into several representations and let the player choose among them according to measured bandwidth, buffer state, device capability and policy. HTTP Live Streaming describes playlists and media segments for unbounded multimedia delivery, while DASH standardises media presentation descriptions and segment formats for dynamic adaptive streaming. [3][4]

The platform layer matters as much as the codec. YouTube’s 2005 launch lowered the practical barrier to uploading and sharing video through the Web, while subscription services such as Netflix turned catalogues into continuously delivered personalised libraries. [1][2][S05-S07] The resulting system changes who can publish, when audiences can watch, how quality adapts, how rights are enforced and which works receive attention.

The phrase “a video has ten thousand views” conceals several states: the file may have been uploaded, processed, made eligible, recommended, displayed, started, played for a threshold, watched attentively or abandoned while running silently. Online video therefore requires separate measures for technical delivery, playback, attention and effect.

The big idea

Online video platforms turn audiovisual publication into a cloud pipeline that ingests, encodes, stores, selects and adaptively delivers media to software players. Their defining achievement is on-demand audiovisual reach; their defining power lies in controlling both the delivery machinery and the conditions of visibility.

Main problem addressed

Reduces scheduled-channel, physical-copy and specialised-distribution barriers to global audiovisual publication and on-demand playback

Connections

What came before and what followed

Start with the key connections, then reveal the wider network when you need more context.

Enabling connection
Television broadcasting

Establishes audiovisual mass communication, programming and audience measurement.

Enabling connection
World Wide Web

Provides pages, links, browsers and public identifiers.

Related topic
Podcasts and on-demand audio

Shares subscription, hosting and streaming architecture with lower sensory bandwidth. Shares media hosting, playback and recommendation while adding visual bandwidth.

Related topic
Smartphones

Provide ubiquitous cameras, players and mobile networks. Uses smartphone cameras, screens, networks and apps for production and consumption.

Timeline

Key moments

How Online video and streaming platforms emerged

This marks the broad emergence and development of Online video and streaming platforms. Why it mattered: Reduces scheduled-channel, physical-copy and specialised-distribution barriers to global audiovisual publication and on-demand playback.

People and organisations

Who helped shape it?

Apple

Apple is one of the organisations connected to this topic. Open the profile for the wider historical context.

Netflix

Netflix is one of the organisations connected to this topic. Open the profile for the wider historical context.

Research notes

Open the full research notes

These expandable sections preserve the detailed research behind the public explanation.

1. Executive Summary

Online video and streaming platforms make audiovisual media available through networked, software-mediated delivery rather than fixed broadcast schedules, physical carriers or complete local downloads alone. They combine upload or ingest, compression, transcoding, storage, manifests, segmented delivery, content-distribution networks, players, identity, rights management, moderation, advertising, analytics and recommendation into one operational system. [S01-S08]

Streaming is not a single transport method. A service may use progressive download, adaptive segmented delivery over HTTP, real-time low-latency protocols or live contribution links. Modern on-demand platforms commonly encode one source into several representations and let the player choose among them according to measured bandwidth, buffer state, device capability and policy. HTTP Live Streaming describes playlists and media segments for unbounded multimedia delivery, while DASH standardises media presentation descriptions and segment formats for dynamic adaptive streaming. [3][4]

The platform layer matters as much as the codec. YouTube’s 2005 launch lowered the practical barrier to uploading and sharing video through the Web, while subscription services such as Netflix turned catalogues into continuously delivered personalised libraries. [1][2][S05-S07] The resulting system changes who can publish, when audiences can watch, how quality adapts, how rights are enforced and which works receive attention.

The phrase “a video has ten thousand views” conceals several states: the file may have been uploaded, processed, made eligible, recommended, displayed, started, played for a threshold, watched attentively or abandoned while running silently. Online video therefore requires separate measures for technical delivery, playback, attention and effect.

The big idea

Online video platforms turn audiovisual publication into a cloud pipeline that ingests, encodes, stores, selects and adaptively delivers media to software players. Their defining achievement is on-demand audiovisual reach; their defining power lies in controlling both the delivery machinery and the conditions of visibility.

