Hi @warren , kindly see this .

I have some initial implementation that i want to show you regarding fetching tweets and replying to tweets automatically . I have hosted that mini-project , should i add that to this proposal as well ? Thanks !

Sure, go ahead and add that in! Thanks!

Can you also link to the relevant portions of OmniAuth issues from the repository?

It'd be great also to link to, and collect ideas from, the relevant milestone: https://github.com/publiclab/plots2/milestones and the associated planning issue if there is one! Thanks!

@warren...I have added the demo project , kindly see that . Also added a youtube link :-)

Also i would like to know whether this Soc idea https://publiclab.org/wiki/gsoc-ideas#Interactive+Project+Maps requires Leaflet-Blurred-Location as a pre-requisite ? Thanks so much !

@mridulnagpal what do you think?

Added some wire-frames for more clarity .

@warren...actually i am interested in doing Social Media Integration idea (not the OAuth-based Login) but as coding period is very long so i am hoping to do part of Interactive Project Maps also . What do you think ? @mridulnagpal

@sagarpreet Interesting. I would like to work with you for the socialisation part. I am proposing an idea for OAuth based login. I want to deal with the facebook for the summer.

Hi @bansal_sidharth2996...Thanks so much :) Yes that would be great !

@sagarpreet @warren Sure I can use some help in upgrading leaflet blurred location. If @sagarpreet is willing to work on it as well, we can extend the features we aim on having plenty of other features as well.

@mridulnagpal...i would love to ! Thanks :) Can you also see the current proposal state . What do you think ?

I would suggest getting a little familiar with leaflet blurred location, and its demo. Also our main task would be adding information to our maps using tags, like project description, # of people, etc. Also we need to maintain subscribers list for regions, or boxes in our case. Please have a look at my proposal and leave any comments on what features we can enhance even further.Also the proposal looks really good to me. Thanks :)

Hi @warren , I have added the flowchart explaining the algorithm to fetch tweets and showing to our website and also replying to tweets .

Also found this interesting article on why Social Media Integration is important : https://economictimes.indiatimes.com/small-biz/sme-sector/power-of-social-media-a-win-win-marketing-strategy-for-smes/articleshow/60231774.cms

Ooh, i really like this item: Profile Map showing Near By Projects with a different marker

More soon when I do a full review, thank you!!!

A few ideas;

If we stopped using IFTTT to generate tweets, and did it via the API somehow, could we get the tweet ID directly that way? And if so, could we link the tweet id to the node with a node tag like twitter:_______ maybe?

Is 10 minutes based on a Twitter API limit? Just curious!

For the project map work, I'm interested in hearing from @stevie and @liz -- they have a lot of ideas about project and page mapping that I hope could be useful! And also see this related but old page: https://publiclab.org/places

Also see this suggestion I put up about how the "reply by tweet" might be similar in some ways to "reply by email"-- https://publiclab.org/notes/gauravano/02-18-2018/gsoc-proposal-email-integration-project#c18697

What do you think? Maybe reach out to @gauravano to discuss!

Thanks! Great list of important projects here!

Hi @warren

1.) I tried very hard to use the Twitter API and twitter gem to directly tweet .
But it seems they have depreciated the API to post images with tweets , only we can tweet text without image .

I even asked for help here https://github.com/sferik/twitter/issues/638 and tried the solution given by some good folks in that issue comments but still no luck . I will try some alternatives to post images as well in the next week .

Also you once said here : when we solved this tricky IFTTT issue :-) that you don't liked the idea of baking new channels of publication into our core code , which makes total sense .

The best solution and the fastest solution i think is just to update the tweets generated from RSS feeds to start with the node-id , for example :

Tweet-15850 via Ag8n: http://ift.tt/2FJ8tlw The difference between medical particle monitoring and your particle monitoring?

1.1) We can quickly parse 15850 and search in our Node-database and add power tag twitter:XXXXX (here XXXXX is the tweet-id) .

1.2) Similarly when someone comment on the any tweet , we fetch this comment => then calculate the parent using parent = client.status(i.in_reply_to_status_id) where i = comment object and then using again the node-id = 15850 which is in parent.title we can add this comment to our Comment database .

1.3) Now for adding comments from PublicLab to Corresponding tweet (this is exciting !) , we can simply use the Power-Tag twitter:XXXXX and Client.update( body of comment , reply_to: XXXXX) we can post the comment on that Tweet .

2.) Why 10 minutes ?

Actually i have hosted the demo website on Heroku and scheduler used is Heroku-scheduler which has minimum running time of 10 minutes . But we will use the Whenever gem , which can even run after every 1 minute ! @namangupta mentioned whenever gem documentation in his proposal here .

