Re-activate langdocstatus browser

[HedvigS]

Just... a general thumbs up and also a (hopefully) not too pushy question: were any of the comments and questions we had about GlottoVis answered?

[xrotwang]

Answered by whom? Or answered in code? Am 02.04.2017 13:11 schrieb "Hedvig Skirgård" notifications@github.com:

Just... a general thumbs up and also a not to pushy question: were any of the comments and questions we had about GlottoVis answered?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/clld/glottolog3/issues/72#issuecomment-290979677, or mute the thread https://github.com/notifications/unsubscribe-auth/AA1HKL38EA2HzEm1IzeOKPO_yJNwdT6vks5rr4J4gaJpZM4Mt90Q .

[HedvigS]

Sometime in 2015 Harald send me and Robert a message about Glottovis, Speckmann's team wanted feedback. I sent back some questions and comments, I think you did too. I didn't hear anything more.

[HedvigS]

Oh, and then there was another message from them in 2016, asking for feedback again. I sent back basically the same comments as 2015.

[d97hah]

I've been communicating with them, I am up to date. But I am not sure what information Hedvig is asking for.

2017-04-03 1:40 GMT+02:00 Hedvig Skirgård notifications@github.com:

Oh, and then there was another message from them in 2016, asking for feedback again. I sent back basically the same comments as 2015.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/clld/glottolog3/issues/72#issuecomment-291023460, or mute the thread https://github.com/notifications/unsubscribe-auth/ADUDyPFHyIV3tKU3ieurg2zL8HYLOGkSks5rsDHigaJpZM4Mt90Q .

[HedvigS]

Looking at Glottovis again now. I think it's great that Thom did this, it looks really cool! If this is gonna go on the site thought I think some things just need to be a bit clearer so it's useful to the general user of Glottolog. As usual, feel free to ignore my opinions. If you're thinking of launching Robert's browser instead, that feels a bit unfair to Thom, but I don't know your arrangements and plans.

We emailed about this in September 2016 and April 2015, looking at it now some things have been changed or I've just found them on my own. I didn't hear back about anything before, but I'm glad Harald's been kept updated.

These things of mine remain:

• the clusters/sunbursts are very useful as they solve the known problem of ordering of dots. There needs to be some kind of pedagogical explanation of them and the colors though if people are going to take to it
• in general, there's a lot of smart things going on here, but since they're new they need to be explained.
• there's a legend for the colors for descriptive status and endangerment status, great! I didn't find this before, but it must have always been there.
• the buttons for stats per continent and time scale might need just simple labels for the axises so that it's easier to understand them right away
• possibility of making pacific centered by scrolling east/west
• a table displaying numbers of languages per description status and per endangerment level globally would be useful
• when you click a language, you get displayed the descriptive status but the endangerment status is not listed. it's obvious from the color, but maybe should be listed explicitly too
• Have it say Glottovis on the website and not just in the URL
• explain why some languoids are not shown (the "!" button)

(* extrawork that is way too late for: time slider that displays the world map and the languoids endangerment level and descriptive status at a certain point in time. This is the best feature of Robert's map, and it's really something people enjoy.)

Now, I don't know what you guys want from all of this. This is clearly a very useful function, but it's also a fancy fun visualisation function for the website and probably not part of the core functions of Glottolog (sharing the content and letting people do what they will with it). I probably care more about such things than you do, and you decide. Do with it as you will.

[HedvigS]

These suggestions of mine are basically all geared towards user-friendliness. There's nothing wrong with the basic underlying structure, and a smart person will figure it out. It depends on the intended audience and reach that you want for this part of Glottolog.

[HedvigS]

Oh noes... status/browser is down ;( ;(

I actually just wanted to send a screengrab of glottolog language dots over new guinea to a friend. Is there any chance that a world map for just the "languages" tab of Glottolog might come into existence? Coloured by family?

[xrotwang]

http://glottolog.org/glottolog/language.map.html?country=PG#6/-6.235/149.747

[HedvigS]

Thanks Robert @xrotwang

I was after the entire island, i.e. the indonesian side too. Ah, it's ok. We found another map.

[HedvigS]

Btw, I realise that rendering a map with all 8,000 dots is cumbersome. The other day on facebook, Claire was asking if there was such a map and today someone else asked for something similar. What I'm saying is it'd be a popular feature, if heavy and a bit superfluous to the core function of Glottolog.