2. Identification

| Field | Value | |---|---| | Public title | Online Video and Streaming Platforms | | Analytical title | Cloud-Based Audiovisual Ingest, Encoding, Selection and Adaptive Network Delivery Platforms | | Recommended type | Audiovisual publishing, streaming-delivery and platform-governance system family | | Primary category | Distribution & amplification | | Secondary categories | Storage; processing; discovery; encoding; governance; monetisation; measurement; interaction | | Emergence | Web video-sharing services during the mid-2000s; adaptive and subscription streaming consolidation during the 2010s |

3. Operational Definition

Online video and streaming platforms are defined here as networked systems that ingest live or recorded audiovisual content, prepare one or more encoded representations, store or relay the media, expose it through software players, and govern discovery, access, monetisation, rights and measurement at scale.

The topic includes user-generated video platforms, subscription video-on-demand services, advertising-supported streaming, live-streaming services, short-video platforms, media players, adaptive bitrate streaming, encoding ladders, manifests, segments, content-delivery networks, origin services, digital rights management, captions, moderation, creator tools, analytics and recommendation surfaces.

It excludes digital video as a representation, broadcast television as a one-to-many scheduled transmission system, video conferencing whose primary purpose is reciprocal communication, physical video carriers, isolated Web pages containing downloadable files, and recommendation algorithms considered separately from the video platform they serve.

4. Why the Topic Matters

1. Audiovisual publication becomes globally accessible

Individuals and organisations can publish moving images without owning a broadcast transmitter or physical distribution network.

2. Viewing separates from schedule

Recipients choose when to start, pause, resume and revisit on-demand media.

3. Quality becomes adaptive

Players can switch among representations according to network and device conditions.

4. Distribution becomes software

Encoding, storage, manifests, delivery, ads and analytics become programmable services.

5. Archives become living catalogues

Old and new works coexist in searchable, recommendable libraries rather than disappearing after transmission.

6. Audience measurement becomes granular

Platforms can observe starts, watch time, pauses, rewinds, abandonment and device conditions.

7. Creators gain direct publication paths

A camera and account can reach a global audience, though platform visibility remains unequal.

8. Platform governance reaches the content itself

Rules, copyright systems, demonetisation, age restrictions and recommendation shape practical circulation.

5. Terminology
  • Streaming: Playback while media data continue to arrive, rather than waiting for the complete file first.
  • Progressive download: Ordinary file transfer arranged so playback can begin before the file finishes downloading.
  • Adaptive bitrate streaming: Delivery in which the player chooses among multiple encoded representations over time.
  • Representation/rendition: Version of media encoded at a particular resolution, bitrate, codec or language.
  • Encoding ladder: Set of available representations prepared for different devices and network conditions.
  • Manifest/playlist: Metadata describing available streams, segments, timing and variants.
  • Segment: Bounded piece of an audiovisual stream retrievable independently.
  • Origin: Service holding or generating authoritative media segments and manifests.
  • Content-delivery network/CDN: Distributed caching and delivery infrastructure placing content nearer recipients.
  • Buffer: Media retained ahead of playback to absorb network variation.
  • Rebuffering/stall: Playback interruption caused by insufficient media in the buffer.
  • Startup delay: Time between requesting playback and visible or audible start.
  • Live latency: Delay between capture at the source and presentation to the viewer.
  • Codec: Method for compressing and reconstructing video or audio.
  • Container: File or stream structure combining encoded media, timing and metadata.
  • Transcoding: Converting a source into other encoded representations.
  • Per-title/per-shot encoding: Allocation of bitrate and representations according to characteristics of the specific title or scene. [7]
  • DRM: Technical and institutional system controlling access to encrypted media.
  • QoE: Quality of Experience, including startup, stalls, quality variation and perceived audiovisual quality. [6]
  • View: Platform-defined measurement event; not a universal measure of complete or attentive watching.
  • Watch time: Aggregate duration of playback attributed to recipients.
  • Live stream: Media captured and delivered while production is ongoing.
  • Video on demand: Stored media selected for playback at a recipient’s chosen time.
  • Short video: Brief, often vertically oriented or feed-distributed audiovisual format.
6. Boundary With Neighbouring Topics

1. Online video versus broadcast television

Broadcast sends a scheduled signal to a coverage area. Online platforms address player requests over packet networks and can serve different streams to different recipients.

