Have you, a sensible developer, noticed how horribly bad LLM-generated code is?

Then you should have a pretty good idea how I, a database administrator, feel when I see the queries your ORM has generated.

#SQL
#Database
#LLM
#ORM

Hibernate ORM 7.0.0.Final Released

#java #orm #relational

https://github.com/hibernate/hibernate-orm/releases/tag/7.0.0

Working on an mini-oxm (orm and odm in one) in rust and tinkering on the way how entity data should be tunneled back to the caller. Normally this would be done by a simple generic but I want the Driver/Backends to be dyn-compatible so the decision which backend to use can be done at runtime.

Currently I have two prototyped designs:

• Either a trait Row { fn try_get(&self, idx: &str) -> Result<&Box<dyn Any>, Error>; }, where each driver must encode the data into an custom row type and is thus memory bound, or
• By an provider-consumer approach that uses two concurrent async tasks (tokio::spawn) and communicates by one shared Arc<Mutex<Option<Box<dyn Any>>>> slot that is blocked until the consumer has consumed it. This is light on memory but has the burden of requiring async (which is already required so no problem) but also a lot more of syncronsation code (Mutex, CondVar etc.)

So at the end it's the old problems of memory vs CPU time; choose one or the other, not both.... I could now microbenchmark it but both approaches will perform VERY differently based on context / environment / application, so benchmarking it dosnt give that much insight I think. Sure I could implement both based on an create feature (or directly let people choose via a low-level API) but would that benefit that users?

What's your opinion?

I wanted to use `models.IntegerChoices` for a field, but I have noticed when I add a new option it creates an additional migration.

If I do something like the screenshot, it kind of works? And no extra migration is created when I add a new option. However, it's messy and not DRY and I wish there was something better.

Am I missing something about IntegerChoices that would skip making a migration? Is there a better approach?

Preview of #LittlePrince, a TUI #Mastodon client. God know why I chose that name:) Just toots, no boosts, and threads! Not quite like #mutt yet (see 2nd s.shot), but we're going steady. A shame I had to fight with #ORM's just to drop them and do all just with SQLite. More on that soon. #Rich will provide the TUI experience, unless someone suggests something else :)

At Touchstone Infotech, our expert team is dedicated to helping your brand/business connect with the right audience through powerful social media strategies.

Why wait? Let's take your business to new heights!
Contact us now!
https://zurl.co/RI6Gj
.
.
.
#digitalmarketing #digitalmarketingagency #touchstoneinfotech #growyourbusiness #branding #instadaily #socialmediamarketing #socialmedia #seo #orm #follow #businessgrowth

Top Online Reputation Management Company in Gurgaon - Digital Upshot

Boost your brand’s online presence and credibility with expert ORM services from Digital Upshot in Gurgaon. Safeguard your reputation with tailored solutions. #ORM #OnlineReputation #DigitalMarketing

https://www.digitalupshot.com/orm-seo/

Well I have this Medium account and never posted anything on it, so I decided to write up my littletable project. littletable was an experiment in ORMs that I started about 15 years ago, but it has really come in handy in dealing with small data sets and CSVs (up to a few thousand items). https://medium.com/@ptmcg/data-in-the-small-python-package-littletable-combines-in-memory-nosql-orm-with-schemaless-setup-54ea0b24ac6a #python #orm #database #littletable

Those who use Prisma in production, what's your rollback strategy?

Do you use shadow database?

Is there a way to generate down migrations along with up migrations?

Is there a way to rollback a particular migration/version?

RT's are appreciated.

Amy #django #orm person:

How to convert the following to a QuerySet (I would love to have objects than values/dict)? #python

```
SELECT "containers_containerbase"."id",
"containers_containerbase"."cid",
"containers_containerbase"."cname",
MAX("containers_containerbase"."ctime")
FROM "containers_containerbase"
GROUP BY cname
```

Asked a #django #ORM question in the forum.

https://forum.djangoproject.com/t/quering-latest-entries-wtih-specific-values/33157

Do you know the solution?

Do you know any good #django tutorials explaining various scenarios of #database handling (structure and data save/retrieval)? Things to read and learn more on this topic?