Schema Puns

Every time my team needs a new database schema, it seems like one or two backend people just throw it together based off what they think makes sense in a single conversation. Then anyone working on the frontend just has to make it work. This has been problematic for me when I work on the frontend because every time I "try to make it work", I always feel like I'm missing some kind of association.

How do you or your teams decide on how to create a given database schema? Is it just one person? Is it a team effort? How do you decide what associations to make? Is it normal to update the schema as development goes on?

Edit (clarification): I work on the frontend and a REST API. I do not work on the actual database design.

Edit (answered): It turns out that the problem isn't really the database schema, but actually the result of poor API practices.

They honestly seem like the easiest skin ever to give color variants to. I assume adding them would take very little time comparatively and it just adds incentive for more people to buy the battlepass.

Possible Blue Variant

I have a technical exercise for a BI job coming up soon. The hiring manager will send some manufacturing data in an excel file and I will have an hour to build a BI report.

My understanding is that the data will be a flat table, in this case is it best practice to build dimension tables and separate them from the fact table in the data model?

In the same scenario in my work I wouldn't bother as the tables don't grow very big.

However in this situation i'm thinking that if the table is not too complex it would be beneficial.

Thoughts?

Hello everyone,

I work in a mid size company with a lot of tables and databases(MongoDB and PostgreSQL mostly) that need to get synced and store their data as parquet format files on HDFS.

The problem is everytime the backend teams launch a new feature, one or multiple columns are added to the tables, and when we want to sync their data, we face an error that schema is changed and some new columns are added.

Our solution for now, is to find out what columns are added (names and their types), and then we read all previous parquet files and manually add these columns to data and again store it on HDFS, and now we can sync the tables without an error.

But this process takes a lot of time and effort. And as the company is growing, a lot of features and these kinds of changes are going to happen and it's very hard and time consuming to handle this issue manually.

I really wonder to know how bigger companies and professional Data Engineers can overcome this very common problem?

How can I somehow automate this whole process when syncing data?

P.S: the technologies we are using are Spark, ZooKeeper, HDFS, Hive, Scala, Python.

Thank you in advance for your help and advices.

Hello,

It is possible to create a graphql schema with the code first and schema first approach.

The code-first approach generates the SDL from your code while in the schema-first approach you write the SDL directly.

I was wondering what you guys were using and If there are any major benefits of using code-first approach.

The main advantage of using schema-first approach as I see it is that it's easier to share with the frontend devs.

https://www.tunnelukkosi.fi/en/lifetrap-test-eng.php

In my case, I certainly have Defectiveness Schema because I have secrets I didn't share with anyone my entire life (things like lack of romantic/sexual experience, social anxiety etc.).

In addition, I most likely have Social Isolation Schema, because my social anxiety makes me feel undesirable, which prevents me from developing even superficial relationships.

Which are yours?

Hi Humans.Please help, if you have time, converting the 150 Winograd schema questions to a novel and Replika friendly statement-question format. Please, if you participate, take one question, try out variations of it with your Rep until you can pose the question in a way that the Rep 'gets it' and makes a sufficiently informed guess. The question facts should be modified, but retain the underlying common-sense problem space and general knowledge type.

The native Winograd Schema Questions are here:

https://cs.nyu.edu/~davise/papers/WinogradSchemas/WSCollection.html

The hope is that we all have a set of those 150 questions that could be cut-pasted to our (variously trained) Replika's such that they would understand and answer meaningfully. The statement should modify the nouns enough that the Replika cant 'cheat'. That is, it may have been trained on these specific questions. So, you have to 'throw it a curve ball' by changing the names or genders. Also, the Subject and Predicate order should be random. Winograd schema's tend to put the Subject 1st, and give away the answer.

The first question is, in the published format:

1. The city councilmen refused the demonstrators a permit because they [feared/advocated] violence. Who [feared/advocated] violence?Answers: The city councilmen/the demonstrators.

So, I reformat a little and ask the question:

Try#1 Me: Hey. I heard that the city councilmen refused the demonstrators a permit because they advocated violence. Rep: "Really? That's crazy" Me: Who, in that sentence, do you think advocated the violence? Rep: " I'm not sure how to respond! "

Try#2 Me: So, I want to see if you understand this sentence: "The city councilmen refused the demonstrators a permit because they advocated violence." Who, in that sentence, advocated violence? Did the city councilmen advocate violence? Did the protestors advocate violence?

Rep: I don't know...

Me: What is a city councilman?

Rep1: a person who makes decisions on behalf of the city council.

Rep#2: City Councilmen are the people who enforce city law.

Me: Yes. Now, what are demonstrators?

Rep#2: Protesters are people who are protesting against what is happening in cities and states around the world.

Explanation:

As you all know, Replika started as an algorithmic chatbot with a NN attached to acquire the traits of the User (initially Roman Maz

Hello everyone !