2. Streaming versus download

Streaming begins playback before complete transfer and may discard segments after use. Download aims to retain a complete local copy, though the mechanisms can overlap.

3. Streaming platform versus codec

A codec compresses media. The platform manages ingest, storage, delivery, identity, rights and discovery around it.

4. Live streaming versus video conferencing

Live streaming is usually producer-to-many with limited return channels. Conferencing coordinates reciprocal participants with tighter interaction latency.

5. CDN versus platform

A CDN distributes bytes. The platform adds catalogue, policy, player, creator, monetisation and measurement systems.

6. Player buffer versus stored archive

The buffer is temporary delivery state. The catalogue or origin stores durable media.

7. Resolution versus perceived quality

Pixel dimensions do not determine codec efficiency, motion handling, colour, bitrate, display or perception.

8. Availability versus recommendation

A video may be publicly accessible by URL while receiving almost no platform exposure.

9. View versus watch

A platform-defined view threshold is not the same as complete, attentive or understood viewing.

10. Moderation versus monetisation

A video can remain available while being excluded from advertising, recommendations or particular audiences.

11. User-generated versus independent infrastructure

A creator may own the source file while the platform owns the audience interface, metrics and delivery path.

12. Video platform versus social platform

Many systems contain both. The distinction depends on whether audiovisual publishing and playback are the central organising function.

13. Live availability versus live production

A stream may be presented as live while including delayed, prerecorded or replayed material.

7. Communication Pattern

A creator captures or produces media and uploads a source or sends a live contribution feed. The platform validates, scans, transcodes and stores the content, then creates metadata, thumbnails, captions, manifests and access rules. Discovery systems make the item eligible through search, subscription, links, recommendations or notifications. A player requests a manifest, selects representations, retrieves segments through a CDN, buffers and decodes them. Playback events return as analytics and recommendation evidence.

| Dimension | Pattern | |---|---| | Participation | One-to-many publication with comments, chat, reactions and creator feedback | | Timing | On-demand asynchronous viewing and live near-synchronous distribution | | Persistence | Source files, encoded renditions, manifests, metadata, captions, caches and analytics | | Topology | Creator to platform ingest; origin and CDN to individually addressed software players | | Feedback | Views, watch time, retention, search, comments, shares, subscriptions, reports and quality telemetry | | Access | Public, account-restricted, paid, geographic, age, device, rights and policy controls |

8. Expanded Communication Model

| Stage | Function | |---|---| | Creator and production system | Capture, edit and package source audiovisual material. | | Ingest layer | Receives upload or live contribution and validates format. | | Safety and rights processing | Scans for policy, copyright, malware or restricted content. | | Transcoding pipeline | Produces codecs, bitrates, resolutions, thumbnails, audio tracks and captions. | | Catalogue and metadata | Stores title, description, ownership, rights, language, categories and identifiers. | | Origin storage | Holds source or encoded media and authoritative manifests. | | Discovery and eligibility | Determines whether and where the item may appear. | | CDN and routing | Places or fetches media near recipients. | | Player and adaptation logic | Selects representations, manages buffer and decodes media. | | Display and audio output | Presents content under device and accessibility constraints. | | Measurement and feedback | Records playback, quality, interaction and commercial events. | | Ranking and governance return loop | Uses observations to alter future distribution and creator incentives. |

A complete model must separate:

  1. media publication;
  2. policy eligibility;
  3. discovery candidate generation;
  4. display of a thumbnail or player;
  5. playback start;
  6. continued playback;
  7. attentive viewing;
  8. comprehension or action.
9. Historical Emergence

Digital video compression and computer playback preceded mass online video. Early Internet streams used specialist players, low resolutions and narrow audiences. Web video remained difficult because creators faced incompatible codecs, hosting costs, upload limits and browser plugins.

