This (see bottom entire messages copyed and pasted) was posted from burst.nets owner i guess in one of the isp lists i subscribe too? Im my opinion another bad business model, thats destined for failure. As you can read below it appears burst.net is attempting to offset its underselling of what servers and bandwidth cost them by trying to sell all their one way bandwidth to isps. While this is not a bad idea, their basing their entire business model of nocster.net on the idea that their going to get enough ISP's to pay for their 250mbps bandwidth soon to be 400 so they can basically sell their servers and service at a loss. This presents a series of challenges:

1. What if they dont find these ISP's, if they dont then it will be hard if not impossible to make a profit.

2. Even if they find these ISP's, what happens if they lose one of them, are they going to raise prices on their customers.

This business model seems even worse then rackshack.net at least with rackshack all their dialup customers are their own, their not betting on a dialup company to survive and staying with them.

When our company prices a product or service, we price it so that if the customer uses every piece of resource we still make money, and we dont rely on other companys to try to offset expenses. All of our bandwith is one way too, and weve been aproached by isp's wanting to buy bandwidth cheap one way, and we say no way Jose(corrected) even if we did sell isps cheap bandwidth we would continue to price solutions based up on if we did not have a isp or customer in place offsetting bandwidth costs.

Is the whole industry moving to the when, if, and but business model. What ever happen to the old buy the product at wholesale, mark it up a bit, and provide great service. I guess thats why weve been in business for over 9 years. I dont see that being the case for any company using the when, if, and but business model. Companys using this model seem to rely on luck and hope rather, then a concrete foundation.

This isnt to put down any company, just interested in getting feedback. I suspect a majority of you are consumers and would love 99.00 cpanel servers at nocster.net but put the price and offering aside, do you really see this as a feasible business model.

See below for the post from burst.net:

Subject: Looking to move one-directional bandwidth to the right
partner.
From: "BurstNET Technologies, Inc President - C.E.O."
<president@burst.net>
Date: Sat, 18 May 2002 18:04:20 -0400
X-Message-Number: 1


Greetings,

First off, I apologize in advance if some of you think this is spam,
such
is not my intention at all.
I am not trying to spam the list to make a sale, but rather find the
right
company who needs the opposite of what we do.
Now that that is done with, I'll get to the point :-)

Our company maintains a data center in Scranton PA USA.
We currently have over 250MBPS of bandwidth coming into the facility,
which
is soon to be over 400MBPS.
The bandwidth is carried over two fiber providers, Verizon and CTSI,
and
will soon have a third PPL Telcom.
The Tier 1 providers are Sprint, AT&T, Worldcom, and soon adding in
Qwest &
Cogent.

We offer primarily web hosting and managed servers, with some dedicated
access and wireless clientele.
Due to our service offering we use our bandwidth at about a 5:1 ratio
(upload/download).
That means 80% of our downstream/download bandwidth goes unused!
My research shows that most ISPs are the exact opposite, with users
downloading alot, but uploading very little.
This means that if we could find several ISPs to work with, we may have
a
great synergy here between our firms.
We could sell this bandwidth very inexpensively, as to us it is just
sitting here unused.
Your company could save a bundle on one of it's highest costs of doing
business - monthly bandwidth costs.

So, any company using bandwidth in the opposite direction our company
does,
and wants to dicuss this further, please contact me via email at
ceo@burst.net so that we may begin a dialog. I should think you would
want
your POP to be based on the East Coast of the US, to keep this
affordable
for any telco loops that would be involved. The Tri-State area would be
optimal, but if looking to purchase enough bandwidth, it may be
economical
as far as Maryland, Virginia, Ohio, etc....we would have to look into
the
loop prices to see.

I look forward to hearing from anyone that this could benefit...

Warm Regards,
S. Matthew Arcus
President/CEO
BurstNET Technologies, Inc.

