Dotcom编年史 (三)：没法做一套双向链接

Footnotes#

1. Paul Jones，《The Story Behind the WWW Hypertext 91 Demo Page and UNC and me》，2013 年 5 月 29 日，博客。事后回忆，距事件二十二年。Jones 当时与 Judd Knott 共同负责北卡大学信息技术办公室的网络服务研究小组；Berners-Lee 来访在 “late Fall of 1991 on his way to San Antonio”；Jones 在来访之前打电话给 “a nearby Computer Science department”，六条意见在那通电话里；演示用 Jones 自己的 NeXT 和 Berners-Lee 带来的 “floptical drive” 上的 Hypertext ‘91 演示页，Jones 注 “NeXT pioneered a read-write optical disk in a case. No one followed”；请求经 CERN 转到 “a database in the next room” 的 WAIS 网关再回来。那位计算机系教授未具名，六条意见按 Jones 的转述，第一条原文 “Links in hypertext must be bidirectional. WWW’s are one way only.”，第五条 “It’s obvious that whoever wrote the hypertext engine doesn’t understand SGML. This HTML is done all wrong.”，末条 “he’s giving this away for free. That means it has no value at all. We have our own version and a strong licensing partner.” Jones 另记 Berners-Lee 行前 “was confident given the immediate acceptance of WWW on newsgroups and as seen in his logs that he would get a good speaking slot at the conference”，同为事后回忆。本文只用作 1991 年行家意见的一个样本，不据此认定其代表性；引文为本文自译。 ↩ ↩²

2. 牛津 1976 年毕业、用烙铁和旧电视机拼第一台计算机、Plessey 两年做分布式事务系统与消息中继、1978 年在 D.G. Nash 写智能打印机的排版软件、1980 年 6 至 12 月在 CERN 做合同软件工程师、1984 年到 CERN 做 “distributed real-time systems for scientific data acquisition and system control” 并 “designed a heterogeneous remote procedure call system”，均据 W3C 网站上的 Berners-Lee 长版履历。这是机构事后整理的履历，本文只取年份与职务，不用它证明他当时怎么想。“超文本是他业余的事”是本文据 1989 年提案“for my own private use”与履历的概括；“见识过一致性保不住”是本文的推想。 ↩

3. 会议为 ACM Hypertext ‘91，1991 年 12 月 15 至 18 日，圣安东尼奥。论文只得到海报与演示位，据 W3C 的 A Little History of the World Wide Web 1991 年 12 月条 “Presented poster and demonstration at Hypertext’91”；评审人 Mark Frisse 的说法出自他 2015 年 4 月 26 日在 Mike Caulfield 博客下的留言，事后自述，本文用其“委员会决定”“像写作系统”“争论链接是否必须双向”与 “Tim had a demo and really rocked the house” 四点；论文集目录见 DBLP，无 Web 条目，Samuelson 与 Glushko 一文在第 39 至 50 页；“反响热烈”出自 Berners-Lee 1992 年 1 月 24 日的项目通讯（www-talk 1992 年第 1 号），原文 “The W3 demonstration had an enthusiastic reception at ‘HyperText’91’ in San Antonio, Texas.”，同一节接着写 “Jean-Francois Groff presented W3 live at the Software Engineering and AI for High-Energy Physics workshop in La Londe, France. We’ve also been asked to demonstrate, as well as present a paper, at the Joint European Networking Conference (JENC3) in Innsbruck, Austria, in May.”，通讯全文未提海报。“不打算在这上头花时间”是本文的读法。SLAC 服务器与演示的关系见 第二篇。证据整理见 Hypertext ‘91 上的 WWW 论文待遇与演示。 ↩

4. Nelson 1960 年得到念头、1963 年选词、1965 年发表，据他 1999 年 8 月 20 日的 Xanadu 官方年表“1960: Ted Nelson gets the idea of hypertext. (Word chosen 1963, published 1965.)”，事后自述。1965 年论文 “Complex Information Processing: A File Structure for the Complex, the Changing, and the Indeterminate”，Proceedings of the ACM 20th National Conference，1965 年，第 84 至 100 页，本文只核题名与出处。“还没有做出来”以 1991 年 Samuelson 与 Glushko 一文称其正由 Autodesk 商业化、1992 年 CERN 页面称代码“随时”发布为据。1960 至 1988 年的来龙去脉见 Dotcom 编年史 - 番外二 代码随时发布。 ↩

