You send a pm to the admin with your question and if it does turn out to be frequent (that's what the 'F' stands for!) it'll be added.

Last modified: Aug. 29, 2009, 6:10 p.m.

Method 1

Add an issue at github and label it "Please hide conlang >name of conlang<". Replace >name of conlang<" with the name of the conlang. In the comment, add an url to the homepage of the conlang at CALS. You have to be a manager of the conlang.

The admin will toggle a hidden flag on the conlang that'll make the conlang invisible for everyone (even the admin), but it will still exist in the database, which means there is an undo if you should ever change your mind.

Method 2

Add an issue at github and label it "Deletion Request". Add the name of the conlang to the title, and an url to the page in the comment part.

Please consider if method 1 is sufficient, as method 2 erases history and needs a lot of scary manual operations directly in the backend. Method 2 does not a happy admin make...

Erasing history is not at all a nice thing to do (and it messes up the calculation of milestones among other things) so a way to hide a conlang is in progress. Each deletion will be blogged so that future researchers will know to stop looking.

Last modified: Jan. 25, 2015, 5:44 p.m.

If you stored your email-address you can reset your password.

If you didn't store your email-address you have a problem. Send a message to kaleissin, on twitter, identi.ca, delicious, flickr, irc (ircnet), en.wikipedia, zbb etc. etc., twitter and irc is checked the most often.

Last modified: Aug. 29, 2009, 5:57 p.m.

A new bugtracker is being looked for/planned, in the meantime, if you're logged in try to send a message to admin, explaining the problem. Remember telling what page you were on! If you can't log in, see How do I get a new password?.

Last modified: Sept. 7, 2009, 2:48 p.m.

Currently, for every feature, the most frequent value is recorded, making up a "meta"-conlang which is the most average. Then this meta-conlang is compared with each conlang in turn. The conlang gets one point per value it has in common with the meta-conlang, and this is divided on the number of features that is recorded for that conlang.

This means that conlangs with few features tend to be more cookie cutter average than conlangs with many features.

Last modified: June 26, 2012, 10:09 p.m.

As long as a user has never added anything else here, like a pm or language or a translation, send a message to admin as that user, and ask for the user to be removed.

It's the same procedure for languages and translations. Send a message as the user that added the translation or added or manages the language, include the url to the page it is in, and ask for its removal. If somebody else manages or edits a language it cannot be removed.

Last modified: Oct. 8, 2009, 7:06 a.m.

It's a glyph in UNICODE, ● 0x25CF BLACK CIRCLE, recently changed from ⚫ 0x26AB MEDIUM BLACK CIRCLE. It looks like a filled disc. If it instead looks like for instance a hollow square, or a square with numbers (and maybe letters ABCDEF) in it, or a question-mark, then all of the below holds:

1. The font you are using lacks that glyph.
2. The browser and/or OS you are using is incapable of fetching that glyph from some other font on your system that do have the glyph, so called font/character substitution.

Fixing either point above, by using a font that do have the glyph (fixing point 1), or by using a different browser/OS (fixing point 2), will show the correct glyph.

The developer uses an OS that does point 2 above very well (though several different fonts are used to show tricky passages of the IPA), so the problem can't even be triggered.

Fonts that are known to have this glyph:

Known affected systems are:

• Nintendo DSi: lacks glyph in font

Last modified: Sept. 25, 2009, 9:12 p.m.

The Leipzig Glossing Rules is basically a collection of conventions for interlinear morpheme-by-morpheme glosses.

Currently, if there is the same number of tokens in the original text as on the gloss-lines, the interlinear can be formatted according to the "Leipzig Glossing Rules" interlinear format. Such an interlinear will reflow (try resizing the browser-window until it it is very narrow) while keeping the columns intact. If there is a different number of tokens in the lines, the interlinear will be saved as monospace and an error message wil be shown. A token is a bunch of letters and symbols separated from other tokens by a space. You could say that in English, most word is one token, while some compound words are written as two or more tokens.

The optional rule 2a in the Leipzig Glossing Rules cannot currently be used here since with 2a, there need not be the same number of tokens per line.

So what to do if a word is written with one or more spaces, but the interlinear would only be one token for that word? Try a dash: A B-C would match up with 1 2. For the opposite problem, there are plenty of rules, see Leipzig Glossing Rules, especially point 4.

Last modified: Jan. 23, 2016, 3:30 p.m.

If it is close enough to a natural language to be described with some of the features from WALS, it is eminently suitable. If it is too strange for that, but have a good homepage, it is also suitable. If there are poems and texts in it, or it has been used for relays or translation exercises, it is suitable. Sketches that are (so far) just a grammar or just a a phonology or just a collection of abstract meanings with no way of using them yet is also okay, since these can grow into a full language and can serve as an inspiration.

If all that is known about it is it's name, it's not suitable. Ciphers like rot-13 and language games like pig-latin are not suitable. Relexes are in a gray area, they better have something interesting in addition, like Nadsat!

Last modified: Sept. 13, 2011, 8:42 a.m.

A lurker is someone who just observe and never participate. For CALS, that means someone, who has never made or edited a language, never made a comment or sent a private message, never earned a badge etc.

Last modified: Sept. 22, 2010, 1:49 p.m.

CSV when used on CALS is short for comma-separated values. It's an export-version of a database table that can be read by for instance spreadsheet software like Excel.

The delimiter is a comma, ,, and all strings are quoted with double quotes, ".None stands in for missing value and maps to SQL NULL. True is binary true and maps to SQL true, False is binary false and maps to SQL false.

Last modified: Dec. 11, 2011, 7:54 p.m.

"Relex" is an abbreviation of "relexified" or "relexfied language". A relexified language is one where you keep the grammar and lexicon but change what each word sounds like, needing a dictionary to look up the meaning. Relexes are also a way to encrypt a text, and is then called a cipher. It's a very low quality encryption though as all it takes is to build a frequency list of a text written in the relex and then compare that to frequency lists of other languages to make a map from relexed word to original word. Don't use a relex to keep important secrets!

A relex is more sophisticated than a substitution cipher like rot-13, a game like pig-latin, or a spelling reform, but not that much more sophisticated.

Last modified: Sept. 13, 2011, 8:22 a.m.

The badges show how active a conlanger has been here. You are much more likely to get an answer to a private message from a be-badged person than one without, for instance.

Last modified: Sept. 11, 2009, 8:57 a.m.

Countach and its siblings countach2, countach3, countach4 etc. are test-users, for when it is necessary to check that making new users works as it should. These users get deleted with no notice.

Test-languages have names starting with "testarossa".

Yes, the admin likes (dreaming about) certain fast, pretty, Italian and ridiculously priced cars.

Last modified: Oct. 11, 2009, 8:31 p.m.

The language needs to have at least 50 A features at WALS to be considered for CALS. The A-features are the original features that were in WALS when CALS was first designed. WALS-features whose numbers end in a letter other than A, like 144K SVONeg Order are subtypes of the A features, in 144K's case 144A Position of Negative Word With Respect to Subject, Object, and Verb.

There are no plans at present to add the non-A features to CALS as the necessary rewrite would be massive.

Last modified: Jan. 25, 2015, 5:33 p.m.