|WikiProject Computing||(Rated Start-class)|
Dict IS a network protocol, as stated right on the first line, and NOT a file format, as it is implied in the rest of the article. The article is for the most part incorrect and confuses readers. See comments down this page.
The Spanish version of the article (http://es.wikipedia.org/wiki/DICT) correctly describes Dict protocol and should be translated into English.
What is "webster protocol"?
Why doesn't Wikipedia support DICT?--22.214.171.124 17:55, 12 July 2006 (UTC)
Though the question remains: Why doesn't Wikipedia support the dict protocol? Wouldn't it be great (and faster) if Wikipedia, Wiktionary etc. would be available through dictionary applications? —Preceding unsigned comment added by 126.96.36.199 (talk) 06:33, 29 September 2008 (UTC)
because it seems dead.
Added one client and a link for the GNU server. Rearranged the dictionary lists and divided them into (monolingual) dictionaries of English and bilingual dictionaries. Rearranged external links to push the protocol's RFC to the top.
I'm not sure whether this:
- Lingvo English-Russian and Russian-English dictionaries are not free, but when purchased, can easily be converted into DICT format
should be listed among free dictionaries, it sounds like a plug.
I don't understand the fragment: There are also programs that read the DICT file format directly. For example (...) Well, don't the other clients read DICT directly?
Why doesn't this protocol hit the world?
Yeah, why, truly?! Because it suffers from not sufficiently documented. WWW.dict.org provides one rfc, and users aren't served by rfcs. I have a hard time finding and setting up database urls. Rursus dixit. (mbork3!) 08:03, 1 April 2011 (UTC)
- It could hit the world, if our Wikimedia projects used it
- Wiktionnary (requires specific microformats to allow complete processing ans separation of definitions by languages and lems), Wikispecies (the "DICT definitions" could a a list of subtaxons used to define a taxon "DICT word").
- Even Wikipedia (article contents can be seen as "definitions" of the article name) could be available under the DICT protocol.
- We could as well enumerate other things than just the main namespace of wikis; notably we could the contents enumerate categories as if they were "DICT definitions", given their parent category name used as a "DICT word"
- In MediaWikis, the "DICT words" are exactly what can fit within a full Wiki page name. As these names use the colon (:) as prefixes for namespaces (or interwikis) and slash (/) to index subpages, but also reserve the question mark (?) and hash (#), these could be used under all existing MediaWiki conventions to give parsable conventions for what can fit in a "DICT word". Thus these characters allow easy derivation creation of microformats, using templates already used in Wikitionnary to format the language sections, and their subsections and lems or to access to examples, external references.
- Wiktionnary could also be able to transclude, via a Lua module and a formatting template inserted in articles, the contents of some other free DICT databases (such as those listed in "http://www.dict.org/")
- As the Lua module would perform external requests, this could add delay to the delivery of pages, so its effect should be to perform requests on demand, possibly using a local cache for remote DICT servers It would most probably need jQuery to modufy the content of pages.
- An interesting option to consider in the implementation is to include MIME support since the begining, to allow "Accept:" headers in queries (so it could return data in JSON or XML format, or in HTML format matching what is already generated in rendered Wiki pages, or plain text format; only the HTML format does not require the integration of parsable microformats in wiki pages), and "Accept-Encoding:" (the returned responses should be compressible in gzip or deflate formats, just like with out current HTML output). For using the "raw" wiki format (visible in wii code editors), the best option is to use a query parameter within "DICT words", such that a DICT query of the "article-name" (may be quoted) will return the definitions of "article-name", rendered as HTML by default, the query of the quoted "article-name?format=raw" would return the wiki code without rendering the wiki page...
- With this extension of MediaWiki, we could then register all linguisitic editions Wiktionnarywithin dict.org and allow easier interchanges (and integration in softwares that already use DICT to look for various databases, to querty definitions, lookup orthographies in correctors, suggest replacement terms, control text style and prefered terminologies in some categorized domains).
- Another option would be to have a separate DICT server (hosted in Wikimedia Labs) querying the normal Wiktionnary server (but parsing the microformats inserted in rendered wiki page), if it's not hosted simply in a "Special:DICT" page of the wiki server.
- Also look at Wik2dict for such attempt to use the DICT protocol on Wikimedia projects and more generally on any Mediawiki server. It failed because its only goal was to convert static dumps of Mediawiki databases to giant static DICT files (to be used then on a separate DICT server). My opinion is that this would have succeeded if if was a true DICT server performing conversions on demand (with a local cache).
- verdy_p (talk) 19:56, 20 September 2013 (UTC)