5. 这一段是本文为读者写的背景，不出自某一封信。Bush 的“路径”（trails）见其 1945 年《As We May Think》，Berners-Lee 等 1992 年春的论文“The Dream”一节即由此起笔（见 universe 脚注）；反向追踪与完整性出自 Kappe 1992 年 6 月 25 日的信（见 demands 脚注）；评审对“引用完整”的关切出自 Frisse 2015 年的留言（见 ht91 脚注）；引用、版税建在链接上，出自 Samuelson 与 Glushko 对 Xanadu 的转述（见 samuelson 脚注）。“唯一的结构”与“地基”是本文的说法。 ↩

6. Tim Berners-Lee，《The ENQUIRE System》手册，1980 年 10 月，W3C 保存的人工转录本。§3 “when altering one part of the system, it is useful to know by which other parts it is used”；§5.2 列出关系类型，成对标为 “Complementary”：uses／used-by，includes／part-of，made／made-by，describes／described-by；同节 “person - those strange unreliable beings on which all systems eventually depend”。手册没有说明反向的那条关系是自动生成还是编辑时手工填入，本文只说“成对”，不说“自动双向”。“正是 Kappe 十二年后要的反向追踪”是本文的对照。1980 年只有他一个人用、从未发表、“not realising that a term had already been coined for the idea: ‘hypertext’”，见 Information Management: A Proposal，1989 年 3 月，“Personal Experience with Hypertext” 一节；六个月合同工的时间（1980 年 6 至 12 月）见 bio 脚注。 ↩

7. Pamela Samuelson 与 Robert J. Glushko，《Intellectual Property Rights for Digital Library and Hypertext Publishing Systems: An Analysis of Xanadu》，Hypertext ‘91 Proceedings 第 39 至 50 页，1991 年 12 月；Glushko 存档的 PDF。同期文献，但是法学与工程角度对 Nelson 设想的转述，不是 Nelson 本人的文字。本文用其摘要与第 41 至 43 页：“By far the most ambitious proposal for a digital library and hypertext publishing environment is Ted Nelson’s Xanadu system”“a vast digital library containing all of the world’s literature”“writing environment, publishing environment, library, and bookstore in one”；按字节收费、一到两成归作者；衍生文档与链接作为作品的版税；已发表文档不可撤回；发表需签字；“Xanadu is being commercialized by Autodesk”；“some aspects of the Xanadu model of author and user behavior may be unworkable”。“引用嵌入”是本文对“compound documents consisting of portions of a number of different documents”的概括，Nelson 自己的术语是 transclusion，该文未用此词。 ↩

8. Tim Berners-Lee、Robert Cailliau、Jean-François Groff、Bernd Pollermann，《World-Wide Web: The Information Universe》，Electronic Networking 第 1 卷第 2 期，1992 年春。“The Dream”一节起于 “Since Vannevar Bush’s article, men have dreamed of extending their intellect…”，有 “The Xanadu system is an ambitious distributed hypertext project”；“W3 Architecture”一节问 “why a global system has not already come into existence”，答以缺少 “A common naming scheme for documents / Common network access protocols / Common data formats for hypertext”，并有 “Most research in hypertext systems (the Xanadu project excepted) has focussed on the user interface and authoring questions, rather than the questions of wide-area and long-term distribution.” 1989 年提案对 Nelson 与 docuverse 的引用见 第一篇。 ↩

9. Jean-François Groff，《Re: Raisch’s Attention link》，1992 年 6 月 30 日（www-talk 1992 年第 129 号），原话 “a read-based charging scheme (which was to be implemented in Xanadu as an answer to the Copyright issues)”“unfortunately not available online to my knowledge”；“付费链接”指 Raisch 提出、Jim Davis 讨论的 attention link。Berners-Lee，《HTTP2: caching and copyright》，1993 年 1 月 8 日（1993 年第一季度第 22 号），原话 “Have you read ‘Litterary Machines?’ That goes into this in a lot of detail, or at least Xanadu should have done.” Groff 说的“据我所知网上没有”指那篇论文不在网上，不指 Xanadu 软件；“只存在于纸上”是本文据 Groff 的 “was to be implemented” 与 CERN 页面 11 月仍称代码 “due out ‘any time now’” 作出的判断。Berners-Lee 原句 “That goes into this in a lot of detail, or at least Xanadu should have done.”，正文译作“或者说，至少 Xanadu 本该把这些做细”。 ↩ ↩²

