R0045/2026-03-29/Q002 — Assessment¶
BLUF¶
Apache was already the dominant web server by hostname count in mid-2001, holding 63% share versus IIS at 20%. This dominance had been established since 1996. However, when counting physical servers rather than hostnames, IIS led significantly because Apache was concentrated at shared hosting providers running many sites per server. By November 2003, Apache ran approximately two-thirds of all web sites. The question of "when Apache became dominant" has a nuanced answer: by hostname count, circa 1996; by enterprise deployment, the picture remained contested through the early 2000s.
Probability¶
Rating: Not applicable (query mode — complex factual question)
Confidence in assessment: High
Confidence rationale: Netcraft survey data is the definitive source for web server market share data from this period. Multiple independent reports of the same survey data confirm the figures.
Reasoning Chain¶
- Netcraft's June 2001 survey received responses from 29,302,656 sites [SRC01-E01, High reliability, High relevance]
- Apache held 63.02% by hostname (18.5M sites) and 62.42% of active sites [SRC01-E01]
- Microsoft IIS held 20.38% by hostname (6.0M) and 26.14% of active sites [SRC01-E01]
- When counting physical servers, Windows led at 49.2% vs Linux at 28.5% [SRC03-E01, High reliability, High relevance]
- The physical vs. hostname discrepancy was explained by Apache's concentration at hosting providers running many sites per server [SRC03-E01]
- By November 2003, Apache reached approximately two-thirds of all hostnames; domain parking companies that had temporarily switched to IIS reverted to Apache [SRC02-E01, Medium reliability, High relevance]
Evidence Base Summary¶
| Source | Description | Reliability | Relevance | Key Finding |
|---|---|---|---|---|
| SRC01 | Netcraft June 2001 via ApacheToday | High | High | Apache 63.02%, IIS 20.38% by hostnames |
| SRC02 | Netcraft Nov 2003 via Slashdot | Medium | High | Apache reached ~2/3 of web by late 2003 |
| SRC03 | The Register analysis | High | High | Physical server count: Windows 49.2%, Linux 28.5% |
Collection Synthesis¶
| Dimension | Assessment |
|---|---|
| Evidence quality | Robust — Netcraft is the definitive source for web server surveys in this period |
| Source agreement | High — all sources cite the same Netcraft survey data with consistent figures |
| Source independence | Low — all ultimately derive from Netcraft surveys; this is inherent to the domain |
| Outliers | None |
Detail¶
The key insight is that Apache's "dominance" depends entirely on measurement methodology. By the hostname metric that was standard at the time, Apache was overwhelmingly dominant at 63% vs. IIS at 20%. But the physical server metric told a different story — Windows ran on more physical machines. This distinction was important for enterprise decision-making: large enterprises cared about their own deployments, not shared hosting statistics.
Gaps¶
| Missing Evidence | Impact on Assessment |
|---|---|
| Gartner's specific assessments of Apache for enterprise use in 2001 | Cannot answer what analysts said about Apache's enterprise viability |
| Enterprise-specific web server deployment data (as distinct from overall internet) | Cannot distinguish enterprise vs. hosting provider Apache usage |
Researcher Bias Check¶
Declared biases: No researcher profile provided.
Influence assessment: The query assumes Apache "became" dominant, implying a transition. In fact, Apache was already dominant by hostname count since 1996. The question may reflect a misconception that Apache was an underdog in 2001.
Cross-References¶
| Entity | ID | File |
|---|---|---|
| Hypotheses | H1, H2, H3 | hypotheses/ |
| Sources | SRC01, SRC02, SRC03 | sources/ |
| ACH Matrix | — | ach-matrix.md |
| Self-Audit | — | self-audit.md |