- Home
- Get A Printable Dictionary
- Search Best Words
- Recent Changes
- How You Can Help
- valsi - All
- valsi - Preferred Only
- natlang - All
- natlang - Preferred Only
- Languages
- XML Export
- user Listing
- Report Bugs
- Utilities
- Status
- Help
- Admin Request
- Create Account
|
Discussion of "cfilyfacki"
Comment #1:
Not lowest scoring lujvo
|
Wuzzy (Wed Feb 26 14:09:02 2014)
|
Jbovlaste only should have the lowest scoring form of a lujvo. Jbovlaste tells you that every time you want to create a new word. The lowest scoring form for this is not “cfilyfacki” but “cfifa'i”.
|
-
Comment #2:
Re: Not lowest scoring lujvo
|
gleki (Wed Feb 26 14:27:06 2014)
|
Wuzzy wrote: > Jbovlaste only should have the lowest scoring form of a lujvo. Jbovlaste > tells you that every time you want to create a new word. The lowest > scoring form for this is not “cfilyfacki” but “cfifa'i”.
You've downvoted a lot of words but how many times you entered alternative solutions?
|
-
Comment #3:
Re: Not lowest scoring lujvo
|
Wuzzy (Wed Feb 26 15:26:29 2014)
|
gleki wrote: > You've downvoted a lot of words Yes, that’s true.
> but how many times you entered alternative > solutions? Not often.
|
-
Comment #6:
Re: Not lowest scoring lujvo
|
gleki (Thu Feb 27 06:08:36 2014)
|
Wuzzy wrote: > gleki wrote: > > You've downvoted a lot of words > Yes, that’s true. > > > but how many times you entered alternative > > solutions? > Not often.
So this leads to denying the existence of some concepts in Lojban, this is almost equivalent to just deleting words. What would others do when they try to find a word for a concept they need and fail to find it?
|
-
|
|
|
Comment #4:
Re: Not lowest scoring lujvo
|
Lorenzo Von Matterhorn (Wed Feb 26 23:54:04 2014)
|
Wuzzy wrote: > Jbovlaste only should have the lowest scoring form of a lujvo. Jbovlaste > tells you that every time you want to create a new word. The lowest > scoring form for this is not “cfilyfacki” but “cfifa'i”.
So you also reject ci'omle because citmle has a better score? What is the benefit of such a policy? If lots of people prefer the sound of ci'omle, why should the score matter?
I don't understand that excessive hardlinerism from you. It's always black or white, right or wrong. "The CLL says X so anything else is completely wrong and evil", "jbovlaste says Y" so suggesting other word forms means we must downvote them all. I find this attitude very destructive.
|
-
Comment #5:
Re: Not lowest scoring lujvo
|
Wuzzy (Thu Feb 27 02:48:09 2014)
|
selpahi wrote: > Wuzzy wrote: > > Jbovlaste only should have the lowest scoring form of a lujvo. Jbovlaste > > tells you that every time you want to create a new word. The lowest > > scoring form for this is not “cfilyfacki” but “cfifa'i”. > > So you also reject ci'omle because citmle has a better score? Yes. I reject it being in the dictionary. However, this does NOT mean that I reject “ci'omle” _in general_.
> What is > the benefit of such a policy? To avoid reduncancy in jbovlaste. Consider the opposite case: You include every possible word form of a given lujvo. But everything is treated in jbovlaste as if it were a completely different word. This would mean you’d repeat the definition over and over again. And if you want to edit something, you would have to edit the other definitions as well. It is very easy to become inconsistent then. To make life easier, the dictionary only includes the lowest-scoring form. It is assumed the user can infer all the other valid forms of the lujvo. The definition applies to all lujvo forms equally.
> If lots of people prefer the sound of > ci'omle, why should the score matter? The score only matters for dictionary purposes. In actual speech, the score does not matter.
What about this compromise?: If you want a certain word form to be included, mention it in the notes instead of creating an entire new entry. Here’s an example: malgli (mentions “malglico”). There is no entry for “malglico” because it is not needed.
> "jbovlaste says Y" so suggesting other word forms means > we must downvote them all. I have justified that. There is a very good reason for this policy.
> "The CLL says X so anything else is completely wrong and evil" and > I find this attitude very destructive. I understand. This is a big topic and perhaps I will explain myself on the mailing list in detail.
|
-
Comment #7:
Re: Not lowest scoring lujvo
|
Lorenzo Von Matterhorn (Thu Feb 27 11:02:10 2014)
|
Wuzzy wrote: > selpahi wrote: > > Wuzzy wrote: > > > Jbovlaste only should have the lowest scoring form of a lujvo. > Jbovlaste > > > tells you that every time you want to create a new word. The lowest > > > scoring form for this is not “cfilyfacki” but “cfifa'i”. > > > > So you also reject ci'omle because citmle has a better score? > Yes. I reject it being in the dictionary. > However, this does NOT mean that I reject “ci'omle” _in general_. > > > What is > > the benefit of such a policy? > To avoid reduncancy in jbovlaste. Consider the opposite case: You include > every possible word form of a given lujvo. But everything is treated in > jbovlaste as if it were a completely different word. This would mean > you’d repeat the definition over and over again. And if you want to edit > something, you would have to edit the other definitions as well. It is > very easy to become inconsistent then. To make life easier, the dictionary > only includes the lowest-scoring form. It is assumed the user can infer > all the other valid forms of the lujvo. The definition applies to all > lujvo forms equally.
Okay, this is a valid point. Under the current rules, every two lujvo that contain rafsi of the same words in the same order are considered the same word, so it would be easier to have them organized more automatically. However, jbovlaste has no such feature, and it also doesn't seem to happen very much that multiple forms are added.
> > If lots of people prefer the sound of > > ci'omle, why should the score matter? > The score only matters for dictionary purposes. In actual speech, the > score does not matter.
Sure, but apparently the scoring algorithm can sometimes prefer words that humans don't prefer. When that happens, maybe we (the speakers) should have a bigger say than an artificial algorithm (whose initial purpose it was to make the shortest and most pleasing forms get the best scores). This sort of thing doesn't happen so often that adding an alternative form causes a lot of extra effort. I can think of ci'omle and je'umlu, which some people prefer because they get rid of the "tml" cluster, or because they don't like cit in the word for "cute".
> What about this compromise?: If you want a certain word form to be > included, mention it in the notes instead of creating an entire new entry. > Here’s an example: malgli (mentions “malglico”). There is no entry > for “malglico” because it is not needed.
That would be an option.
> > "The CLL says X so anything else is completely wrong and evil" > and > > I find this attitude very destructive. > I understand. This is a big topic and perhaps I will explain myself on the > mailing list in detail.
Okay. It's not a very fun topic, though. :)
|
-
|
|
|
|
|