Conlang Random Thread

[bradrn]

Alright, I’ve had another look. Can you do the following steps for me?

1. Press ‘Download’. (No need to actually go all the way through with the download, just pressing the button should be enough.)
2. Open the developer tools by right-clicking in the middle of the webpage and then pressing ‘Inspect’.
3. Near the bottom of the screen which comes up, you should see something starting <a id="dl". Right-click on that and select ‘Edit as HTML’.
4. This should give you some text, which you can copy. What text do you get?

[Raphael]

<a id="dl" style="display:none;" href="data:text/plain;charset=utf-8,V%3Daeiou%0AL%3D%C4%81%C4%93%C4%AB%C5%8D%C5%AB%0AC%3Dptcqbdgmnlrhs%20%20%20%20%20%0AF%3Die%0AB%3Dou%0AS%3Dptc%0AZ%3Dbdg%0Alh%7Clj%0A%5Bsm%5D%2F%2F_%23%0Ai%2Fj%2F_V%0AL%2FV%2F_%0Ae%2F%2FVr_%23%0Av%2F%2FV_V%0Au%2Fo%2F_%23%0Agn%2Fnh%2F_%0AS%2FZ%2FV_V%0Ac%2Fi%2FF_t%0Ac%2Fu%2FB_t%0Ap%2F%2FV_t%0Aii%2Fi%2F_%0Ae%2F%2FC_rV%0Alector%0Adoctor%0Afocus%0Ajocus%0Adistrictus%0Ac%C4%ABvitatem%0Aadoptare%0Aopera%0Asecundus%0Af%C4%ABliam%0Ap%C5%8Dntem%0A%0A" download="temp2.txt">Download output lexicon</a>

[bradrn]

<a id="dl" style="display:none;" href="data:text/plain;charset=utf-8,V%3Daeiou%0AL%3D%C4%81%C4%93%C4%AB%C5%8D%C5%AB%0AC%3Dptcqbdgmnlrhs%20%20%20%20%20%0AF%3Die%0AB%3Dou%0AS%3Dptc%0AZ%3Dbdg%0Alh%7Clj%0A%5Bsm%5D%2F%2F_%23%0Ai%2Fj%2F_V%0AL%2FV%2F_%0Ae%2F%2FVr_%23%0Av%2F%2FV_V%0Au%2Fo%2F_%23%0Agn%2Fnh%2F_%0AS%2FZ%2FV_V%0Ac%2Fi%2FF_t%0Ac%2Fu%2FB_t%0Ap%2F%2FV_t%0Aii%2Fi%2F_%0Ae%2F%2FC_rV%0Alector%0Adoctor%0Afocus%0Ajocus%0Adistrictus%0Ac%C4%ABvitatem%0Aadoptare%0Aopera%0Asecundus%0Af%C4%ABliam%0Ap%C5%8Dntem%0A%0A" download="temp2.txt">Download output lexicon</a>

Now that is really weird. Because when I take this and use it to download the same file you have, I don’t see any HTML line breaks.

By the way, what method are you using to view the temp2.txt file you get?

[Raphael]

Wait, wait, wait...

Perhaps we've been talking about different download links all the time? Do you mean the Download link in the SCA's main menu? The one that gives you all the data from all the fields? I mean the "Download output lexicon" link that appears near the bottom of the link after you hit "Apply", and that downloads only the output lexicon.

Even hovering my mouse over it and hitting "copy link location" gives me

data:text/plain;charset=utf-8,leitor%3Cbr%3E%0Adoutor%3Cbr%3E%0Afogo%3Cbr%3E%0Ajogo%3Cbr%3E%0Adistrito%3Cbr%3E%0Acidade%3Cbr%3E%0Aadotar%3Cbr%3E%0Aobra%3Cbr%3E%0Asegundo%3Cbr%3E%0Afilha%3Cbr%3E%0Aponte%3Cbr%3E%0A

[bradrn]

Wait, wait, wait...

Perhaps we've been talking about different download links all the time? Do you mean the Download link in the SCA's main menu? The one that gives you all the data from all the fields? I mean the "Download output lexicon" link that appears near the bottom of the link after you hit "Apply", and that downloads only the output lexicon.

Even hovering my mouse over it and hitting "copy link location" gives me

data:text/plain;charset=utf-8,leitor%3Cbr%3E%0Adoutor%3Cbr%3E%0Afogo%3Cbr%3E%0Ajogo%3Cbr%3E%0Adistrito%3Cbr%3E%0Acidade%3Cbr%3E%0Aadotar%3Cbr%3E%0Aobra%3Cbr%3E%0Asegundo%3Cbr%3E%0Afilha%3Cbr%3E%0Aponte%3Cbr%3E%0A

Oops… yes, I was talking about the main ‘Download’ button! And I can replicate the bug when I press ‘Apply’! I’ll have yet another look at the code to see if I can find the cause of this bug.

[bradrn]

I found what’s causing the bug. The issue is this part of the code, in the DoWords() function:

Code: Select all

	document.getElementById("olex").innerHTML = olex;

	// Allow download of output 
	var dle = document.getElementById("dl");
	if (dle != null) {
		dle.setAttribute('href', 'data:text/plain;charset=utf-8,' + encodeURIComponent(olex));
		dle.style.display = "block";
	}

Here, olex is encoded using HTML, so it can be assigned to the #olex <div>. But then this HTML-encoded variable is assigned to be the ‘Download’ link — hence the bug. I must admit that I’m not entirely sure what the best way of fixing this bug is. One easy way is just replace <br> with a newline, but that’s treating the symptom rather than the cause. (And it doesn’t fix any related bugs which may appear: e.g. when ‘show differences from last run’ is checked, the downloaded file gets <b>s inserted into it as well.) Alternately, olex appears to be accumulated word by word, so just maintain another variable alongside it using newlines rather than <br>s. But it’s late and I’m too tired to have a go at implementing this right now… maybe tomorrow.