3.) And @warren , @liz , @stevie , please check the demo and let me know if anything needs to be added / modified .

Thank so much for feedback !

@namangupta , @gauravano , @vidit what do you all think about this ? Thanks !

On important thing to note in Twitter API is that we cannot get replies to a Tweet directly !

So the alternative is to fetch recent replies to @PublicLab on twitter and then follow the above flowchart in the proposal . @warren kindly see the demo in the proposal and let me know if anything need to be changed or modified .

Thanks !

@sagarpreet yes you are right that we cannot get the tweet's reply directly but i have some implementation regarding this. @warren told me to get into this reply-by-tweet concept so is it okay with you if i get into this problem?

https://publiclab.org/notes/namangupta/02-17-2018/gsoc-proposal

Sure ! Go Ahead ! It would be wonderful to work together :)

Do take a look on how i am implementing in the demo . Also it would be great if you can figure out a way to tweet a node using Twitter API (with IMAGE !) .

Thanks !

i will call you today or tomorrow to discuss about the feature if its fine with you and you are not busy?

Ok...I will share my personal number with you on gitter . But it would be great if you could call me on Thursday as i am a little busy with my University examinations .

Is it ok with you ? Thanks !

@sagarpreet we already have numbers ( coding ninjas august alchemy batch 2017 :p ). And lets connect on thursday and All the best for your exams.

Awesome ! And thanks :)

Hi @mkashyap354 ...I see you are a potential mentor for Social Media Integration idea (listed here ) . It would be great if you can review this proposal . Thanks !

hi @sagarpreet I go through you proposal and I have a question for you that why we use only twitter although we have other social site also like facebook etc. thanks

Thanks @mkashyap354 !

I asked the same question to @warren and @liz a long time ago :)

I guess the answer is that Facebook networks People while Twitter networks ideas and topics.

Hi @warren , @liz , @stevie !

I would really need your guidance on what Map related features should i focus in my proposal ?

Which current issues on Github related to Map integration should i add here ?

Also kindly see my above comment (the big one) on Twitter integration .

Thanks so much ! I really appreciate your help here :)

Hi, all -- still trying to keep up with the proposals... so many good ones!

As to Twitter vs. FB or others (just to explain for @mkashyap354), we hope to do integration with several, and I think it's fine to focus on one at a time. We've seen lots of people respond helpfully to questions from the site which we post (now automatically) to Twitter, and getting some of those replies to show up back on the site would be great.

I LOVE the flowchart and background work you've done on Twitter. I think this is pretty great. One idea i had though -- tell me what you think -- is to use the nid in the tweets we generate as part of the URL maybe? Or, as a hashtag -- then search for that hashtag -- in @publiclab tweets, like #15744 for this node. Could that make things slightly easier? Just trying to brainstorm a bit.

Regarding map related features, let's see... I think there are a lot of different projects going on in maps, including new layers, ways to show different types of content (users, nodes) -- I wonder if we ought to create a JS library for maps on Public Lab -- a way to generate a Leaflet map and pass it node or user ids, and have it produce the Leaflet map instance? That would help standardize mapping implementation across the site, too. We could spin this out as a sub-library in JavaScript, kind of like how we've done for:

• https://github.com/publiclab/PublicLab.Editor/
• https://github.com/publiclab/inline-markdown-editor/

I hope that's helpful feedback!

Thanks so much !

1.) The hashtag idea is great ! Yes we can do that .

2.) And regarding the JS library , I really think we should do that to standardize the use of maps at all places . Saying that , TBH i would need your help as to what will be the flow of code . I am comfortable with Leaflet library , Vanilla Javascript and Jquery , have basic knowlege of NodeJS and NPM and gruntJS and Jasmine Framework for testing .

By studying this awesome library : https://github.com/publiclab/PublicLab.Editor , I am able to understand how this works . I am sure with the time we have before the summer and with your guidance i can learn how to create consistent and maintainable JS library .

This will be exciting and i would love the opportunity to create a JS library from scratch for PublicLab ! Thanks :)

I really liked your timeline. It has given priorities to each issue according to the need of issue. I also like your flow charts a lot they explain the logic very easily.

Thank you @ bansal_sidharth2996 !

For maps-layer JS library , i was thinking of making independent modules like :
1.) Layers module .

2.) Full-Screen module

3.) Near-By projects module

4.) Near-By people module

5.) And a Main-Module whose instance will be made .

Also including features like ( Using Leaflet library (i have some initial implementation for these) ): * Refresh Map on panning and zooming * Finite Scrolling and panning of maps * Location - Security

CODE FLOW / STEPS :

1.) Creating src , dist , test and example folder .

2.) Creating independent modules in src folder .

