# Specere Milli Conversations
###### tags: `Documentation`
## Tools and Wikis:
## User Stories
Since the last one started as user story and became a epic. Let's define user stories which are short and simple things in the below format, scoped to science archives, or other some specific scope could be suggested. if we can add more stories, it should help us with a list of features which we can prioritize by story points that we ll assign after having a small sample of stories.
> As a [type of user or role] i want [some goal] so that [some reason]
### Milli User Stories
* As an "Science student" i want better finding aid so that i will be able to browse the archival metadata descriptions of different archives
* As a research associate i want to add tag to a particular archive item metadata description, so that i can browse my tags later
* As a researcher i want to add text comments to a particular archive item metadata description, so that i may quote in my research paper.
* As a science journalist i want to link two archival items with a custom relationship so that i can analyze emerging patterns for stories
* As an archivist i want to create categories so that i can organize the archival research
* As milli platform owner i want to create new Sites / Accounts so that the contracts, policy of interaction or integration with other institutions or archives or formal orgs can be managed
* As milli Technical admin i want to setup background processes or jobs for data exchange, parsing or any other adapters
* As milli site (Account) admin i want to create and manage users
* As a milli site admin i want to approve or reject annotations(tags, comments) so that there is no spams
### Specere User Stories
* As a research intern i want to write exhibits on archival artefacts so that audience can access the material
## Features - derived from user stories exercise
The major user personas that came out from the user stories exercise are,
* research associate
* science journalist
* milli platform owner
* milli Technical admin
* milli site (Account) admin
Above can be boiled down into the following roles that would interact with the platform
* User (without any registration)
* Researcher (All curators, story tellers)
* Site Admin
* Add / remove users
* Add / remove annotations
* Platform Owner
* Add / remove sites (via. contracts)
* create issues for platform admin related to technical problems
* Platform Admin
* Manages integration and other technical requirements and creates software development contracts
## Architecture Discussions
* Around the Diagram Venkat shared with functions flow of the platform
* Need to de-structure the diagram in terms of services and dependencies (whiteboard pic to be added)
* Need also a User Story to work on datamodel
* Too many moving pieces
* Annotation based data model to be able to de-couple from the domain dependencies
* Archival description is a global standard. It is adopted by all the 100s of countries around the world and there is a manual for this: https://www.ica.org/sites/default/files/CBPS_2000_Guidelines_ISAD%28G%29_Second-edition_EN.pdf
* You DON'T need to read it. Just saying that this is the book that guides most archival description implementations all around the world.
* Tags need to conform to what are called archival 'access points' and 'name agents'. Each object can have access points, that is, tags related to names of people, organizations, locations, and subject areas. For example, have a look at the archival access points on the right column on this page: https://digitalcollections.ohs.org/coll-189-vancouver-avenue-first-baptist-church-collection-1940-2015
* We can have two sources for these access points (tags).
* One can be the large source from Library of Congress:
* (a) Names of people, places: http://id.loc.gov/authorities/names.html
* (b) Subjects: http://id.loc.gov/authorities/subjects.html
* Keep in mind that for these, LoC has already created a tech space to download and use this data on any local system: https://id.loc.gov/techcenter/
* Second can be the internal Milli sheet of material based on user input. E.g.: Indigenous communities have populated material with their own way of thinking and describing. It may have the same ontology, or it may have other kinds of indexes. eg: Instead of describing an object by name, place, subject, date, etc, they may describe objects by 'emotion it induces', 'colour thought it sparks', etc.
by prasoon on architecture
## Frontend discussion
- Browser extension: to be able to annotate while browsing
- Dashboard: To Manage Annotations and workflow
- Roles: ["Admin", "User"] - User will be able to add annos to one anno collection, let's call `Pending Review Collection` admin will be able to move / add anno to all anno collection `Approved Collection` or `Rejected Collection`
- Index Front Page - Finding Aid (Discovery & search), Show Annotations
by bhanu on features / mockup
## Server side discussions
Model Reference https://www.w3.org/TR/annotation-model
Potential candidate for annotation server - https://github.com/dlcs/elucidate-server
Hasura engine with Postgres and Graphql
Models and diagrams - TBA
Papad is the audio interface
papad backend gitlab url https://gitlab.com/servelots/papad-api
~~Due Date: 24th April~~
Due Date: 14th May
Papad Frontend - features - tagging, adding a audio, creating a channel, add cover image
Papad Backend - Vignesh is working on a Django prototype,
- Above milestones with something showable before 24th
- Testing and ticking through the features lists of front end
- Testing backend for performance and scale
> [name=bhanu] Date: 12/05/20
> Update: Though the frontend was ready, we were unable to deploy due to delays in backend development
Due Date: TBA
* changes to layout and styling
* Upgrade to audio player to allow timestamped tagging metadata, and the pipeline that comes from One's testing
* Upgrades to data model to support time-stamp annos
* Depends on testing from one
> [name=bhanu] Date:12/05/20
> The integration with backend in real time a dependency, blocked progress on this one. have spoken to both athithya and vignesh on how to loosen the dependency from backend. and this week am planning to spend some time in getting this milestone moving.
### Graphql backend
Due Date: TBA
Prasoon is able to get his code working to provide graphql endpoints for the discovery platform metadata serializing and more tba
### Tags index for auto complete search filter
Having a service which can be configured to fetch from topic model services like LOC or dbpedia
How to do it? what more will it take?
p: should not be hard to build api for static data.
find link or api for LOC
## Minutes of meeting and brainstorming
Summary: a short report on our vocab discussions to come up.
### call on April 9th
* Started with general discussion
* Satakshi and shafali have presented story telling UI page
* Prasoon facing some issues in server side. Venkat asked prasoon to contact Janastu people
### call on April 6th
* Micah and shafali presented Shatakshi story telling in UI design perspective view
* Shatakshi explained we need to explore #tag exploration
* Venkat comment "If I havent shared it before, have a look at Nieman Storyboard. Some things to adopt, drop from there. I think you'll see some minor similarities.
* Dinesh said we have to think like provisional way to end user
* Three tasks:
* Tell a good story
* Show connection between objects
* Show how stories are unique space to think about seeing objects
* venkat comment 'satakshi, one key thing: just make sure that all objects you use are coordinated with Prasoon for back-end ingestion. They should all have the same metadata fields like the rest of the ingest data. Some objects are not in the initial partner ingest. Story objects should be watertight since they will have highest visibility.'
* venkat said "this is the 'video' we were talking about from earlier. Something generic that will have permanence. https://pelagios.org/
But it doesnt need to be video. It can be static sketch art. Just something better than the clunky blue flowchart here:
* High level new #deadline #milestone , wherever this needs to go
* Mar 30: Story design deadlines set (fill in dates below)
* Apr 6: Story handover
* Apr 15: server-UI integration, front end for designed discovery/annotation pages
* Apr 22: Vocabulary integration
* Apr 24: UI vocab integration
* Apr 27: sanity Check D0-D9
* Apr 29: Five new users, with tasks
* May 20: User testing complete
* May 27: Stop code, only publicity elements hereon
* June 3: Milli 1.0 is Live for public use as part of Milli Sessions 2.
#### Shared Link
* Satakshi comment 'Talk to Nikhil Nagaraj <
email@example.com about sound files. He is a brilliant sound engineer. He may have ideas. He usually composes his own sound for most situations, but maybe he has a database of his own work to share if you tell him open source, public project, etc.'
### call on March 26th
* General discussion with Team
* Shafali and Shatakshi explaining shatakshi story telling
* venkat comments on hubs would work well for this, but i worry about bandwidth and such.
* I wonder if this can be a 'made' version of panellum/spacetour, bhanu?
* Bhanu replied "Depends on what kind of media content we have, ex. 360 images needed for space tour"
* slight disagreement, Prasoon. Seems like non-linear might work well. Would these be our intents from the story, in this order?
- Tell a good story around a random walk, a meandering through archives.
- Show connections between objects in different archival repositories
- Show how Milli Stories are a unique space to think/rethink how we see historical objects.
- All this said, I am not sure I got what linear or non-linear mean here. A linear walk through the history of random walks? It's a rumination on random walks, isnt it, and so what is the linearity/non linearity in play here?
- Prasoon comment "That's why the AB testing to test the hypothesis. Or we could decide based on general agreement here, it's a diverse group, we could certainly do that."
- Micah comment Linearity in the interface of human and story
- Venkat comment on what is everyone's take on this as being goals for the story:
- Tell a good story around a random walk, a meandering through archives.
- Show connections between objects in different archival repositories (Are we doing this?)
- Show how Milli Stories are a unique space to think/rethink how we see historical objects.
- Venkat comment Here is a vague spec for Milli Stories (going forward) and the prototype story: Milli shows how a user can browse through its archival objects, annotate them, add them to their storytelling carts, find patterns among the objects through their annotations, and just come up with a story that connects very different objects in different archives. Milli is what a user makes of it.
- In all seriousness, if prototype story can do this, we are done: SHOW patterns among the objects through their annotations, and come up with a story that connects very different objects in different archives.
### call on March 23rd
* Joel have presented Milli page to team members
* Prasoon comment on Db-application connection is what I'm trying to sort. DB is live on a docker. Will be able to deliver authentication/authorisation before we end this week. I can assist with integrating with the UI right away (might have to run the app in a local docker considering the security risks without proper testing).