[Raphael]

Congratulations, great work!

As an aside, IMO it's amazing how much functionality the online SCA fits into a fairly small number of kilobytes of javascript.

[bradrn]

Congratulations, great work!

Thank you!

As an aside, IMO it's amazing how much functionality the online SCA fits into a fairly small number of kilobytes of javascript.

It definitely is. But for comparison, I wrote another SCA, mostly compatible with SCA² plus a couple more features; the sound-changing code alone is 401 lines long, which actually looks like a pretty similar length to the corresponding part of SCA². If you want, you can see the truly horrendous code here. (It’s successor is far, far worse though… I still use it pretty often, but I have decided that if I ever need to change it, I will create a new SCA from scratch rather than attempt to modify that code.)

[zompist]

Didn't see this till now.

Brad, there's a link at the top of the help file on the new features. The other one is intermediate results.

I'm aware of the <br> bug. It's just laziness. I add the <br>'s for the HTML output, and didn't bother to do something with them. I did figure they're easy for the user to get rid of (with replace all), but I should look at it.

[bradrn]

Brad, there's a link at the top of the help file on the new features. The other one is intermediate results.

I know — earlier Raphael linked to your blog post on the new features.

I'm aware of the <br> bug. It's just laziness.

Sounds familiar…

[Raphael]

In one of my naming languages, "kirg" means "hand", and "zi-" is a plural prefix, but I haven't got a word for "human" (or, more accurately, "member of the species to which the speakers of the language belong") yet. Would it be plausible to use "zikírg" as the singular of "human", and "zizikírg" or maybe "zizkírg" as the plural of "human"?

Bump.

[Raphael]

In one of my naming languages, "kirg" means "hand", and "zi-" is a plural prefix, but I haven't got a word for "human" (or, more accurately, "member of the species to which the speakers of the language belong") yet. Would it be plausible to use "zikírg" as the singular of "human", and "zizikírg" or maybe "zizkírg" as the plural of "human"?

Bump.

Ok, I'm now tending towards combining it with "túdu" "two" into sing. Túduzkírg, pl. Zitúduzkírg, probably with some shortening of the latter to come later.

[Raphael]

Ok, I'm now tending towards combining it with "túdu" "two" into sing. Túduzkírg, pl. Zitúduzkírg, probably with some shortening of the latter to come later.

Which leads me to my next question: would it be plausible to have an inflectional affix that not only adds a phoneme or phonemes, but completely replaces a phoneme or phonemes under certain circumstances? That is, in the case of the concrete example I'm currently thinking of, if the singular of a word is "Túduzkírg", and my usual plural prefix is "zi-" or "z-", would it be plausible to have a plural "Zúduzkírg" in this case?

[Richard W]

Ok, I'm now tending towards combining it with "túdu" "two" into sing. Túduzkírg, pl. Zitúduzkírg, probably with some shortening of the latter to come later.

Which leads me to my next question: would it be plausible to have an inflectional affix that not only adds a phoneme or phonemes, but completely replaces a phoneme or phonemes under certain circumstances? That is, in the case of the concrete example I'm currently thinking of, if the singular of a word is "Túduzkírg", and my usual plural prefix is "zi-" or "z-", would it be plausible to have a plural "Zúduzkírg" in this case?

Yes. In spoken English, the 'ed' morpheme has such an allophone, e.g. taught from teach, thought from think and, less impressively, bent from bend. (The vowel alternation is more complicated in the first two, but historically, it is the past tense marker that has caused the loss of the final consonants.) And some morphemes are particularly vulnerable - both has and had from have.

[Raphael]

Yes. In spoken English, the 'ed' morpheme has such an allophone, e.g. taught from teach, thought from think and, less impressively, bent from bend. (The vowel alternation is more complicated in the first two, but historically, it is the past tense marker that has caused the loss of the final consonants.) And some morphemes are particularly vulnerable - both has and had from have.

Ah, thank you!

[mae]

-

[bradrn]

thinking about doing an a posteriori Sinitic conlang where some people during the tang dynasty end up way out west during its expansion there. not sure where it would go though. Could I get it all the way in modern kazakhstan/kyrgyzstan?

The Mandarin dialect Dungan sounds like what you want:

The Dungan language (/ˈdʊŋɡɑːn/ or /ˈdʌŋɡən/) is a Sinitic language spoken primarily in Kazakhstan and Kyrgyzstan by the Dungan people, an ethnic group related to the Hui people of China.

So: yes, you can definitely do this, as it is not only plausible but attested. (Dungan was formed from migrations in the 1870s, so if you’re going all the way back to the Tang dynasty then you could probably go much further than Kazakhstan/Kyrgyzstan.)

[mae]

-

[Richard W]

I had heard about Dungan before but just completely forgot about it lol. Great news for me then.

I'm not sure it's much of a precedent though. It's younger than the Liverpool Chinatown!

[bradrn]

I had heard about Dungan before but just completely forgot about it lol. Great news for me then.

I'm not sure it's much of a precedent though. It's younger than the Liverpool Chinatown!

I’m having trouble seeing how that’s relevant.