ARC (file format)
ARC (file format)
Main page
A

ARC (file format)

logo
Community Hub0 subscribers
Read side by side
ARC (file format)
View on Wikipedia
from Wikipedia
ARC
Filename extension
.arc, .ark
Internet media type
application/octet-stream
Uniform Type Identifier (UTI)public.archive.arc
Developed bySystem Enhancement Associates
Type of formatData compression

ARC is a lossless data compression and archival format by System Enhancement Associates (SEA). The file format and the program were both called ARC. The format is known as the subject of controversy in the 1980s, part of important debates over what would later be known as open formats.

ARC was extremely popular during the early days of the dial-up BBS. ARC was convenient as it combined the functions of the SQ program to compress files and the LU program to create .LBR archives of multiple files. The format was later replaced by the ZIP format, which offered better compression ratios and the ability to retain directory structures through the compression/decompression process.

The .arc filename extension is often used for several unrelated file archive-like file types. For example, the Internet Archive used its own ARC format to store multiple web resources into a single file.[1][2] The FreeArc archiver also uses a .arc extension, but uses a completely different file format. Nintendo uses an unrelated "ARC" format for resources, such as audio, or text, in GameCube and Wii games. Several unofficial extractors exist for this type of ARC file.[which?][citation needed]

History

[edit]

In 1985, Thom Henderson of System Enhancement Associates wrote a program called ARC,[3] based on earlier programs such as ar, that not only grouped files into a single archive file but also compressed them to save disk space, a feature of great importance on early personal computers, where space was very limited and modem transmission speeds were very slow. The archive files produced by ARC had file names ending in ".ARC" and were thus sometimes called "arc files".

The source code for ARC was released by SEA in 1986 and subsequently ported to Unix and Atari ST in 1987 by Howard Chu. This more portable codebase was subsequently ported to other platforms, including VAX/VMS and IBM System/370 mainframes. Howard's work was also the first to disprove the prevalent belief that Lempel-Ziv encoded files could not be further compressed. Additional compression could be achieved by using Huffman coding on the LZW data, and Howard's version of ARC was the first program to demonstrate this property. This hybrid technique was later used in several other compression schemes by Phil Katz and others.

Later, Phil Katz developed his own shareware utilities, PKARC and PKXARC, to create archive files and extract their contents. These files worked with the archive file format used by ARC and were significantly faster than ARC on the IBM-PC platform due to selective assembly-language coding. Unlike SEA, which combined archive creation and archive file extraction in a single program, Katz divided these functions between two separate utilities, reducing the amount of memory needed to run them. PKARC also allowed the creation of self-extracting archives, which could unpack themselves without requiring an external file extraction utility.

Following the System Enhancement Associates, Inc. vs PKWARE Inc. and Phillip W. Katz lawsuit, SEA withdrew from the shareware market and developed ARC+Plus.[4] This version included a full-screen user interface, with the last known version being 7.12.[5] SEA was eventually sold to an unspecified Japanese company in 1992.[6]

The ARC format is no longer common on PC desktops, but most antivirus scanners can still uncompress any ARC archives found in order to detect viruses within the compressed files.

Lawsuits

[edit]

In the late 1980s a dispute arose between SEA, maker of the ARC program, and PKWARE, Inc. (Phil Katz Software). SEA sued Katz for trademark and copyright infringement. An independent software expert, John Navas, was appointed by the court to compare the two programs, and stated that PKARC was a derivative work of ARC, pointing out that comments in both programs were often identical, including spelling errors.[7]

