Contest For a Better Open-WRT Wireless Router GUI 217
Reader RoundSparrow sends word of a contest, with big cash prizes, being mounted by a commercial vender of open source Open-WRT routers. You have 10 months to come up with "the most impressive User Interface/Firmware for Ubiquiti's newly released open-source embedded wireless platform, the RouterStation." Entries are required to have open source licensing and will all be released. First prize is $160,000, with four runners-up receiving $10,000. RoundSparrow adds: "Could be built on top of existing X-WRT or LuCI OpenWRT web interfaces. OpenWRT Kamikaze 8.09 was just released. Now is perfect timing for OpenWRT to get some kick-ass interface and usability ideas. I'm not affiliated with the contest vendor."
X-WRT? (Score:5, Insightful)
What's wrong with X-WRT?
OpenWRT is something you set up, then forget. It doesn't need "themes" or "skins", or 3d effects. This is not "pimp my router".
I have a suggestion ... (Score:5, Insightful)
Please.... (Score:5, Insightful)
Oh god no.....no Web 2.0 Crap. the router GUI is supposed to be fast, small, and compatible with EVERYTHING.
DDWRT has a problem with Firefox on the latest builds because of the stupid Web2.crap to make things more flashy instead of working right.
I am really tired of the "ooh shiny" becoming far more important than functionality lately in both projects.
Re:X-WRT? (Score:5, Insightful)
It is amazing how negative everyone is about this (Score:5, Insightful)
It's not like it's your money! I currently use Tomato on one of my routers. I love the interface. I don't log in very often, mostly to check those fantastic real-time usage stats.
But when I do log in, it is nice to be able to find things quickly. I respect developers who take into account usability and style. In fact, I have basically no respect for those who discount it.
You probably can code circles around me. But in the end, the customer or user only sees the interface. They only see those "useless" graphics, and that "Web 2.0 Crap". Yet, a well designed interface will allow new users to appreciate the product faster, and hopefully keep them around.
Just because the majority of web developers suck at designing "web 2.0" interfaces doesn't mean that the problem lies with the "web 2.0" part. We'd have a lot less technology if we used that metric to measure a tools value.
open source and usability (Score:5, Insightful)
Re:Please.... (Score:3, Insightful)
I am really tired of the "ooh shiny" becoming far more important than functionality lately in both projects.
I do netmgt for a living (client and server side, both) and I've turned down jobs that emphasized glitz over actual *needed* functionality.
I still maintain that a simple forms/cgi interface with NO javascript is all you need to get the job done. I wrote an entire NMS on form/cgi (1998 era) and it didn't have 'active stuff' but who the hell cares! the only lacking I had was no dynamic graphs - and I solved that by computing .gifs from the GD library and that had the benefit of being able to do a simple save-as for the graphs!
the requirements for form/cgi is SO simple. even a toaster can support it.
but noooooo. no one wants simple anymore. they want jscript, they want non-bookmarkable pages, they want LAYERS of complex code (job security?) - but they don't want basic, reliable, platform-independant code.
the trend saddens me. netmgt is supposed to be THIN. somewhere, the old school netmgt guys must have all retired or died out and the new ones never learned about the lean-n-mean nature of what netmgt is (used to be) about.
Re:needs an easy way to edit firewall rules (Score:5, Insightful)
Not everyone should be administering a network either. Any literate adult meets all of the requirements necessary to learn how to do so but there are plenty of people who should not perform this task. The system is self-correcting however; the ones who shouldn't have done so are the ones who have most of the problems. If you a) don't know how to properly do something and b) refuse to learn how to properly do it, then it makes sense to ask (or hire) someone else to do it for you.
The people who think that this simple observation somehow does not apply to them, or that getting pissed off at someone like me who points that out is going to change the reality of the situation, well, I bet they wonder why they have such bad "luck" with these things. You attempted something that you don't actually know how to do and experienced undesired results; what a surprise, it must have been those evil elitist geeks! Seriously though it's amazing how upset people get sometimes when you dare to suggest that there are tasks which require a bit of skill and that doing them without that bit of skill can cause problems. You'd think that this were some kind of highly controversial position for which there was no conclusive evidence.
To more directly answer your post, I think iptables itself is rather irrelevant. The story is about a router GUI, which would probably be a front-end to iptables. There are some very nice GUI tools available for iptables; if Open-WRT's offerings are on a par with them, then they would provide a way to edit firewall rules that's about as easy as it's going to get. I do think that a firewall is one of those few applications where there is some inherent complexity that cannot be made much simpler without severely compromising the device. It's like that Albert Einstein quote (paraphrase): "Things should be made as simple as possible, but no simpler."
