Software versioning
Software versioning is the process of assigning either unique version names or unique version numbers to unique states of computer software. Within a given version number category (e.g., major or minor), these numbers are generally assigned in increasing order and correspond to new developments in the software. At a fine-grained level, revision control is used for keeping track of incrementally-different versions of information, whether or not this information is computer software, in order to be able to roll any changes back.
"Version number" redirects here. For versioning of other products, see Mark (designation).Modern computer software is often tracked using two different software versioning schemes: an internal version number that may be incremented many times in a single day, such as a revision control number, and a release version that typically changes far less often, such as semantic versioning[1] or a project code name.
History[edit]
File numbers were used especially in public administration, as well as companies, to uniquely identify files or cases. For computer files this practice was introduced for the first time with MIT's ITS file system, later the TENEX filesystem for the PDP-10 in 1972.[2]
Later lists of files including their versions were added, and dependencies amongst them. Linux distributions like Debian, with its dpkg, early on created package management software which could resolve dependencies between their packages. Debian's first try was that a package knew other packages which depended on it. From 1994 on this idea was inverted, so a package that knew the packages it needed. When installing a package, dependency resolution was used to automatically calculate the packages needed as well, and install them with the desired package. To facilitate upgrades, minimum package versions were introduced. Thus the numbering scheme needed to tell which version was newer than the required one.[3][4][5]
Software examples[edit]
Python[edit]
The Python Software Foundation has published PEP 440 – Version Identification and Dependency Specification,[23] outlining their own flexible scheme, that defines an epoch segment, a release segment, pre-release and post-release segments and a development release segment.
TeX[edit]
TeX has an idiosyncratic version numbering system, an unusual feature invented by its developer Donald Knuth. Since version 3.1, updates have been indicated by adding an extra digit at the end, so that the version number asymptotically approaches the number π. (This is a form of unary numbering; the version number is the number of digits.) As of February 2021, the version number is 3.141592653. This is a reflection of TeX being very stable, and only minor updates are anticipated. TeX developer Donald Knuth has stated that the "absolutely final change (to be made after [his] death)" will be to change the version number to π, at which point all remaining bugs will become permanent features.[24]
In a similar way, the version number of Metafont asymptotically approaches Euler's number, e.[24] As of February 2021, the version number is 2.71828182. Metafont was also devised by Donald Knuth as a companion to his TeX typesetting system.
Apple[edit]
During the era of the classic Mac OS, minor version numbers rarely went beyond ".1". When they did, they usually jumped straight to ".5", suggesting the release was "more significant".[a] Thus, "8.5" was marketed as its own release, representing "Mac OS 8 and a half", and 8.6 effectively meant "8.5.1".
Mac OS X departed from this trend, in large part because "X" (the Roman numeral for 10) was in the name of the product. As a result, all versions of OS X began with the number 10. The first major release of OS X was given the version number 10.0, but the next major release was not 11.0. Instead, it was numbered 10.1, followed by 10.2, 10.3, and so on for each subsequent major release. Thus the 11th major version of OS X was labeled "10.10". Even though the "X" was dropped from the name as of macOS 10.12, this numbering scheme continued through macOS 10.15. Under the "X"-based versioning scheme, the third number (instead of the second) denoted a minor release, and additional updates below this level, as well as updates to a given major version of OS X coming after the release of a new major version, were titled Supplemental Updates.[25]
The Roman numeral X was concurrently leveraged for marketing purposes across multiple product lines. Both QuickTime and Final Cut Pro jumped from version 7 directly to version 10, QuickTime X and Final Cut Pro X. Like Mac OS X itself, the products were not upgrades to previous versions, but brand-new programs. As with OS X, major releases for these programs incremented the second digit and minor releases were denoted using a third digit. The "X" was dropped from Final Cut's name with the release of macOS 11.0 (see below), and QuickTime's branding became moot when the framework was deprecated in favor of AVFoundation in 2011 (the program for playing QuickTime video was only named QuickTime Player from the start).