* Prasoon shared docker to Bhanu. Bhanu having some doubts on endpoints level , he clarified with Prasoon
* Dinesh and Venkat have discussed subject headings in annotation and cryptography
* Bhanu added his point annotation database from the backend that is object and for the frontend we have to query the object
* Prasooon and Venkat have discussed about Search functionality and user profile search. Prasoon said some of the queries will respond later so that from frontend we have to make the query and make the responsible more faster
* Bhanu comment on "Omeka search query Re. 13 ways"
* Teams have discussed about 26th March deliverable
* Bhanu expected to complete the static sides to complete by Joel and prasoon have explained what he have deliver¬ deliverable within this date
* Bhanu comment "Right now on our server we have .env files"
* Dinesh said dont change the deliverable plans except date
* Prasoon said we need participate for making the doc from the tech side
* Bhanu explaining the delivery plan: Test server-UI integration and UI by Mid-April.
* Update about the backend:Application is dockerized, will be available on the docker hub after fixing some DB session leaks. Was planning to be done by now but, I've accumulated a lot of tech debt in the last 2 cuz of the rush so removing some obvious code smells.
- Items to be delivered over the next 3 weeks:
- Test data population
- Milli vocabulary (authority and subject headings) if we can freeze the discussion
### call on March 16th
* Bhanu not ok with T-10 prototype because we need to do HTML page and interactive things to the static page
* Joel getting clear with Milli design
* Prasoon checking the Data and will make the csv file and also for the admin login we can create 1admin account while developing the application
* Shafali an Prasoon discussed about plus mail service and prasoon said that it will limit to gmail user
* Bhanu comment on "Designers will do desktop prototype for t-10 & Dev shud see how much of it can be responsive"
* Venkat, Joel and Micah have discussed object Page and Annotation part
* Venkat suggested for the prototype1 we will do data dump
* Prasoon said that we should have data authority file
* Prasoon, Bhanu and Joel have discussed User, admin role
* Bhanu have explained the close functionality with toaster example
* micah, venkat have discussed about Story telling part
##### Shared link
### call on March 9th
* Sathakshi have updated Salim Mali history on archive perspective view
* Satakshi shared her experience to read the salim mali article.
* Venkat asked how to deployed sathakshi miro board work , written work on milli for the story telling behind the stories
* Micah explained how to annotate the document with action of words, years, keywords
* Shafali have explained **Get involved** UI page
* For the donate page once end users wants to know where it goes
* Joel, bhanu and prasoon have discuss documentation for all things.
* Joel confirm that will start with tech this weekend and bhanu also wants to contribute
* These are the two articles that me and shafali wrote for Compost which is a new publication about digital commons
### call on March 5th
* started with general discussion
* Venkat explaining narrative concept to Joel
* prasoon asked someone will look into the front end develop code from API point view
* Satakshi updating shaliman alli archive on narrative perspective view
* Venkat comment "This could be a story about inflection points.How small moments lead to bigger ideas. Sometimes, we look back and say 'that was THE moment that changed things'. We want these moments to mean something. It gets into the idea of meanderings, and allows us to 'technically' explore objects in the archive. Working title: A random walk"
* Micah and Shafali looked into Object page, now they are designing Partner Page
* firstname.lastname@example.org : Dan Gillean, AtoM Program Manager
(About: Little MVC PHP/Framework) used by AToM.
### call on Feb 23rd
* Started with General Discussion
* Venk comment : Draft Milli metadata (standard sheet for all levels) * = ISAD essential, & = Milli essential
* Level of description [Drop down, fixed list]
* Extent (cubic metres, use number. If single object: NA)
* Creator (tie to Access point: Name; alphanumeric)
* Date (acceptable: YYYY-MM-DD, YYYY, YYYY-MM)
&*Partner Reference Code (alphanumeric)
&Partner or Repository (alphanumeric)
Milli Reference Code (TBD)
Type [Drop down, fixed list]
Language [Drop down, fixed list]
Description/scope and content (alphanumeric)
Rights [Drop down, fixed list] (this will be a useful filter going forward)
Milli URI (url validator)
Access points: Name (comma separated values?) (alphanumeric)
Access points: Subject (comma separated values?) (alphanumeric)
Partner URI (url validator)
Unsure:Physical characteristics/dimensions (alphanumeric)
* Examples of archival description fields:
* In case nobody saved it, **Decided filter fields:** General (small search box for creator/title/scope and content: type query for refinement. or have drop down for specific field search from these three)
Date (sliders, bucket years, etc)
Type (img, audio, video, etc)
Access point: Name (places and people. Pull from controlled vocabulary: similar to LoC subject name authority)
Access point: Subject (pull from controlled vocabulary: similar to LoC subject headings)
Language (drop down list / name with counter)
Annotation type (toggle: user/ golden record)
Partner (drop down list / name with counter)
* Examples of consortia search filter options:
- Oregon does exactly what we are discussing: 'person' filter in two different fields:
- Europeana, nice UI maybe?
- Good ole DPLA:
* Prasoon said, for injection properties there are several ways of representing, users dont have to interact with Json
* Bhanu said copying from omeka, auto mapping with headers and controlling vocabulary
* no need to upload data file
* D1 search UI discussion
* will start D1 UI development
### call on Feb 12th
* Team member suggested will use #tag in signal to maintaining the important messages and links shared by the team Milli Group
* Micah suggested #millilink #millireflection and venkat suggested #mixie
* Micah have question how we are going to handle users stories
* Joel explained the user stories role
* prasoon said we have the epic to divide the task
* Venkat, Micah and Shafali what are all the data should annotate
* Venkat explained authority names and subject to Micah and shafali using NCBS archive https://archives.ncbs.res.in/node/801
* For the filtering, we can filter by date, filter by author, filter based on category
* If we see the milli vocabulary we have so many community words
* If we click the author name , it should redirect to particular author other objects also
* **Deliverable D0:**
* Prasoon said API for EAD3 and Data Ingestion from the XML is done. CSV export is pending
* For the Header protocol to automapping the Omeka method. instructions to parse through json. lambda functions and stuff.
* 6main everything as an annotation allows for scaling to any number of fields. For those who consume API, not much is different other than scale.
* CRUD UI may be not sure, has not started
* **Deliverable D1**
* Searching through are done
#### Shared link:
* Authority Name: https://id.loc.gov/authorities/names.html
* Subjects: https://id.loc.gov/authorities/subjects.html
* Back to basics, let's not reinvent. https://dp.la/search?q=india
* Look at this http://pdf.oac.cdlib.org/pdf/caltech/benzer.pdf and scroll through pages 2-4. At page 4 or 5, see field called 'subjects'
### call on Feb 9th
* started with general discussion about miro,mail Id
* satakshi explained 1975s cartoons archives in miro board
* satakshi said there are two schematics way to narrative 1. cartoon, images 2. cities space and changes in environment
* Venkat personally struggling on how to think about narrative
* micah comment : mixing what venkat and prasoon are saying, you doing the comparision between the two approaches would probably give us a lot of vocabulary for microinteractions in the storytelling and annotation layer
* Dinesh said if we take a story structure as abstract how we will think as milli
* Venkat comment :Thinking thru....what should be the tone of the story?
1. Explanatory / historical this, then this, then this
3. Reflective: Essay-ish in tone
4. Straight up pageturner: Conflict, narrative arc, then what happened, how did it make you feel?
I think this prototype Milli story needs to have the tone of a reflection, a musing. That is what we can pull off.
This reflection is around some ideas, and how the storyteller is forging connections across very different objects.
* on the otherside of the world.. CS people struggling to make sense of semantic models
* movement as a better storytelling instrument other than conflict
* Venkat comment: Micah, I agree with you. However, the reason I suggested writing was because I saw it as a starting point for a prototype. Playing to strengths. And we could then, well have three ways of telling the same story as our prototype: a Miro-like visual, a written essay with one image, and a peopleplex.
* Prasoon comment : I did not think of Miro as an investigation into a future widget. I was just talking about vocabulary/semantics of the data exchange. UX and I are miserable together.
* micah comment: also satakshi look at korsakow as a way to think about meta-narrative process for us
##### shared links
### call on Feb 5th
* General Discussion
* Micah sharing the Miro board to explaining the UI
* They also created Object UI and annotating page
* Venkat suggest for the Item level we can keep image
* Micah,venkat, shatakshi discussing how annotating text should display once the text is annotated
* Prasoon and venkat suggest will place some icons over the annotated text
* venkat comment : I dont still get how we are annotating a single word in the field 'title'. What am I missing? I guess I am wondering how this neatly feeds back in discovery.
* Dinesh suggest the private and public annotating document. If the user login to his/her own account then he is having the 2options to annotate the document 1. private 2. public annotation document
* Micah comment : personal annos as a downloadable file? that can be uploaded again to use on the platform when needed
* Joel will comeup with Image Annotation on Tuesday
* Prasoon, Venkat Dinesh said Linking is necessary but hard part of this linking to another archive or bookmark the document is difficult
* Prasoon said if we are showing Recommendation link we should be careful there
* Prasoon and bhanu said that target would have selectors
* Venkat said title,date - annotation to the object & link should be the annotation to another object
* Prasoon said API is already there I have to make scalable
* Bhanu said once home page design is confirmed will start the developing
* Shafali arises the question Are we think about Mobile responsive for the Milli website?
