~ / topics / Reference
$man articles-and-sources

Articles and Sources

In brief: This is a reading list for the Year 2038 problem. Each entry names the source, says who published it and when, gives the link, and explains in a sentence or two why it is worth reading. The entries are grouped into news, technical deep-dives, academic research, government and standards material, talks, tools, and historical Y2K sources. All entries were located during research for this wiki on or before 2026-10-21.

Risk level n/a (bibliography)
Status 81 entries; checked 2026-10-21
Last reviewed 2026-10-21

Overview

How to use this list.

Verification notes. Most entries were fetched directly. Entries marked (seen in search results) could not be fetched (paywall, HTTP 403 or offline). Their details come from search-engine snippets or from citations in other fetched sources. Dates are publication dates unless marked otherwise.

News & journalism

  1. "Need 32-bit Linux to run past 2038? When version 5.6 of the kernel pops, you're in for a treat". Richard Speed, The Register, 2020-01-30. https://www.theregister.com/2020/01/30/linux_5_6_2038/. Reports Arnd Bergmann's statement that Linux 5.6 is the first kernel that can serve as a base for a 32-bit system running past 2038, along with his caveats about interfaces that cannot change.
  2. "Linux 5.6 Is The First Kernel For 32-Bit Systems Ready To Run Past Year 2038". Michael Larabel, Phoronix, 2020-01-29. https://www.phoronix.com/news/Linux-5.6-32-bit-Past-Y2038. A short news item on the same milestone. It notes that user space must also be rebuilt with a 64-bit time_t C library.
  3. "The Time Namespace Appears To Finally Be On-Deck For The Mainline Linux Kernel". Michael Larabel, Phoronix, 2020-01-14. https://www.phoronix.com/news/Linux-Time-Namespace-Coming. Explains why time namespaces don't virtualise the realtime clock, which matters for container-based 2038 testing.
  4. "Curious tale of broken VPNs, the Year 2038, and certs that expired 100 years ago". Thomas Claburn, The Register, 2024-02-09. https://www.theregister.com/2024/02/09/it_incident_report_the_clock/. A first-hand incident report from 2010, in which a CA's 32-bit date overflow and NTP bugs took down VPNs across about 100 retail sites.
  5. "Y2K38 'Epochalypse': Das Jahr-2000-Problem kehrt zurück". Nadine Dressler, WinFuture, 2024-03-06. https://winfuture.de/news,141592.html. Covers a demonstration that 32-bit Windows 10 loses core functions when its clock is set to 2038. In German.
  6. "Risky Bulletin: The looming Epochalypse". Catalin Cimpanu, Risky Business, 2025-03-24. https://www.risky.biz/risky-bulletin-the-looming-epochalypse/. Argues that 2038 is harder than Y2K because many devices cannot be updated, and calls for regulation that mandates 64-bit time.
  7. "The Epochalypse: It's Y2K, But 38 Years Later". Zoe Skyforest, Hackaday, 2025-07-22. https://hackaday.com/2025/07/22/the-epochalypse-y2k-but-38-years-later/. An accessible explainer covering OS adoption dates and filesystem limits (ext4 2446, XFS 2486). Its mortgage example is not tied to a specific incident.
  8. "Debian isn't waiting for 2038 to blow up, switches to 64-bit time for everything". Gareth Halfacree, The Register, 2025-07-25. https://www.theregister.com/2025/07/25/y2k38_bug_debian/. Reports Debian 13's move to 64-bit time_t on 32-bit ports, the i386 exception, and the 6,429 packages that reference time_t.
  9. "The Unix Epochalypse might be sooner than you think". Richard Speed, The Register, 2025-08-23. https://www.theregister.com/2025/08/23/the_unix_epochalypse_might_be/. Describes how volunteers at The National Museum of Computing found a PDP-11/73 that crashes at the start of 2037, a year before the rollover.
  10. "The Y2K38 Bug Is a Vulnerability, Not Just a Date Problem, Researchers Warn". SecurityWeek, 2025 (exact date unconfirmed). https://www.securityweek.com/the-y2k38-bug-is-a-vulnerability-not-just-a-date-problem-researchers-warn/. Summarises the Darley and Umbelino research showing that GPS spoofing or NTP injection can trigger 2038 bugs early (seen in search results).
  11. "Exchange Server – Email Trapped in Transport Queues". Guy Bruneau, SANS Internet Storm Center, 2022-01-02. https://dshield.org/diary/28204. Contemporary coverage of the "Y2K22" Exchange failure, a signed 32-bit overflow with the same shape as the 2038 bug.
  12. "25 Years From Today: A Time For Bugs" (title per URL slug). InformationWeek, 2013-01-19. https://www.informationweek.com/it-infrastructure/25-years-from-today-a-time-for-bugs. An early mainstream piece reporting 2038 limits in 32-bit Android devices (cited by Wikipedia; not fetched).