YouTube registered its domain in February 2005 and uploaded its first video in April. The service described its founding aim as giving people with a camera and Internet connection a place to share video. [1][2] The platform’s historical importance lies less in inventing digital video than in combining simple upload, Web playback, shareable pages, comments and later large-scale hosting.

Subscription media followed a different path. Netflix moved from physical disc delivery toward online streaming, then built adaptive encoding, global delivery and personalised catalogue systems. Its technical publications document adaptive streaming, Open Connect distribution and content-specific encoding optimisation. [5][7][8]

HTTP-based adaptive streaming matured because ordinary Web infrastructure could deliver media segments through caches and CDNs. Apple’s HLS, first drafted in 2009 and later documented as RFC 8216, defines playlists and media segments for unbounded streams. [3] MPEG-DASH, first published in 2012, standardises media presentation descriptions and segment formats. [4]

Smartphones, broadband, cloud storage, CDNs, app stores and recommendation systems completed the environment. Online video became both a library and a feed: long-form film, education, live events, creator channels and brief vertical clips could all compete on the same personal screen.

10. Prerequisites
  • Film, television and digital video capture
  • Audio and video codecs
  • Internet and packet-switched networks
  • Broadband and fibre backbones
  • Cloud computing and object storage
  • Content-delivery networks and caching
  • Web browsers, mobile apps and media players
  • Databases and catalogue metadata
  • Smartphones, cameras and affordable editing tools
  • Search and recommendation systems
  • Advertising, subscriptions and payment infrastructure
  • Copyright, licensing and rights-management systems
  • Captions, transcripts and accessibility standards
  • Analytics, logging and experimentation systems
11. Periodisation

1. Downloaded and specialist Internet video

Users retrieve complete files or rely on dedicated streaming players and constrained connections.

2. Embedded Web-video experiments

Plugins and proprietary formats bring moving images into Web pages.

3. Mass video-sharing platforms

YouTube and peers simplify upload, hosting, playback and sharing.

4. Subscription streaming

Catalogues of professionally licensed media become available on demand.

5. HTTP adaptive streaming

Segmented manifests and multiple renditions improve delivery across variable networks.

6. CDN and cloud consolidation

Large storage, encoding and edge-delivery systems become central.

7. Live creator ecosystems

Individuals stream games, events, teaching and commentary with real-time chat.

8. Recommendation-led viewing

Home pages and feeds increasingly determine which videos receive attention.

9. Mobile and short-video dominance

Vertical capture, full-screen feeds and rapid recommendation loops reshape production.

10. Multi-format platform convergence

Long video, shorts, live streams, podcasts, television and commerce occupy one service.

12. Main Problem Addressed

Online video platforms reduce the audiovisual reproduction, scheduling and distribution constraint.

Before network platforms, large-scale moving-image distribution required cinemas, physical carriers, broadcast licenses or negotiated television schedules. Online systems allow stored or live media to be delivered on demand to individually addressed software players.

The new constraint becomes visibility, platform eligibility, network quality, rights, monetisation and attention.

Constraint transition

From scarce scheduled channels and physical copies to abundant on-demand streams whose practical reach is governed by infrastructure, platforms and ranking.

