3 ms·
It looks like you wrote more about the database advantages here than on the main website https://didgets.com/ https://didgets.com/ A feature matrix or direct co
by mtmail 4y ago
It looks like you wrote more about the database advantages here than on the main website https://didgets.com/ https://didgets.com/ A feature matrix or direct comparison would be great. For any advantage in one area (e.g. query speed) I suspect there will be downside in another area (e.g. slower inserting of new data).
When comparing database I'm used to look at feature lists, e.g. read vs write performance, ability to compress data, scale among multiple servers or such. For example https://en.wikipedia.org/wiki/ClickHouse#Features https://en.wikipedia.org/wiki/ClickHouse#Features or https://en.wikipedia.org/wiki/Apache_Cassandra#Main_features https://en.wikipedia.org/wiki/Apache_Cassandra#Main_features Then later I look at the ecosystem, e.g. how to do backups, client libraries (linking to a shared library vs direct vs HTTP API) and ease of use.
I don't mind creating indexes. I maintain database that have almost as many indexes (measured by data size) than data records.
- didgetmaster 4y agoThe project is just getting going so there are still some features yet to be implemented before it will compete directly with other DBs feature for feature. (e.g. it only does a couple types of JOIN operations so far). The key-value stores I created were actually invented to be used as tags for the object store I created to compete with conventional file systems. It wasn't until I discovered how fast the tags worked before I decided to benchmark them against popular DBs like PostgreSQL, Microsoft SQL Server, or MySQL. When I discovered that queries against the same data sets were about 4-10x faster on my system, I turned my focus to the DB features. As you pointed out, there is still a lot of work to be done on the marketing and sales side if this is going to be a successful startup. I haven't done much benchmarking on insertion speed. I can insert CSV or Json files at about 100K rows per second, but I have not done a direct comparison against row oriented systems for that.