For that reason, I question the type of "easy to use" to which you refer. If you have a solid working knowledge of TCP/IP, then you should be able to handle any firewall and "easy to use" would mean automating what can be automated to save you some keystrokes and to avoid some unnecessary tedium. If you don't have a solid working knowledge of TCP/IP, it would probably mean dumbing things down to make up for your lack of understanding, which of course would result in a less thorough or a less accurate configuration.
Given the security issues that can arise from a misconfigured firewall, I would suggest that this is one area where enabling people who don't really understand what they're doing is asking for trouble. You're not really doing the less-knowledgable any favors by setting up a situation, in the name of convenience, where they are likely to have problems that they won't know how to solve. The good news is that there is abundant documentation on both TCP/IP and iptables, so anyone who is interested and motivated can easily learn how they work.
Re:It is amazing how negative everyone is about th (Score:5, Insightful)
In fact, I have basically no respect for those who discount it.
You probably can code circles around me. But in the end, the customer or user only sees the interface.
Actually you've hit on a major problem of programers that we don't like to talk about (well, except me, obviously..). The thing is, GUI design is a complex art, one that takes a long time to learn to do well, so its hard to be good both at visual interfaces and the often very complex code that they control.
I know this from my own work. I'm a pretty good coder (gosh, how modest of me). I can write code to just about anything, and charge a pretty penny to do so, but my ability to code a user interface is rather poor. Sure I know all the theory, but there's something extra you need, that 'eye for the visually pleasing' thats hard to cultivate unless user interfaces are what you do all the time.
I've used plenty of applications where the guy who wrote the backend code also coded the gui, and as a rule the gui is somewhat lacking. This is't just restricted to single coder projects, it also occurs when a project is full of able back end coders, and they build the gui to suit their own level of ability to use the code.
You can see this if you use Emacs. Nice though that software is in features, the interface is godawful, and actively prevents anyone new to computer usage or programing from using it.
"Simplicity and intuitiveness for the end user" (Score:5, Insightful)
"Simplicity and intuitiveness for the end user (both newbie and expert)"
Maybe this will be won by the most blinged-up interface but there's hope here that the competition organisers get some well thought out entries which help guide the users through the configuration of their routers.
Some installs are jargon heavy and just assume you know what all the options mean, little to no explanation or help. I've spent many hours sweating over some WRT GUIs that have (to me as a relative beginner) had meaningless options. I really really want to use these excellent installs but I get really put off by zero-to-poor documentation or explanations of what all the options are.
A simple interface with excellent documentation and guidance would be worth the prize.
Re:X-WRT? (Score:5, Insightful)
This isn't a theme competition, it's a user interface competition - usability counts much more than the style of the buttons.
Also, it's a genious move. When the clients are released, you'll have plenty to choose from. Also, being open source. you can merge the best bits of all the clients into one really good one.
Whoever came up with this idea should get a massive pay rise.
BTW, we could adopt the methodology in other areas too.
Re:Please.... (Score:4, Insightful)
In theory, yeah.
In practice, I just checked my mail on my phone, saw a slashdot reply notification, and clicked the link. Then stared at the screen in disbelief, as the phone showed me that to show me a message maybe 2KB in size it had to download 1MB worth of crap.
Reloading the page is something that takes a very small fraction of a second, when the server isn't doing any heavy lifting, and the page doesn't have half a megabyte of javascript.
And a much larger amount of devices can show that simple page. The router interface shouldn't be flashy. It should be neat and simple, something you could deal with from a cell phone if it was needed in an emergency. I can't tell you how much I hate the websites of various device manufacturers that require hunting the link for the drivers download in the source, because the only available browser I had was lynx, and the link is impossible to find in it.
IMO, don't bother with the flashy stuff. I'll be the one who is going to mess with the router settings, and I want it stable, functional, and usable on all devices. If it prefers form over function I'll go with another product.
Re:I have a suggestion ... (Score:4, Insightful)
I second the GP poster. Tomato rocks. Clean interface, lots of functionality, good documentation.
I've been many times at the OpenWrt. It sure looks like a full featured linux dist, but they sure forgot to put clear simple instructions to get my router running it. They seem to try to be so many things that they forgot to cater to, what seems to me to be, their most plentiful potential users: Linux users that would like to run OpenWrt in a router.
No, I don't want to edit /etc/network files, I want a simple GUI that does the job.
Re:needs an easy way to edit firewall rules (Score:2, Insightful)
Re:Please.... (Score:4, Insightful)
CGI doesn't scale well.
and since when does netmgt need to support LOTS of concurrent connections?
see, that's my point - unless you know the *problem space*, you are likely to over-engineer the solution.
most netmgt solutions are WAY over-engineered. they are not going to be hit by thousands of clients. likely they'll be polled frequently by a few NMS's and by frequently I mean a few packets of query every 15minutes, TOPS.
you do not need 'scaling' at the client. you just don't for this problem space.