Technical deep-dives (LWN, kernel docs, glibc docs, Debian wiki)

  1. "System call conversion for year 2038". Jonathan Corbet, LWN.net, 2015-05-05. https://lwn.net/Articles/643234/. Explains why the kernel chose 64-bit timespec64 over a nanosecond counter, and outlines the staged syscall migration.
  2. "The (initial) glibc year-2038 plan". Nathan Willis, LWN.net, 2015-11-19. https://lwn.net/Articles/664800/. The origin of the _TIME_BITS feature-macro design, including Rich Felker's objections based on the LFS experience.
  3. "Summary of the DebConf 2038 BoF". Steve McIntyre, via LWN.net briefs, 2017-09. https://lwn.net/Articles/732794. A concise list of every layer that needs fixing, and why embedded and statically linked software is the biggest risk.
  4. "Approaching the kernel year-2038 end game". Jonathan Corbet, LWN.net, 2019-01-11. https://lwn.net/Articles/776435/. Describes the _time32/_time64 syscall split, COMPAT_32BIT_TIME and the new socket options. Essential background on the kernel.
  5. "Solving the Year 2038 problem in the Linux kernel". Deepa Dinamani, Opensource.com, 2019-01-21. https://opensource.com/article/19/1/year2038-problem-linux-kernel. A first-hand account of converting VFS inode timestamps to timespec64, including a Coccinelle "flag day" patch across 80+ files.
  6. "[GIT PULL] y2038: core, driver and file system changes". Arnd Bergmann, LKML, 2020-01. https://lkml.iu.edu/hypermail/linux/kernel/2001.3/05499.html. The pull request that completed the kernel side for 5.6. It includes the "first release that can serve as a base" statement (seen in search results).
  7. "[RFC][PATCH] timekeeping: Limit system time to prevent 32-bit time_t overflow". Miroslav Lichvar, LKML, 2015-04-15. https://lkml.rescloud.iu.edu/1504.1/05251.html. A rejected but instructive proposal to stop the clock a week before overflow. It shows the risk of NTP or PTP pushing clocks forward.
  8. "GNU C Library 2.34 released". Carlos O'Donell, libc-alpha via LWN.net, 2021-08-01. https://lwn.net/Articles/864920. The authoritative release notes for _TIME_BITS=64, the LFS requirement and the Linux 5.1 minimum.
  9. "Debian looks forward to 2038". Jonathan Corbet, LWN.net, 2023-07-17. https://lwn.net/Articles/938149/. Covers Steve Langasek's transition plan, the t64 renaming scheme and the debate over excluding i386.
  10. "The perils of transition to 64-bit time_t". Michał Górny, Gentoo blogs, 2024-09-28. https://blogs.gentoo.org/mgorny/?p=2218. The clearest explanation of why mixing time32 and time64 objects corrupts memory, and how a source-based distribution can migrate safely. LWN discussed it at https://lwn.net/Articles/992120/.
  11. "ReleaseGoals/64bit-time". Debian Wiki (living page). https://wiki.debian.org/ReleaseGoals/64bit-time. The full record of Debian's transition: dates, abi-compliance-checker statistics, abi=+time64 build options and known risk areas.
  12. "y2038". Arnd Bergmann et al., KernelNewbies, last edited 2019-01-18. https://kernelnewbies.org/y2038. The historical project page for the kernel effort. It is now mostly of archival value.
  13. "Linux 5.6". KernelNewbies, release 2020-03-29. https://kernelnewbies.org/Linux_5.6. Release notes confirming that the time namespace arrived in Linux 5.6.
  14. "time_namespaces(7)". Linux man-pages project (man7.org). https://man7.org/linux/man-pages/man7/time_namespaces.7.html. Documents which clocks a time namespace can offset (not CLOCK_REALTIME) and the timens_offsets interface.
  15. "Avoiding the year 2038 problem". GNU Gnulib manual. https://www.gnu.org/software/gnulib/manual/html_node/Avoiding-the-year-2038-problem.html. How Autoconf projects can enable 64-bit time, or require it with year2038-recommended.
  16. "musl 1.2.0 released". Rich Felker, musl mailing list (Openwall), 2020-02-21. https://openwall.com/lists/musl/2020/02/21/1. Announcement of musl switching all 32-bit architectures to 64-bit time_t.
  17. "gmtime, _gmtime32, _gmtime64". Microsoft Learn. https://learn.microsoft.com/en-us/cpp/c-runtime-library/reference/gmtime-gmtime32-gmtime64?view=msvc-170. Documents the Windows CRT's 32-bit and 64-bit time functions and warns against _USE_32BIT_TIME_T.
  18. "The DATE, DATETIME, and TIMESTAMP Types". MySQL 8.0 Reference Manual, Oracle. https://dev.mysql.com/doc/refman/8.0/en/datetime.html. The primary source for MySQL's TIMESTAMP upper limit of 2038-01-19 03:14:07 UTC.
  19. "krb5 commit: Add y2038 documentation". MIT Kerberos, 2017-05. https://mailman.mit.edu/pipermail/cvs-krb5/2017-May/014236.html. A real-world example of the unsigned-reinterpretation strategy ((time_t)(uint32_t)ts), valid until 2106.
  20. "The NTP Era and Era Numbering". David L. Mills, University of Delaware. https://www.eecis.udel.edu/~mills/y2k.html. The authoritative explanation of NTP eras and the 2036 rollover.