On August 2, 1988, the plaintiff and defendants announced a settlement of the lawsuit, which included a Confidential Cross-License Agreement under which SEA licensed PKWARE for all the ARC-compatible programs published by PKWARE during the period beginning with the first release of PKXARC in late 1985 through July 31, 1988, in return for $62,500, which at the time was an undisclosed payment amount. In the agreement, PKWARE paid SEA to obtain a license that allowed the distribution of PKWARE's ARC-compatible programs until January 31, 1989, after which PKWARE would not license, publish or distribute any ARC compatible programs or utilities that process ARC compatible files. In exchange, PKWARE licensed SEA to use its source code for PKWARE's ARC-compatible programs. PKWARE also agreed to cease any use of SEA's trademark "ARC" and to change the names or marks used with PKWARE's programs to non-confusing designations. The remaining details of the agreement were sealed. In reaching the settlement, the defendants did not admit any fault or wrongdoing.[8] The Wisconsin court order showed defendants were ordered to pay damages to plaintiff for defendants' acts of infringing Plaintiff's copyrights, trademark, and acts of unfair trade practices and unfair competition.[9]

The leaked agreement document revealed under the settlement terms, the defendants had paid plaintiff $22,500 for past royalty payments, and $40,000 for expense reimbursements. In addition, defendants would pay plaintiff a royalty fee of 6.5% of all revenue received for ARC compatible programs on all orders received after the effective date of this Agreement, such revenue including any license fees or shareware registrations received after the expiration of the license, for ARC compatible programs. In exchange, plaintiff would also pay a commission in the amount of 6.5% of any license fees received by plaintiff from any licensee referred to plaintiff by defendants, whether before or after the license termination date.[10]

After the lawsuit, PKWARE released one last version of his PKARC and PKXARC utilities under the new names "PKPAK" and "PKUNPAK", and from then on concentrated on developing the separate programs PKZIP and PKUNZIP, which were based on new and different file compression techniques and archive file formats. However, following the renaming, SEA filed a lawsuit against PKWARE for contempt, for continually using plaintiff's protected mark ARC, by turning ARC from noun into verb in the PKPAK manual.[11] The U.S. district court of the East District of Wisconsin ruled SEA's motion was denied, and the defendant was entitled to recover the legal cost of $500.[12]

The SEA vs. PKWARE dispute quickly expanded into one of the largest controversies the BBS world ever saw.[13] The suit by SEA angered many shareware users who perceived that SEA was a "large, faceless corporation" and Katz was "the little guy". In fact, at the time, both SEA and PKWARE were small home-based companies. However, the community largely sided with Katz, due to the fact that SEA was attempting to retroactively declare the ARC file format to be closed and proprietary. Katz received positive publicity by releasing the APPNOTE.TXT specification documenting the ZIP file format, and declaring that the ZIP file format would always be free for competing software to implement. The net result was that the ARC format quickly dropped out of common use as the predominant compression format that PC-BBSs used for their file archives, and after a brief period of competing formats, the ZIP format was adopted as the predominant standard.

In an interview, Thom Henderson of SEA said that the main reason he dropped out of software development was because of his inability to emotionally cope with what he claimed was the hate-mail campaign launched against him by Katz.[14]

See also

[edit]

References

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

ARC (file format)

View on Grokipedia
from Grokipedia
ARC is a lossless data compression and multi-file archival format developed by Thom Henderson of System Enhancement Associates (SEA) in 1985 as part of the ARC utility program for MS-DOS and CP/M systems.[1][2] The format combines files into a single archive while applying various compression algorithms, including adaptive Huffman coding, Lempel-Ziv-Welch (LZW), and packing methods, to reduce storage needs and facilitate file distribution over early networks like bulletin board systems (BBS).[3][4] Initially shareware, ARC gained widespread adoption in the mid-1980s for its efficiency in handling executables and text files on limited floppy disk storage, predating more standardized formats like ZIP.[5] However, its prominence waned after legal disputes: in 1988, SEA initiated lawsuits against Phil Katz and PKWARE, alleging trademark infringement and unauthorized replication of ARC's compression output in the competing PKARC and PKZIP tools, which initially supported reading and writing ARC files.[6] The conflicts, including claims of reverse-engineering proprietary packing algorithms, prompted PKWARE to develop the incompatible ZIP format, which eventually supplanted ARC due to broader compatibility and avoidance of litigation risks.[1] Modern support for ARC remains limited to legacy tools and emulators, underscoring its role as a foundational but transitional technology in computing history.[3]