10. Autodesk 1992 年 8 月 20 日新闻稿，收于 John Walker 编《The Autodesk File》第五版，网络版 Farewell, Xanadu：“Autodesk, Inc. announced today that it will divest itself of two companies … the American Information Exchange (AMIX) and Xanadu Operating Company”“they do not fit into Autodesk’s core business”，过渡期约三十天。同期一手文件，出自撤资方。CERN 页面两个版本：1992 年 11 月 3 日快照写 “The Xanadu Operating Company is now owned by Autodesk, but the project continues. Code is due out ‘any time now’.”，并已有 “(I have a copy of LM 90.1 - TBL)”；后来的版本写 “was owned for a while by Autodesk, but later dropped. (Not clear in 1993 what is going to happen there)” 及 “(I have a copy of LM 90.1 -TBL)”，后一版本页面本身未注日期，本文按括号内的“1993”系年；“我有一本 LM 90.1”两版皆有。“把一套不会断的链接从头到尾想了一遍”是本文对《Literary Machines》所述 Xanadu 设计的概括，依据是 Samuelson 与 Glushko 的转述，本文未读该书原文。撤资后的去向本文不写，同期报道未核。整理见 Autodesk 1992 年撤出 Xanadu 与 CERN 同期对 Xanadu 的描述。 ↩

11. Robert Raisch，《Links and Type》与《Links, Types and Documents (Third time’s a charm)》，1992 年 5 月 21 日、6 月 22 日（第 65、104 号），原话 “ever since I first had the pleasure of hearing Ted Nelson speak in 1977”“Inclusion — when the document is created, elements from other documents are collection to be included in the representation of the current document”“When might we expect extensions to WWW that support collaboration?”。Frank Kappe，《HTML DTD and related problems (rather long)》，6 月 25 日（第 118 号），原话 “backwards-tracing of links, which is essential for maintaining the integrity of the hypertext”。Berners-Lee，《Re: Links that refer to a range of text, not just a point.》，6 月 25 日 22:50，回复的是 Jim Davis 关于链接指向一段文字的问题，非直接回复 Kappe（第 120 号），“in principle, there is no reason why one shouldn’t have overlapping anchors, or at least nested ones. But not now.” ↩ ↩²

12. Berners-Lee，《Re: HTML DTD and related problems (rather long)》，1992 年 7 月 15 日 00:33（第 149 号），开头 “I am replying very late to a message of Frank Kappe, dated 25 Jun 1992.”，论交叠锚 “This of course is counter to the SGML philosophy of strict nesting (SGML people can get quite religeous about this, but I can’t.) I think it is useful to be able to refer to two separate overlapping anchors.”，原话 “the ‘link database’ model you propose is the Intermedia model of Norman Meyrowitz et al.”“If you expand that system to more than one web server, and scale it to global, then you find that the same problems of ensuring consistency occur between the multiple link databases when before they occured between dcouments. You can’t set up a system of bidirectional links for example, which you could in the non-distributed case.”“Link databases have their advantages, though. A model I am rather keen on atthe moment is of servers which are both link databases and source code control (sccs, rcs etc) systems”，末段 “you can go a small extra step and make the server coordinate simultaneous edits by many people, which gives you a group editor”。“压了二十天”按 6 月 25 日至 7 月 15 日算；“列表上的问题差不多都归他一个人答”是本文据 1992 年存档的印象，未统计；“写成了他下一个想做的东西”是本文的读法。《Re: Citations Using Message Ids for Notes in Logs》，7 月 14 日（第 145 号），“a really useful project which is up for grabs right now is a server which hoards mail, inexes it, and fills in all the back-links … This would be a cool excercise in the management of unruly information … Any takers?”，发出时间 7 月 14 日 20:51。Edward Vielmetti，《saving documents, losing links》，10 月 31 日（第 247 号），“all of the cross reference links were just gone”。“没有一个系统替你保证链接两头都在”一句是本文的解释，不出自信件。 ↩ ↩²

