Help me with my new website!

[FlatAssembler]

I have to check support for the FEATURE, rather than the browser.

Well, in the compiler web-app, I abstained from browser-sniffing as much as possible. Sometimes detecting a lack of particular feature and mitigating it is easy to do, like here:

Code:

if ((typeof document != "undefined") && !document.getElementsByClassName)
    document.getElementsByClassName=function(str)
    {
        var niz=document.getElementsByTagName("*");
        var ret=[];
        for (var i=0; i<niz.length; i++)
            if (eval("/(\\s|^)"+str+"(\\s|$)/").test(niz[i].className))
                ret.push(niz[i]);
        return ret;
    }

or here:

Code:

if (!Array.prototype.includes)
    Array.prototype.includes = function (x)
    {
        for (var i = 0; i < this.length; i++)
            if (this[i] == x)
                return true;
        return false;
    }

Detecting and mitigating the inconsistent implementations of "innerText" (which are probably unintentional bugs in browsers) also wasn't very hard, all I had to do was this:

Code:

var kod = document.getElementById("assembly").innerText;
if (!kod)
{
    alerted = 1;
    alert("Emitter error: Your browser doesn't appear to support the JavaScript 'innerText' directive. Can't send the assembly code to the server. Please use either Internet Explorer or some modern browser.");
    document.getElementById("AJAXmessage").style.display = "none";
    return;
}
if (/\n\n/.test(kod))
    kod = kod.replace(/\n\n*/g, '\n');
if (!/\n$/.test(kod))
    kod+='\n';

However, sometimes detecting a particular feature isn't easy. For instance, the code I've used to detect SVG support is this:

Code:

if (!document.createElementNS)
{
    alert("Diagram error: Can't access the SVG elements from JavaScript. You appear to be using a browser that doesn't support that. If you are using Internet Explorer 11, make sure it isn't running in the compatibility mode.")
    alerted = 1
    return
}

Is that correct? It's conceivable that some particular browser supports "createElementNS", but can't render JavaScript-generated SVGs, right?
The CSS Box-Model bug in Internet Explorer 6 also isn't trivial to detect. The code I ended up using, after lot of experimenting, is this:

Code:

if ((document.getElementById("zatvori").offsetWidth<18 || document.getElementById("zatvori").style.width=="21px") && document.getElementById("zatvori").style.display!="none") ie6=true;

The "onmouseleave"-bug in Android Stock Browser 4.1 doesn't appear to be detectible via JavaScript, does it? Neither is the Android Stock Browser itself easy to detect via JavaScript, since it uses almost the same user agent as Safari (which doesn't suffer from that bug). I mean, the work-arounds I've ended up using look like this:

Code:

if ((typeof navigator != "undefined") && /(M|m)obile/.test(navigator.userAgent))
   document.getElementById("content").onclick = function () {
      document.getElementsByTagName("nav")[0].style.display = "none";
};

That assumes all mobile browsers suffer from that bug, this is not true and it will be even less true in the future.
The "z-index"-bug in Android Stock Browser 4.1 also appears not to be easily detectible via JavaScript. I've mitigated it similarly, but I suppose that if I managed to find a reliable way of detecting Android Stock Browser, my web-app would run smoother in other mobile browsers.
And, the only work-around around Internet Explorer 11 ignoring the CSS "overflow" property on the JavaScript-generated SVG elements on my homepage is browser-sniffing, right?

Code:

if (!(navigator.userAgent.indexOf("Trident")+1))
    crtanje.appendChild(krug);

OK, maybe using SVG polylines would be slightly better, but this is far easier to do, isn't it?

I suspect that the WAY you said things might have been part of your banning, not only your political position.

If I remember correctly, I posted the following on a day of an anniversary of the Genocide of Vukovar: "Čemu plakati za Vukovarem? Po svemu što znamo, Bitka za Vukovar mogla bi biti samo mit!". That means: "What's the point of crying about Vukovar? For all we know, the Battle of Vukovar may be just a myth!".

---

By the way, which browser do you think will become the most hated among web-developers once Internet Explorer, Microsoft Edge and Android Stock Browser (all of which are either deprecated or about to become deprecated) disappear? I think it's going to be Firefox. I've had to mitigate Firefox-specific behavior two times in the compiler web-app:

Code:

if (document.getElementById("zatvori").style.left.substring(0,document.getElementById("zatvori").style.left.length-2)
-document.getElementById("content").style.width.substring(0,document.getElementById("content").style.width.length-2)>10)
    document.getElementById("content").style.width=
        document.getElementById("zatvori").style.left.substring(0,document.getElementById("zatvori").style.left.length-2)-3+"px";

Code:

if (!window.event) {
    isFirefox=true;
    window.event=new Object();
}
if (isFirefox)
{
    window.event.clientX=e.clientX;
    window.event.clientY=e.clientY;
}

For comparison, I've had to mitigate Opera-specific behavior once, and I've never had to mitigate Chrome-specific behavior (I was developing in Safari).