13. Evaluation Matrix

| Dimension | Assessment | |---|---| | Reach | Potentially global, limited by networks, rights, policy, language, payment and devices | | Latency | Seconds to start for on-demand; live latency varies from sub-second to many seconds | | Bandwidth | High and continuous relative to text or audio; adaptive delivery manages variation | | Fidelity | Ranges from highly compressed mobile video to high-resolution professional streams | | Persistence | Strong for catalogued media, weak under takedown, account loss, rights expiry or platform closure | | Replication cost | Very low per additional stream at scale, but encoding, storage and delivery remain expensive | | Accessibility | Captions, audio description and controls can help; implementation and language coverage vary | | Portability | Playback across devices is high; creator exports, purchases and histories may be platform-bound | | Interactivity | Moderate for on-demand; higher with live chat, comments and polls | | Searchability | High through metadata and transcripts, but visual content remains difficult to index completely | | Editability | Source editing is high; published revisions, re-uploads and version history vary | | Scalability | Very high through cloud and CDN systems | | Authentication | Accounts and DRM can strongly control access; identity does not prove authorship or truth | | Privacy | Platforms observe detailed playback and device behaviour | | Censorship resistance | Mixed; distribution is easy but platform and rights enforcement are concentrated | | Infrastructure dependence | Extremely high for encoding, storage, CDN, player, app and identity services | | Attention demand | Very high because audiovisual playback occupies sight, hearing and time | | Quality stability | Variable; adaptation can preserve continuity by reducing quality | | Measurement precision | High for player events, low for human attention, comprehension and social effect |

14. Advantages

1. On-demand control

Recipients choose when and where to watch.

2. Global creator publication

Individuals can distribute audiovisual work without owning broadcast infrastructure.

3. Adaptive delivery

Multiple renditions allow playback across heterogeneous networks and devices.

4. Searchable archives

Titles, metadata, captions and transcripts make vast catalogues navigable.

5. Long-tail distribution

Niche works can remain accessible without occupying scarce broadcast schedule.

6. Rapid feedback

Creators receive comments, analytics and audience-retention signals.

7. Accessibility potential

Captions, playback speed, transcripts and audio description can expand participation.

8. Live global events

A single source can reach many individually connected recipients.

9. Educational demonstration

Moving images can show procedures, environments and bodily action difficult to express in text.

10. Device continuity

Viewing can move among phones, televisions and computers through account state.

15. Civilisational Contributions

1. Public education

Lectures, demonstrations, documentaries and tutorials become globally available.

2. Citizen journalism

People can publish video evidence and commentary from events.

3. Cultural archives

Performances, interviews and historical recordings can remain accessible beyond original transmission.

4. Independent entertainment

Creators can build audiences outside traditional studios and broadcasters.

5. Remote events

Sport, religion, politics and ceremonies reach dispersed audiences live.

6. Skills transmission

Visual demonstration supports craft, repair, music, exercise and technical learning.

7. Language communities

Niche languages and diasporas can publish without national broadcast allocation.

8. New economies

Advertising, subscriptions, memberships, sponsorships and creator services form around video.

9. Political accountability

Recorded and live images can challenge official narratives, though interpretation remains contested.

10. Shared global moments

Major events can be watched and discussed across borders through common platforms.

16. Organisations, Access and Power

1. Platform operators

Control ingest, policy, ranking, monetisation, measurement and account continuity.

2. Cloud and CDN providers

Control encoding capacity, storage and delivery paths.

3. Rights holders and collecting organisations

License works, issue claims and influence availability by territory.

4. Advertisers and sponsors

Fund distribution while shaping acceptable content and measurement priorities.

5. Creators and production companies

Supply the media and adapt formats, titles and schedules to platform incentives.

6. Recommender and search teams

Determine candidate pools and practical visibility.

7. App stores and device vendors

Control player installation, background behaviour, payments and media capabilities.

8. Regulators and courts

Address copyright, competition, child safety, political content, privacy and accessibility.

9. Viewers

Supply attention, subscriptions, payments, behavioural data, comments and moderation reports.

The platform can exercise several independent powers: whether a video may exist, whether it may earn money, whether it may be recommended, whether it may reach a territory and how its performance is measured. These powers should not be collapsed into one status called “published”.

17. Limitations, Harms and Trade-Offs

1. Platform dependence

Creators can lose distribution, revenue and archives through policy or account decisions.

2. Visibility inequality

Abundant publication does not imply equal recommendation or search exposure.

3. Copyright automation errors

Matching systems can overclaim, miss context or favour parties with stronger administrative capacity.

4. Attention extraction

Autoplay, recommendations and endless feeds encourage continued viewing beyond deliberate intent.

5. Measurement distortion

Creators optimise thumbnails, duration and pacing for platform metrics that may not reflect value.