History

Origins and Development

The ARC file format originated in 1985, when Thom Henderson, founder of System Enhancement Associates (SEA), developed the ARC utility program for MS-DOS systems.[5][1] This shareware tool combined file archiving with data compression, addressing limitations of earlier standalone utilities like SQ (Squeeze) for packing and basic archivers for grouping files.[7] ARC's initial implementation employed Lempel–Ziv–Welch (LZW) compression, an adaptive dictionary-based algorithm that improved efficiency over prior methods, making it suitable for distributing software via limited-bandwidth dial-up bulletin board systems (BBS).[7][2] SEA marketed ARC as a commercial product while distributing it under shareware terms, requiring users to register for full features after a trial period.[5] The format quickly gained prominence in the BBS community from 1985 to 1989, becoming the dominant standard for compressing and archiving files due to its balance of compression ratios and extraction speed on era hardware.[3] Early versions supported multiple packing methods, including a "trimmed" Huffman coding variant for enhanced performance on repetitive data.[8] Development progressed through iterative releases, with SEA issuing updates that refined the file header structure for better metadata handling, such as file dates, sizes, and CRC checksums for integrity verification.[1] By late 1989, version 7.x introduced optimized compression schemes, though the core format remained backward-compatible to maintain interoperability across SEA's ecosystem.[8] These enhancements solidified ARC's role as a precursor to later formats, influencing subsequent archivers amid growing demand for efficient file distribution in pre-internet computing.[2]

Evolution and Updates

The ARC utility, initially released in 1985 by System Enhancement Associates (SEA), saw iterative updates that primarily enhanced compression efficiency through new algorithms while maintaining core file structure compatibility. Early versions supported basic methods such as "packing" for uncompressed or lightly reduced storage and "squeezing," an LZW variant for dictionary-based compression. By version 6.02, released on October 3, 1988, the program included refinements to existing schemes like "crunching," which combined statistical modeling with Huffman coding for better ratios on certain data types.[9] Version 6.00, documented in January 1989, formalized these capabilities in an updated manual, emphasizing multi-file archiving with optional password protection and self-extracting variants.[10] The final major release, version 7.x (e.g., ARC712.EXE), emerged in late 1989 or early 1990 as part of the ARC+Plus package, introducing "Trimmed" (method ID 10) as the default compression. This scheme integrated RLE90 run-length encoding to preprocess repeats, a 4KB LZ77 history buffer for match distances up to 4096, and adaptive Huffman coding on 314 symbols (including literal bytes and length codes from 3 to 59), yielding superior performance over prior methods without altering the overarching header format.[8] Subsequent development ceased amid competition from PKZIP, with no format-altering updates beyond v7; later efforts like SuperARC variants for platforms such as Atari 8-bit were derivatives based on v5.0 code rather than official SEA evolutions.[9] The format's flexibility in per-file method selection (via one-letter codes in headers) allowed backward compatibility, but its evolution prioritized algorithmic gains over structural overhauls.[8]

