From this perspective, Docker might just be the greatest invention of all time โ turning โit works on my machineโ into a global standard. Pure genius.
asromafanisme
16 days ago
It doesn’t work on my machine
But it works on production
Now it also doesn’t work on production
After investigation: what kinds of magic make it work on production previously, the code makes no sense
Ximidar
16 days ago
Honestly package managers and build pipelines don’t get enough credit. They are modern marvels that just work. (Most of the time)
Porsher12345
16 days ago
Someone pls explain
DogOk6506
16 days ago
Name of the film?
9xl
16 days ago
Assuming the machine has the same cpu architecture.
SaltMaker23
16 days ago
The “work on my machine” crowd will take all roads installing 50 libaries/services on their host rather than running Docker or any other VM.
“It’s faster”, “it works better”, “I don’t understand docker”, “it’s easier this way, less complex”. Running everything differently the only shared part of the system’s setup is only the small portion of files that are part of the git repo.
Ofc everything runs on a different version, setup, or a whole different thing (eg: file caching instead of Redis, because obviously if you don’t use the provided docker setup, a well designed Redis system that integrates as expected with our system can become an impossible task to solve when you refuse to use things that are available)
Until the ultimate “why doesn’t it work in pipeline/staging/production ? it’s fine when I run it locally”
From this perspective, Docker might just be the greatest invention of all time โ turning โit works on my machineโ into a global standard. Pure genius.
It doesn’t work on my machine
But it works on production
Now it also doesn’t work on production
After investigation: what kinds of magic make it work on production previously, the code makes no sense
Honestly package managers and build pipelines don’t get enough credit. They are modern marvels that just work. (Most of the time)
Someone pls explain
Name of the film?
Assuming the machine has the same cpu architecture.
The “work on my machine” crowd will take all roads installing 50 libaries/services on their host rather than running Docker or any other VM.
“It’s faster”, “it works better”, “I don’t understand docker”, “it’s easier this way, less complex”. Running everything differently the only shared part of the system’s setup is only the small portion of files that are part of the git repo.
Ofc everything runs on a different version, setup, or a whole different thing (eg: file caching instead of Redis, because obviously if you don’t use the provided docker setup, a well designed Redis system that integrates as expected with our system can become an impossible task to solve when you refuse to use things that are available)
Until the ultimate “why doesn’t it work in pipeline/staging/production ? it’s fine when I run it locally”