Tag Info

It is an option, not a design principle. Sharing domain objects is usually more concerned with business capabilities and validation (methods) than data structure (properties). The question is whether ...

A good rule of thumb for deciding the number of partitions would be twice the number of executor cores available

Thank you for sharing your valuable insights! I really appreciate it, as it's hard to find in-depth discussions about practical DDD out there. I have a follow-up question regarding your point: You ...

The main bottleneck in your implementation is use of a deque to implement the A* priority queue. Having to (re)sort that deque on each iteration of the outer loop comes with a performance impact. Use ...

I participated in writing many systems intended for banks and insurance companies. In all of them, critical calculations were performed using double variables. However - and this is important - all ...

How to load only what I need? Are there good patterns to run business logic on an Aggregate Root without loading the whole object tree, while still keeping the business rules safe? You are allowed to ...

First of all, pygame.key.get_pressed()[pygame.K_RIGHT] and pygame.key.get_pressed()[pygame.K_LEFT] are recomputed len(vertices) times while this is not actually needed. While you can pre-compute it ...

First, make a backup and make sure you can restore it in case you fat-finger something. Even better, try out all these optimizations on a staging site before using them on a live site. I wrote a free ...

Since you need to get at various parts of the JSON, you should make columns for those parts in the table. Extract them when receiving the data. Also, do cononical transformations as desired (remove ...

I have total control of the JPEG encoding parameters, including subsampling ratios and quantization matrix. I would start with 444(no subsampling) and q=100 and revisit subsampling and q<100 later ...

There are obvious trade-offs between accuracy and efficiency here. As opposed to the computation of exp(x)-1, where the most significant term is 0.5*x, which can be computer error free in binary ...

Let's stay in Python, and keep using the test data you offered. This is not a greedy solution, it's an optimal one. LP works, and you can reduce the number of nodes down to 4 and all pods will still ...

Can't speak for EF, but NHibernate offers the ability to filter collections. This provides the ability to load an aggregate root but only load related entity based on some criteria. IIRC, UML has the ...

This in, indeed, a question too abroad with no real data or any efforts tried. I guess you want to create those 10 groups based on different variables (your data). You could take a look to clustering ...

You don't need to worry about memory implications in this case. $result is an object of type mysqli_result. When you pass an object into a function, only its identifier is passed and not the whole ...

The Table bars should have the primary key/index bar, depends on how many distinct values there are. If its a small number it could be the table never leaves the db-cache and the index is not used...

Is this distinct column indexed in any way?

You could create a table bars with columns bar and bar_count, fill it initially with SELECT bar, count(*) from foo group by bar. Then you need to let INSERT/UPDATE/DELETE TRIGGER on foo handle the ...