On 03/03/2014 04:34 PM, Tres Finocchiaro wrote:
User agent spoofing is common enough that we should take it in account. 

@Vesa,

I'd like to see some evidence in support of this claim. UA spoofing generally requires a 3rd party add-on installed in the browser.  It may be common for a handful of our developers, but our developers aren't necessarily using our home page to download the software, so the argument is still invalid.

UA spoofing is used widely enough in the Linux community. What do you want, a survey? I'm not aware of any such survey being conducted... There are tons of tutorials all over the web describing how to spoof UA. There's now been what, 3 people posting in this thread only who've stated they do UA spoofing, and I'm not even sure if all of them actually are developers (at least not currently active ones).

UA spoofing is used as a security measure, and we certainly should not be making usage of internet any harder for people who care about security. We should not legitimize any practice that depends on users compromising on their security or privacy to function.

there's been a lot of mentality lately to try to somehow shove the "Linux" part under the carpet, in some kind of fear of alienating windows users. I don't really know what to make of that. I think part of the problem is that many windows users just see LMMS as a "free replacement for FL studio"

This claim is preposterous.  You could make the same claim that people choose Ubuntu a free replacement for Windows or OSX.  You have NO RIGHT judging the motives of people's ability to choose.  You have no right making broad claims about the LMMS community's motives for involvement based on platform.  Furthermore, you do not know the lengths Stian, unfa, Uros have gone to increase involvement.  These type of words, such as "shove linux under the carpet" and "many windows users" are the very seeds in which plant discriminatory dialog.  Let's not take this dialog any further in this thread.


There's nothing "discriminatory" about this. Being a "windows user" is a self-inflicted condition, it's not like something you're born with and can't help being.

It's simply a statistical fact that the average windows user brings less to the table of an average free software project, than the average Linux (or Mac) user does. Part of this is due to different software culture and expectations of the users. I'm sure once some commercial Linux distro achieves real mainstream popularity (Tizen, maybe..) this situation will be different. But right now, it isn't, and it's not going to change by denying it.

And yes, I'm sure that many people DO choose Ubuntu purely because it's free, and see it as a "free windows" or "free mac os". In fact I've seen it in practice... And that's fine really. We can't control what people think - if users want to consider LMMS a "free FL studio", then they're going to and we can't change that, but we shouldn't **encourage** that kind of attitude.

I think we have a pretty solid divide between a fundamental assumption in site design... The use of JavaScript or not.  I've seen Johnathan recommend Bootstrap many times talking about redesigning the site, so if JavaScript is off the table (and I'm not sure who decides this) then it should be determined now.

The #1 rule to decide whether a site makes use of JavaScript or not is the site's users.  I would be conformable representing the vast majority of our site visitors and say that they do have JavaScript enabled.  (This assumption is due to the fact that the Facebook page doesn't work well without it).

If we vote to rule out JavaScript, I'm afraid LSP2 might hit some brick walls.

Using Javascript is entirely fine when there's actually a reason to use it - when you actually need it for some functionality. When it's just there to facilitate a flashy gimmick, I don't see the point.

We can implement functions that make use of Javascript, but the absence of Javascript should not break down the site - it should still be usable without Javascript, you should be able to look at the front page, read the news, look at screenshots and download packages without Javascript - none of those things **require** JS to implement.

As for the OS guessing, I already presented a perfectly viable suggestion, which I'll recap here:

When Javascript is available, display the guessed OS first, the other packages after it. No extra link or button, display all at once, but display the guessed OS first, maybe a bit separated from the rest for emphasis.

When Javascript isn't available, or it fails to decipher the user's OS (or the OS is in the "other" category), just display all of the options equally.