* Bhanu said we have to consider mobile view also
* We have to thing about motivation for the field/value - 2ways to motivate 1. Description 2. commenting
* Private and Public annotation
#### Shared Link
* **Milli main:** https://www.figma.com/file/Gc02lSEv5QmWZWqiSLl1tY/Milli-Main?node-id=0%3A1
### call on Feb 2nd
* Started with General Discussion
* Venkat comment: Digression stmonk - Free and open source self-hosted newsletter and mailing list manager https://listmonk.app/
* Prasoon comment "I would volunteer to maintain a list serv if people are alright with a few unschedued downtime"
* Dinesh shared list [WSMDiscuss] The birth of this list, fifteen years ago, January 10 2006 https://lists.openspaceforum.net/pipermail/wsm-discuss/2021-January/005229.html
* lists.openspaceforum.net Mailing Lists
* Dinesh said we stop mailing list service becausse of our server comes down, liability
* Bhanu will call today Joel
* Dinesh explained BCP annotation, protocol, renarration BCP renos – Dropbox Paper https://paper.dropbox.com/doc/BCP-renos-7nb3aY5SrGGsvegMLlfWW
* shafali suggested to add the content "How papad will helpful for Biotech protocol?"
* Dinesh gave example for the renarration Eg: Raika Bio-cultural Protocol http://mitan.in/bcp/raika/
* Dinesh thinking how to relate this reneration document, structural dicument, BCP document
* Dinesh updated Janatu **Film Festival** upcoming event
#### shared links:
* make - Mailing Lists [Savannah]
* Kailash Nadh's personal homepage - Developer / Tinkerer / Absurdist
* Dropbox Paper
* BCP renos – Dropbox Paper
* Dropbox Paper
### call on Jan 29
* General discussion about Gardening, Vaccination
* Prasoon explaining shared links on Signal
* Prasson explained Muckrack.com functionality
* Dinesh said we have to keep mind that successfulll searching algorithm(Ranking algorithm)
* Prasson comment: "We need an alternative to page rank"
* Bhanu explained the singular and plural tag searching/filtering isue on papad
* Shafali said basically am looking on the object for tagging and comments, we will have tag inside the comment
* Venkat comment "True twitter popularize it but IRC had it too?"History of Hashtags: How a Symbol Changed the Way We Search & Share
* Prasoon said twitter having #tags. We have to think like from the markup
* Shatakshi said how data should accessible to data. By using keyword to search the article in different period, she is thinking how to interconnect this too.
* Shathakshi said am trying to compare schematic way of NCBS archive and long emergency most of the article on Education, Muslim kids, religious based on Eucation and population space have to think about it.
* Venkat said we have to thing about Nodal graph to connect the node from 1975 archive to NCbs archive and also he said we will figure out permission later
* Shafali said we did object UI page, and she said ant design framework does not have some features like icons, link icons, small button
* Bhanu said we can do with css
### call on Jan 22nd
* Bhanu liked antdesign https://www.antdv.com/docs/vue/introduce-cn/
* Micah said we have to choose brand color. Team decided to go with Green and white
* Prasoon and Bhanu said we should go with object portrait we make it as option or space now will catch up later
* Shathakshi look into NCBS archives and long emergency archives
* Venkat explaining metadata with different timing .
* This, for instance, is a narrative around a set of objects with the user annotation: 'processes'. The original archive may not have this annotation.
* Prasoon comment: It is a meta-cognition capturing. Might not be very easy to do.
* Joel Introduction and he explaining understanding of NCBS
* Micah and Bhanu explaining where we need front end developer help
* Joel, want to have a look at this info architecture developed thanks to Shafali and Micah. Any thoughts? Not a trick question. Just asking.
And thanks, all, for teaching me a new phrase, tech debt.
* Joel, you can also look at
(the doc file there) to see some high level concepts and specifying for Milli
Q: Why is this email a pre-determined number sentences or less?
Ant design: https://www.antdv.com/docs/vue/introduce-cn/
### call on Jan 12th
* Prasoon will come Bangalore 19Jan
* Bhanu shown servelots office
* Bhanu and Venkat discussed about tomorrow meeting with HS how to go. Venkat not available tomorrow.
* consuming records prasoon wants to add to the deleiver0
* We need different types of resources, CRUD operation, filter functionality
* For scaling we need there is a standard vocabulary and also not standard vocabulary. We will play with standard vocabulary
* Bhanu suggested for the demo think we will go with Auth0 service. Auth0 service will give 7000users as free service Unique Id.
* Venkat suggested we need annotation toggle on and off event for annotation
* For annotation, instead of toggle Micah suggested pad.ma
* Two kinds of story outputs from theme '1975' in prototype1. (a) A blogpost that has audio, video, image objects pulled from milli objects: Copy this.
* (b) Node-edge visualization that uses a spreadsheet input and gives a node graph output.
* We need story for the modelling
* Prasoon comment : How many clicks from Object/annotation to story? Explainer wont be more than 15 seconds.
* Venkat shown the Milli story telling presentation to team
* Shafali and Micah asking intern for the animation part. Venkat asked them to recruit from juniors students
* Venkat comment I think we can rehash a few phrases from this, https://bit.ly/Milli-tech
* For login, we have to manage Id from backend - we have to think about User roles and permissions
* D0 - Data ingest
* CRUD - https://en.wikipedia.org/wiki/Create,_read,_update_and_delete
* Oregon Historical Society ref that Sh made:
* https://pad.ma/grid/title/list==shaina:Slum_Bombay_%281991%29_:_The_Footage_and_the_Film - discovery level
* https://pad.ma/CDH/player - object level
### call on Jan 8th
* Consuming omeka API, converting resources to Milli resources.
* Shreyas - BHC archives
* Dates have to be decided, scaling roadmap
* Milli.link have to host examples, prototypes in text, visual
* The link between annotations and archive descriptions - crosswalks between the two
* Blog, micro-blog, twitter, (byte-size)
* T minus 11 week - March 26
* Milli week was last year June, we should be able to do it again because it was successful
* March will allow us to finesse it
* **Decision** -
* prototype 1 is the current partner
* smaller repos are also fine; sizable amount to search through
* #UIUX - bring annotation language to search
* D0 - Data ingest and parsing
* D1 - also utilise annotation vocabulary in filter, Filters for different types of resources; search through golden records and annos db; choose vocab that allows for scaling;
* D2 - Search Result page has an option to annotate - at different levels and motivations
* D3 - - login, key, persona(?)
* D4 - Description page - annotatable, Look at user annos, modify , add , edit, delete
* D5 - d4 -> d1; feedback loop; backend for it: annotation to discovery
* D6 - 1975 story as a demo, saved a story pin, album, using a storytelling widget called peopleplex; see the story; your saved story pins will show "also featured on story x" ; Deconstructed version of the story; Proof-of-use; demo of what could be done if the user could use it
* D7 - About Milli, explainer
* D8 - Text Narrative about web annos and archives and storytelling
* D9 - Easter Egg about Satakshi being User 0
* **Scaling to P2 and P3** -
* How do we deal with quality and moderation (people are approaching it with good faith - prasoon against)
* User & ID management - Decentralized and distributed ID provider
### call on Dec 29th
* https://docs.google.com/spreadsheets/d/1Nk65trki8uix2RI6_vMmDNPKLKVBXXj2A9HbLK5ulTo/edit?usp=sharing cv link
* Venkat, bhanu and Dinesh discussing CV updates
* First will proceed the Interview process with Hariom Soni, Shivani Thakur, Swaroopa
* Will come up with some exercise for the Interview persons
* Bhanu suggested instead of giving UI or Frameworks will give some exercise to do.
* Athithya comment my suggesstion for the interview question Fizz Buzz problem
* On Jan 5th 2021 will conduct Interview for them at 3pm
* room 1 - prasoon, bhanu, venkat, shafali will take it forward
room 2 - athithya, dinesh, micah, priyanka,satakshi
will take it forward
### call on Dec 22nd
* Dinesh have updated Durgadahalli Works, Village Panchayat Elections
* Venkat and Dinesh updating CV for the front end developer
* https://thoughtput.in/ , https://altcampus.io/ - interview person reference
* This is the hasjob post:
* CV for the Front End Developer:
* Prasoon and Bhanu discussing the Interview Selection process.
* Bhanu will share the hasjob link to Jaga people also https://www.jaaga.in/about
### call on Dec 18th
* Taiga link: https://tree.taiga.io/project/prasoon-milli/kanban
* Exploring Taiga with team members
* Bhanu invited all members on Taiga
* Venkat confirm the FronDeveloper recruit content to the team
* Figma design link https://www.figma.com/file/gDqIy5VF6JAPAUZGGenfkJ/Milli-Future-Design
* Bhanu confirming the Informtion Archive UI flow with micah
* Bhanu assigning frontend todolist on Taiga
* Venkat comment: Search, select, annotate, narrate
* Narrating will be slightly different: You select, but you effectively create a new archival record by creating a story.
* Dinesh, venkat, prasoon discussed Narration part
* Prasoon said at any point we need notification alert
* Micah asked we can scaleup the prototype timings
* Bhanu explaining the flow. Frontend will make wthout backend support after UI develop will integrate the API
### call on Dec 15th
* Bhanu and Dinesh discussed Elastic search Issues
* Dinesh updating fibre and wifi connectivity on village.
