Hacker News Reader

Comment by willcipriano | Hacker News Reader

Curl supports: "DICT, FILE, FTP, FTPS, GOPHER, GOPHERS, HTTP, HTTPS, IMAP, IMAPS, LDAP, LDAPS, MQTT, POP3, POP3S, RTMP, RTMPS, RTSP, SCP, SFTP, SMB, SMBS, SMTP, SMTPS, TELNET and TFTP. curl supports SSL certificates, HTTP POST, HTTP PUT, FTP uploading, HTTP form based upload, proxies, HTTP/2, HTTP/3, cookies, user+password authentication (Basic, Plain, Digest, CRAM-MD5, SCRAM-SHA, NTLM, Negotiate and Kerberos), file transfer resume, proxy tunneling and more."

I could see creating quality implementations of each of those taking that long.

https://curl.se/

View on HN

MQTT even?

Yup: https://curl.se/docs/mqtt.html

That sounds like a bad architecture and a bloated app. It's like if VSCode decided "we don't do plugins, everyone commit your idea directly into our main repo"

A sensible plugin architecture would let you do either

    curl --plugin=LDAP

or a naming convention would load the appropriate plugin based on the protocol.

Yes I know, armchair code design. And I also know it's nice not to have to deal with configuration, a large app "just works" (though FYI, it does have 15-20 dependencies which are similar to plugins, it has 3 direct and more indirect dependencies). But, 2hrs a day * 23years = 16790hrs!!!!! Something seems wrong for that much time to be spent on "curl"

Just to be clear, first, thanks so much for curl! And second, of course having not touched the code ever I have no idea why it required 16k hours of dev time and continues to require more, it could totally be justified.

I'm only saying that at a glance, 2hrs a day, 23years sounds surprising and it makes me wonder if there isn't some ways to reduce the load.

I doubt re-architecting into modules would make the bugs in each of those protocols go away. The bugs would still be in the modules.

You seem to be conflating curl as only an app that deals with http/https. Curl really does a lot more.

It's not about the bugs. It's about one person putting 16k hours time into it instead of 100 people each putting in 160hrs each (yes I know that's not how it works but something is arguably wrong if it takes 16k hours of one person's time for something like curl

Mainly it is also a library that others can use for all of those protocols and SSL backends.

Reply on news.ycombinator.com