CowboyTeal writes:
"Windows 8 is still being disputed as either the product of a genius or a nerdy sadist but that doesn't mean Windows 9 isn't in the works. That said, how would you guys improve Windows if you could change anything about it? Has windows 8 improved or degraded your overall experience of the Windows platform? If you're not a Windows user, what features would you like to see in Windows for possible assimilation?"
(Score: 1) by tangomargarine on Monday March 03 2014, @10:42PM
Honestly I'm kind of curious why Linux doesn't just give up and default to CRLF in everything. I mean, CRLF should work perfectly on everything, basically, shouldn't it?
Hmm...I didn't realize that Windows is basically the only major consumer system that uses non-LF. Classic Macs used CR and all the rest is historical or embedded/technical stuff the average consumer could give a flying fuck about.
"Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
(Score: 2) by istartedi on Wednesday March 05 2014, @04:51AM
A lot of the upper layer Internet standards like POP3 use CRLF. HTTP uses it. The interesting thing is that there is what they call the "robustness principle" in these protocols. Developers were advised to "generate conservatively and accept liberally". So. At one point I had to write an analyzer that accepted bare LF from a server written by Netscape. That, BTW, was not my only bad experience with code written by Netscape. I've heard some people say that the robustness principle actually leads to things not being robust. I think more people now are of a mind that you should reject anything that is malformed, thus compelling people to generate conservatively. I guess you could say the robustness principle is half broken: "generate conservatively" == good. "accept liberally"==bad. Trouble is, people didn't want implementations that would display error messages. They wanted it to "just work", so you get bloat to deal with all the crap that people generate, a larger attack surface for malware, etc... and the rest is history.
Anyway, I digress. CRLF is in a lot of RFCs. I don't know who lead the charge on that. Perhaps nobody. Perhaps it's because there was a culture of graybeards that actually tested early protocols on line-printers and it just stuck.
Appended to the end of comments you post. Max: 120 chars.