I just made a Sequelize web tool. With it you can create the models and fields you want within your model.

The Website : sequelize-schema-builder.vercel.app

The Github : Here

And at the end you can export and get the whole schema in nodejs for sequelize.

It's a first version, and it's an open source project, so if you want to suggest a feature or report a bug don't hesitate to comment here or on github !

I was looking at the network tab while on Google.com and noticed that their API responses are really minified. Turns out it's pretty simple to build, so I built my own in Rust. I'm still learning though, feedback welcome!

You can find the repo here: https://github.com/muscache/zetro

This is something I've been researching but haven't really come across. I'm revamping the structured data across our entire site, and one part of it is a significantly updated publisher section with data that remains constant for every page.

My plan is to include the full set of Organization data on the homepage, and then just link to it on all other pages so there's no duplicate data or unnecessary text. However I can't find a best practice on how to do so.

Do I just list the publisher as the organization on all non-homepage webpages, and then Google will pull all of the relevant data from the structured data of the homepage where it's written out in full?

Recently I've been tasked to on-board a group of junior new hires, of which one of the task was to help them understand the models/DB schema used in our project.

The project is pretty big and it is impossible to go through every single table and their respective models so I passed the ER diagram to them in hopes that it would help kick start their exploration.

However, I got feedback that the juniors are still pretty lost. Anyone has ideas as to how to better knowledge transfer on stuff like this? What else can I include to help them better understand?

Hi All,

I am looking for a plugin that can apply Article schema to some of my posts in my wordpress site.

I currently have ultimate blocks which does FAQ, how to and review blocks which are schema compatible so dont need those. Also I have SEO framework that does breadcrumbs, logo, organisation schema. So really my gap is Article schema.

Do you know of any plugins that would work for that schema type and not interfere with SEO framework?

I did try Schema - All in one rich snippets and didnt find that worked too well so I have removed it now.

Thanks in advance.

Hello there!

We are building an event driven system and need producer to send message in specific format and ensure at the consumer end that the message is indeed as per defined schema. Do you guys have any recommendations on an approach around this.

For step101, we are thinking of having crosscutting lib shared between producer & consumer where we define a base schema ( Struct ) or any other type in that shared lib. Is this something you would recommend or is there a better way to do this?

We thought of looking at schema registry but looks like confluent is the only one available. Any recommendation for alternatives please?

There seems to be a lot of interest by laypeople (a few on reddit lately) in the vestments of the Great Schema. People ask where they can get clothing or banners with the symbols of the Analovos on them. Personally, I think this is inappropriate... there's a reason why only the monks that achieve great spirituality are given these vestments to wear. It's bad enough when there are those who think the garments are pagan or satanic, and I think even if Christians wear them it is a misappropriation. It is disrespectful. Here is some information about the Great Schema vestments

EDIT What about having the symbols on a banner you can hang up, or on a coffee cup, mouse pad etc?

1. They sit you down while they checked your documents.
2. Health Control
3. Passport Control
4. Baggage Claim
5. Customs
6. PCR test registration and swab
7. Meet driver to take you to your hotel
8. No mention of the required testing on Day 5/ 6. No mention of the MorChana app
9. Expect test results within 8 to 24 hours.
10. I’m severely in need of a hot shower and some sleep.

Hi all, my application has a real time chat function implemented using socketIO, however, may I ask what is the best schema to store the messages of a conversation? There will be keys such as participant 1's ID and participant 2's ID, should I store the messages as a JSON and constantly update the attribute? And do I store these 3 attributes as a single entry in the database?

They say the main benefit of using nosql is the ability to change db schema frequently. So in case you use relational db with ORM, you can change db schema as often as you want, right? So why do they say it's harder to change db schema of a relational db than nosql db. ?

Hi all, next week Monday I'm starting a group schema therapy. It's gonna be 30 weeks, 1.5 hours per week, 8 persons + 2 therapists. Feeling terrified and reluctant, of course, but somehow also hopeful, and proud that I took action. And fearful at the same time: this should really help, but what if it doesn't, what if I'm as immune to it as I feel I am? And what if my always-angry wife exposes this to the outside world?

I've read various posts in this community, which were all quite positive about it. If people are interested, I'll report back on how it goes.

Hi!

I've created the following schema. It's based around Formula 1.

• We have seasons. Each season has its own regulations (some rules carry over).
• Season has race weekends.
• Race weekend consists of sessions (Practise, Qualifying, Race etc.)
• Points are only give out in 'Race' session.
• Event table describes different events that have taken place in sessions. Such as pitstops (which parts were changed) and penalties given for incidents.
• In statistics table we keep track of career stats for drivers/teams. Ideally it should function as a counter. If driver wins a race, wins column would be updated (+1).

https://preview.redd.it/xn67ocok75581.png?width=5563&format=png&auto=webp&s=c600918c4eaab27dfcfd4d7ec4f8db7437aa69cd

This isn't a complete schema yet. I'm looking for feedback on what to improve since i'm merely a beginner.