Technical Specifications

File Structure and Headers

The ARC file format employs a sequential structure without a central directory, consisting of one or more archive entries (members) followed by an end-of-archive marker comprising the bytes 0x1A 0x00.[11] Each entry begins with a fixed 26-byte header that encapsulates metadata for the stored file, including its compression details and attributes.[12] The absence of a global header or index requires extraction tools to parse entries linearly from the file's start. The header commences with the signature byte 0x1A at offset 0x00, serving as the entry delimiter, immediately followed at offset 0x01 by a single byte denoting the compression method—common values include 0x03 for packing (run-length encoding), 0x04 for squeezing (RLE combined with Huffman coding), and 0x08 for crunching (RLE with LZW dictionary compression up to 12-bit codes).[4][12] Offsets 0x02 through 0x0D hold the filename as 12 ASCII characters, typically padded with nulls or spaces if shorter, accommodating MS-DOS 8.3 naming conventions with an implicit extension period.[12] The compressed data length follows as a 4-byte little-endian unsigned integer at offsets 0x0E to 0x11, succeeded by a 4-byte MS-DOS timestamp (dword) at 0x12 to 0x15—encoding year (offset from 1980), month, day, hour, minute, and seconds/2 in bit-packed format. A 16-bit CRC checksum of the compressed data occupies offsets 0x16 to 0x17 (using polynomial 0x1021 or equivalent for validation), and the original uncompressed file size concludes the header as another 4-byte little-endian value at 0x18 to 0x1B.[4][12]
Offset (hex)Size (bytes)FieldDescription
001SignatureFixed value 0x1A marking entry start.[12]
011Compression methodByte ID for algorithm (e.g., 0x08 for LZW-based crunching).[4][12]
02–0D12FilenameASCII name, padded if necessary.[12]
0E–114Compressed sizeLittle-endian length of following data.[12]
12–154MS-DOS timestampBit-packed date/time (year-1980:7 bits, month:4, day:5, hour:5, min:6, sec/2:5).[12]
16–172CRC-16Checksum of compressed data (polynomial 0x1021).[4]
18–1B4Original sizeLittle-endian uncompressed file length.[12]
Subsequent to each header lies the compressed file data, whose length matches the specified compressed size; informational or extended records (e.g., comments or OS-specific attributes in ARC 6.0 implementations) may append in some variants but are not universally present.[12] Later ARC versions introduced subtype-prefixed informational items (e.g., type 0x14 for archive details), prefixed by a 2-byte length and 1-byte subtype, though core structure remained backward-compatible with early releases from 1985.[12] Self-extracting archives prepend executable code before the first entry, while multi-part archives concatenate without additional markers.[11]

Compression Algorithms

The ARC file format supports multiple lossless compression algorithms, applied independently to each archived file to minimize overall size, with the method selected based on empirical testing for the best ratio. The compression type is specified by an 8-bit identifier in the file header, allowing flexibility across different data types. Early versions of the ARC utility, released by System Enhancement Associates starting in 1985, relied on simpler techniques like run-length encoding and Huffman coding, while later iterations from 1989 onward incorporated dictionary-based methods inspired by LZ77 variants.[4][8] One foundational method, known as "Packing" (compression type 3), uses run-length encoding (RLE) with 0x90 as a special marker: a literal 0x90 is encoded as 0x90 followed by 0x00, while repeats of the prior byte are encoded as 0x90 followed by the count (up to 255 additional occurrences). This pre-processing step reduces redundancy in repetitive data but offers limited compression for diverse content.[4] "Squeezing" (type 4) builds on RLE by applying static Huffman coding to the encoded stream, where the code tree is prefixed with a 16-bit node count followed by an array of 16-bit node indices (negative values indicating leaf symbols). Decoding proceeds bit-by-bit in least-significant-bit order, making it suitable for files with skewed symbol distributions post-RLE.[4] "Crunched" (type 8) extends RLE with Lempel-Ziv-Welch (LZW) dictionary compression using an 8-kilobyte buffer, starting with 9-bit codes that dynamically expand to 12 bits; a reset code (0x0100) clears the dictionary when full, with bit widths adjusting accordingly. This approach, akin to early adaptive dictionary schemes, improved ratios for textual and structured data but was computationally intensive for the era's hardware.[4][13] Advanced methods in extensions like PAK (version 2.0, July 1989) include "Distilled," which integrates LZ77 sliding-window matching (8KB buffer, up to 60-byte matches, supporting prehistory offsets) with dual Huffman codebooks: a dynamic one for literals (0-255), match lengths (3-60), and stop codes, plus a fixed one for offsets derived from decompressed byte contexts. The stream is bit-packed, with offsets computed variably (0-7 extra bits based on recent data).[14] "Trimmed," default in ARC version 7 (late 1989), preprocesses with RLE90 before LZ77 (4KB buffer, matches 3-59 bytes) encoded via adaptive Huffman trees over 314 symbols (literals plus length codes, with 256 as stop). This hybrid yielded competitive ratios, outperforming simpler modes on mixed workloads while remaining decode-efficient.[8]

