Slashdot Mirror
Wikipedia Needs $20K
TaranRampersad writes "Wikipedia's server is crashing off and on, and Jimmy Wales has posted a letter requesting some assistance from anyone out there with a dollar burning a hole in their pocket. Let's face it, you really don't need that candybar anyway ..."
Wikipedia is trying to offer information to help the community. I can't understand why the slashdot community doesn't want to help out a dying webserver, but wants to buy air bazookas over at thinkgeek.
Like Sweepstakes? Try out my service @ http://www.yourpowersweeps.com -- Free 21 day trial, no cc needed.
Oh, please, don't be such an asshole.
This money is going to a FREE project that anyone can contribute to. It's not going to a site with pop-ups and banner ads. It's a non-profit (as far as I know) resource for everyone.
It's only fair to pass the hat around. This isn't some company's or kid's project to fill their own pockets.
This isn't just "someone"'s website, it's "everyone's" resource. That's part of the whole wiki philosophy, isn't it?
---
Never criticize religion on Slashdot. You will be modded down for "Troll" no matter how factual it is.
You know, they have three servers. Two are cheap-failing-hardware-nuked, and then's the one always pulling off the job. IIRC the webserver still serving, never has had a faliure like the other two... And btw, give me a *star* for donating. :)
Giving money to wikipedia is, IMHO, more useful and a much better idea than giving 4000$ to the first guy that will port Mozilla on the Amiga platform.
But hey it's my own opinion mod me down if offtopic but no flame please.
Iraq: war to save the U
Wikipedia isn't just some other site begging for money, and they aren't asking for money for their content (though it's worth something, certainly, it's free to all - and Free too, I think) - their load is so huge, they really need thousands of dollars for their servers. I'd rather give them my 10 bucks than deal with the unpleasant alternatives, like ads plastered everywhere, or seeing wikipedia go away.
Maybe instead of giving cash, donate some of your old SUNs, SGIs etc, and help building it on a distributed architecture with really deep redundancy, where each component by itself isn't very reliable, but all together form a really strong cluster?
Asking for money is always the easiest way, and because of the number of people asking (Just look at all those PayPal Donate banners!) the chance of success is nearly null. What about taking a different path?
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
It's always hard to just request donations for a dollar sum as an open source project. I think they'd have much better luck requesting hosting/hardware donations. It's much easier for a corporation to donate hardware (they get to write off their cost retail even though the actual cost to them is far below that) than money.
When its just hosting needs, being able to massively farm out helps to. A lot of university groups look to help host a few different things. One group may not be able to satisfy all the needs but ten groups might.
Just my two cents...
int func(int a);
func((b += 3, b));
Why are there so many flames in this thread? Slashdotters are really showing a lot of hypocrisy here: we want everything to be open source and free, but when an open source project asks for a little help, we turn our backs on them. I'm sure we wouldn't see the same kind of comments if Linus Torvalds was asking for help.
Thanks for bringing up the $20k issue. I was wondering about the price as well, but then figured out that it's just a made up number. If he asks for $20k maybe he'll get $1k. But the idea bothers me.
servers are cheap these days. really. I've found p3-666 machines in the trash a few years back, and other people are finding nice rack mount servers with drives,etc. I can't afford much more than my rent, and yet I can come up with more server power when I need it, just by using a bunch of old P300s or whatever.
The Wiki things are cool in a way, but too filled with unqualified opinion.
And Britannica isn't? Yes, everyone says "You can't trust the Web because everyone has an agenda to push", but that's true for *all media*. In fact, the more a source claims to be "balanced" the more likely it is biased. Rather than trying to find an unbiased source, learn to read between the lines and figure out what the biases are.
"Good to have" is an understatement. Wikipedia as a resource gains a lot of its value by being always available, so that last 5% of uptime is worth $20k in donations IMHO.
I can't agree enough with the poster above.
There is an enormous amount of negativity that I have seen thrtown around in this thread.