Apple's next macOS release, provisionally numbered 10.16,[26] was officially announced as macOS 11 at WWDC in June 2020, and released in November 2020.[27] The following macOS version, macOS Monterey, was released in October 2021 and bumped its major version number to 12.[28]
Microsoft Windows[edit]
The Microsoft Windows operating system was first labelled with standard version numbers for Windows 1.0 through Windows 3.11. After this Microsoft excluded the version number from the product name. For Windows 95 (version 4.0), Windows 98 (4.10) and Windows 2000 (5.0), year of the release was included in the product title. After Windows 2000, Microsoft created the Windows Server family which continued the year-based style with a difference: For minor releases, Microsoft suffixed "R2" to the title, e.g., Windows Server 2008 R2 (version 6.1). This style had remained consistent to this date. The client versions of Windows however did not adopt a consistent style. First, they received names with arbitrary alphanumeric suffixes as with Windows Me (4.90), Windows XP (5.1), and Windows Vista (6.0). Then, once again Microsoft adopted incremental numbers in the title, but this time, they were not versioning numbers; the version numbers of Windows 7, Windows 8 and Windows 8.1 are respectively 6.1, 6.2 and 6.3. In Windows 10, the version number leaped to 10.0[29] and subsequent updates to the OS only incremented build number and update build revision (UBR) number.
The successor of Windows 10, Windows 11, was released on October 5, 2021. Despite being named "11", the new Windows release didn't bump its major version number to 11. Instead, it stayed at the same version number of 10.0, used by Windows 10.[30]
Other schemes[edit]
Some software producers use different schemes to denote releases of their software. The Debian project uses a major/minor versioning scheme for releases of its operating system but uses code names from the movie Toy Story during development to refer to stable, unstable, and testing releases.[31]
BLAG Linux and GNU features very large version numbers: major releases have numbers such as 50000 and 60000, while minor releases increase the number by 1 (e.g. 50001, 50002). Alpha and beta releases are given decimal version numbers slightly less than the major release number, such as 19999.00071 for alpha 1 of version 20000, and 29999.50000 for beta 2 of version 30000. Starting at 9001 in 2003, the most recent version as of 2011 is 140000.[32][33][34]
Urbit uses Kelvin versioning (named after the absolute Kelvin temperature scale): software versions start at a high number and count down to version 0, at which point the software is considered finished and no further modifications are made.[35][36]
Internal version numbers[edit]
Software may have an "internal" version number which differs from the version number shown in the product name (and which typically follows version numbering rules more consistently). Java SE 5.0, for example, has the internal version number of 1.5.0, and versions of Windows from NT 4 on have continued the standard numerical versions internally: Windows 2000 is NT 5.0, XP is Windows NT 5.1, Windows Server 2003 and Windows XP Professional x64 Edition are NT 5.2, Windows Server 2008 and Vista are NT 6.0, Windows Server 2008 R2 and Windows 7 are NT 6.1, Windows Server 2012 and Windows 8 are NT 6.2, and Windows Server 2012 R2 and Windows 8.1 are NT 6.3, however the first version of Windows 10 was 10.0 (10.0.10240). Note, however, that Windows NT is only on its fifth major revision, as its first release was numbered 3.1 (to match the then-current Windows release number) and the Windows 10 launching made a version leap from 6.3 to 10.0.
Pre-release versions[edit]
In conjunction with the various versioning schemes listed above, a system for denoting pre-release versions is generally used, as the program makes its way through the stages of the software release life cycle.
Programs that are in an early stage are often called "alpha" software, after the first letter in the Greek alphabet. After they mature but are not yet ready for release, they may be called "beta" software, after the second letter in the Greek alphabet. Generally alpha software is tested by developers only, while beta software is distributed for community testing.
Some systems use numerical versions less than 1 (such as 0.9), to suggest their approach toward a final "1.0" release. This is a common convention in open source software.[37][38] However, if the pre-release version is for an existing software package (e.g. version 2.5), then an "a" or "alpha" may be appended to the version number. So the alpha version of the 2.5 release might be identified as 2.5a or 2.5.a.
An alternative is to refer to pre-release versions as "release candidates", so that software packages which are soon to be released as a particular version may carry that version tag followed by "rc-#", indicating the number of the release candidate; when the final version is released, the "rc" tag is removed.
