User:Jouster/open/VPP

Obfuscation

Take a look at the image to the right. Is obfuscation, via RTL, ROT13, excessive and roundabout use of {{templates}}, etc., appropriate for the Wikipedia project and in the spirit of the GFDL? Note that your opinion may vary based on whether the obfuscation is occurring in User space, Template space, or article space (though I, personally, do not feel it's valid for it to vary). Perhaps most importantly, is a sufficiently-obfuscated page fundamentally different from "stealth protection"? Jouster (whisper) 17:29, 5 July 2007 (UTC)

This might be besides the point of the discussion you would like, but I can't reproduce that screenshot; I see perfectly valid templates that aren't called in obfuscated manner. I haven't looked at the templates themselves. SamBC 17:54, 5 July 2007 (UTC)

It depends on your browser's ability to support RtL text. Feel free to ignore the actual user page referenced by the image; it's just used as an example and might change at its user's whim, anyway. The image, and what it portrays, are what I'm concerned about, along with the other examples (ROT13, crazy template hijinks, &entity; codes, etc.). Jouster (whisper) 18:01, 5 July 2007 (UTC)

Okay, so my initial gut reaction... if this is occurring outside the user space, then it's a very bad thing contrary to the point of wikis in general, and definitely wikipedia in particular. In the user space... meh. I'm not sure if a general point could be made for the user space, I mean, I'd prefer obfuscation not happen at all, but if it's used in the user space for any purpose other than to get away with things that are bad in themselves, I can't get worked up about it. I'll try to put a more coherent thought together later, preferably after more input from others. SamBC 19:03, 5 July 2007 (UTC)

Well, at the risk of being harsh, how is the User space so very different? Users don't own their own space; it's there as an extension of the Encyclopedia. Even if they did, they very specifically and freely copyleft their contributions in both user and article space. The point of the GNU and the open-source movement in general, from which cultures the GFDL arose, was to avoid receiving data that you could not modify as you saw fit. After all, that's the only difference between an executable program and its source code—one is easy to edit, one is not. With sufficient time, resources, and skill, you don't need the source code to modify a computer program; you can do it on the bare machine code. Here, we're deliberately making the "source code" harder to edit, starting from the simplest measure of reversing the text order, and running up to the extreme measure of heavily-interlinked templates that literally encode the page source in, for example, ROT13. Is there a fundamental difference between doing that and "compiling" the source code into nigh-unreadable machine code, as in a program? Would not the ethos of open source—let alone those of Wikipedia, which promote world-wide-editable source—be then violated?

In summary, I find little excuse for there to be permission to vary from the absolute commandment of readability and editability in the name of vandalism prevention, let alone vanity. This seems like a cut-and-dried case to me. Can you explain your reasoning for feeling otherwise? Jouster (whisper) 02:49, 6 July 2007 (UTC)

I guess one way to put it is that it's just as wrong, in an abstract objective moralaity, in every case, but it's less important, or less harmful (less wrong in a utilitarian sense), in user space. SamBC 13:06, 6 July 2007 (UTC)

You've complained about my userpage too many times. You even have uploaded 4 pictures (and one on an external website) about how my CSS — when exactly positioned — ruined the site. It's time to move on. —« ANIMUM » 13:33, 6 July 2007 (UTC)

Okay, if this is a recurrent head-to-head between you two, how about you try to get someone else in to arbitrate or mediate or something? Jouster, if this is a real problem that this is just an example of, can we have more exapmles? SamBC 13:51, 6 July 2007 (UTC)

Magnus is of the opinion that everything I do is part and parcel of my agenda to discredit him, or something. As for other examples, see AzaToth, for sure, and presumably others as well. I'd love to hear something on-topic from Magnus (like, for example, a defense of the merit of this type of encoding, since he apparently thinks it helps) rather than this continued belief that everything I do is about him. Jouster (whisper) 14:00, 6 July 2007 (UTC)

And is that page another example of the rtl that doesn't show up in my browser, or is there a different complaint for that one? Apologies if this sounds suspicious, I'm just trying to figure out what the problem is supposed to be. SamBC 14:10, 6 July 2007 (UTC)

No worries! Yes, it's another RTL obfuscation. A little longer, too, so you can really get an idea how difficult it would be to edit it. Jouster (whisper) 14:53, 6 July 2007 (UTC)

So, let's for a moment assume that I know nothing at all about RtL text on computers (I actually know a little, but let's just assume for now). How, exactly, would a user cause their user page source to look like that? What support is required in a browser for it to then look that, and what browsers have this support? If a user sees a page source like that and tries to edit it, will their browser edit box behave RtL as well? SamBC 15:11, 6 July 2007 (UTC)

"How, exactly, would a user cause their user page source to look like that?"

By adding the Unicode character for right-to-left text at the beginning of the text.

"What support is required in a browser for it to then look that, and what browsers have this support?"

RtL support, in general, is required; you can get kinda-crappy versions of it in older versions of browsers, but I've confirmed these changes are effective in FireFox 2.x and Internet Explorer 7.x. It's probably fair to suspect that most other modern browsers support at least some subset of the features these two do.

If a user sees a page source like that and tries to edit it, will their browser edit box behave RtL as well?

Yes, which can be very disorienting. Typing "Hi" on a RtL-formatted text input box, for example, will cause "H" to appear, and then an "i" to appear to its left, with the cursor coming to a rest at the left of the "i". The Unicode RtL character does not display in the text box in either FireFox or IE, though you can copy the text out of the page into a "dumb" editor, like Windows' Notepad or extensions-disabled Vim, and remove it. Copying it into a more-sophisticated editor will often result in that application (correctly, to be fair) deciding the text is RtL, and switching modes appropriately. Jouster (whisper) 15:35, 6 July 2007 (UTC)

Examples of obfuscation. Try to remove the red text from each. Please revert to my version when you've succeeded (or given up!):

Simple (RtL): Example 1
Text spam: Example 2
Complex templating: Example 3
Extreme: Example 4

For each example, please indicate where/when you feel its use is appropriate, and feel free to add any thoughts you have. Jouster (whisper) 16:59, 6 July 2007 (UTC)

Well, I have to admit that the RtL thing still utterly boggles me, on several levels. In any case, I say that none of those are in any way appropriate, ever, just that it matters less in user space; it's still wrong. I would certainly support the principle of a policy against this, but I wouldn't expect or support its being persued heavily in userspace. SamBC 17:43, 6 July 2007 (UTC)

(The content above is originally from Wikipedia:Village pump (policy), and is included here for context. It may be most appropriate to continue the discussion there.)