But to my mind, Wikipedia is one of the gemlike projects out there that has an enormous amount of unadulterated MERIT.
Many of the posts decrying the cry for funds fall into two camps:
1) What the hell, $20k for your website infrastructure? Plan better, you assholes!
or
2) Wikipedia is useless/not worth it.
Many of the posts SOUND like 1), but are driven by a strong desire to demonstrate 2) -- for instance, the large number of posts that are claiming that "Wikipedia has become too political".
I don't think that people realize what the real issue here is. The issue is nothing less than total freedom of information.
Articles on wiki are moderated by public opinion -- and while this has a moderate negative influence in HIGHLY CHARGED, HIGHLY CURRENT topics -- political ones, particularly -- the bottom line is that wikipedia provides an incredible way for the truth to be heard and recorded. Everyone can contribute to this record of defined "truth", and if a revelation is made, it can be judged on its merits by millions of people.
Essentially, in this age of enormous uncertainty, slanted polls, (corrupt?) (liberal? conservative?) "corporate media", in this age where the visible "barometers" of world opinion (polls/interviews/random tests/scientific research) -- the informational underpinnings of representative democracy! -- may be subject to large-scale manipulation, and freedom of information is being decided for years to come, Wikipedia provides, if not "absolute truth", a body of information that has been thoroughly bathed in the democratic process. It may not be as white and pure as if it were written by the existing information aristocracy/meritocracy, but it is most assuredly free to all, and as unbiased as that process can make it.
Support of Wikipedia is, in a sense, support of the principles of democracy/communism itself -- support of the idea that fairness is most reliably and safely accomplished by even "unqualified" consensus. It's everyone's information. This just makes the process transparent, and rips the lid off of "true" and "false", right and wrong, belief and disbelief, and transfers the power to the people.
Go Wikipedia! If there ever was a project with real, LONG-TERM value . . .
Yes, we're about even in traffic with slashdot these days.
Chu vi parolas Vikipedion?
Thanks for bringing up the $20k issue. I was wondering about the price as well, but then figured out that it's just a made up number. If he asks for $20k maybe he'll get $1k. But the idea bothers me.
servers are cheap these days. really. I've found p3-666 machines in the trash a few years back, and other people are finding nice rack mount servers with drives,etc. I can't afford much more than my rent, and yet I can come up with more server power when I need it, just by using a bunch of old P300s or whatever.
One word: reliability.
Sure, any geek can make a computer out of toothpicks and bubble gum and run Linux on it and call it a "server", but these guys are trying to _reduce_ the amount of downtime they're seeing on some high-load systems. So it looks to me like they're trying to buy reliable hardware: new systems, lots of redundancy, and none of this "buying off eBay" or "systems assembled from parts pulled from the trask" junk which some people are suggesting. Real server hardware costs real money, presumably at least a few thousand per system.
Given, I'm talking about a _small_ commercial service company, but $20K USD is not a huge amount of money for us to look at for hardware. Running big websites (and especially other app servers aswell) can require very high-quality hardware (read "expensive") and (potentially) licensing to do it reliably.
Seriously, these guys (I've never used the Wikipedia) could easily spend $20K on hardware and not have that great of a system...but they claim that it will fix their problems...if they provide a good FREE service (no-one seems to disagree with that,) then I trust their judgement. I think I'll go donate $10 just cause they're trying to do something good for the "Internet Community".
[BTW - I'm a partner in an Internet based business that actually makes a profit...$20K for hardware doesn't sound like much to ask. We've looked at single units - not complete systems mind you - that cost more than a quarter million USD.]
Kendell
Second, you probably have no idea how big wikipedia is. Others have posted about the traffic, so I won't go into that. Its not a single sever that's needed. We need a load balancer and several servers.
Third, a wiki is a lot harder to run than simple static pages (in terms of CPU and disk).
