Sharing and Embeding ideas

(Chut Ko) #1

In this topic, i would like to discuss ideas and desired features for network embedding and sharing.

the core idea of elvis is that we give people a tool they can use to analyze public spending data, come with conclusions and share them with public.

nowadays, (data)journalists and bloggers are the most desired niche we would like to address. they use content management systems with wysiwyg editors to write articles full of structured text, links, embedded images and chunks of html code.

elvis needs to become part of these articles as an evidence to support their conclusions about public spending. therefore we need to make the visualised data available in form of shared link, static images and most importantly, interactive embedded networks (quite analogic with google maps sharing options).

(Chut Ko) #2

here are some old ideas from the initial system design:

my current ideas:
link sharing
this is a simple one, we provide an html input to copy the address of the public network from. would be cool to have some URL shortening algorithm

image sharing
user could make a snapshot of current network (with current zoom and coordinates) and get the URL of and image (where to host the images?) to share via html tag

network embedding
this is a tricky one. there are few options of embedding:
a) embeding the network view (need to restrict some interaction - sidebar, detailed view etc)
b) detailed view - embed supplier/procurer/relationship detail
c) sidebar - embed sidebar with lists of procurers, suppliers and relationships

there should be an option where user can set the width and height of the embedded network. therefore, all the components (a,b,c) will need to be fully width and height responsive.
i suppose each of these components will need to have separate template files so the embed backend logic can run smooth.

also, in the future, changing colour schemes will be useful (to merge elvis visual with the visual of the server the article is being published at)


would be cool to have some URL shortening algorithm

too bad that is taken … perhaps ? That one is free. We can then add just some gibberish after the slash/. Is this easily doable? @victor @georgiana_b @ca1yps0

user could make a snapshot of current network and get the URL of and image to share via html tag

how do you share images through a HTML tag?
I would rather opt for just simply downloading images. That way we will prevent having to think of storage or god forbid saved images views. User can upload it the image wherever her/imself.

network embedding

For starters, I would go for just option a) . We don’t know if anyone would use b) or c) yet, but it’s good to have it in the back of our minds.

there should be an option where user can set the width and height of the embedded network.

perhaps also a zoom level? You don’t want the embed to zoom in and out on mousewheel scroll, it is one of the most annoying things.

in the future, changing colour schemes will be useful
yes! Future :slight_smile:

(Chut Ko) #4

yep, that is better :slight_smile:

there should be an option to turn the zoom on/off, but i would not expel it totally. on tiny proportions zoom can be useful. also, its rather matter of UX to make it not annoying.

(Georgiana B) #5

I think network embedding b) and c) can be omitted for now if we provide a way to open the embedded network on Elvis in a new tab for people who want to take a closer look.

Let’s say you have the network of a market and you want to embed it but with a focus on a certain subnetwork. I think both zooming level and position are needed to achieve this.

Simple embed, no position and zoom:

Focus on a particular subnetwork:

Being able to zoom in/out of the embed is a separate feature and may be useful, especially if we end up doing :point_up:.


After exploring some options on embedding,
we agreed that for now we will only do an “export to image”.
Embed requires a longer traject in respect to large networks, responsivity etc.

For now we will allow the user to take a screencap in hers/is prefered part of network.

Looking into possibilities on how to do this best

(Chut Ko) #7

i will come with solution. in the meantime, others could write down ideas they mentioned during the call so we dont forget them for the future


I imagined the functionality as something like the gnome imgur shell extension does (or gnome natively for that matter) extension github

You chose a part of the screen and it takes a screenshot, In the case of above it uploads it to imgur. See demo in this video

In our case it would trigger some modal window I believe, with the screenshot.
On the screenshot should be the elvis logo. What else?

Also we should offer a link to the network somewhere to copy easily.

And what about that URL shortener? ^^


yes! @victor mainly :slight_smile:

(Victor Nițu) #10


But… but… why do we need one? I can’t remember!


URL shortener? Because @Chut_Ko said that it “would be cool to have some URL shortening algorithm” :slight_smile:

so … to be cool?
I can imagine sharing the humongous URL on Twitter could pose problems.


@Chut_Ko for the image sharing … I have no more ideas than I already shared (about how I think it should work).
Just an elvis watermark in the corner :slight_smile:

(Chut Ko) #13

using a “crop” area might bring several problems, i would go for something simple.
this solution is similar to google maps embed and seems intuitive to me:

  1. you click share/export, modal window comes up
  2. in a modal window, you set up dimensions, map center and zoom
  3. click download, you have the image with watermark. system also generates (and displays) URL for you so you can access the same settings (network, center and zoom) with a link.

(Chut Ko) #14

this is my take on the question of whether to copy the canvas.

do you guys know what i mean?

visually, i think it would be possible to create an overlay rectangle to display what is behind the dialogue. but i have to check