* In the meantime, dinesh, generally know of Tarana wireless?
* Prasoon explanining API Architecture
* We have to keep mind on NioxsEntity link to the Annotation
* Prasoon trying to create Entity(object) and then that link to annotation on Backend
* An annotaion represented link/ comment
* GraphQl is a 1code we can reuse it. GraphQl can query whatever we need based on the type
* Bhanu explaining the Graphql with Messaging app.
* Prasoon said if we need specific result to filter and make a developer life easy Graphql is helping smart
* Neo4j is running as an Embeded Database. Now prasoon using Neo4j to access the DB.
* Prasooon updating his GraphQl issues. Graphql does not allowing characters
* Venkat updated shathasi updates
* Dinesh thinking using GraphQl will look into the Graph Traversal, pattern
* Prasoon said Good example is filtering the comments thumbnails with like symbol
* Venkat discussing Milli Annual Report and he thought to technical platform and prasoon suggested techincal discussion
### call on Dec 8th
* Prasoon almost completed Annotation API its working, maybe next week he will give demo.
* Prasoon saying from the UI whatever you are adding comment will go as annotation to the backend
* Micah said we are following W3c model, so text wi ll go as annotation
* Micah said, if user visit the translate page then backend does n't know the translation
* Prasoon said backend will take care about this
* Venkat and Micah discussed item level vanila UI. Micah and Shafali thinking to add the comment and suggest the comment on flipping animation mode
* Team discussed with UI flow
* Micah said comment comes like anno to anno,object to anno, object to object, anno to object in different color
* Exploring Omega collection UI and Import & Export features https://digitalcollections.ohs.org/371n5278
* Venkat wont available on Friday NCBS conference is there https://www.ncbs.res.in/events/apls-apls-20201211-library-sundar
* Venkat talked shathasi and he shared figma link. She is interest
### call on Dec 4th
* Shathakshi shared CV https://docs.google.com/document/d/1Oknscf95IGyZIso89Qjo9dUTRHDzJQWarIUtT9gxOMg/edit
* Venkat comment on
* 1. Our personal motivations
* 2. Defining structure/views for Milli
* 3. Milli: Society, Trust or Section 8 non profit
* 4. Milli initially: Inflow/outflow of money and projects
* 5. Social Alpha support toward taking on people specifically for a short contract. So: archivist, software architecture? This may help start getting other sources?
* Not quite Milli tech. But I want to discuss a few points with you regarding network and administration of Milli. And I'd like to get your thoughts and concerns and advice.
For Point # 3, I think we need a separate Milli Tue/Fri discussion.
Listed in order of immediacy:
1. A listserv for Milli. I think we need a few of these. Milli tech is running. Another for milli members (people who submit material, etc. Think of this as a partner list). And a third for general milli announcements -- seminars, workshops, events, links. I think the first two should allow two-way communication and the third can be a one-way broadcast (to reduce spam).
2. For our chat tomorrow, Nov 17, can Janastu put together a workflow for the next 'product', and how it fits within the pipeline of the blue flowchart?
I imagine we are talking about a release in January 2021?
We also need to identify some 'deliverable' from our NCBS work (the contract that was delivered in March 2020) just so we know how to define what was done with the contract. Will need for auditing, etc. Maybe this Jan 2021 thing can fit as the 'product'?
3. Setting up Milli as an independent non profit Section 8 company. This will allow it to have some independence from any of the incubating spaces, including NCBS. And it will also help define a structure for flow of money in and out of Milli. I have spoken to a few people who have laid out the pros and cons of Trusts, societies, companies and I think the Section 8 company approach may be best.
I have got verbal yesses from about 5-6 institutions for being a member. I think we can look into a model where member institutions give a nominal fee to be a member? We'll have to figure out what that should be. But to start, we could have an MoU or an agreement (especially necessary for the larger govt-type institutes) that defines what they intend to be part of and what they agree to give to a Milli platform. I have been wondering what the governance structure of a Milli should be like and how to build in inclusivity + structure + efficiency. But to start, we do need this signed document from institutions. I think it will give Milli more credibility and pathways for grants. I have also now received a verbal yes for a small grant toward Milli tech work starting early next year.
- Grants: govt and quasi govt (various world govts 😃)
- Philanthropic sources
- Curriculum projects
- Individuals donations through an option on the website
Envisioned outflow (as you will see, there is more outflow than inflow 😃:
- Administration of Milli
- Annual conference (maybe June every year?)
- Tech team
- Archiving and standards team
- Contracts for conservation
- Travel and logistics
* Bhanu shared his experience on Milli world
* We are solving an interesting problem, using tech that's hopefully gonna have a big impact in determining where the Internet is heading.
* Shafali shared her experience. Milli make easier to archive world
* Prasoon said our product its like mastadon. People can use our instance they can develop their own
* Micah comment: another strategy that shreyas is following at LLN is signing MoUs with govt Colleges to bypass fcra which has a synergy with getting govt. archives
* Bhanu and Dinesh like the idea to connect the people from outside so they can also tell their idea
* Prasoon said what are the area we need advice from other people to contribute and also we send mail to some other people. We have to take the decision for the product not other members apart from open talk he agreed
* Micah said we have to register the milli as a company and we have to proceed it will make sense for me.
* Micah comment: i don't see an immediate need because funding can always come from partnering collectives.
* prasoon comment: I used to get wads of cash from an organization that wasn't the recipient of the funding that was funding me
* Venkat said we need a structure for the Milli
* Prasoon comment: I don't agree. We need a charter as soon as we can. Registration can happen as it happens.
* Micah comments: i agree with charter but as a living doc that we should publish and workshop constantly with our larger groups
* Micah comments on his experience: it interests me to see the contradictions of structure and un-structure which exists in my head and in our discussions, also very interested in a future of largely personal archives and how we makes sense of each others pasts, +1 on internet futures
* Prasoon starting the doc for milli structure
* venkat comment:
* 1. Our personal motivations
* 2. Defining structure/views for Milli
* 3. Milli: Society, Trust or Section 8 non profit
* 4. Milli initially: Inflow/outflow of money and projects
### call on Dec 1st
* Regular start on Jaipur lit fest, chomsky's comment, indo german, roland barthes...
* Recap of 27th conversation
* Micah explaining wireframe
* Prasoon comment: For liniking Visualization we have to think about type for the Item
* Venkat said for single Item kumu graphical representaion make sense
* Dinesh said we always known what is the properties for annotation
* Bhanu liked the user view level
* Venkat comments: Two things I forgot to mention on Friday:
* 1. Please replace 'creator' with'person', we then need to add relationships. It will have 'creator person' 'associated with person', etc
* 2. Both archive and user will have multiple subjects and multiple persons (in diff roles, as creator and other associations). So, an archive item with no user annotations will still have multiple values for the field 'person' and field 'subject'.
* Venkat explaining the Milli entry with examples http://archives.ncbs.res.in/node/801
* Dinesh said a person have different types of attributes
* Prasoon explained the Data structure issues with date example.
* Venkat comment on UI , we have to think about OHR model, https://digitalcollections.ohs.org/371n5278
* Bhanu said we have to look into how we are going to add interruption.
* Micah asked everyone to comment on that figma page(if changes need)
* Friday will do conversation about Milli admin page
### call on 27th Nov
* Milli Glossary - https://hackmd.io/Q4L0-ssNTDie3EusWjlhrQ
* Prasoon need server for implementing recogito project
* Bhanu and Prasoon discussing about next 4months plan for the delivery
* Venkat said we get a spreadsheet for Discovery phase and Narration phase and we will start the work on that.
* Prasoon suggest Tiga for maintaining to do list
* micah questioning do we need Vocabulary as text field?
* Prasoon doing exercise past last week - annotating data model
* Bhanu and Micah discussing about code editor https://editorjs.io/getting-started
* We have to make it decision for the mobile view
to display sidebar, popup modal
* Have a look at: http://archives.ncbs.res.in/node/631
* Page 9 of https://www.ica.org/sites/default/files/CBPS_2000_Guidelines_ISAD%28G%29_Second-edition_EN.pdf Will talk to explain.
* 6 essentials on ISAD(G)
* a. reference code;
* b. title;
* c. creator;
* d. date(s);
* e. extent of the unit of description;
* f. level of description.
* Non annotatble
* a. reference code;
* e. extent of the unit of description;
* f. level of description.
* Dont annotate type, collection, format, citation.
* b. title;
* c. creator;
* d. date(s);
Title, Author: Creator, Date, Type, Collection, Location/Place, Subject(s), Language, Format, Description, Citation
* Emergency thread: http://archives.ncbs.res.in/node/561telling
* Venkat took it up to research 3 or more archives for story telling data
* Emergency in 13 Ways, tiny threads: http://stories.archives.ncbs.res.in/search?query=emergency&query_type=keyword&record_types%5B%5D=Item&record_types%5B%5D=File&record_types%5B%5D=Collection&submit_search=Search
For Mid-End of Feb 21
* Common environment, where anyone can login and look at api n stuff
* Annotation modelling phase (priority high): assuming the annotation is not coming from different server,
* Annotation api endpoint - hasDependency to above point
* EAD endpoint for mockdata (priority med) -
* Dependency for further development of the EAD mock data on Frontend
* CSV module (priority low) - A seperate service module
* W3C SpecificResource
* usecase one archive exporting in a defined format, imported to milli DB, which is discoverable, query through the metadata ex. `discover/?q=map`
* Annotate a name of the city (location, person, event, ) using W3C SpecificResource mapping to LOC name files for subject headings and these annotations create new access points for discovery.