Our business model is NOT based on selling opposite direction bandwidth to ISPs.
We just happen to have a ton of bandwidth just sitting here unused (result of 5:1 ratio), and figured why let it go to waste when we can sell it for exteremly low prices and gain additional revenue from it. Do we need the money? No. Would it be nice to get additional revenue from it? Of course. So if an ISP can save alot of money on getting bandwidth from us we don't use, it is a win-win situation for both companies. We already have several interested parties, including a regional telco provider that signed a contract to purchase 10MBPS from us.

Also, BurstNET has been getting into selling dedicated access lines and wireless internet in our regional market, to dump this leftover downstream bandwidth as well. We are also in the process of getting contracts for regional univertsities and tech centers to supply bandwidth to them. There is more going on at our company than just managed dedicated servers and web hosting :-)

Besides, we ran the numbers, and we could still make money selling NOCSTER servers on non-Cogent providers we currently have, at the same pricing level. Our NOCSTER product line is not a result of Cogent, and will continue to exist regardless of whether Cogent remains. We are just at the economy of scale point where it is not costing us as much to build, service, and serve each server as it did with just a couple hundred machines. We are now able to negotiate better terms, lower pricing for hardware, etc...

So to sum this all up, our NOCSTER plans:
1. are NOT based on Cogent's pricing
2. are NOT based on selling opposite direction bandwidth
3. and BurstNET is not going out of business anytime soon, nor accepting any buy-out offers WHATSOEVER.


Sean R.
BurstNET


PS - Our company was founded June 1 1991....long before 99% of the hosting companies out there. We will be celebrating our 11th anniversary come June 1 2002 - The NOCSTER release date.

Funny that you didn't mention the $99 managed servers in the bad business model. Hey if they can do it more power to them, but frankly there is a reason rackshack offers completly unmanaged servers for $99/month, and a reason people complain even about their unmanaged support. Thats their gig though, they're up front about it and thats what their customers know theyre going to get.

If you figure it out it looks like rackshack servers push around 50gb of bw a month on average so we'll use that number in my little theoretical hosting company setup here, nurst.bet, offering a $99 mangaed server.

income:
server life over 2 years, $2400

expenses:
cost of server - $1000
50gb bandwidth average for 2 years - $600 (50 cents a gig im being very generous)
1/100 of a 52k/year admin (assuming he can manage 100 servers 24/7) - $1040

That sticks you in the red over the course of two years by around $240 , even more in the red shorter term, without calculating in any of the expenses of running a datacenter like electricity, power, a/c, security etc. Keep in mind to cut costs were doing this under the assumption that 1 admin at 52k a year is going to provide halfway decent managed support to 100 servers 24/7/365.

I was always under the impression people went into and expanded their business to 'make' money not to lose it. Somewhere something is missing in this equation and I don't know quite where it is yet, but I imagine we'll find out soon enough from their customers when they start moving those boxes.

Hope I didnt 'burst' anyones bubble here...Maybe I'm wrong, if I am I'm gonna do exactly what they're doing myself and make millions, but I don't think I am ;)

-Brendan

Originally posted by cbaker17
...and we say no way hosay(bad spelling :))
Out of that entire post, I found that to be the best part. Anyone else find that funny and ironic? Hehe...

On topic: I don't think it's a "bad" business plan, but I don't think it's a wise one either. I agree with your "if our customers use up all their resources we still make money" bit, which is what we base our prices on. The development of Nocster will be an interesting one.

nurst.bet LOL!!!

Well maybe there is a bigger picture.. these situations arise in many industries, not just hosting/IT..

Simply put, if your company gets to a particular size, you can take advantage of 'economies of scale' and your business model becomes more refined [hopefully]

So theoretically, at a particular level you can buy your wholesale resources or 'parts' or whatever, at better prices and on better terms [things that may help the company in areas other than gross revenue, like taxation etc]

So let's say you have a deal that means that your company has a level of business which is still profitable, but because you recently made a huge purchase of your resources or parts [on very favourable terms and conditions], your margin slims down until you generate more of the business that you can now supply.... get that sales team motivated!!

As a consequence of this position.. you may find that you can increase the volume of business by offering things at or below their apparent cost, because you have a huge stock of resources/parts which might otherwise sit idle while your sales department pulls their finger out.. some money is better than no money or even temporary storage costs!!