Software and Compatibility

Original SEA ARC Program

The original SEA ARC program was developed by Thom Henderson and released in 1985 by System Enhancement Associates (SEA) as shareware software primarily for MS-DOS systems on IBM PCs.[15][5] Written in C using the C86 compiler, it enabled users to bundle multiple files into a single .arc archive while applying lossless compression to reduce file sizes, facilitating efficient storage and transfer over dial-up connections.[5] The program quickly gained traction among operators of electronic bulletin board systems (BBSes), which dominated pre-Internet file sharing, as it significantly decreased download times for software distributions—for instance, version 5.12 compressed an MS-DOS 3.20 installation archive by over 20%.[15][5] Key features included command-line operations such as adding files (a), extracting (e), listing contents (l), and deleting entries (d), with support for multiple compression methods indicated by a method byte in the file header, including storage (method 0), packing (method 1), and squeezing (method 2) for varying data types.[1] Later updates introduced additional algorithms like crunching (method 3) and squashing (method 4), alongside optional password protection for archives, though forgotten passwords could render files irretrievable without recovery tools.[5][1] SEA released the program's source code in 1986, promoting ports to platforms like Unix and Atari ST by developers such as Howard Chu in 1987, which extended its utility beyond DOS environments.[15][1] By 1989, ARC had evolved to version 6.00, incorporating refinements for better compatibility and performance, but the core program remained a staple for archival tasks until displaced by faster alternatives.[10] Its design emphasized simplicity and effectiveness for resource-constrained 1980s hardware, supporting features like file commenting and multi-volume archives, though it lacked native encryption beyond basic passwords.[1] The program's shareware model, requiring registration for full access, contributed to its widespread adoption while generating revenue for SEA.[5]

Third-Party Implementations

PKARC, developed by Phil Katz in 1987, was a prominent third-party implementation designed for compatibility with the ARC format, offering faster compression and extraction speeds than the original SEA software.[16] It supported the same file structure and compression methods, allowing users to create and manipulate .ARC archives, but its release prompted a copyright infringement lawsuit from SEA alleging unauthorized code duplication.[17] Other contemporary third-party tools, such as QUARK and SQUASH, also utilized the ARC format for archiving and compression on MS-DOS systems, extending compatibility without direct affiliation to SEA.[17] These implementations maintained the core header and packing mechanisms of ARC, enabling interoperability in the BBS and shareware ecosystems of the late 1980s. In modern contexts, compatibility is primarily through extraction tools rather than full read-write support, as the format's proprietary elements limited widespread adoption post-lawsuit. IZArc, a free Windows utility, can decompress legacy ARC files alongside numerous other formats.[18] Command-line tools like unarc, available on Linux distributions, provide extraction capabilities for older ARC archives by parsing the 1A header and file entries.[19] These tools focus on read access for archival preservation, reflecting the format's obsolescence for new creations since the early 1990s.[20]

Modern Tools and Emulation

IZArc, a free Windows-based compression utility, supports extraction and creation of files in the legacy SEA ARC format, allowing users to handle these archives on modern operating systems without requiring DOS emulation.[18] Other tools, such as certain Linux command-line utilities like the arc package, provide compatibility for unpacking older ARC files, primarily for archival and retro computing purposes.[21] For scenarios where native support is absent or incomplete, DOS emulators enable execution of the original SEA ARC program (ARC.EXE) from the 1980s. DOSBox, a widely used x86 emulator, runs the utility accurately on Windows, Linux, and macOS, facilitating direct extraction via commands like ARC e filename.arc in an emulated MS-DOS environment.[22] Enhanced variants like DOSBox-X offer improved compatibility for period-specific behaviors, such as handling compression algorithms like LZW or Huffman coding inherent to early ARC versions.[23] Online platforms like PCjs provide browser-based emulation of IBM PC hardware, permitting users to upload and process ARC files interactively without local installation, preserving access to self-extracting archives and multi-volume sets.[5] These emulation approaches maintain fidelity to the format's original specifications, including variable-length headers and packing methods, though they may require sourcing authentic ARC executables from trusted retro archives to avoid compatibility issues with modified versions.[5]