* we will use one of the 6 isdag field say Description field, where users can contribute annotations to the description field
* Woah! so we allow for users to add fields itself? yes, venkat just said this and thats 2 mins after i wrote the question
Design brainstorm Micah
Item level usecase img, audi, vid, text
has 6 or more isdag fields
which can be having multiple values
But how do we see this on screen?
One is sidebar and the other is pop-up
V: Toggle buttons and map filter type styling to filter annotations
Toggle for seeing annotation on the field itself,
and toggle for whole page
toggle for adding, like a + button
And about the free text, accha is title and description also annotatable access points?
Everyone agrees on this
### call on 24th Nov
* Micah and shafali created document for Milli Glossary https://hackmd.io/@miximon/milliglossary/edit
* Prasoon working on schema for the Data model and he will share the schema for the backend and annotation document
### call on 20th Nov
* https://medium.com/pelagios/pelagios-in-the-indian-subcontinent-94fef92ad4a5 pelagios
* Prasoon explaining Scala app workflow
* some Recogito reviews and background intent:
Andrew Lindley is the developer behind Recogito, it seems.
* Micah updating Milli control access
* Bhanu said Milli controller is the user platform
* three kinds of metadata: technical, administrative, descriptive. What Milli puts out in the public space would be the descriptive. The other two also get captured as part of our process, but for backend.
* Venkat and Prasoon said we need to come up with Internal language
* the developer of recogito, the most conntributions is from this guy https://github.com/rsimon
* Bhanu introducing Kumu to team https://kumu.io/
* rhumbl.com, https://networkx.org, https://socnetv.org/
### call on 17th Nov
* Bhanu explaining Recogito to team
* Venkat: Big archives might just stay within their archives software like AtoM or homegrown archive software. But as long as it exports to excel, it fits the description that Prasoon made. So, milli can say it only receives csv or spreadsheet files. ?
* Micah:Makes sense. We'll have to see how this will work for everyone else other than partners
* Can we invite a possible partner to the call and just talk to them about our vision and see where they buy in
### call on 10th November
* Bhanu explaining Biome Projects to team
* Bhanu, Dinesh! Please advertise in your Emergency circles! Talk at 4pm this Friday.
check out garikuyyr video for girish @ kotagiri
### call on 6th November
* https://vangarderen.net/ Peter Vangarderen Blog
* https://www.youtube.com/watch?v=Yth7O6yeZRE&t=06h04m32s Vangarderen talk
* Micah and Shafali working on UI for Milli and Hackmd document
* Micah and Shafali explaining Milli flowchart
* bhanu, prasoon,Dinesh and micah updating commenting facility
* aside: comments and spider - https://github.com/StanleyLsx/app_comments_spider
* prasoon said will avoid this comment complexity
* Comment is only text-blob for now #decision
* Apache 2.0 License https://github.com/pelagios/recogito2
* This is great stuff. I am excited. For next call, could you guys expand on this annotation idea and see how to tweak a recogito (from a UI perspective) to make us go from 'web annotation: describe' to 'adding a sentence in the comment field' to 'highlighting portions of it to tie it to archival description fields and relationships'.You can use
https://www.ica.org/sites/default/files/CBPS_2000_Guidelines_ISAD%28G%29_Second-edition_EN.pdf and/or http://archives.ncbs.res.in/node/801 for field names
* In addition to the ICA and NCBS link above, you can use this very simple implementation of ICA's description:
### call on 3rd November
* Prasoon made document Lifecycle of an Object https://notabug.org/prasoon/niosx/wiki/Lifecycle+of+an+Object
* Micah explaining wireframe for viewing annotation
* shafali explaining wireframe for annotation with text
* Venkat saying we have to careful about file sharing while using Milli annotating tool
* MILLI as a researcher's tool, research trails (record), cherrypick annos, links stored as part of breadcrumbs,
### call on 28th October
* Bhanu explained Codex. Codex is an editor tool where we can annotate the code.
* Micah explaining mapping
* More on codex editor - interesting take on modelling interpretation layers... "The Codex – an Atlas of Relations | ZfdG - Zeitschrift für digitale Geisteswissenschaften" http://www.zfdg.de/sb004_008
* Checkout djit.su they use this editor, you can simply sign up. Related reading.. https://medium.com/djitsu/djitsu-the-problem-with-web-development-and-what-to-do-about-it-2c1925b0c2b1
* @micah I looked up codex editor, code is open source "GitHub - flexscss/codex.editor: CodeX Editor 2.0 " https://github.com/flexscss/codex.editor
* It's js application with ui components, let me know if you have any ideas to play around this one.
* And one last thing for today, not sure if this is oss, but for reviewing
* This tweet vibes with my take on the "labour of curation"
* Related another kool project https://getmemex.com/
* There is java, scala, lisp all of this in this repo https://github.com/pelagios/recogito2,
* Each subject heading has its own purl with a bunch of fields:
* A way to show where a subject heading has been used, variants of heading, first usage date, etc
* https://www.w3.org/TR/swbp-image-annotation/#mOnto Schementic diagram for micah and shafali to create schemantic diagram
* Archival Science Commons http://network.bepress.com/social-and-behavioral-sciences/library-and-information-science/archival-science/
### call on 20th October
* Shafali and Micah looking into miro and W3C
* Micah explaining UI flow
* Figma - https://www.figma.com/file/gDqIy5VF6JAPAUZGGenfkJ/Milli-Future-Design?node-id=23%3A6
* https://mitpress.mit.edu/books/rogue-archives - Rogue Archives
* Shafali wants to clear about accesing part
* Dinesh, prasoon, venkat and Bhanu adding their points on the Ui flow
* is this ok "are any archivists working with w3c web annotation? or annotations groups working with any archival standards?"
* bhanu asking is this ok "are any archivists working with w3c web annotation? or annotations groups working with any archival standards?"
* thats too easy to miss by many who many not be tech enuf. but a good start
* Sample set of digital objects (item level description):
* Excel sheet that backs it:
Can develop an annotation/interpretation for this, and develop a workflow to get annotations back into the excel sheet.
* Link to all the digital objects:
### call on 13th October
* Annapoorna talking about CEPT archives http://www.ceptarchives.org/, http://www.modernistmagazines.com/
* https://www.ideasofindia.org/ - Ideas of India
* https://dsal.uchicago.edu/ - Digital Access South Asia Library
* https://www.indianmemoryproject.com/ -Indian Memory Project
* http://www.archives.godrej.com/ - Godrej Archive
* Shafali: Time also translated into holistic thing
*V: Thanks, Annapurna. It was a little difficult for me to articulate what I was trying to say in the moment. In the survey of various archives you shared, what struck me was this variety of sources and possibilities. I am also struck by how forgiving the world of archives has been. 1/n
* The archive is political, of course. It is exclusionary, authoritative as well. And yes, it erases more than it keeps. But it does aspire to be useful to a diversity of people, and toward a variety of stories. 2/n
* Which brings me to the next point and that is about the interplay between curation, collection, archive, interpretation and narration. At any given point, I feel that the 'object', for lack of a better word, can occupy various spaces. And we, as conveyers of the idea of the object, can play the role of curator, collector, archivist, interpreter, narrator. And in each moment, we will see a different view of the object. 3/n
* I state these to say that I am a little more forgiving of the CEPT archives :ic) If it makes attempts to be useful, if it has built a certain ethos around its material, if it has paid attention to origins/context/process, then maybe it is still doing something. 4/n
* Something I mentioned during the Milli sessions sort of comes to mind. I feel archives need structure, and shine light on the context and process of <individual/community/corporation/family/movement/theme…>. What an archive looks like is quite diverse today, as it should be. 5/n
* But one rule of thumb about whether something starts to resemble an archive that might last could be: when you are able to see material described well, see its relation to the <individual/community…> and have the ability to constantly revisit it to catapult it into a variety of future narratives. 6/n
* If we approach an archive as viewers of the goods, but not renarrators, then I think it's worth relooking at the structure of the archive. The archive has to be an active, breathing element in the production of knowledge, not a silo of knowledge seen in a glass case. 7/n
* P: We need a formal structure for the Archive
* Bhanu: Limitation for the architecture
* Venkat Srinivasan said to Micah yes, every piece of art (and every piece of scientific data is art, too) can be seen as part of an archive. We are living through pieces of an archive. We are the archive. But as Prasoon is pointing out, we need some methods to navigate the 'archive' within a consortium framework. We are not making judgments on what is an archive. We simply want to find ways for it to be visible, re-usable.
* I have struggled with this question of what is an archive for a while now. So, reupping this one rule of thumb I took up to set up a litmus test: An archive needs to have material described well. It needs structure/taxonomy. One should see the object's relation to the <individual/community…>. And most importantly, in my opinion, a user can revisit it and find new, future narratives. These are often easily done with cataloguing protocols. But there might be other ways to do it.