The net effect is that the company's revenue increases, and the ACTUAL costs, when viewed from the whole company's perspective [bottom line], of things like $99 server deals is not what it would be if you were to take an isolated incidence like you did Brendan..

The flow on effects are also the new attention that's paid to the company which in turn should generate further sales, of non-$99 products.. so there is a marketing benefit also..

This is very thin and basic as a scenario, but is a very real one and common in industry generally..

The purpose of a company is to make profit and there are many ways to do that, and the more money you play with the more options you have for sophisticated tricks and turns..

Having said all that.. I agree with the basic model that plodding away and building a loyal client base over a longer period of time is a more attractive business plan for me.. but of course there are many, many ways to make it happen and just because it doesn't suit you or you don't understand it, doesn't mean it BAD..

If it works .. it's good.. and if it works for a long time it's excellent..

Cheers..

Perhaps burst you feel im attacking you, this is not the case at all, its just mind boggling, i mean believe being in the market for over 4 years and in business for over 9 years i know exactly what you pay for bandwidth. So this is why im perplexed...

If you can pull it off more power too you, but ive crunched the numbers and unless you playing numbers game, and the maybe, if, and but game its impossible, and thats even if i consider you have an arrangement with cpanel to get the license for completely free, which i doubt very much if thats the case. Perhaps you would like to lay the numbers out on the table that you yourself have crunched showing out you plan to make a profit based on you not getting a single customer to use your incoming bandwidth. And one other thing i never even brought cogent into the equation, you did. Thats a whole nother story i wont even go into.

How the heck is jose (corrected) spelled anyways??

I wish them luck.

However I only dream of the day that Cogent (if) goes under! I can't wait to see the shackdown. (And there will be a shack down.)

I see far too many people who run their colo business like many RS reseller host run their hosting company. They sell at a lost longterm. We all know you can't make money selling $25.00/yr accounts. Even at RS. We now see the same principals applied to colo or dedicated. RS hasn't proven it can work. Sorry if you think a year and a half proves success, but I do not. Look at all the giants that are now dwarfs. It takes longer then a year and a half to come out in the wash.

I think the loss of Cogent will seperate the men from the boys right now (substitute "women from the girls" if you are offended). I know for a fact many who push their services here, have NO back up to a loss of Cogent.

OK, I'm done. :) . Good luck Sean. Glad to see you back here.

Felix your right but your basing your assumption on a stable business model of buying in volume, getting a good price, marketing it up a little and reselling, as was presented this is not the business model burst is using. Even if they were using strictly cogent (which their not) they would pay 30.00/meg then the lowest ive heard of cpanel is 25.00/license, that is 55.00 then a comparable server lease is going to run at least 50.00 and thats for a low end server, so at this point their losing money on every server sold. Now even this could be considered profitable in volume, becauseyou can assume not every customers going to us all their allotted bandwidth, but then you figure in the fact they also use tier 1 providers which cost say 100-300.00/mbps then your whole model regardless of what your getting servers for, regardless of what your getting licenses for, and regardless of how much the customer uses is thrown to the wayside.

Originally posted by cbaker17

How the heck is hosay spelled anyways??

Sorry for the off-topic, I think that you mean Jose.

Charles,

I know you were not attacking us :-) Just trying to respond to some of your thoughts with some explanation of how we can make a profit doing this....


Warm Regards,
Sean R.
BurstNET

THANK YOU i shall correct it, was bugging me :)

Originally posted by Jedito


Sorry for the off-topic, I think that you mean Jose.

allera, thanks to this, I finally understand what you mean. Holy crap, it took me a few minutes to stop laughing after catching this. :D

That was the funniest thing I have read in a while. Then he posted "How the heck is hosay spelled anyways??" but edited it to what it is now.

Sorry, Charles, that was just too funny. Thanks man. :)

I would love to see Cogent go out of business as well...that would put alot of our competitors out of business...and get market prices back to where they were 2-3 years ago...when our normal product line was considered inexpensive. Can you imagine the fallout in the industry from that? ...or atleast see all other backbones get down well below $100/MBPS to compete with Cogent...but that is just not gonna happen.