Academic & research papers

  1. "The tourism industry and the risk of 'the year 2038 problem': Long-term thinking and the illogical cycle of ICT risk management". Y. Ram, International Journal of Tourism Research, 2022. https://onlinelibrary.wiley.com/doi/abs/10.1002/jtr.2517. A peer-reviewed risk-management view of why organisations defer long-horizon ICT risks like 2038 (seen in search results).
  2. "Detecting and Analyzing Year 2038 Problem Bugs in User-Level Applications". Suzuki, Kubota & Kono, 2019. https://www.researchgate.net/publication/338509499_Detecting_and_Analyzing_Year_2038_Problem_Bugs_in_User-Level_Applications. An empirical study of 2038 bugs in user-space software. The full text was not accessible, so the venue is unconfirmed (seen in search results).
  3. "Solving the Year 2038 Problem in 32 bit Linux by Converting Time in Clock Synchronization Protocols". Yabuki, Okabe, Horii & Toyama, IEICE FIT 2019. https://www.ieice.org/publications/conference-FIT-DVDs/FIT2019/data/pdf/B-013.pdf. Proposes converting time inside NTP or PTP synchronisation, rather than changing kernel or libc, for long-lived industrial systems. Mostly in Japanese (seen in search results).
  4. "Addressing the Year 2038 Problem: Transitioning to 64-Bit Time Representations for Sustainable Computing". ResearchGate preprint, 2024-12. https://www.researchgate.net/publication/387262809_Addressing_the_Year_2038_Problem_Transitioning_to_64-Bit_Time_Representations_for_Sustainable_Computing_Abstract. A survey-style preprint on origins, risks and mitigations. It has not been peer reviewed (seen in search results).
  5. "Forward to the Past: The Y2K38 Problem Ahead". Pedro Umbelino, Bitsight TRACE, 2026-01-18. https://www.bitsight.com/blog/what-is-y2k38-problem. Industry research with internet-scan exposure estimates for NTP servers, Boa web servers and tank gauges, plus replacement-cost modelling.
  6. "Epochalypse Now — or how to deal with Y2K38". Stan Kaminsky, Kaspersky, 2026-01-19. https://www.kaspersky.com/blog/mitigating-y2k38-vulnerability-in-organization/55150/. A practitioner's programme guide covering inventory, prioritisation, vendor engagement and safe testing.