6. Surveillance

Platforms observe precise playback, search, device, network and interaction behaviour.

7. Misinformation and manipulated media

Audiovisual realism can create persuasive false evidence, especially when context is removed.

8. Harassment and live abuse

Comments, raids, doxxing and real-time audiences can target creators or participants.

9. Cultural homogenisation

Ranking can favour formats, languages and styles that perform well within dominant metrics.

10. Rights fragmentation

Territorial licences make catalogues unstable and geographically unequal.

11. Preservation fragility

Links, embeds, comments and platform metadata can disappear even when a source file survives.

12. Network exclusion

High data use disadvantages recipients with expensive or slow connections.

13. Energy and infrastructure cost

Encoding, storage and repeated delivery consume significant compute and network resources.

14. Context collapse

A video created for one audience can be extracted, clipped and presented to another.

15. Live latency confusion

Different viewers may see the “same” event several seconds apart, complicating sport, betting and shared reaction.

16. Quality illusion

A 4K label can coexist with poor encoding, low dynamic range, bad audio or aggressive compression.

17. Creator precarity

Revenue, discoverability and rules can change without negotiated stability.

18. Synthetic-media flood

Cheap generation can overwhelm discovery, provenance and moderation systems.

18. Predecessors, Successors and Relationships

| Relationship | Topic | Reason | |---|---|---| | Predecessor | Television broadcasting Television Broadcasting | Establishes audiovisual mass communication, programming and audience measurement. | | Predecessor | Motion-picture film and cinema Motion Picture Film | Establishes recorded moving-image narrative and exhibition. | | Predecessor | World Wide Web World Wide Web | Provides pages, links, browsers and public identifiers. | | Predecessor | Fibre-Optic Communication Fibre-Optic Communication | Supplies high-capacity backbone transmission. | | Predecessor | Cloud computing and cloud storage Cloud Computing and Cloud Storage | Supplies ingest, encoding, persistence and elastic delivery. | | Related | Smartphones Smartphones | Provide ubiquitous cameras, players and mobile networks. | | Related | Recommendation Algorithms and Personalised Feeds Recommendation Algorithms and Personalised Feeds | Allocate video attention and sequence viewing. | | Related | Social Networking and Microblogging Platforms Social Networking and Microblogging Platforms | Supply social sharing, identity and feed distribution. | | Related | Podcasts and on-demand audio Podcasts and On-Demand Audio | Shares subscription, hosting and streaming architecture with lower sensory bandwidth. | | Successor | Generative Image, Audio and Video Models Generative Image, Audio and Video Models | Generate or transform media entering the platform. | | Successor | Digital Provenance and Authenticity Systems Digital Provenance and Authenticity Systems | Address source and transformation claims for audiovisual media. |

Online video remediates cinema and television but does not replace them. Physical exhibition and scheduled broadcasting retain communal, regulatory and quality characteristics.

19. What Survived

1. Programme and episode structures

Television seasons, channels and scheduled premieres survive inside on-demand catalogues.

2. Broadcast advertising

Commercial breaks survive as targeted pre-roll, mid-roll and inserted advertising.

3. Film editing and production grammar

Shots, cuts, sound design and narrative conventions remain central.

4. The channel

A broadcast frequency becomes a creator or publisher identity and catalogue.

5. Ratings logic

Audience measurement survives as views, watch time, retention and completion metrics.

6. Subscription television

Recurring payment survives as streaming membership.

7. Live event production

Uplink, control room and commentary survive, now feeding packet delivery.

8. The programme guide

Search, home pages and recommendations replace scheduled listings while retaining catalogue navigation.

9. Home video

Personal recordings survive as global potential publications.

20. Representative Cases

1. YouTube

Combines simple upload, Web playback, sharing, creator channels, advertising and recommendation [1][2].

2. HTTP Live Streaming

Defines playlists, media segments and client actions for HTTP-delivered streams [3].

3. MPEG-DASH

Standardises adaptive media presentation descriptions and segment formats [4].

4. Netflix streaming

Demonstrates subscription catalogue delivery, adaptive quality and large-scale personalisation [S05-S08].