Legal Controversies

Dispute with PKWARE and Phil Katz

In April 1988, System Enhancement Associates (SEA), the developers of the ARC archiving utility, filed a lawsuit against PKWARE Inc. and its founder, Phil Katz, accusing them of copyright infringement, trademark violation, and misappropriation of the ARC program's user interface and "look and feel" through Katz's release of PKARC in 1987.[6] SEA alleged that PKARC, which provided faster compression and extraction while remaining compatible with ARC files, incorporated substantial portions of ARC's proprietary source code rather than being a clean reverse-engineered implementation.[24] Katz maintained that he had disassembled ARC binaries to study its algorithms legally, denying direct code copying, but forensic examination of PKARC revealed structural and functional similarities beyond what independent reimplementation would typically produce.[25] The conflict stemmed from PKARC's rapid adoption in the bulletin board system (BBS) community, where it undercut SEA's market dominance by offering superior speed without requiring users to abandon existing ARC archives.[26] SEA, a small firm led by Thom Henderson, viewed the software as a derivative work infringing on their intellectual property, particularly since ARC's compression methods and file header structures were central to its functionality.[6] Katz's approach, while innovative in optimizing algorithms like LZW-based packing, blurred the line between permissible reverse engineering and unauthorized replication, a distinction pivotal in early software copyright disputes.[27] Prior to full resolution, the parties entered a confidential cross-license agreement on July 31, 1988, allowing temporary distribution of PKARC under SEA oversight, but escalating tensions led to contempt motions by August.[28] The broader dispute highlighted tensions in the shareware ecosystem of the late 1980s, where proprietary formats competed fiercely amid limited standardization, ultimately pressuring Katz to pivot away from ARC compatibility.[29]

Lawsuit Details and Claims

In April 1988, System Enhancement Associates (SEA), the developer of the ARC archiving program and format, filed a lawsuit against Phil Katz and his company, PKWARE, Inc., in federal court alleging multiple forms of intellectual property infringement related to Katz's PKARC utility.[6][30] SEA specifically claimed that Katz had engaged in trademark infringement by using the "ARC" designation, which SEA asserted as its proprietary mark for the file format and associated software.[30] Additionally, SEA accused PKWARE of copyright violations, contending that PKARC appropriated the "look and feel" of the ARC program, including its command-line interface, and directly copied portions of SEA's source code without permission, transforming a shareware tool into a commercial product.[6][30] The complaint further detailed that Katz had reverse-engineered the ARC program but exceeded fair use by lifting code elements, such as compression algorithms and file handling routines, to create a faster version in assembly language that competed directly with SEA's offering.[6] SEA argued this not only undermined their market position in the burgeoning BBS and shareware ecosystem but also violated the shareware distribution model by leveraging non-public code for profit.[31] In support of these claims, SEA sought comprehensive remedies, including impoundment of all PKARC and PKXARC copies, an injunction barring further sales or distribution by Katz, recovery of all profits derived from the programs, treble damages for willful infringement, maximum statutory damages under copyright law, and reimbursement of attorney fees.[30] The lawsuit highlighted early tensions in software compatibility and reverse engineering, with SEA positioning PKARC as an unauthorized derivative that eroded incentives for original development in file archiving tools.[6] While Katz maintained that PKARC was a clean-room implementation based on public disassembly of ARC files, SEA's verified complaint emphasized evidentiary links to code similarities and unauthorized commercial exploitation.[6]