Government & standards

  1. ICSA-25-261-07: "Dover Fueling Solutions ProGauge MagLink LX4 Devices". CISA, 2025-09-18. https://www.cisa.gov/news-events/ics-advisories/icsa-25-261-07. Advisory for CVE-2025-55068, a Unix-time overflow triggered by changing the system time, with fixed firmware versions.
  2. ICSA-25-296-03: "Veeder-Root TLS4B Automatic Tank Gauge System". CISA, 2025-10-23. https://www.cisa.gov/news-events/ics-advisories/icsa-25-296-03. Advisory for CVE-2025-55067, which describes the clock wrapping to 1901 and the operational consequences.
  3. VDE-2025-007: "WAGO: Year 2038 problem". CERT@VDE, 2025-04-15 (rev. 2026-08-07). https://certvde.com/en/advisories/VDE-2025-007/. The first CERT advisory dedicated to 2038 in industrial controllers (CVE-2025-0101).
  4. VDE-2025-020: "WAGO: Switches affected by year 2k38 problem". CERT@VDE, 2025-06-02. https://certvde.com/en/advisories/VDE-2025-020. CVE-2025-1235. The impact is limited to time and logs.
  5. CVE-2025-55068. CVE Program. https://www.cve.org/CVERecord?id=CVE-2025-55068. The canonical CVE record for the Dover ProGauge time overflow (cited by Kaspersky; not fetched).
  6. "Time Guidance for Network Operators, CIOs, CISOs". CISA, rev. 2020-12-17. https://www.cisa.gov/resources-tools/resources/time-guidance-network-operators-cios-cisos. General guidance on time integrity. No 2038-specific content was found on its landing page.
  7. Austin Group issue 0001462: "Require (at least) 64-bit time_t". Geoff Clare, The Open Group, 2021 (closed 2024-06-11). https://austingroupbugs.net/view.php?id=1462. The accepted POSIX change requiring time_t to be at least 64 bits wide.
  8. RFC 7530: "Network File System (NFS) Version 4 Protocol". IETF, 2015. https://www.rfc-editor.org/rfc/rfc7530#section-2.2. Defines NFSv4's 64-bit seconds timestamps, in contrast to NFSv3's unsigned 32-bit values (cited by Wikipedia).
  9. "Homeland Security provides info about 2019 GPS rollover event". GPS World. https://www.gpsworld.com/homeland-security-provides-info-about-2019-gps-rollover-event/. Coverage of the DHS/NCCIC memo for owners and operators. It is the closest US government precedent for a rollover advisory (seen in search results).

Note: No dedicated Year 2038 publication from NIST, ENISA or UK NCSC was found as of 2026-10-21. See Risk-Assessment-Checklist.

Talks & videos

  1. "The end of time, 19 years to go". Arnd Bergmann (Linaro), Linux Plumbers Conference, 2018-11-13. Abstract: https://lpc.events/event/2/contributions/57/; video: https://www.youtube.com/watch?v=NkUJJb4kDcc. A status report on kernel, filesystem and syscall work by the effort's lead developer.
  2. "It's The End of the World... (in 21 Years)". Steve McIntyre, DebConf17 BoF, Montréal, 2017-08. Slides: https://einval.com/~steve/talks/Debconf17-eotw-2038/; video: http://meetings-archive.debian.net/pub/debian-meetings/2017/debconf17/it-s-the-end-of-the-world-in-21-years.vp8.webm. The session that started Debian's planning.
  3. "2038 is gonna be epoch!". Pedro Umbelino & Trey Darley, Hack.lu 2025, 2025-10-23. https://archive.hack.lu/hack-lu-2025/talk/TLVK3W/. Controlled experiments on IoT, ICS and OT devices showing 2038 failure modes and attack vectors. No recording link was listed.
  4. "Pulling 32-bit time_t Asbestos out of the Open Source Ecosystem". Trey Darley, FOSDEM 2026 BoF, 2026-01-31. https://archive.fosdem.org/2026/schedule/event/KTFEWJ-2038-rollover-open-source/. Introduces the "2038-Class Risk Exposure Matrix" for coordinating open-source remediation.
  5. "HS106: Planning for the Epochalypse". Johna Till Johnson & John Burke, Packet Pushers Heavy Strategy (podcast), 2025-06-17. https://packetpushers.net/podcasts/heavy-strategy/hs106-planning-for-the-epochalypse/. A management-level discussion of embedded-system inventory, procurement clauses and board escalation.
  6. "Dealing with 2038". John Stultz, Linux Kernel Summit 2014 topic proposal. https://www.kernel.org/pub/linux/kernel/people/paulmck/LKS/LKS2014Topic/000117.html. An early kernel-community call to plan the transition (seen in search results).

