That script is incorrect as you well should know.
Yes, as I said, semi-colon where there should be a comma. Where for many that minor typo is effectively invisible hence:
Even if there are multiple errors then clearing the first error usually clears other errors because the previous syntax was incorrect and affects following lines.
Except the first error they list is three lines later! Hence you're not looking anywhere NEAR where the actual error is! It took me multiple passes and several minutes to find that semi-colon even with the validation report because the first of three errors was significantly
AFTER the actual error location! As such you're not quickly finding a damned thing. When the first error reported is the derpy validator claiming that :background is an invalid psuedo-state
(when it's a huffing property that's completely valid) that ceases to be a useful debugging tool.
Laugh being if I had run my normal checks -- via Edge's document inspector that also reports CSS errors -- it would have been caught in a meaningful manner by reporting the ACTUAL invalid semi-colon.
I'm worried that blink based Edge is going to lose that feature.Introducing your own syntax
You've said this a few times and I have to say,
WHAT introducing my own syntax? This is just me disagreeing with things they added in HTML 5, and REALLY disagreeing with things they removed in HTML 5.
See the derpy switch of TFOOT and TBODY order, the removal of projection,tv from the spec, reporting errors for
once VALID vendor created media targets
(the entire POINT of the media attribute) because they now have some dumbass magical list you have to match.
The media targets of the media attribute should be none of the HTML specification's bloody business, and under 4 Strict that's exactly how they worked. It's the job of UA makers to determine what values to support and how... JUST like there should be NO default appearance for HTML elements in the HTML specification, because that's not what HTML is for or why it was even created in the first place!
If it were 4 strict both projection and tv were valid, as were ANY other values because it was up to vendors not the specification to create them so as to best handle any conceivable devices past, present, or future. Their changing that is one of the MANY things in HTML 5 that makes it total 100% herpafreakingderp.
Hence my zero confidence in the WhatWG or W3C in terms of the direction they're dragging HTML... right back to the ugliest and worst of 1990's browser wars era practices.
The PHP Framework rendered page is faster than the single html page and there are no slowdowns?
Sorry but bullcookies. Over ten runs the two versions off your server (both of which seem to be PHP) trade blows +/- 300ms here, though the average of 350ms for the original page and 240ms for the one in the subdirectory tends to put that to the lie.
But again those are pingdom's numbers and they provide anything BUT consistency in testing... never have with anywhere form 200ms to a full on second of difference across runs not being at all unusual.
How many runs did you make? What were the best, what were the worst, what was the average. If you only ran the test once, it probably doesn't count.
Perhaps not good but at least it does not time out
I think you completely missed my meaning. Accessing it directly via it's URI
1) should be blocked. It should 404. I should NOT even be able to access that file.
Take for example:
https://cutcodedown.com/template/common.template.phpThe file is there. You're NOT allowed access to it via http/https. That's a good thing.
(something I need to fix with this SMF install in fact...)2) if I AM able to access it, as an include it should output NOTHING since that can become a security risk depending on what's in the include. In this particular case there are no nasties, but it's bad practice to make includes that are raw code. Particularly those that form the document <head>... hence the content of such an include should be wrapped in functions.
But again, <?php only once, no ?>, and any/all output wrapped in echo...
Though to be fair, I'm the nutter who thinks that heredoc, nowdoc, php shorttags, and double quoted strings should be stricken from PHP entirely. Just as includes should block direct output and instantly break local scope.Hence my using this little bit of derpitude all the time:
function safeInclude($filename) { include($filename); }JUST to break the scope where an include occurs so no locals are passed... but then I isolate scope -- PARTICULARLY of the database connection -- in ways most people never even think of.
