Perfectly balanced.. as all things should be 😑

Dread it, run from it, destiny arrives all the same.

Reality is often dissapointing.

well you must love hasura ahah

Coded graphql for awhile now front and back, loved it before. Now its a pain.

Hasura vs typeorm what do you prefer?

As a new web dev I can honestly say I dont understand both

@@mohabedr5030 you are talking about different things... Hasura is api gateway, while typeorm is ORM wrapper around database for typescript...

Any drawbacks or problems/issues was encountered while using it?

@@mohabedr5030 Both are two different things and approach to a different problems.

What Bangalore ? Wow we indians are stepping up the game

Well im not surprised

Have been to their office in Koramangala 😄

yep the founder is from IIT Madras

Kinda guessed it by the name, Asura means demon 👿 , I guess that's why that logo

CSV is better

freak in the streets, google in the sheets 🤷‍♂️

Just store your information in notepad job done!

@@stephencooper8710 and save it as CSV an uninstall MS Office

who needs a database when you can have one big json file

That's a good way to look at it

@@Fireship what do you think of prisma? Versus others such as Sequelize

Right, as did Prisma 2 haha

This aint new Node, Ruby, Python, Java, Php and .Net eco systems already have tons of code generators / db resolvers pre written for generic usage with DBAL in frameworks. The thing is they will only work out of box with 1-1 mapping, meaning you wanna keep entities / table names and resource name same? It's good, the problem becomes a challenging when you have things like async io, messaging and data aggregation involved in your logic, they are separate technologies on their own and requires specific learning curve. With more and more complex data and underlying systems not just being monotonous graphql can be a pain to implement when you are creating resolvers by yourself with such systems. That is the main usecase that it unifies different systems at large scale, where you have different kind of systems for different purposes and still they are able to talk at unified front. This adds burden on backend devs to implement resolvers for each and every usecase of the data / resources and becomes a huge task where you have very different kind of input data, storage data and output data abd then theres responsibility of keeping data consistant with processing systems. We already have plenty of Json based RESTful protocols and xml based protocols in past. The only appealing thing about graphql is it's ease of use since it's front end layers are designed for consumers. So yeah this is yet another fancy tool for tinkering till you get serious and make a system that handles real traffic and real usecase. You never wanna expose your db structure like these anyways and there's always other systems involved.

@@codeaperture Used to be a big fan of sequelize, until I tried out Prisma 2 one day and that was it... Never looked back since. I love sequelize but, sorry to say, Prisma 2 is simply better in every concievable way

I fully agree with you on this. It's also very important to know when to use a specific tool or service and if there are a million options to choose from then you don't have the time to compare the pros and cons for all of them... In the end the decision you make probably won't even matter that much in your project since due to the flexible nature of SW you can still "make it work" somehow; the only ones who suffer in the end are the servers and end user apps doing needless additional calculations; all these mistakes add up (that's why browsers today take so much RAM just to display some picture/text and maybe a video or just play some music). This is the kind of meta-programming decisions that I hope AI will be able to help take and optimize in the future somehow because it's clear that people don't give a $hit and who can blame them when fast solutions/iterations is where the money is. /end rant :P

@@DesyncX I'll just say, interesting note on the AI. Oh, and I guess I'm reminded of tools that help you make those decisions and deal with the side effects of them like package size.

@@DesyncX Love that view. Something I explain all day :)

I feel the same way bro

@@FS-yq9ef index.html and script.js in a folder is best i can do ok.

You do know that Hasura is open source and you can host it yourself :)

They have a docker-compose file that you can use. Just docker-compose up -d and visiting localhost:8080 will bring you to the console

I'm guessing his job is to keep up with latest tech and teaching that stuff. 😄

Hasura configuration in the web console is exportable as metadata files (hasura metadata export), so you can commit human-readable yaml files to your git repo. You can apply it to other server instances, similar to the db migration files.