I know the "official" reason, but I'm curious if there's any correlation between the blocking of mibbit and axod's post about how much revenue mibbit was making here (http://axod.blogspot.com/2009/06/revenue-browser.html) - the timing is awfully suspicious.
I mean, a day after the post they "magically" have their own web client up that they would rather you use.
I had a look at the qwebirc client they implemented, default installation right now (looks like it was just thrown up) no ads yet, wondering how long before that changes?
EDIT - well it seems that in part some of the argument from freenode's perspective is about money.
(00:44:02) axod: nenolod: and please explain what incentive I would have to falsify such information
(00:44:17) nenolod: axod: assistance with ban evasion in exchange for money
(00:44:26) axod: nenolod: seriouslly?
(00:44:30) nenolod: axod: seriously.
(00:44:33) axod: lol
I've actually spoken to several users who found that to be true. Even if silly, from the point of view of someone who knows what IRC is. Which is far from everyone.
//edit: don't care much about karma, but I am curious as to how pointing out that not everyone knows what they're doing on IRC gets me downvoted.
Oh. Thanks for explaining. I didn't mean to imply that mibbit users are retarded. Just that they don't quite know (or care?) as much about IRC as other people do.
I take offense at that. Do you have enough data to make that judgement? Out of the millions of people using Mibbit, you've done a survey and measured? Or is it just more likely that you'll notice if someone causing abuse is using mibbit?
Erm, might I suggest you reread what I said here, in this particular thread. Which amounted to saying that not all Mibbit users know or care much about IRC. Which is a very silly thing to be taking offense to, let alone demand to be proven.
You should probably point out here that nenolod was one of the non-freenode-staff members in #freenode at that point, presenting his personal opinion. It's about as germane to this discussion as me pulling up random person B's comment about how mibbit is <insert sleazy thing here>.
Moving past the heated conjecture, the more accurate statement from a freenode staffer:
njan: Xenos, for what it delivered, mibbit i)
disrupted a large number of our network users, and ii)
used a disproportionate amount of staff time.
njan: Xenos, we chose a solution that fixed those two
problems - using staff time and disrupting our 50k users
- whilst giving the mibbit users an alternative.
Freenode doesn't need -- or have any responsibility to support -- Mibbit. Moreover, we've banned mibbit entirely from our channels without receiving any complaints or noticing any significant drop in channel traffic.
That's a ridiculous statement. I was in the #freenode channel for a few hours, I saw many people (other than the people coming from #startups) who were logging in to ask why Mibbit wasn't working.
Quite a few of my development peers sit behind such firewalls and yet can and do still connect using their desktop clients. It's not something we discuss, so I don't know what method they're using, but as a member of the community in question it's fairly readily assumed that you'd know how to handle such a situation. We're under absolutely no responsibility to provide "IRC accessibility" to all comers.
To rephrase things another way, would you ban mIRC if you determined that 100% of disruptive users used it as their client? Even though there are plenty of legitimate users of mIRC?
It seems more likely to me that there is some sort of objection to 'web clients' because desktop clients are the 'one true way' to access IRC. Note that I say this as someone that prefers to IRC from console clients like BitchX and irssi through a screen session.
I known plenty of people that have used IRC that don't 'know' what IRC really is other than how to use the graphical elements of their client to access things for the 'chat room'.
And in this case you are not 'providing' IRC accessibility. Mibbit is providing that. The only thing that you're doing is restricting accessibility by banning Mibbit. Now if Mibbit were banned by default on your IRCd and you had to 'grant access' to them you could argue the other case.
If mIRC, by it's nature, resulted in a disproportionate maintenance cost as compared to other clients, then yes, I would seriously consider banning it. Mibbit requires additional support from Freenode: trust of its declared client IP addresses, and additional abuse support from the Freenode staff.
I wouldn't demand AOL expend any effort whatsoever to support my 3rd party AIM service without remuneration, and regardless of remuneration offered, it would remain their perogative to say "no".
Why is Freenode, as an independent market actor, not allowed the same prerogative? Is it because Freenode operates as a non-profit, while Mibbit is a small startup?