3.) Creating example HTML files in example folder to test different map modules .

4.) Using Jasmine framework for testing to be put in test folder .

5.) Using gruntJS plugins for concatenating and compressing modules into one single JS file .

6.) Publishing the library to npm .

7.) For increasing code - maintainability and usability , we can use : https://coveralls.io/ and https://travis-ci.org .

I was thinking of starting the work before the actual coding period starts like from 24 April as i am already a contributor for PublicLab from a long-time , hence skipping the community-bonding period exercise , only if you all agree with this :)

What do you think @warren ? Thanks !

Thank you @ bansal_sidharth2996 !

For maps-layer JS library , i was thinking of making independent modules like :
1.) Layers module .

2.) Full-Screen module

3.) Near-By projects module

4.) Near-By people module

5.) And a Main-Module whose instance will be made .

Also including features like ( Using Leaflet library (i have some initial implementation for these) ):

• Refresh Map on panning and zooming

• Finite Scrolling and panning of maps

• Location - Security

CODE FLOW / STEPS :

1.) Creating src , dist , test and example folder .

2.) Creating independent modules in src folder .

3.) Creating example HTML files in example folder to test different map modules .

4.) Using Jasmine framework for testing to be put in test folder .

5.) Using gruntJS plugins for concatenating and compressing modules into one single JS file .

6.) Publishing the library to npm .

7.) For increasing code - maintainability and usability , we can use : https://coveralls.io/ and https://travis-ci.org .

8.) Integrating with plots2 .

I was thinking of starting the work before the actual coding period starts like from 24 April as i am already a contributor for PublicLab from a long-time , hence skipping the community-bonding period exercise , only if you all agree with this :)

What do you think @warren ?

Thanks !

I also do agree with @sagarpreet, I know public lab since last yr I drafted my first proposal(eventhough was not selected at that time for GSoC) and contributing since October, so we guys don't require any community bonding period. We can start the work prior to that if the mentors would wish and we got selected.

I agree with @sagarpreet and @bansal_sidharth2996 , we can start coding earlier too.

:-) of course you can!

Regarding the leaflet library -- i think it's best to think about it in terms of the inputs and outputs. So the docs would say something like:

Something like that -- make sense? Then we just have to add a few simple lines to each place we make/use a map.

Maybe looking at an example of another Leaflet map library would make this even easier, like addTo(map) syntax: https://gitlab.com/IvanSanchez/Leaflet.GridLayer.GoogleMutant

Yeah this is a cool idea.

Just for learning purposes i made this demo JS library :-)

https://github.com/sagarpreet-chadha/Leaflet_learning

• Contaning only single module SkyTruth Alert .
• Following modular approach .

@warren ... kindly see this . Thanks !

Also should i update my proposal , the maps part ? Showing doc/API of JS library maybe ?

Oh wow, cool! If you make a gh-pages branch and (on that branch) check in your node modules, we can try it out! There are some gotchas so give it a try and I can help you along the way. Don't get too frustrated!

Sure, this is a great addition!

Thanks so much @warren !

Ok so i tried hosting using the gh-pages branch . The link to Webpage is :

https://sagarpreet-chadha.github.io/Leaflet_learning/

However i have changed the cdn links to the actual links and not to the one's locally stored in node_modules folder in index.html file , i think node_modules folder is very large and hence not uploaded to github server or maybe it is still uploading ?

If the node_modules folder is still uploading , then this webpage should work after some time , https://sagarpreet-chadha.github.io/Leaflet_learning/example/example1.html , right ?

What do you think ?

And will try to update my proposal today . Thanks !

I have updated the proposal here , added :

• Mockups
• API/Doc of the JS library

And updated my timeline here :

https://drive.google.com/file/d/1n-eZ9_BBfhiTAFsWr479G9miKVa3iFWT/view?usp=sharing

Kindly see the above PDF link so that i can submit my final Proposal , the timeline consists of approx. 2 months work on maps and remaining 1 month work on reply by twitter feature .

Thanks :-)

Great proposal @sagarpreet . The best!!

Thanks @vidit ! Means a lot !

Hi @warren !

So i found this JS library https://github.com/MatthewBarker/leaflet-wikipedia/blob/master/source/leaflet-wikipedia.js
with demo here :
http://matthewbarker.github.io/leaflet-wikipedia/

This fetches from wikipedia API and shows markers on map ! Also it removes marker outside the current bounding-map .

I will try to learn from this library and try to understand how the Leaflet's classes are getting extended here and how the Leaflet class functions are getting overwritten .

And the demo i made above got 43 downloads in 4 days on npm ! Yay ! Thank you !!!

Wow, very cool! Yes, this looks like a great library to study. Great work!