| Age | Commit message (Collapse) | Author |
|---|
|
Now any exception that is not related to sockets will be logged
Threads should still clear themselves even after exception
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Fixed bug with file data stream fetching, as when file reach the end they return `b''` and not `None`
Bug was causing threads to run forever, effectively completely halting the website from functioning
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Use semaphore to limit core usage slightly (maybe)
Also changed default configs a bit
|
|
Thread list was not cleared, so it was accumulating a ton of dead threads, now it should be fixed, as each thread removes itself at the end of request
Added limit to how much data the client can send, as currently client data streams are unsupported
Do minor renaming and allow to switch between HTTP and HTTPs
|
|
|
|
Probably useful to have it
|
|
|
|
Odd because I don't know why I coded it that way
|
|
|
|
|
|
When a client decides to update a page a bunch of times, it sends this error to a server. There's nothing server can do about this
|
|
|
|
Previously _client_thread was checking for new data from client, which is very much useless, because the client will not send any data back twice.
This commit removes the useless while loop that was there, as well as the thread that was started
So currently, it's 1 thread per 1 request, which is how it should probably be
|
|
|
|
|
|
Instead of sending full files, and therefore storing them in ram (which is not always possible), send data in chunks of 64KiB
|
|
|
|
Removed timeout because it interrupts client's downloads
|
|
|
|
|
|
|
|
|
|
|
|
|
|
There is a weird issue when a user tries to load a page, doesn't get a response (because that wasn't implemented yet), and when the webserver is attempted to be shutdown, it still tries to fetch user's request
|
|
|
|
|
|
|
|
Problem was that I wanted to do things actually in parallel, and wanted to launch a task but without blocking the thread, asyncio was for whatever reason blocking the thread which was causing issues when 1 user was downloading a file and trying to do something else on the page (like load a new page or something)
Threading ACTUALLY does do exactly what I want via `threading.Thread(...).start()`, so that will be used going forward
Also, it seems to 3x faster that way, because with asyncio the page took 1 sec to load, with threading it takes 280 ms, which is weird even if we are doing things one at a time
|
|
|
|
|