Tools & projects

  1. libfaketime. wolfcw/libfaketime project, GitHub. https://github.com/wolfcw/libfaketime. LD_PRELOAD library that fakes the clock for a single process. It names 2038 testing as a use case.
  2. cppcheck y2038 addon. Cppcheck project, GitHub. https://github.com/danmar/cppcheck/blob/main/addons/y2038.py. Checks _TIME_BITS and _FILE_OFFSET_BITS and flags Y2038-unsafe identifiers.
  3. y2k38-checker. cysec-lab, GitHub. https://github.com/cysec-lab/y2k38-checker. A Clang Static Analyzer plugin detecting time_t downcasts and filesystem timestamp handling.
  4. "RFC: Y2038 check for lossy time_t conversions". LLVM Discourse, 2026-06-12. https://discourse.llvm.org/t/rfc-y2038-check-for-lossy-time-t-conversions/91060. A proposed clang-tidy check, still under discussion as of 2026-06-30.
  5. Yocto/OpenEmbedded 32bit-time QA check. OpenEmbedded-Core, insane.bbclass. https://github.com/openembedded/openembedded-core/blob/master/meta/classes-global/insane.bbclass. Build-time ELF scan for about 110 glibc 32-bit time APIs.
  6. Yocto Project 4.3 migration notes. Yocto Project docs. https://docs.yoctoproject.org/5.1/_sources/migration-guides/migration-4.3.rst.txt. Records the switch to 64-bit time_t for 32-bit targets (seen in search results).
  7. abi-compliance-checker. lvc, GitHub. https://github.com/lvc/abi-compliance-checker. The ABI diff tool Debian used to find libraries affected by the time_t change (seen in search results).
  8. QEMU -rtc option. QEMU documentation. https://www.qemu.org/docs/master/system/invocation.html. How to start a VM with its real-time clock set to any date, for full-system 2038 tests.
  9. The Epochalypse Project. Community project (Umbelino, Darley et al.). https://epochalypse-project.org/en/faq/. Advocacy and coordination site. Its FAQ notes the absence of any regulatory mandate for 2038.
  10. Gentoo tc-has-64bit-time_t helper. Michał Górny, gentoo-dev list, 2024-06. https://public-inbox.gentoo.org/gentoo-dev/usexgss9h@gentoo.org/t. An eclass helper that lets packages skip post-2038 tests on time32 systems (seen in search results).
  11. Coverity Y2K38_SAFETY findings in BIND 9. ISC GitLab issue #3497. https://gitlab.isc.org/isc-projects/bind9/-/issues/3497. Shows the commercial checker's real output on a major codebase (seen in search results).

Y2K historical sources

  1. "Year 2000 Computing Crisis: Costs and Planned Use of Emergency Funds". US GAO (AIMD-99-154), 1999. https://www.govinfo.gov/content/pkg/GAOREPORTS-AIMD-99-154/html/GAOREPORTS-AIMD-99-154.htm. Puts US federal Y2K cost estimates at about $7.5 billion, and found that only 7 of 24 agencies tracked actual costs separately (seen in search results).
  2. "Y2K Aftermath – Crisis Averted: Final Committee Report". US Senate Special Committee on the Year 2000 Technology Problem, 2000-02-29. https://www.govinfo.gov/content/pkg/GOVPUB-Y1_3-PURL-LPS90964/pdf/GOVPUB-Y1_3-PURL-LPS90964.pdf. The Senate committee's closing assessment of what went wrong and what was prevented (seen in search results).
  3. "The Journey to Y2K: Final Report of the President's Council on Year 2000 Conversion". President's Council on Year 2000 Conversion, 2000-03-29. Description: https://itlaw.fandom.com/wiki/The_Journey_to_Y2K:_Final_Report_of_the_President%27s_Council_on_Year_2000_Conversion. The executive-branch retrospective. Only a secondary description was located (seen in search results).
  4. OMB "Progress on Year 2000 Conversion", 6th Quarterly Report. US Office of Management and Budget (Clinton White House archive). https://clintonwhitehouse3.archives.gov/omb/inforeg/y2k6q.html. An example of the periodic government progress reporting that a 2038 programme could imitate (seen in search results).
  5. "Unix Tick Tocks to a Billion". Wired, 2001-09. https://www.wired.com/2001/09/unix-tick-tocks-to-a-billion/. Contemporary coverage of the 2001 "billennium", a small dress rehearsal for Unix-time limits (cited by Wikipedia; not fetched).
  6. CNN/IDG report on Y2K-related heating-system failures in Korea (archived, 2000-01-03, per URL). https://web.archive.org/web/20221127191633/http:/www.cnn.com/2000/TECH/computing/01/03/korea.heat.y2k.idg/index.html. One of the real Y2K failures that Kaspersky cites when arguing that 2038 failures will be real too (cited by Kaspersky; not fetched).
  7. "Time formatting and storage bugs". Wikipedia. https://en.wikipedia.org/wiki/Time_formatting_and_storage_bugs. A catalogue of Y2K, GPS, NTP, 2038 and 2106 issues, with links to primary sources. Useful as a starting point.