* Bhanu: We are not look into labour but we have to look labour for our project
### call on 6th October
* Venkat: will think about redesign the papad annotation
* Micah : me and shafali will t
* V: Atom is a tool to access the memory
* https://drive.google.com/file/d/1us4DVX9J71l5tKIRAUN5JrxXpozIkUJb/view - Atom related Video
* https://www.accesstomemory.org/en/ - Official Website
* https://drive.google.com/file/d/1lFrNlZ3BHObUyVmJXIkUBXeZvlVqMwkj/view - message recording Atom
* Alternative vocaubulary -https://snaccooperative.org/
* https://snaccooperative.org/view/83424623 - Mahatma Gandhi
* https://snaccooperative.org/visualize/connection_graph/83424623/11315063 - connection Graph
* https://www.w3.org/TR/annotation-model/#motivation-and-purpose - w3c annotation model,refer to motivanational purpose
* s: we came up with Discovery layer
* B: will look into Interpretation layer
### call on 2nd October-
* D: Typically our Repository
* Micah: Will categorise Mind map data
* open linked data
* https://geocities.restorativland.org - Geocities Website
### call on 29th September
* https://cryptpad.disroot.org/slide/#/2/slide/view/PQg7MBKaXtm7qP9tIP1UYleGA4DQXGBsPyrxUpWhgH8/present/ - prasoon presentationon Decentralized, federated tech
* https://en.wikipedia.org/wiki/Turtle_(syntax) - Turtle
* P: base,prefix wiki and prefix magic start and endwith <>
* <Q92991> id assign to the wiki data
*Q5 HUman - https://www.wikidata.org/wiki/Q5
* library of congress this url
*Q1 - Univers, Q2 - Earth, Q3 - Life, Q4 - wiki labrl: Death
* reposting Turtle https://en.wikipedia.org/wiki/Turtle_(syntax)
* RDF - https://www.w3.org/RDF/
* concepts are something that have a name
* "Function" of a concept is due to a concept's place in a strucutured metatdata container
* Artefacts => SomeStructure
* Himalaya - https://www.wikidata.org/wiki/Q5451
* Banana has energy,taste. banana -
* what is meta class - class here -
* Concepts has the name,name could be anything
* EAD - https://www.loc.gov/ead/EAD3taglib/EAD3.html
* sample ead 3 from line no.61 https://notabug.org/prasoon/niosx/src/master/app/repository/MockDataRepository.scala#L61
* author according to schema.org
* Reference implementation https://en.wikipedia.org/wiki/Reference_implementation
* V: Eg for the Archival https://digitalcollections.ohs.org/371n5278
* Each node has Data structure
* peer concept store - A data structure
that maps concepts to nodes
* kinlist - https://www.urbandictionary.com/define.php?term=Kinlist
* Json linking data- https://json-ld.org/
* This is our input data at NCBS for one collection, before it went on ArchivesSpace catalogue: https://drive.google.com/file/d/1h-LurCrCNHEanNzG91MryYSCFsQpwTDQ/view
* ArchivesSpace catalogue for same:
* M: What is the inter connection 1 page to another page
* song - https://youtu.be/0vvBXdHfOvU
annotated lyrics -https://genius.com/Purandara-dasa-mullu-koneya-mele-annotated
* more annotated Source :
* B: more archhives links https://www.loc.gov/ https://dp.la/
* V: Revisit https://bit.ly/Milli-tech
* V: Browse through Day 1 of the Milli sessions to see an overview of archives in the country:
### call on 25th September
* Dinesh: Gnutella for peer to peer network protocol
* Shafali and Micah came up with annotation model design https://www.figma.com/file/gDqIy5VF6JAPAUZGGenfkJ/Milli-Future-Design -demo
* P: when we say access point,Acces point is the index,here we select some text and make it is annotate then it will look like chunk
* Sha: Index point means its like searching keyword?
* Bhanu: Archive word is not one domain.There are many domain - cultural domain,..
* P: many people having different ideas about Vocabulary. Monday Prasoon will present.
* Dinesh: What is the problem in that Figma
* Micah: We dont know the End of the project,we discuss 3layers and came up with annotation design
* Dinesh and Bhanu: ROM technology
* Dinesh and Bhanu: Hypothesis we can annotate the document with text,image,voice mesage but there is no link inter connection
* Dinesh: There is a problem with Miro annotation, some object changed then it be difficult
* Bhanu: Alipi is a different service
### call on 22nd September
* Dinesh: explaining about archive annotation
* https://drive.google.com/drive/folders/1Z-ez-fXP8yCPa_mCoMC7h926kxsorFpY - list out of archive reading books
* Bhanu: explained Annotation
* Dinesh: Replicability and precision with web search as example
* Bhanu: explained replicable and precision via gitlab an github
* Bhanu: explained Papad Channel and tag filtering
* V: We always got the question from What is archive
* Bhanu: will discuss Gracphical Data when prasoon is available
* V: we have to create UI for Audio annotation, Describe naration in Structural way
* "w3act" annotation and curation
* https://ewaconference.com/virtual-programme/ - Engaging with Web Archives
* http://www.openarchives.org/ - good place to see archive dictionary
* Digital archive storytelling 2016-17: bit.ly/archivelets
### call on 15th September
* Ps: Access point - Index format
* https://roamresearch.com/ - roamresearch
* Ps: I need segment of the Item. AD doesn't give that
* http://bit.ly/Milli-tech is a document in progress around specs for Milli, etc. From earlier this year. But it has an intro to archival standards, broad intent, structure, and ends where we start talking about UI for the annotation interface.
is a link that we share with all people who join the Archives at NCBS. Academic papers on archives, oral histories, how-tos, all archives specific. No tech/UI.
* https://hackmd.io/0ogY6lp6THqOhZdLpzrr8A?view#Architecture-Discussions for the location on hackmd where we summarize current status of ideas for UI/architecture for annotations.
* V: 10minutes overview the Architecture of Milli
* Micah: https://dorian.fraser-moore.com largely references of data-based websites
* P: explained Metadata Store Architecture - User API, Federertion Layer, GraphQl Server
* P: wants to working on Reseach,I am in the process of abstracted.I have timelines ready and will discuss after 2days.
* P: will slowly moved into linked data for the annotation
* Bhanu: We have 1 annotation server Mongo DB, Python
* https://drive.google.com/file/d/1Nk_3sHMinslOdRm8jPNISDYz-V8MnD8c/view - Prasoon Architecture Update
* P: There is a L-designer,they talk about scaling problem
* V: sep 18 -2020 we are hosting a public lecture:
* Bhanu,Dinesh,Prasson - talked about web annotation
### call on 8th September
* Bhanu: will call yatarth 1day for meeting and we will ask his suggestion
* Venkat - I felt Yatharthsi is good time to work in Janastu
### call on 4th September
Bharath from AUP talking about City achive
* https://www.taciturban.net.in/kolkata-urban-archive/ - Kolkata Urban Archive
* Hyderabad Urban Lab: https://hydlab.in/
### call on 1st September
Open Archives link
* Discussing about Interview People Saasi,
* P: Discussed about Designer Payment
* D: I have to do mailing for thisira and about exercise
* D: Will invite thisira to join meeting in Jitsi
* https://gather.town/ - national conference meeting happening in gather
### call on 25th August
* Micah Introduction
* https://www.figma.com/file/lGZklLSlFA2yjPxmcyykud/Karwaan?node-id=1%3A2318 - micah UI design updates
* Pj: Concermous of archive,
* Venkat: By using archive to tell the story
* Our intern we have 3layer
* Discovry Layer - Catalogue,Build protocol
* Annotation/interpretation Layer -Team Janastu connect together(How to build UI,backend)
* Naration Layer
* We do not know who is the user, donor
* M: My interest is in Discovery Layer
* https://roamresearch.com/ - Roam Research
### call on 18th August
* portfolio discussion - UI Design
* Papad API changed into MongoDB
* Yssars doing Radio Course
* The isro project, Indian Sonic research organization
* https://drtc.isibang.ac.in/research/living-knowledge - living knowledge
* Audio annotation will make it Independent
* Discovery - Milli platform
* Annotation data will go Annotation Server
* Bhanu: https://thestore.swtr.in/ is a old annotation server, we were working on - this is our initial audio annotation
* http://bit.ly/papad-annos - papad anotation (Venkat will come up with new idea for this adio annotating tag)
### call on 11th August
* P: Refactoring cleaning things up and have to dockerize
* Availability update
* Pri: Archivel experience
* Dinesh : Community meeting community knowledge
* Reactive type
* Learn Haskell for great good - forlearning haskell
### call on 7th August
* Dinesh: What is the role of Larger part of Library
* Madhn: Library is a knoweledge center, collecting idea from people to people. We will deal with overall knowledge.Focus to embed teacher - Improve the old education system.We have a expert teacher. It is a foundation. We are going to have 2 more campus.
* We can create small archives,it will be useful for researcher.
* Demo -> Schoolbooks archive http://schoolbooksarchive.azimpremjiuniversity.edu.in/
* How to define a textbook? is a big challenge
* What is the SBA URI?
* We want to give a persistant site.We dont want to loose the user
* Atom we used for different repository
* Dinesh: Right now governmet suggested way to go Doordashan,my question is how will set the library in large scale?