13. Eliot Kimber（署名 drmacro@vnet.IBM.COM），《Explicit Linking is Impossible》，comp.text.sgml，1993 年 5 月 4 日，经 Marc Andreessen 5 月 5 日转至 www-talk（1993 年第二季度第 225 号），原话 “cross-enterprise linking … is so clearly impossible as to be not worth considering”，Andreessen 附注 “[emphasis added -marc]” 与 “Didn’t you lose $5 billion last year?”；aronsson@lysator.liu.se 5 月 5 至 6 日（第 226、229 号），“next to impossible”“like pulling the rug away from under my feet”“ISHN”；Vielmetti（第 227、236 号），“pre-load a search”“spidey”“(you hope)”；Tony Sanders（第 230 号），“I agree that massive explicit linking is impossible, but it isn’t desirable either”；Andreessen（第 228 号），“Limitations of current technology should not be extrapolated to judge possibility or impossibility; I would have thought the Wright brothers taught us that principle”。Rob Raisch，《Re: group annotation server》，6 月 17 日（第 539 号），“Ahhhh… Dreams of XANADU. (Where’s my laudanum?)”“They might well change your life. I know they changed mine.” 线程整理见 1993-05 www-talk 线程 Explicit Linking is Impossible，第 225 至 236 号内没有 Berners-Lee 的发言；“他不必说”以下是本文的读法。Kimber 帖中另有 “Even if you apply tools at one order of magnitude to make the impossible possible (say a database system to track links and manage changes), the next order of magnitude increase will be impossible”，可见他不认为工具能解决；“Xanadu 承诺的正是把这件事从人手里拿走”是本文的判断，Kimber 本人未提 Xanadu。Raisch 6 月 17 日原句 “Ahhhh… Dreams of XANADU. (Where’s my laudanum?)”，鸦片酊一语指柯勒律治《忽必烈汗》一诗的来历，正文略去。Andreessen 生于 1971 年 7 月，1993 年 5 月时 21 岁，据通行传记记载；“莱特兄弟不用维护链接”是本文的话。 ↩ ↩² ↩³

14. Christopher McRae 语经 Kevin Altis 转引，《Re: URN/URL file links》，1993 年 9 月 14 日（第三季度第 896 号），“HTML does support two-way links. See the discussion of the REV attribute … Of course, that doesn’t mean that anybody’s using them. :)”；frans van hoesel 同日（第 903 号），“can someone explain to me how to use them … Is it supported in browsers like Xmosaic?”，本文检索该线程后续未见对 Xmosaic 一问的回答，“没有人回答”以存档为限。Berners-Lee，《Re: URN single or multiple variants》，9 月 21 日（第 982 号），“(I think these are basically Ted Nelson’s ‘tumblers’)”，括号为原文所有。tumblers 的解释据 Samuelson 与 Glushko 一文所述 Xanadu 的文档编号，非 Nelson 原文。 ↩

15. Roy T. Fielding，《RFC: Multi-Owner Maintenance robot (MOMspider)》，1993 年 12 月 6 日（第四季度第 710 号），原话 “referenced objects which cannot be accessed (broken links)”“recursively traverse each owner’s web from a specified ‘top’ document down to each leaf node”“build a special index document for each owner”；Berners-Lee 同日回复（第 729 号），20:48 发出，“I’ll play devil’s advocate here”；Fielding 说用 LINK 写主人是 “a contrived reverse LINK relationship”，Berners-Lee 答 “The LINK elementt was designed to define any two-ended relationship, or binary predicate. It was designed for relationships like Jack loves Jill / Jill loves Jack / Jack likes pie / Jill makes pie / Jack eats pie / Jack adicted to pie / Jack needs pie”。“心情不坏”是本文的推想；“没有反对”指该信未反对蜘蛛本身，只议标签。结尾所引病历一句出自 Fielding 同一封信：“maintenance is rarely or inconsistently performed and the infobase eventually becomes corrupted”。此前列表上有零星的死链讨论，未逐一核对。“今天网站用的死链检查工具做的就是这件事”一句是本文的对照，不出自信件。末尾“剧透”两句是本文的话，“天生残疾”是借教堂山那六条与评审意见的口气，不是任何一位当时人的原话。 ↩ ↩²

16. Roy T. Fielding，《Maintaining Distributed Hypertext Infostructures: Welcome to MOMspider’s Web》，第一届国际 WWW 会议论文集，日内瓦，1994 年 5 月，第 147 至 156 页；修订版载 Computer Networks and ISDN Systems 第 27 卷第 2 期，1994 年 11 月（ACM 数字图书馆条目）。本文只用它证明程序做了出来并在半年后发表，未读全文；“第一个专门为断链写的程序”限于本文读到的 www-talk 材料，1992 年 7 月 Berners-Lee 提议的邮件反向链接服务器是提议，未见有人接手。 ↩