Primary sources added from the consolidated findings review

These entries were added on 2026-10-21 after reviewing date-time-rollover-findings.md. Each was fetched and checked before it was added, except where marked.

  1. "Feature Test Macros". GNU C Library manual 2.44. https://www.sourceware.org/glibc/manual/2.44/html_node/Feature-Test-Macros.html. The authoritative description of _TIME_BITS=64, the paired _FILE_OFFSET_BITS=64 requirement and the fallback to legacy kernel interfaces (not re-fetched; the NEWS file cited in OS-Linux covers the same facts).
  2. <sys/types.h>. POSIX.1-2024 (The Open Group Base Specifications Issue 8). https://pubs.opengroup.org/onlinepubs/9799919799.2024edition/basedefs/sys_types.h.html. The normative text requiring time_t to be an integer type at least 64 bits wide. This is the edition-pinned URL.
  3. "Package time: UnixNano". Go standard library. https://pkg.go.dev/time#Time.UnixNano. A good example of a conversion-specific limit: the result is undefined before 1678 or after 2262, although time.Time itself has no such limit.
  4. "millis()". Arduino Language Reference. https://docs.arduino.cc/language-reference/en/functions/time/millis/. An uptime counter that wraps after about 50 days. It belongs to a different class of bug from calendar-epoch overflow.
  5. "Considerations in adopting RHEL 9, Chapter 11: File systems and storage". Red Hat. https://docs.redhat.com/en/documentation/red_hat_enterprise_linux/9/html/considerations_in_adopting_rhel_9/assembly_file-systems-and-storage_considerations-in-adopting-rhel-9. Confirms that RHEL 9 enables XFS bigtime by default (such file systems will not mount on RHEL 8) and that ext4 needs inodes larger than 128 bytes.
  6. "GPS Week Number Rollover". Martin Burnicki, Meinberg Knowledge Base, 2024-07-10. https://kb.meinbergglobal.com/kb/time_sync/gnss_systems/gps_week_number_rollover. A vendor explanation of fixed-pivot failures on arbitrary weekends, RTC-seeded extended week numbers, and why rollover testing needs a GPS simulator.
  7. "Update your iPhone or iPad software if you're experiencing issues with location, date, and time". Apple Support (archived). https://support.apple.com/en-us/108341. A first-party consumer-device GPS rollover advisory, with affected models, fix versions and a hard deadline of 2019-11-03.
  8. RFC 1119: "Network Time Protocol (Version 2)". David L. Mills, 1989-09. https://www.ntp.org/reflib/rfc/rfc1119/rfc1119b.pdf. Shows that the 2036 overflow and the need to tell 1900 from 2036 were documented in 1989.
  9. "Year 2000 Computing Crisis: FDIC's Efforts to Ensure Bank Systems Are Year 2000 Compliant". US GAO (T-AIMD-98-73), 1998-02-10. https://www.gao.gov/products/t-aimd-98-73. Shows how a regulator drove readiness, testing and contingency planning across thousands of supervised firms, a model for sector regulators facing 2038.
  10. "Leadership emerges as Y2K's lesson learned". Tanya Ballard Brown, Government Executive, 2000-09-15. https://www.govexec.com/technology/2000/09/leadership-emerges-as-y2ks-lesson-learned/7062/. Contemporary coverage of GAO's AIMD-00-290 lessons report. The full report PDF is at https://www.gao.gov/assets/aimd-00-290.pdf.

See also

References

  1. Year 2038 problem — Wikipedia — used to locate several primary sources listed above (InformationWeek, RFC 7530, Oracle and MySQL release notes).
  2. Time formatting and storage bugs — Wikipedia — used to locate the Wired billennium article and GPS/NTP primary sources.