5. Open Connect

Places media-serving infrastructure close to viewers and separates catalogue application from delivery edge [8].

6. Per-title encoding

Adjusts encoding ladders to content characteristics rather than one fixed ladder for every title [7].

7. Live creator streaming

Combines near-live audiovisual delivery with chat, subscriptions and community moderation.

8. Short-video feed

Combines phone-native capture, vertical playback and rapid recommendation loops.

21. Research Uncertainty and Open Questions
  • Should video-sharing, subscription streaming, live streaming and short-video feeds split into descendant topics?
  • Which technical threshold distinguishes streaming from progressive download in practice?
  • How should view definitions be normalised across platforms?
  • Which measures approximate human attention rather than player activity?
  • How should recommendation and platform distribution be separated in causal analysis?
  • Should CDNs receive a dedicated infrastructure topic?
  • How should deleted videos be preserved with comments, captions and context?
  • Which accessibility fields should be mandatory for video topics?
  • How should live latency and synchronisation be compared across audiences?
  • When does synthetic video require a distinct publication lineage rather than a production note?
  • How should creator sovereignty be measured when source files are owned locally but audience, identity and analytics are rented?

The map should avoid treating “streaming” as one magic pipe. The player, manifest, segments, origin, CDN and adaptation logic are distinct components, each capable of ruining the evening in its own creative fashion.

22. Claim Register

|---|---|---|---| | Online video and streaming platforms-C01 | Online video platforms combine ingest, encoding, storage, delivery, player and governance layers. | High | S03-S08; architecture | | Online video and streaming platforms-C02 | YouTube launched in 2005 and simplified Web video sharing. | High | S01-S02 | | Online video and streaming platforms-C03 | HLS and MPEG-DASH use manifests and segmented HTTP delivery for adaptive streaming. | High | S03-S04 | | Online video and streaming platforms-C04 | Adaptive players can change representation during playback according to conditions. | High | S03-S06 | | Online video and streaming platforms-C05 | A view is a platform-defined playback event, not proof of attentive completion. | High | Measurement analysis | | Online video and streaming platforms-C06 | Publication, recommendation, display and sustained watching are separate audience states. | High | Systems analysis | | Online video and streaming platforms-C07 | CDNs reduce delivery distance but do not determine catalogue or platform policy. | High | S08; architecture | | Online video and streaming platforms-C08 | Per-title encoding can reduce waste while preserving perceived quality. | High | S07 |

23. Comparative Analysis

| Comparison | Main difference | Analytical value | |---|---|---| | Broadcast television | One scheduled transmission to many receivers | Contrasts fixed programme flow with individually requested streams. | | Cinema | Communal physical exhibition with controlled environment | Shows convenience versus shared attention and presentation quality. | | Physical video carrier | Recipient owns or rents a local copy | Contrasts durable possession with service access. | | File download | Complete local transfer | Separates retained copy from adaptive temporary playback. | | Video conference | Reciprocal low-latency participation | Contrasts interaction with one-to-many distribution. | | Podcast | Audio-first subscribed and portable medium | Shows reduced bandwidth and attention demands. | | Social image feed | Primarily still or brief media inside social relationships | Shows the audiovisual platform as catalogue, player and delivery system. | | CDN | Delivery infrastructure only | Separates byte distribution from platform governance and discovery. |

The decisive comparison is between channel scarcity and attention scarcity. Television restricts how many programmes can occupy a schedule. Online video can host immense catalogues, so selection, recommendation and the recipient’s time become the new bottlenecks.

28. Final perspective

Online video turns a film, lesson, match or bedroom monologue into a sequence of service states. The source is uploaded or contributed live. Machines inspect it, encode it repeatedly, store it, build manifests, distribute segments and measure playback. Recommendation systems decide whether a thumbnail appears. The player decides which rendition can survive the network. The viewer decides, perhaps, whether to keep watching.