Why is Freenode, as an independent market actor, not allowed the same prerogative? Is it because Freenode operates as a non-profit, while Mibbit is a small startup?
I'd suggest that freenode's arguments seem to be fairly hollow given that freenode as a network is populated by tech savvy users and if they want to cause strife on the network, then the client they use wont make much different, they will find a way.
I personally find it incredibly convenient that axod writes a post about mibbit's revenue that is posted to HN and then the next day, mibbit is banned from freenode but that's ok, freenode has implemented their own webchat client that you can use. Great timing no?
Given we've established that people trolling freenode will find a way to do so regardless, that could be construed as a little fishy, no? Especially since it's just a default installation of an open source client - it wouldn't surprise me if ads pop up on the freenode webchat client in the next few months.
I also find it interesting that even though axod was in channel asking the #freenode ops what he can do to fix the issues and the general response was "sorry, it's done now" - stonewalling him only lends weight to my argument.
Don't get me wrong, I'm all for freenode trying to make money to support their servers, but at least be transparent about it.
I'd suggest that freenode's arguments seem to be fairly hollow given ...
The point is, isn't this entirely their perfectly reasonable prerogative as free actors? Given that my own minimal channel 'work-load' has increased due to Mibbit, I'm inclined to take their position at face value.
However, if I were to entertain the notion of a secret plot for IRC monetization, I still wouldn't begrudge them that choice.
... freenode as a network is populated by tech savvy users and if they want to cause strife on the network, then the client they use wont make much different, they will find a way.
Mibbit's user base is relatively self-selecting. I don't believe it's inconceivable that the Mibbit user base is much more inclined to cause strife.
Whenever I go hiking in a relatively populated area, I always am bemused at how litter and graffiti disappears just a mile or so down from the trail head -- it seems that anyone willing to leave behind beer cans and spraypaint trees is also generally unwilling to walk more than a couple miles to do it.
> However, if I were to entertain the notion of a secret plot for IRC monetization, I still wouldn't begrudge them that choice.
I think people are taking issue with them being secretive about it. "We're banning Mibbit users because they abuse our network" vs "We're banning Mibbit users because we want them to use our client instead so that we can make money" are a little different from each other. If the latter is true, then they should at least be open about it.
Openness is preferable, but there is absolutely no direct evidence to support the conclusion that they wish to monetize Freenode via web chat -- the timing is only circumstantial.
"Your honour, I thought he might object to being shot, but after I did it, he never ever raised the question! If that's not silent agreement, I don't know what is!"
If a project developer, active user, or someone else with genuine stake in the project community complained, we'd pay attention ... but nobody has, despite the numerous venues available to them to do so. Your honour.
It's not our responsibility to account for, be accountable to, or be concerned for outliers.
I'm a project developer (Django + other things), IRC has been blocked at my last two places of work and I've often used mibbit, which I found to be an excellent client - better than many desktop clients in fact. Please consider this a complaint.
Then if the Django IRC channel has banned Mibbit, I suggest you speak with them -- assuming Freenode decides that expending volunteer services to support a commercial service is a worthwhile venture.
Otherwise, there's the freenode web IRC interface.
You just said you would pay attention if someone "with a genuine stake in the project community complained". And then someone did, and you brushed him off.
"Our channels". As I've said before, we've been forced to regularly ban Mibbit on our channels (as in, software projects with which I am involved). I simply don't represent freenode, or django, or web developers in general.
[edit]
I see Mibbit fans are downvoting any negative comments en masse, not to mention the vitriolic dialog with freenode staffers on IRC. Yeesh.
Honestly, freenode is loaded up with tech savvy users. This will solve nothing. If users want to cause strife on channels, the tool they use is inconsequential. Below are two scenario's off the top of my head.