Settlement Outcomes

In 1989, System Enhancement Associates (SEA) and PKWARE reached an out-of-court settlement resolving the copyright, trademark, and trade secret claims related to PKARC's development and distribution. PKWARE agreed to immediately cease selling and distributing PKARC, its ARC-compatible archiving utility, and Phil Katz committed to never again producing software compatible with the ARC format. Katz also covered SEA's legal fees as part of the agreement, amid his inability to sustain prolonged litigation expenses.[6][32][33] The settlement did not result in a formal admission of code infringement by Katz, whose PKARC had allegedly incorporated reverse-engineered elements from ARC's compression algorithms, including Lempel-Ziv variants. However, the terms effectively ended PKWARE's involvement with ARC-derived technology, shifting focus to independent development. No broader licensing of ARC to PKWARE occurred, as Katz rejected such proposals during negotiations.[31][34] This resolution accelerated the decline of ARC's market dominance, as PKWARE pivoted to create the ZIP format—initially released via PKZIP version 1.0 on April 24, 1989—as a non-compatible alternative emphasizing faster compression and broader usability. ZIP's shareware model and algorithmic improvements, such as deflate, facilitated its rapid adoption over ARC, which SEA continued supporting but struggled to evolve amid the dispute's fallout.[29][6]

Impact and Legacy

Influence on Archiving Standards

The ARC format, developed by System Enhancement Associates in 1985, established early conventions for combining lossless data compression with multi-file archival storage, particularly on MS-DOS systems and bulletin board services (BBS).[3] It introduced structured headers containing file metadata—including names, original and compressed sizes, timestamps, and cyclic redundancy checks (CRCs)—followed by compressed data blocks, which facilitated reliable extraction and verification.[3] This approach addressed the limitations of prior single-file compression tools, enabling efficient distribution of software and data bundles, and set a precedent for subsequent formats by prioritizing interoperability in unpackaging archived content.[5] ARC's market dominance from 1985 to 1989 spurred competitive innovations, most notably influencing Phil Katz's creation of PKARC as a reverse-engineered compatible tool, which exposed demands for enhanced compression ratios and licensing flexibility.[35] Katz's subsequent ZIP format, released in 1989 via PKZIP 1.0, retained ARC-like header designs and multi-algorithm support (e.g., adapting elements of ARC's "Reduce" and "Squeeze" methods) while introducing deflate compression for superior performance.[36] This evolution transformed ZIP into the de facto standard, as evidenced by its integration into Windows since version 98 and widespread adoption in archiving tools, effectively supplanting ARC's direct use but perpetuating its foundational principles of metadata-driven, verifiable archives.[37] Although ARC lacked formal standardization efforts—unlike ZIP's later documentation in PKWARE's APPNOTE specification—its role in normalizing compressed archiving influenced institutional practices, such as BBS file libraries and early internet distribution, where compatibility with ARC unpackers became a baseline expectation until ZIP's ascendancy.[3] The format's emphasis on modular compression (supporting interchangeable algorithms like LZW variants) prefigured modern standards like those in TAR/GZIP combinations, though ZIP's public-domain dedication and legal resolutions accelerated broader industry convergence on open, extensible archival structures.[35]

Criticisms and Achievements

The ARC format achieved prominence as one of the earliest systems to integrate lossless data compression with multi-file archiving in a unified process, enabling efficient bundling and size reduction of files for distribution over bulletin board systems in the mid-1980s.[38] Introduced in 1985 by System Enhancement Associates, it supported selectable compression methods—including packing (run-length encoding variants), squeezing (dictionary-based), crunching (LZW-like), and others—tailored to different data types, which allowed users to optimize for specific workloads like executable files common in shareware.[12] These features, combined with 16-bit CRC checksums for each file, improved data reliability during transfers on slow, error-prone dial-up connections, contributing to ARC's widespread adoption in the MS-DOS ecosystem before standardized alternatives emerged.[12][38] Criticisms of ARC center on its technical limitations relative to subsequent formats, particularly weaker compression ratios achieved through its fixed set of algorithms, which proved less effective than LZW-based methods in competitors like early ZIP implementations for diverse file types.[39] The format's header structure, while compact (starting with a 1Ah identifier followed by method code, filename, sizes, date, and CRC), lacked support for modern necessities such as encryption, archive splitting, or Unicode filenames, restricting its scalability for larger datasets or secure applications.[12] By the late 1980s, these shortcomings—exacerbated by the proprietary nature of its full implementation—led to rapid supersession by more adaptable, publicly documented rivals, rendering ARC inefficient for ongoing use in evolving computing environments.[12][38]