* Radioactive version of learn addition and subtraction
* Md: Audio and video recording is not clear.
* NIC, Information
* CSIR -
* P: Few things we have to thing about it, 1. Harvesting mechanism,Reconsellation - particular structure index ,not like peer to peer system its like a federated system,Its not be centralised authority.
* Places that do interoperability work, sourced from Madhan: https://urdip.res.in/, https://www.nic.in/, https://inflibnet.ac.in/
* https://conf.activitypub.rocks/#talks - activityPub protocol
* 2setup protocol 1 for discovery, 2 for transport layer
* PS: Roughly I have activitypub architecture will go forward with this
* Dinesh: Will go with activitypub
* Bhanu : Activitypub is a good for web annotation
* one of the early use of both
* Activity pub https://www.w3.org/TR/activitypub
* indicative link re spectrum
### call on 31st July
Madan from APU speaking about Bhargava papers collection, interoperability, metadata harvesting.
* https://islandora.ca/islandora-installations - islandora
* CCMB: https://en.wikipedia.org/wiki/Centre_for_Cellular_and_Molecular_Biology
* APU's AtoM test interface: http://18.104.22.168/index.phphttps://timesofindia.indiatimes.com/blogs/men-and-ideas/think-of-the-spectrum-like-the-village-commons/
* Mainly focusing on OAP(Open Archives Initiative
* APU's Islandora Test interface: http://ec2-13-58-91-216.us-east-2.compute.amazonaws.com/drupal7/islandora/object/islandora%3Aroot
* Eprints: https://www.eprints.org/uk/, DSpace:
* Discovery tool - Public Knowledge Project:https://pkp.sfu.ca/, Vufind(https://vufind.org/vufind/)
* Jayashree reference: QAMRA Queer Archive for Memory, Reflection and Activism: https://qamra.in/
* Madhan: Next week will introduce Atom
* For discovery we have to start
* annotation model:https://www.w3.org/TR/annotation-model/
* For milli will create a repository
### Call on 28th July
* https://limn.it/issues/the-total-archive/ - total Archive
* Tbd: Helavaru community http://dinesh.servelots.com/helavaru, j.mp/unbox-janastu, https://docs.google.com/presentation/d/e/2PACX-1vT62TNItI55ZrUveMlSvu7_Zcq0pWNQTE48KyIavxdNTtQI8qFlfVhi4HPitP-xBQE3xwmuBZkDhhvq/pub?start=false&loop=false&delayms=30000#slide=id.p
* Tbd: Invite DWeb Meetup (Virtual): 2020 Community Views Around the Globe https://www.meetup.com/dweb-berlin/events/272120947/
* Tbd: Looking on archon
* Papad Update http://papad.test.openrun.net/pages/papad
* V: question about tag (User logging)
* V: able to add Pythagorous theorem.If user wants to type subject,
* Bhanu: accurated tags,permanent tags will be come on home page
* V: Filetering the tags -> semantic way filtering
### Call on 14th July
* ArchivesSpace follows the American implementation (called 'Describing Archives - Content Standard') of ISAD(G) - the intl standard for archival description. http://files.archivists.org/pubs/DACS2E-2013_v0315.pdf
* And true enough, 'creator' field is required only if known.
* V:Obtion b-Send it back 2 user
* Dinesh: consistency check
* P: Follow Reactive Program,clickable elements to move the url
* D: Commercial think
* V:How do we send the data token
* P: We need to identify older
* V: Way to upload all Url
* P: Listed out 3approches in wikipedia
* D:What is the collabarative and Milli2?
* p: Data coming from different sources
* P: Using End to End encryption,Database side
* https://www.safe.com/ deals with similar issue.
* V: Early stage will test out with CSV
* Bhanu: Our backend is going to save EIAD.loosing the user.2nd column will give Dropdown
* D: Funding Email
* V: User Interface is very important for grant
* D: We are looking UI designer
* Looking for an Interaction Designer/UI designer to help us ideate and develop UI for an annotation tool. Work involves converting our user stories into delightful experience for our target audience.
* Dinesh: 1)Structure part
* https://www.ncbs.res.in/events/apls-apls-20200717-partition-sarin-jain: Archives meeting on Friday
* Will start with current.milli2 template
* Bhanu: Will start with modelling
### Call on 10th July
* Dinesh: Schema.org common schema
* pj:SQos also mention annotation which also looking exactly what we are looking
* https://ld42020.sched.com/ Linked Data conference
* Question from Venkat: Parallel convertion about annotation,archive through worldwide
* Dinesh: Everybody is always annotating.
* Bhanu Update:not done about Ui screen.Have a answer about last talk.Partner Organisation
* Venkat: 1947 archive eventually collecting the material
* Why is donor annotating?Why is younger updating? form bhanu
* Dinesh:2kinds of annotation (1)Fwd compatability,(2)Own need .
* P- For the Institution archeive or Independent there is a different way to submission.Low Range have differnt flow
* Motivate low range people
* will seperate Annotation and UI model
* Is it running runtime or locally? from Bhanu
### Call on 07th July
* Annotations canvas framer prototype shared for comments
* V - Web annotation motivations and purpose might be too elaborate for this stage of the project
* Lot of machine learning annotations gives options to
* Multiple Vocabulories
* https://hipstas.org/audiannotate/ for approach
* dinesh: (abstract wikipedia project kicked off): (1) the editors collectively define "constructors" for sentences that are "abstract" in that they capture notions / concepts which are then strung together in tree structure ways to represent sentences. (2) someone (linguists / editors) define renderers for different languages that take constructors and spit out sentences in that language.
* Layout reference http://archives.ncbs.res.in/node/918 for milli annotation
### Call on 03rd July
* Dinesh shared links to naming concepts from wiki data https://lists.wikimedia.org/pipermail/wikimediaannounce-l/2020-July/001618.html
* Prasoon will be deploying Graphql with archives @ NCBS and APU test data and vocabulory endpoints in a few weeks
* V: Discover, Search, results, with high level description, with annotate option
* Transcription service https://transcription.si.edu/
* Annotations paper https://repository.asu.edu/attachments/135630/content/Research%20Paper%20v3.pdf
* Interactive prototype https://www.framer.com/
### Call on 30th June
* Mocking with NCBS test data
* Use LoC vocabulary or standard vocabulary for tagging
* User should be able to annotate, tag a person for ex or an emotion
* What is the target of the annotation, how to know that fro the URL or URI
* Venkat Here is an example of Stanford's purl for a map: https://purl.stanford.edu/fs844yc9264
* What is an archival object? is it the file object? or the metadata link is the object?
* Tarunima on Tattle annotation workflow
* All annotations exist, but how do they present is the matter of moderation rather than letting something exist or not
* Democratic archiving, Archive everything (is the vaccum cleaner approach)
* dinesh https://www.amazon.in/Multimedia-Ontology-Representation-Santanu-Chaudhury-ebook/dp/B010ACW672
* This is the mastodon account prasoon mentioned
* Venkat Here is an example of Stanford's purl for a map: https://purl.stanford.edu/fs844yc9264
* Without curation there is no archiving, power equations of who is visible and who is not
#### Follow up tasks
* Preparing the mockup with NCBS dataset for papad ingest and curation
* Using named entity recognition auto suggest for tags entry
### Call on 26th June
**About EAD format and interoperability strategies**
- PJ: 2task for EAD
- Shortlisting the protocol
- Identify the name for the fields
- What will be the Data model of Milli, are we saying that we will only choose the fields we prescribe and throw away the rest? Venkat feels that we should keep all the metadata but depend on a few fields
- If there is additional information beyond the scope of Milli Data model, then how do we address the parsing issue
- Most archival description dont use a field as "emotion" like fear, sadness so on but a community archives around those ideas
- but fields such as emotions are usecase for annotations so it's extensible
- Pj: we can have all EAD fields for parsing purposes going by EAD 3 format, along with that we will have a minimum required fields which depends on our presentation logic
- How do we allow for failures, and not create adoption barriers, and avoid to and fro in updating data
- Milli Data model will have a minimum requirement, the milli presentation logic is a subset of fields from the minimum requirement
- All Organization accounts will have an ID, each EAD object will be an ID related to the org accounts
- Let's consider that all the fields of EAD 3 format, and minimize requirements for the presentation logic
- Compare 2 different EAD (NCBS and APU) from two different archives as test data, to put out a graphql api
- We will start with EAD 3 and then start looking at crosswalks and later look at the Spreadsheet
- May take 2-3 weeks for test data parsing
- DPLA - put it as human readable format
- Crosswalks: https://guides.lib.utexas.edu/metadata-basics/crosswalks
- 26 fields. And six essential fields: https://www.ica.org/sites/default/files/CBPS_2000_Guidelines_ISAD%28G%29_Second-edition_EN.pdf
- DPLA's metadata application profile: https://pro.dp.la/hubs/metadata-application-profile
- Online archive of california https://oac.cdlib.org/
- DPLA vocab descriptions https://drive.google.com/file/d/11KyNt2NwXMDrB1Qpvwm1Yxva3Qg06hYB/view
- Product development document, prasoon will do the architecture document and Bhanu will put the product description
- ncbs test data is on spreadsheet and ready for testing with papad
- saloni UX person, prasoon's fren from sweden looking for interesting project (follow up)
- Online Archive of California: How to contribute to common platform:
- https://oac.cdlib.org/help/ about contribs
### Call on 23rd June
- iaw 2020 videos will go on youtube
- with a link to play list or embed on milli.link
- PJ: EAD is very elaborate and includes presentation into it. May be we start mocking the structure
- Tarunima and Denny on tattle, archive and annotations, using android accesibility api for scraping, take sql dump, export chat feature, whatsapp business api
- format shared by madan from APU, and need for us to look at what is milli needs for interoperability and may be start documenting
### Call on 02nd June
- 13th session on Milli2 talks (by team), discussing schedule
- Webinars standards / rules, documentation,
- milli.link live? email domain and testing
- email@example.com - V will manage mailbox responses, emails will be forwarded to D, B and P for incoming or keeping an eye
- chat on video conferencing app options https://bigbluebutton.org/
- forms and data collection https://enketo.org/#tools
- http://twine.unhcr.org/app/ http://mapyourworld.org
### Call on 29th May
- Venkat - on the write up,
- D - on Chaha Dasara https://docs.google.com/document/d/1ODracOspj1mPKbBwy2_swt_9zYLii9spNbDFK8bypwI/edit?usp=sharing also can check on concepts here http://newsrack.in/about/faq
- Discussions on vocabulory (at philosophical level), at some time we need to resolve / evolve this conversation for the milli platform
- do we start with a controlled vocab, or each user will create their own, how to go about?