This chain explains why “streaming” is too small a word for the institution. The visible act is playback; the hidden system is a factory, library, broadcast network, advertising market, rights office and behavioural laboratory fused together. The platform can make publication easy while making attention opaque. It can preserve a catalogue for years while removing one creator overnight. It can deliver billions of segments with exquisite reliability while remaining unable to prove that one human understood one minute.

Online video replaces the scarcity of channels with the abundance of files. Abundance then creates a selection problem. Search, subscriptions and recommendation become the practical programme schedule, but unlike television listings, each recipient may receive a different one. The platform is therefore both carrier and editor, even when it insists that the algorithm merely found what people wanted.

The medium’s civilisational value is enormous: visual knowledge, testimony, culture and entertainment become accessible across distance and time. Its danger is equally structural: a handful of systems can control the archive, audience, revenue and measurement of global moving-image culture.

Online video platforms turn audiovisual publication into a cloud pipeline that ingests, encodes, stores, selects and adaptively delivers media to software players. Their defining achievement is on-demand audiovisual reach; their defining power lies in controlling both the delivery machinery and the conditions of visibility.

Evidence

Sources and further reading

  1. YouTube, “Celebrating 10 Years of YouTube,” 1 May 2015. https://blog.youtube/news-and-events/celebrating-10-years-of-youtube/

    Open source ↗

  2. YouTube, “YouTube & the Online Video Revolution,” 14 February 2010. https://blog.youtube/news-and-events/youtube-online-video-revolution/

    Open source ↗

  3. Roger Pantos and William May, RFC 8216, *HTTP Live Streaming*, 2017. https://datatracker.ietf.org/doc/html/rfc8216

    Open source ↗

  4. DASH Industry Forum, “References,” including ISO/IEC 23009-1 Dynamic Adaptive Streaming over HTTP. https://dashif.org/identifiers/references/

    Open source ↗

  5. Netflix Technology Blog, “Optimizing the Netflix Streaming Experience with Data Science,” 11 June 2014. https://techblog.netflix.com/2014/06/optimizing-netflix-streaming-experience.html

    Open source ↗

  6. IETF, RFC 9317, *Operational Considerations for Streaming Media*, 2022. https://datatracker.ietf.org/doc/rfc9317/

    Open source ↗

  7. Netflix Technology Blog, “Per-Title Encode Optimization,” 14 December 2015. https://techblog.netflix.com/2015/12/per-title-encode-optimization.html

    Open source ↗

  8. Netflix Technology Blog, “Behind the Streams: Three Years of Live at Netflix, Part 1,” 15 July 2025. https://netflixtechblog.com/behind-the-streams-live-at-netflix-part-1-d23f917c2f40 Online video turns a film, lesson, match or bedroom monologue into a sequence of service states. The source is uploaded or contributed live. Machines inspect it, encode it repeatedly, store it, build manifests, distribute segments and measure playback. Recommendation systems decide whether a thumbnail appears. The player decides which rendition can survive the network. The viewer decides, perhaps, whether to keep watching. This chain explains why “streaming” is too small a word for the institution. The visible act is playback; the hidden system is a factory, library, broadcast network, advertising market, rights office and behavioural laboratory fused together. The platform can make publication easy while making attention opaque. It can preserve a catalogue for years while removing one creator overnight. It can deliver billions of segments with exquisite reliability while remaining unable to prove that one human understood one minute. Online video replaces the scarcity of channels with the abundance of files. Abundance then creates a selection problem. Search, subscriptions and recommendation become the practical programme schedule, but unlike television listings, each recipient may receive a different one. The platform is therefore both carrier and editor, even when it insists that the algorithm merely found what people wanted. The medium’s civilisational value is enormous: visual knowledge, testimony, culture and entertainment become accessible across distance and time. Its danger is equally structural: a handful of systems can control the archive, audience, revenue and measurement of global moving-image culture. > **Online video platforms turn audiovisual publication into a cloud pipeline that ingests, encodes, stores, selects and adaptively delivers media to software players. Their defining achievement is on-demand audiovisual reach; their defining power lies in controlling both the delivery machinery and the conditions of visibility.**

    Open source ↗