Release train[edit]
A software release train is a form of software release schedule in which a number of distinct series of versioned software releases for multiple products are released as a number of different "trains" on a regular schedule. Generally, for each product line, a number of different release trains are running at a given time, with each train moving from initial release to eventual maturity and retirement on a planned schedule. Users may experiment with a newer release train before adopting it for production, allowing them to experiment with newer, "raw", releases early, while continuing to follow the previous train's point releases for their production systems prior to moving to the new release train as it becomes mature.
Cisco's IOS software platform used a release train schedule with many distinct trains for many years. More recently, a number of other platforms including Firefox and Fenix for Android,[39] Eclipse,[40] LibreOffice,[41] Ubuntu,[42] Fedora,[43] Python,[44] digiKam[45] and VMware[46] have adopted the release train model.
Modifications to the numeric system[edit]
Odd-numbered versions for development releases[edit]
Between the 1.0 and the 2.6.x series, the Linux kernel used odd minor version numbers to denote development releases and even minor version numbers to denote stable releases. For example, Linux 2.3 was a development family of the second major design of the Linux kernel, and Linux 2.4 was the stable release family that Linux 2.3 matured into. After the minor version number in the Linux kernel is the release number, in ascending order; for example, Linux 2.4.0 → Linux 2.4.22. Since the 2004 release of the 2.6 kernel, Linux no longer uses this system, and has a much shorter release cycle.
The same odd-even system is used by some other software with long release cycles, such as Node.js up to version 0.12 as well as GNOME and WineHQ.[47]
Sun also dropped the first digit for Solaris, where Solaris 2.8 (or 2.9) is referred to as Solaris 8 (or 9) in marketing materials.
A similar jump took place with the Asterisk open-source PBX construction kit in the early 2010s, whose project leads announced that the current version 1.8.x would soon be followed by version 10.[48]
This approach, panned by many because it breaks the semantic significance of the sections of the version number, has been adopted by an increasing number of vendors including Mozilla (for Firefox).
Version number ordering systems[edit]
Version numbers very quickly evolve from simple integers (1, 2, ...) to rational numbers (2.08, 2.09, 2.10)
and then to non-numeric "numbers" such as 4:3.4.3-2. These complex version numbers are therefore better treated as character strings. Operating systems that include package management facilities (such as all non-trivial Linux or BSD distributions) will use a distribution-specific algorithm for comparing version numbers of different software packages. For example, the ordering algorithms of Red Hat and derived distributions differ to those of the Debian-like distributions.
As an example of surprising version number ordering implementation behavior, in Debian, leading zeroes are ignored in chunks, so that 5.0005 and 5.5 are considered as equal, and 5.5 < 5.0006. This can confuse users; string-matching tools may fail to find a given version number; and this can cause subtle bugs in package management if the programmers use string-indexed data structures such as version-number indexed hash tables.
To ease sorting, some software packages represent each component of the major.minor.release scheme with a fixed width. Perl represents its version numbers as a floating-point number; for example, Perl's 5.8.7 release can also be represented as 5.008007. This allows a theoretical version of 5.8.10 to be represented as 5.008010. Other software packages pack each segment into a fixed bit width; for example, on Microsoft Windows, version number 6.3.9600.16384 would be represented as hexadecimal 0x0006000325804000. The floating-point scheme breaks down if any segment of the version number exceeds 999; a packed-binary scheme employing 16 bits apiece breaks down after 65535.
Political and cultural significance of version numbers[edit]
Version 1.0 as a milestone[edit]
The free-software and open source communities tend to release software early and often. Initial versions are numbers less than 1, with these 0.x version used to convey that the software is incomplete and not reliable enough for general release or usable in its current state. Backward-incompatible changes are common with 0.x versions.
Version 1.0 is used as a major milestone, indicating that the software has at least all major features plus functions the developers wanted to get into that version, and is considered reliable enough for general release.[37][38] A good example of this is the Linux kernel, which was first released as version 0.01 in 1991,[49] and took until 1994 to reach version 1.0.0.[50]
The developers of the arcade game emulator MAME do not ever intend to release a version 1.0 of the program because there will always be more arcade games to emulate and thus the project can never be truly completed. Accordingly, version 0.99 was followed by version 0.100.[51]
Since the internet has become widespread, most commercial software vendors no longer follow the maxim that a major version should be "complete" and instead rely on patches with bugfixes to sort out the known issues which a solution has been found for and could be fixed.