- if n users use same interpretation - can we equate to call it a common vocab
- how to program this? ex one image of Dasara, 4 users annotate. user 1 annotates with x, 2-> y, 3 & 4 -> x
- Authority or reputation? what gives value to the anno
- P - ex: Dasara & Durga Poojo will have their own name space, key = dasara and value ="something / text / image / ..."
- P - A milli namespace for controlled vocab managing
- V - then what didnt make it to the milli vocab, then what is happening to the contributed annotations
### Call on 26th May
Invited Arun to join, Venkat will be unable to join.
* What needs to be tested if papad audio repository should integrate with milli discovery
* Once persistence is added then the service will store data in a database
* now we can test with the mock data repository
* arcGIS story maps - nice authoring tool interface for multimedia / maps storytelling https://storymaps.arcgis.com an example story https://storymaps.arcgis.com/stories/4fdc0d03d3a34aa485de1fb0d2650ee0 & pixelfed
* Persistence layer for milli discovery in 1 week
* currently the structure should be in the mock data repository
* Discussion on https://notabug.org/prasoon/niosx/wiki/Specification+for+Milli+data+discovery+and+exchange
* Annotations part of document or description?
* Authority on adding annotations v/s 3rd party annotations representation
* Loosely couple the annotations from the document and it's metadata itself
* Annotations - 2 parts, target & bodies
* Annotations service is dynamically updated, while the document and it's metadata are edited with authority
* Example: a suggestion / correction to metadata, where a service understands irrespective of the controls metadata
* Priority / curation algorithm, which looks at the context in which the document is seen to prioritise the annotation rendering by the client / service?
* Metadata is definitely the part of the Document or it's descriptors
* A query is derived from the context, so as to fetch the annotations from the annotation servers
* Spend some more time on the data model
* PJ - needs to complete the specs to expand annotation model
### Call on 22nd May
D - on webinar doc, for code of conduct or rules of access
Conversations on internet infrastructure, it's fragility, p2p protocols, Briar,
Briar messaging- also for w4p
Continuing on internet, V stumbling upon Tim berner Lee's mother's oral history https://ethw.org/Oral-History:Mary_Lee_Berners-Lee
Prasoon, has committed backend
Build.sbt - all build
/conf - all configs
conf/routes - controllers.HomeController.graphQl
> another end point to be added, for xml endpoint
/app - everything happening
> endpoints connected to controllers
/generated - ignore it
Everything happening in below folders
/graphql - graphql server plumbing, api or endpoint level fields / interface to db mobdel with graphql
schemaDefinition.scala - all fields declaration for graphql interface. any changes to the api, like name of the field should be done here.
> woah! it commited a swo file
/models/ead3.scala - all models (global scope)
Further conversations, too fast, am dropping, dropped...
### Call on 19th may
8 V - Time segment annotation, how images are connected to audio file
tagging archival object - nuanced ways,
to add tags
* D - Server wont save files, but generate thumbnail or previews and use cache, dash.swtr.us/
file upload? - use existing services or running one is quite a thing clean user interfaces to extract url's and identifiers for image and text tags
* using vocabularies https://ukat.aim25.com/thesaurus/f8/mt805/7453
not sure of the direction
* PRM: suggested local archive specific terms
* PRM: Covid 19 disease embodied exp and stories for research on recovery and co-morbidities (ex. asha workers and other local stories)
* V - https://www.pixstoriplus.com/ https://www.pixstori.com/
* PRM - https://incite.columbia.edu/covid19-oral-history-project covid oral history
* V https://sites.google.com/iu.edu/covid-19oralhistoryproject/about
### Call on 05th May
* milli.link deploy with content and timeline
* Timeline - @sagesalus will send in a day or two and @prasoon will update about the discovery tool timeline
* Papad demo - delaying due to testing local vs remote
* Backend Prasoon follow up -
* Sample covid relief sheet from BCE as testcase for abstract data model https://docs.google.com/spreadsheets/d/1CrfkIv98d49nxIkY0TgZ0QAQIjCv5QsiJvnQJq0ilDg/edit#gid=868253577
### Calls on 10th and 15th of April
Decided to get on call twice (tue & fri) a week. Discussed re. the user stories, persona, roles, frontend screen mockups. our focus will slightly move towards taking audio as the first media type to be supported in the milli annotation canvas to add tags, images to audio or a part of audio.
Added images in Architecture discussions and frontend discussions of this doc.
### Before the 25th call
We had spent some time too setup madoc platform on local server. Tested it with some IIIF dataset, and annotations. However, the later feeling seems that we might get locked into too many of the moving parts (services) here and not be able to / slow down in pushing upgrades / modifications.
Instead would be better for us to focus on one main fundamental thing required to achieved milli / specere and add pieces as we move along.
### Call on the 25th feb
TODO: Add minutes of meeting
* Most archives just have metadata descriptions and a url to the same
* May have a the digital material linked
* Fundamental unit would be a url with some context as target of annotation and body could be workflow tags, comments, links, relationships.
* Lack of Vocab standards makes it hard to support for wide varieties however, prasoon has already written some code to be able to parse & import EAD export format
* Storytelling app (Specere) will be a seperate app from the annotations (Milli) where user will give a modelled spreadsheet to render a story like timeline, spacetour and peopleplex from 13 ways sandbox examples.
### since 10th march
label: After prasoon's visit
### Basic Features (or most needed)
### Optional Features (for future)
* [Slides from Venkat an Overview of Milli](https://mail.google.com/mail/u/0/#inbox/FMfcgxwHMGLSCbFLBkVlCtvnqcnrhmxF?projector=1&messagePartId=0.1)
* http://voice.gramvaani.org/vapp/mnews/1129/show/tags/coronavirus/ (see tags)
* google is indexing the fragments
* papad.pantoto.org audio tagging by community
* Library of Congress name authority (can do bulk download of tags): http://id.loc.gov/authorities/names.html
* Library of Congress subject headings (can do bulk download of tags): http://id.loc.gov/authorities/subjects.html
* milli.link to be populated
## Archived Sections
### User Stories actually became a Epic
Let's first take a few simple use cases
1. Bob is browsing through the catalogue of Archives at NCBS, coming across some images in `File — Box: 3, Folder: 23 Identifier: NI005-7-1-3-23` [Research work -- M Panicker, G Hasan, U Bhalla](http://catalogue.archives.ncbs.res.in/repositories/2/archival_objects/1255) and immediately remembers from his memory that M. Panicker also mentions about this image in his [Oral History Interview : Mitradas Panicker ](http://catalogue.archives.ncbs.res.in/repositories/2/digital_objects/16) that he heard yesterday.
Bob now wants to link this materials with a short commentary.
In the above use case,
Taget = [Research work -- M Panicker, G Hasan, U Bhalla
File — Box: 3, Folder: 23
Comment Text = "Curatorial commentary sample"
Relation = isPartOf LinkTo
LinkTo = [Oral History Interview : Mitradas Panicker ](http://catalogue.archives.ncbs.res.in/repositories/2/digital_objects/16)
This annotation will be sent to annotation server posted to an annotation collection that'll be determined by the workflow tag.
2. Archives at NCBS Admin regularly checks the annotation collection for any new annotation and notices Bob's annotation in `Pending Review Collection` admin will be able to move / add anno to all anno collection `Approved Collection` or `Rejected Collection`. To keep the annotation itself light and not keep status withing anno and make workflow annotation collections.
### Domain name Suggestions
[milli.link](https://milli.link) --> confirmed
VS: I'd say stay close to Milli. It gets at meta meaning of found and between space.
VS - How about omil.li or imil.li or mymil.li
> [name=TBD on domain name, suggests] milli.link as available
> [name=bhanu on ui/ux research] after confirming the domain, we need to put up a landing page. some content we need to write about the project and also may be share with peers to collect some user feature request or wishlist that we can add to our user stories.