Scenario #1 - people will be using proxies and using freenode's webchat client = banning mibbit was pointless
Scenario #2 - people will be using their own irc clients and using open proxies = banning mibbit was pointless
So banning of mibbit hasn't actually solved the problem, but like I theorised above, has the potential to funnel revenue to freenode and given the timing of axod's post about revenue, with the banning of mibbit only a day or so after followed by the awfully convenient default webchat installation that freenode put up, it smells a little fishy (even if it's tin foil hat territory, you still have to admit it's bad timing on freenodes part).
EDIT - Scenario #3 - lowtech, just reconnecting off your router to refresh your IP and connect from home = banning mibbit was pointless
I think the tool required to access a service is pretty darned important in determining your audience. Case 1: September 1993 (AOL on USENET). Case 2: Hacker News (intentionally plain design compared to, e.g., Reddit or Digg).
Scenario #0: people will keep using mibbit, because it's convenient and what they used before, and harass IRC channels on other networks instead.
(Sure enough, freenode's web client looks like basically a teletype in a web browser, while Mibbit's is big and flashy and web 2.0-ish. It's the HN to Mibbit's Digg.)
Ok, so you're suggesting that there is a correlation between how attractive a user interface is and the number of users/trolls?
Well then let's break it down to its most simplistic form, the Operating System, seeing as that affects every service.
How do you account for Windows and OSX, seeing as most of your trolls would come from Windows based machines and yet most people would agree that OSX has the superior/most attractive GUI?
I think you're over-generalizing. I'm only claiming that interface has a big effect on behavior you see in a particular place, not how the behavior got there. I certainly don't believe (and didn't mean to imply) that seeing an "attractive" interface causes a behavior change in an individual.
I also disagree that comparing operating systems is "simplistic". What operating system a person is using today is based on a huge number of factors, of which "attractive" is but one tiny piece -- a far more complex issue than what web chat you use.
What you're observing is what I'd call the same effect. People are lazy about sticking with an interface they know. AOL opening up USENET meant a bunch of jerks could screw things up, using the same interface they'd had. Digg/Reddit predate HN, and the jerks there aren't going to learn a new interface to be jerks to a smaller crowd here. Similarly, you're not going to change OS (and with it, interface) easily, so other effects dominate, and I'd expect most jerks, overall, to be using the most popular OS, overall. I don't know if that's true, but it fits my mental model.
Well then the effect you're describing has more to do with population density than interface?
When things start getting crowded, people start generally becoming rowdy - it seems to be human nature. You can have a small home party with 10-20 people and it's fine. Throw in a hundred gate crashers and it becomes a nightmare.
The thing with the internet is the anonymity that is afforded by such communities means that there is less of a barrier for people to become jerks. Look at YouTube or 4chan.
Youtube has a huge population and the comments are filled with volatile diatribe. Conversely, look at Vimeo - it's a much nicer interface, yet it doesn't have the same community issues that Youtube has.
Now let's look at 4chan. It has a simplistic user interface, but with extreme anonymity and a large enough population, it's the cesspool of the web.
I just don't think your "pretty interfaces = bad" argument has any merit.
Digg wasn't half bad when it FIRST came out, but as more people flocked to it, the quality went down. Ask a few of the old HNers and they'll argue the same about HN since it has grown in size, they argue that the quality has gone down. I mean, we see a thread complaining about it almost every other week it feels like.
Old scenario - people are using mibbit to very easily banevade - crap.
Scenario #1 + #2 - the proxy will be banned or the user refused connection. There is no distinction because they can trust the code they're running to report the IP correctly.
Scenario #3 - "ah crap, you mean I have to reset my router instead of just refreshing mibbit?"
From freenode.net/policy.shtml : freenode may block access to users whose IRC clients run on hosts with open proxies, IIS servers or other categories of software determined to present special risk to our server environment.
nenolod is not a freenode staffer. The impression I got was that freenode felt they were going above and beyond in offering an alternative webchat to mibbit users, which they didn't need nor particularly want to do.
I mean, a day after the post they "magically" have their own web client up that they would rather you use.
I had a look at the qwebirc client they implemented, default installation right now (looks like it was just thrown up) no ads yet, wondering how long before that changes?
EDIT - well it seems that in part some of the argument from freenode's perspective is about money.