Current Relevance

The ARC format maintains niche relevance in digital preservation and retrocomputing, where tools enable extraction of legacy archives from MS-DOS-era software distributions. Utilities such as IZArc continue to support reading ARC files, allowing users to access compressed bundles containing historical executables, documents, and data otherwise incompatible with contemporary systems.[18] This capability aids forensic analysis, emulation projects, and archival efforts by organizations maintaining vintage computing artifacts, though creation of new ARC files remains impractical due to inferior compression ratios and lack of multi-platform standardization compared to successors like ZIP.[40] In specialized contexts, such as emulating early bulletin board systems (BBS) or analyzing pre-1990s file transfers, ARC's structure—featuring support for multiple compression algorithms like LZW and Squeeze—proves useful for verifying authenticity and recovering unaltered payloads from era-specific media.[5] However, its adoption has waned since the early 1990s, with no evidence of active development or integration into modern operating systems as of 2025, rendering it obsolete for routine archiving.[41]

References

  1. ARC File Format
  2. History of Lossless Data Compression Algorithms
  3. ARC (compression format) - Just Solve the File Format Problem
  4. The ARC file format - virtualdub.org
  5. The Joys of Old ARC and ZIP Files - PCjs Machines
  6. LAWSUITS: SEA vs. PKWARE - The BBS Documentary Library
  7. http://bbsdocumentary.com/library/CONTROVERSY/LAWS...
  8. ARC's “Trimmed” compression scheme - Entropymine
  9. Down loaded ARC files... - Atari 8-Bit Computers - AtariAge Forums
  10. ARC - Atari Wiki
  11. ARC (compression format) - Just Solve the File Format Problem
  12. The ARC Archive File Format - FileFormat.Info
  13. What are these compression types in the ARC file format?
  14. ARC/PAK's “Distilled” compression scheme - Entropymine
  15. Phil Katz - ESVA.net
  16. There were a lot of encoders at the time using this general scheme ...
  17. arc - File-Extensions.org
  18. IZArc - Free Zip/Unzip Files Utility
  19. How to extract xyz.arc file in ubuntu? [duplicate]
  20. How to open ARC file (and what it is) - File.org
  21. arc man - Linux Command Library
  22. DOSBox, an x86 emulator with DOS
  23. DOSBox-X - Accurate DOS emulation for Windows, Linux, macOS ...
  24. https://www.wsj.com/articles/SB961363319756539141
  25. ARC/ZIP Perspective by Dean W. Cooper - ASCII by Jason Scott
  26. Mr. Phillip Walter Katz | IT History Society
  27. LZ compression - by Bradford Morgan White - Abort Retry Fail
  28. contempt.txt - BBS: The Documentary
  29. How ZIP Became the World's File Format – And the Bitter War Behind It
  30. SEA vs. PKWARE - Higher Intellect Vintage Wiki
  31. The Rise and Fall of a Software Star - Phil Katz Loved Code
  32. Phil Katz - Pointless Database - TheAlmightyGuru.com
  33. Behind the Wires: Tales of internet history - Tedium
  34. FEBRUARY UPDATE> MIKE'S TRAIN HOUSE DCC LAWSUIT ...
  35. Zip Files: History, Explanation and Implementation - hanshq.net
  36. Shrink, Reduce, and Implode: The Legacy Zip Compression Methods
  37. PKZip | Definition, History, & Facts - Britannica
  38. How does file compression work - PeaZip
  39. ARC File Format | .arc Extension - GeeksforGeeks
  40. ARC File Extension - How to open .arc files - FileInfo.com
  41. What Is an ARC? - Computer Hope
User Avatar
No comments yet.