Version numbers as marketing[edit]
A relatively common practice is to make major jumps in version numbers for marketing reasons. Sometimes software vendors just bypass the 1.0 release or quickly release a release with a subsequent version number because 1.0 software is considered by many customers too immature to trust with production deployments. For example, as in the case of dBase II, a product is launched with a version number that implies that it is more mature than it is.
Other times version numbers are increased to match those of competitors. This can be seen in many examples of product version numbering by Microsoft, America Online, Sun Solaris, Java Virtual Machine, SCO Unix, WordPerfect. Microsoft Access jumped from version 2.0 to version 7.0, to match the version number of Microsoft Word.
Microsoft has also been the target of "catch-up" versioning, with the Netscape browsers skipping version 5 to 6, in line with Microsoft's Internet Explorer, but also because the Mozilla application suite inherited version 5 in its user agent string during pre-1.0 development and Netscape 6.x was built upon Mozilla's code base.
Another example of keeping up with competitors is when Slackware Linux jumped from version 4 to version 7 in 1999.[52]
Overcoming perceived marketing difficulties[edit]
In the mid-1990s, the rapidly growing CMMS, Maximo, moved from Maximo Series 3 directly to Series 5, skipping Series 4 due to that number's perceived marketing difficulties in the Chinese market, where the number 4 is associated with "death" (see tetraphobia). This did not stop Maximo Series 5 version 4.0 from being released. (The "Series" versioning has since been dropped, effectively resetting version numbers after Series 5 version 1.0's release.)
Significance[edit]
In software engineering[edit]
Version numbers are used in practical terms by the consumer, or client, to identify or compare their copy of the software product against another copy, such as the newest version released by the developer. For the programmer or company, versioning is often used on a revision-by-revision basis, where individual parts of the software are compared and contrasted with newer or older revisions of those same parts, often in a collaborative version control system.
In the 21st century, more programmers started to use a formalized version policy, such as the semantic versioning policy.[1] The purpose of such policies is to make it easier for other programmers to know when code changes are likely to break things they have written. Such policies are especially important for software libraries and frameworks, but may also be very useful to follow for command-line applications (which may be called from other applications) and indeed any other applications (which may be scripted and/or extended by third parties).
Versioning is also a required practice to enable many schemes of patching and upgrading software, especially to automatically decide what and where to upgrade to.
In technical support[edit]
Version numbers allow people providing support to ascertain exactly which code a user is running, so that they can rule out bugs that have already been fixed as a cause of an issue, and the like. This is especially important when a program has a substantial user community, especially when that community is large enough that the people providing technical support are not the people who wrote the code. The semantic meaning[1] of version.revision.change style numbering is also important to information technology staff, who often use it to determine how much attention and research they need to pay to a new release before deploying it in their facility. As a rule of thumb, the bigger the changes, the larger the chances that something might break (although examining the Changelog, if any, may reveal only superficial or irrelevant changes). This is one reason for some of the distaste expressed in the "drop the major release" approach taken by Asterisk et alia: now, staff must (or at least should) do a full regression test for every update.
Non-software use[edit]
Some computer file systems, such as the OpenVMS Filesystem, also keep versions for files. Versioning amongst documents is relatively similar to the routine used with computers and software engineering, where with each small change in the structure, contents, or conditions, the version number is incremented by 1, or a smaller or larger value, again depending on the personal preference of the author and the size or importance of changes made.
Software-style version numbers can be found in other media.
In some cases, the use is a direct analogy (for example: Jackass 2.5, a version of Jackass Number Two with additional special features; the second album by Garbage, titled Version 2.0; or Dungeons & Dragons 3.5, where the rules were revised from the third edition, but not so much as to be considered the fourth).
More often it's used to play on an association with high technology, and doesn't literally indicate a 'version' (e.g., Tron 2.0, a video game followup to the film Tron, or the television series The IT Crowd, which refers to the second season as Version 2.0). A particularly notable usage is Web 2.0, referring to websites from the early 2000s that emphasized user-generated content, usability and interoperability.
Technical drawing and CAD software files may also use some of kind primitive versioning number to keep track of changes.