Re:Show HN: Fast C-based HTML 5 parsing for Python
by exabrial on Monday June 19, 2017 @02:25PM
(#14589155)
Opinion: permissive parsing is bad for the industry. Speed should be king. It's not difficult to write correct html!
[ Parent
| on HN ]
Re:Show HN: Fast C-based HTML 5 parsing for Python
by kbutler on Tuesday June 20, 2017 @08:43AM
(#14594127)
Be strict in what you produce and liberal in what you accept.
Strictness works well when you have a small number of active producers and consistent, constantly renewed content.
The web is the opposite - a huge number of producers and widely varying content, some of which is 1-2 decades old and will never be updated.
You can design and build a great, strict HTML parser - it just won't work well for a significant portion of the www.
And if you make one that is popular enough, you can change the direction of evolution of the web! (See mobile safari and flash)
Re:Show HN: Fast C-based HTML 5 parsing for Python
by pwdisswordfish on Monday June 19, 2017 @02:50PM
(#14589372)
Isn't it because 'incorrect HTML' has been practically defined out of existence?
Re:Show HN: Fast C-based HTML 5 parsing for Python
by RUTHLESS_RUFUS on Tuesday June 20, 2017 @12:59PM
(#14596275)
Opinion: the industry doesn't equate with reality. Pointy-haired MBAs with some PHP background would tell you otherwise, but there's a whole wide world out there.