If you don't believe the $20000 amount, you're welcome to join the mailing lists and find out for yourself.
Always, always, always run some kind of burn-in suite on new hardware before putting it into production. Even if your manufacturer does stress testing, your server was likely handed over to a shipping company that mistreated it (guaranteed). (VA-)CTCS is a good burn-in suite. If a machine survives one week of CTCS, in our environment it means that it will be worry free for at least 18 months (so far)--not counting inevitabilities like a hard disk mechanical failures (sigh).
We installed seven new servers at a colo in order to migrate a growing web site off of a shared server. CTCS discovered bad RAM in what was to be our database server, a faulty storage controller(!) in our file server, and a bad disk in one of the web servers.
None of these issues were apparant from the get-go. Most of the servers revealed problems within the first 36 hours of CTCS burn-in, with one holding out for 47 hours. If we hadn't run CTCS these issues would no doubt be hounding us for months.
So, two rules of thumb:
If possible, have your server built by a local vendor. The ability to walk into an office and scream at someone is a powerful resource, and you can completely bypass abrasive shipping companies by delivering the server to a colo yourself.
In any event, get your vendor to run CTCS before shipping and run CTCS again once it's on the rack. 72 hours minimum on both ends.
Don't learn these lessons the hard way. The extra time you spend up front can literally save you months of headaches in late night colo phone calls and other consequences of unplanned downtime.
I would like to donate, but I do not donate to causes that try to hide what they will do with the money. And it looks like that is what is being done here.
Yet you pay taxes!
At least with Wikipedia you can have a direct impact. So if you can't contribute money, then contribute some knowledge. Write an article, show the world what you know.
Times are tight, but $1.00 from each Slashdotter would do it.
and raise you one... I donated too.
C'mon ya geeks... everyone knows and has probably read Wikipedia; stick a crowbar in your wallet and cough up a few bucks. Yes, I know it's the holidays and everyone's tapped out, but really... who doesn't have a paypal account with a couple of extra bucks sitting in it? (if you're genuinely broke, relax... I'm not talking to you)
I see people in my medical practice all the time who tell me how they can't afford their antibiotics (even the cheap generic) or other medicine, yet they smoke two packs a day, have a cell phone AND pager, and manage to find beer-drinking money every week. Not to put too fine a point on it, but it often comes down to priorities.
If you've got the cash, why not part with a few bones? I'm sure Wiki would appreciate it if the community would ante up.
Even if a man chops off your hand with a sword, you still have two nice, sharp bones to stick in his eyes.
If you don't want to donate, please don't feel any pressure to do so.
If you would like to help, please feel free to roll up your sleeves and get involved like any other volunteer and make some solid, experience-based recommendations.
Chu vi parolas Vikipedion?
Why don't you email the marketing and Unix/Linux departments of IBM, Sun, and HP?
.com, Sun Microsystems", etc.
/. effect fine if they have an expensive but solid switch. Maybe they might be nice and throw one in as well.
Tell them about Wikopedia ( which they may already know) and mention how many users and hits you get a month and that you need new hardware.
Its likely they will make a deal with you if you agree to advertise for them or put a "Powered by Power5 AIX" or "Powered by the . in
IBM wants to bring Linux to the power4 and power5 servers and is releasing a new blade powered by them that runs on Linux. X86 stuff is garbage. Things like guinine risc and backplaned motherboards like those in Sun and IBM do wonders where pc's fail. Running your site is what the hardware is tailored to be doing.
Still even if you can get a free 2-4 smp x86 Xeon system, take it! A switch sounds like it may need to be upgraded. They cost big bucks though but many limited servers handle the
Yahoo has the powered by HP logo for Yahoo.com and its quite normal.
IBM would be my first pick and would gladly gloat about how much load their Linux based blades can carry. Your site is a perfect example. Reason being is that many IT managers view Linux on anything non intel as garbage. IBM also has big pockets and your server room is pocket change to them.
http://saveie6.com/