Re:I have a suggestion ... (Score:3, Insightful)
I don't think anyone here has enough computers at home to need IPv6. I mean, if you have more than 253(*) computers in your house, you have more problems than needing IPv6.
(*) Number of possible nodes on a Class C subnet minus the router itself.
The only reason I could see myself having to move to IPV6 would be if my ISP forces me to.
Know what SOHO means? (Score:2, Insightful)
The majority of the posts seems to be sneering down in elitism to these poor folks that don't know how to setup a router in the CLI and, god forgive them, try to setup their home network by themselves, without paying a sysadmin to do the work.
Yes, a sysadmin that can't configure a router without a good GUI should be hanged by the neck until death, but last time I checked, the majority of the routers supported by OpenWRT were SOHO ones.
Do you really expect people that just want to setup a minimal network of maybe a printer, one desktop and two lappies to read RFC 1123?
One shouldn't need to. Yes, networking is interesting and useful, but not everyone wants to be an expert on it.
I have seen a physics PhD setting up a wireless net for his lab with WEP, because the list of protocols was sorted alphabetically and it came before WPA. Well, he should have told the lab's undergrad to do the job but, nevertheless, a good GUI would at least put an (deprecated) near the option.
He wasn't being dumb. Would be if he put a short common word as password. Even if he didn't knew about dictionary attacks, common sense would have told him that they're easy to guess. Actually he chose some interval of the digits of pi because it had high information entropy. But how could he know that the router would offer an insecure option to him or which 3-letter acronym was better?
Re:X-WRT? (Score:2, Insightful)
You are assuming that good ideas are harder than good integration.
More and more, I don't think so.
Naw... (Score:5, Insightful)
Maybe the task is needlessly complex? Why should somebody connecting to a access point need to configure which encryption protocol to use? Why should the damn WiFi card and the access point negotiate automagically and *pick the damn best one they both support*?
^ this space reserved for replies telling me that the protocol doesn't allow it. Well you know what? The user isn't broken, the protocol is. 802.11/whatever is horribly insecure because the protocol doesn't fucking handle the encryption for you. Let me pick a key on the router, type it into the laptop, and the damn things can pick WAP, WPA, WEP, WAZOO or whatever the fuck some encryption dudes dream up all on their own. As long as it works and is secure, I dont really care what the hell protocol is used.
Blame the user is arrogant, stupid and is increasingly a black mark on employment. The world expects the programmer and the designer to do everything for them. That is the market. Deal with it. Programmers who design usable stuff will find their skills highly desired. Programmers who say users are spoiled children will rapidly find nobody wants to hire them.
You know who to blame? Blame the designers and programmers for not understanding what the goals and tasks are. Blame the designers (or lack thereof) and programmers for not removing all the roadblocks in the way of accomplishing said goals and tasks.
Honestly, the real reason programmers get bent out of shape (*cough*aside from the weird vibes coming from those still in the computing stone age like RMS*cough*) is that programming a usable interface is very hard. Lots of edge cases and you can't catch all of them. Really, the shit is hard and no silver bullet will ever be found that makes the hardness go away.
Re:Naw... (Score:0, Insightful)
your sig says it all... it's chrome's fault that you can't spell and can't be bothered to learn how.
it's a protocol's fault that everything doesn't automagically connect you the the internets.
it's a programmer's fault that computer interfaces are too hard for you to learn.
i bet you're pissed off a lot. stupid car that won't drive itself, damned bed that doesn't make itself, fucking pr0n that won't jack you off...
Re:I have a suggestion ... (Score:1, Insightful)
I don't have enough computers at home to need ipv6.
Either you have an unusually small number of computers, or your ISP is providing an unusually large number of IPv4 addresses. If you don't have an IP address for every unit you want connected to the internet, then you have a problem. There exist workarounds that almost works (NAT/masquerading), but the only actual solution for that problem is IPv6.
Re:I have a suggestion ... (Score:4, Insightful)
Personally, I don't have enough computers at home to need ipv6.
It's not just about your computers at home. Your home is connected to the Internet. The Internet is growing at an exponential rate. It only has 4 billion addresses. Thus, IPv6 is useful if you plan to continue connecting to things as we approach the not-so-distant future.
Re:X-WRT? (Score:4, Insightful)
People who think "user interface" is synonymous with "themes, skins or 3D effects" are a large part of the reason so much software has a terrible user interface.