[d97hah]

@Robert: I thought one can give in to demand and make Voronoi polygon based maps, perhaps as part of Glottolog. For example, given a node, the partition induced by its immediate children could be coloured. The bottom node of glottolog (or some hypothetical version similar to it) would then give the world family map. There's some version of the four-colour problem to solve though (there are always heuristics and infinitely many shades if nothing else) but with Voronoi regions many individual dots could be reduced to a small set of polygons (usually one since (sub-) families are often continuous) which would make it much less heavy on the browser than 8000 dots. Have you ever pondered this?

2017-04-20 13:33 GMT+02:00 Hedvig Skirgård notifications@github.com:

Btw, I realise that rendering a map with all 8,000 dots is cumbersome. The other day on facebook, Claire was asking if there was such a map and today someone else asked for something similar. What I'm saying is it'd be a popular feature, if heavy and a bit superfluous to the core function of Glottolog.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/clld/glottolog3/issues/72#issuecomment-295696543, or mute the thread https://github.com/notifications/unsubscribe-auth/ADUDyG0S2_vnoc6wHoCv9o9uCsxOqyxkks5rx0JsgaJpZM4Mt90Q .

[SimonGreenhill]

@d97hah -- I, @xrotwang and @LinguList were discussing this a few weeks ago, and @kirbykat did it last week for the DPLACE data.

Nice and relatively simple visualisation, only concern is what to do with e.g. Polynesia, where the wide spread points of the Polynesian triangle will give a massive and overwhelming cell. Also assumes no overlaps, which is false, but unsolvable at the moment.

[d97hah]

And what if you just restrict the polygons to non-water territory?

2017-04-20 15:29 GMT+02:00 Simon J Greenhill notifications@github.com:

@d97hah https://github.com/d97hah -- I, @xrotwang https://github.com/xrotwang and @LinguList https://github.com/LinguList were discussing this a few weeks ago, and @kirbykat https://github.com/kirbykat did it last week for the DPLACE data.

Nice and relatively simple visualisation, only concern is what to do with e.g. Polynesia, where the wide spread points of the Polynesian triangle will give a massive and overwhelming cell. Also assumes no overlaps, which is false, but unsolvable at the moment.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/clld/glottolog3/issues/72#issuecomment-295738793, or mute the thread https://github.com/notifications/unsubscribe-auth/ADUDyCM23aqyzFhgHFq0Eyhjy_UDJ1lYks5rx13CgaJpZM4Mt90Q .

[xrotwang]

I think whether that will lead to better performance in the browser or not, is not completely clear to me. The dots are simple markers in Leaflet, which I guess the library is somewhat optimized for (considering it is the typical use case). Polygons on the other hand may be less optimized. The 345 eco regions in Tsammalex seem be a lot of work for the browser (although I admit that the Voronoi polygons will probably have a lot less points defining them).

[SimonGreenhill]

@d97hah -- well, I guess you're left with lots of empty water. It's such a nuisance that cultures aren't distributed evenly on the globe!

[HedvigS]

In the meantime, is is easy to render something like what was in the descriptive status browser but coloured by family?

[HedvigS]

Robert made a neat map for me with just language dots coloured by family, and permitted me to share it. Claire Bowern wrote back something relevant to this thread.

"Thank you Hedvig and Robert, this is great. We were talking about meshing this with the endangerment stats from ethnologue and endangeredlanguages.com. An effective (if simplistic) graphic might be to dyanmically fade the dots that are highly endangered. Claire"

[HedvigS]

I think trying to show to many things at once in one map is not good, it makes it too busy. I like the coupling of endangerment and descriptive status, as in Robert's browser and Glottovis. I don't think it should be mixed with top-genetic unit.

[xrotwang]

The status browser is up again - with the same status as before, i.e. not fit for publication yet!

[d97hah]

And until it is, don't post url:s to it here...

2017-04-26 12:49 GMT+02:00 Robert Forkel notifications@github.com:

The status browser is up again - with the same status as before, i.e. not fit for publication yet!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/clld/glottolog3/issues/72#issuecomment-297352019, or mute the thread https://github.com/notifications/unsubscribe-auth/ADUDyL4_OMilBJiLy8MKB_4AIt2Swu5Tks5rzyFLgaJpZM4Mt90Q .

[HedvigS]

I keep it in my mind, and in my cookies. and nowhere else :)