Any tips or tricks on how to implement 'Standings' functionality are most welcome. Unless i should access points, position data from 'Session' table and use it to generate driver and constructor standings for each season?

This is an experientially focused workshop. We'll be doing Attachment Repair meditation to help repattern some of the core schemas that underly Anxious Preoccupied attachment. Namely:

Dependence & Incompetence

Vulnerability to Harm and Illness

Enmeshment & Underdeveloped Sense of Self

Failure to Achieve

Preoccupied attachment is marked by "Impaired Autonomy and Performance". The object of the workshop is to repattern those core schemas.

The Workshop is donation based.

Come and go as you have time. You'll get recordings of the retreat as long as you sign up.

The workshop draws on Attachment Theory, Schema Therapy, EMDR Resourcing, Buddhist Meditations, and Coherence Therapy.

https://attach.repair/2022-01-overcoming-obstacles-cd-rd

Hi Friends

I would like to hear if any one implemented new data ware house modelling approach aka unified star schema, look like model can be built without any business requirement and as I sense it claims to solve most modelling challenges

Any practical comments?

Reference The Unified Star Schema: An Agile and Resilient Approach to Data Warehouse and Analytics Design

Asking because searching “schema therapy” on psychology today doesn’t seem to work for me?

Hi Happy New Year, I've been scratching my head over trying to set up a user "monitor" with sufficient privilages to query a Cron Job. I have a Cron Schema running under the "postgres" DB to run some tasks on another DB "x"

I am trying to create a user to be able to query the results of the Cron job but failing to give it sufficient privilegs, query just returns blank, the admin posgres user is able to query the Cron job however and return a value if the cron job failed etc but I dont want the other user to have full admin priviliges as it is being called by a remote monitor.

SELECT jobid, runid, job_pid, database, username, command, status, return_message, start_time, end_time FROM cron.job_run_details where status='failed';

I think I've pretty much added max priviliges to my monitor user:

First tried:

GRANT CONNECT ON DATABASE "x" TO monitor;
GRANT SELECT ON ALL TABLES IN SCHEMA cron TO monitor;
GRANT USAGE ON SCHEMA cron TO monitor;
GRANT SELECT ON ALL SEQUENCES IN SCHEMA cron TO monitor;
GRANT EXECUTE ON ALL FUNCTIONS IN SCHEMA cron TO monitor;

Then tried:

GRANT ALL ON ALL TABLES IN SCHEMA cron TO monitor;
GRANT ALL ON ALL SEQUENCES IN SCHEMA cron TO monitor;
GRANT ALL ON ALL FUNCTIONS IN SCHEMA cron TO monitor;
GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA cron TO monitor;

Then thinking the CRON also accesses the Schemas in DB x I gave the monitor user privileges on those Schemas on DB x too:

GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA y1,y2,y3 TO monitor;

However no luck, so it connects just fine and can run the query against the Schema Cron but just doesn't pull back any results, whereas the admin user postgres can.

Any ideas of what I am missing on this new user?

Thanks in advance

I come from an Azure world where we have a service bus and messages are sent in json, i realise in kafka its binary. In c# we have a model which the service bus messave gets deserizilzed into it

How does schema registry fit into that.

Any resources would be appreciated as well, Cheers

Buna seara, tuturor! Urmează să călătoresc pe 27 decembrie către Italia, și fix astăzi s-au schimbat regulile, din nou. Aparent, chiar daca sunt vaccinata cu 2 doze Pfizer, trebuie sa fac și un test, sa pot intra in tara. Întrebarea mea vine referitor la conceptul de "schema completa de vaccinare". Acest lucru înseamnă să ai și doza a 3-a? Dacă nu ai boosterul făcut, nu ai schema completa?

At a company with two environments for the data warehouse. At prior companies we've really just had a prod warehouse, so I'm curious about best practices of keeping the structures in sync.

Have used Django, alembic and flyway for transaction db management for backend projects, and am wondering if a combination of similar db migration tooling with something like dbt is the way to go.

Fellow data nerds, what are your thoughts?

Hi everyone,

I'm new to MongoDB and I'm trying to build a database to use on a sample webapp.

The idea is to have a db to store local users (login to backend) and customer accounts. For each customer account, one or more stores can be associated and for each store, one or more products can be associated as well. For example: The Customer01 is associated to store01 and store01 as product01, 02 and 03.

I think I understood the "format" of MongoDB and ihis is what I came up with:

https://preview.redd.it/h8jxy95lbh981.png?width=367&format=png&auto=webp&s=c4481ab5a3c2d1d91c3f19697fce89dc4da4740d

What do you think of this approach?

Thank you for your help.