LogFAQs > #884031611

Yellow posted...

Right click this page and press "inspect" or "view source". See all that URL randomly generated gibberish?

*does so*

What I see is the sort of typical generated markup common to large sites like this with various templates filled out by the server.

It's not gibberish to me, because I know what I'm talking about.

In fact, I saw nothing that would indicate any sort of client-side content filtering. I even disabled the scripts on GameFAQs and turned on the content filter just to check, and the fucking fuck fucks were censored. This indicates a server-side operation.

Yellow posted...

Do you think you can filter the word "tit" from there with a "global filter"? No, you would break things.

But that's not where the filtering is happening, and no one's saying it is.

Also, there's no reason you can't do that kind of filtering client-side as long as the script knows what fields to check. This is why I knew to check it by disabling the scripts, because this breaks client-side content filtering on other sites (sites I can't mention openly). I can't link to the site, but in the page source you get something like:
<script id="user-blacklisted-tags" type="application/json">
["string_1", "string_2", "string_3"]

I've redacted some private information, but you get the idea. What that line does is define a set of strings associated with certain elements on the page to be stripped out. A script running in the browser (not the server) matches those strings against metadata assigned to every relevant element, and if it finds a match, that element isn't displayed. A similar method could easily be done ahead of each post in a GameFAQs thread, but instead of simply stripping out the elements completely, it does a character replacement. Of course, this would require GameFAQs to post every word in its filter in every page, which is kinda counterproductive when you're trying to keep those words from getting served to the end user.

And that wouldn't be a "global filter" per se, but it's really two different discussions. I'm talking about a global server side filter, one that checks all user text coming in, and modifies it going out. Not something that does a hard and blind (and frankly so brain-dead no one would ever actually do it) 'find and replace' on raw HTML.

Okay so serious question here - since I can prove you're 100% wrong on this by pointing to working implementations of a process you outright said "would break things", why should I take anything else you have to say seriously? I mean, I don't have a background in web development. I'll be the first to admit that I don't know much about it.

But what I do have is 20 years of programming experience. I've worked on MUDs, some of the most complicated text serving applications outside of academic natural language processing, and know what a fucking string is. But I'm not an HTML coder, so maybe you don't want to hear it, never mind that I can identify differences in content filtering styles from one site to the next just by reading the page source. >_>
---

This album is not available to the public.
Even if it were, you wouldn't wanna listen to it!