<< Good luck Sean. Glad to see you back here. >>

I never left, I just stopped posting...and started behaving :-)
Nah, we have another guy that handles such things now, who doesn't take things as personally as I do ...

Seems to be working out...

Not many complaints about our support service lately...just when we have network problems....part of the business.

Sean R.
BurstNET

Yeppers you know us kansan foke dont know how to spell, they dont teach us them things on the farms....

Originally posted by cbaker17
Yeppers you know us kansan foke dont know how to spell, they dont teach us them things on the farms....

That explains why the little farm I bought promptly turned into a desert.. I thought it was because I had no clue about farming.. when all along it was because I could spell correctly.. <?!>

:bawling:

Hey felix your in luck you dont even have to farm anymore, they pay farmers now to bale up some hay and stick the bales every couple of yards on their land and you get paid every year for doing nothing but having bales of hay or what ever it is sit on some land. Some grassland conservation thingy or something. Sounds like a sweet gig to me. Wonder if i could pull that off in my front yard :) neighbors might not like it much....

You people have been preaching about Cogent going under for so long now.. give it a rest. Just reading the old posts on this board, tons of people saying it would go under in 6 months or less. (About a year ago or more.)

I don't like Cogent, although I do think it's good for audiostreaming.... as we do use it for that.

Cogent has nothing to do with this thread, for some reason burst.net brought it up but it has nothing to do with what were discussing. I started this thread and purposely did not mention it because i don't care to have another cogent thread where people who don't like cogent and people who do bitch about it. This is a thread regarding business models, farm land, and spelling.

PLEASE keep it on track.

Originally posted by cbaker17
Hey felix your in luck you dont even have to farm anymore, they pay farmers now to bale up some hay and stick the bales every couple of yards on their land and you get paid every year for doing nothing but having bales of hay or what ever it is sit on some land. Some grassland conservation thingy or something. Sounds like a sweet gig to me. Wonder if i could pull that off in my front yard :) neighbors might not like it much....


Well that sounds like 'another' good business model.. LOL <scratches head>

Look in the past posts. Cogent was mentioned.

Selective memory is awesome!

Again jayjay it was mentioned, but i didnt bring it up burst did, might want to get that memory checked out :) hehe

Oh and since i started the thread i get to pick what we talk about and its not cogent :)

har har har

Aren't you the funny one? : )

Ok, you win. *hands you a muffin*

He brought it up.. you didn't.. Let the frost elves take over the planet! :(

Is there a website for these so called frost elves, we will pretend like they work on a farm, so the thread is still on topic :)

<< income: server life over 2 years, $2400 >>

When you add in setup fees, extra services, figure 20% of clients will not order the cheapest server...you are looking at more of a $3000/2yr revenue

<< expenses: cost of server - $1000 >>

Not even close...Our setup fees are covering all costs of server builds! And yes we are using name brand parts...Intel, Asus, Maxtor, IBM, Seagate, AMD, Adaptec, etc...
We are buying in bulk from the right places at the right price....and building ourselves, which saves alot of money.


<< 50gb bandwidth average for 2 years - $600 (50 cents a gig im being very generous) >>

50GB a month is a fair assumtion on average
usage.

We are not paying 50 cents a GB though...not even close for our combined NOCSTER backbones. Cogent does have something to do with that though. Remember, we have OC3 lines now, and our own loop down to 401 N Broad in Philly. We have enough providers now, that we have the leverage to pit one against another to get them to offer the least expensive price to obtain our business. You would be surprised how negotiable some bacbones are in order to get your business, and keep it away from one of their competitors.

<< 1/100 of a 52k/year admin (assuming he can manage 100 servers 24/7) - $1040 >>

If you pay $52k a year, then you are paying your admins waaaay too much. You have to hire them young, train them yourselves, move them in from desolate or pink-slipped parts of the country, and treat them well. Being in a very low cost of living city, Scranton PA, helps salary levels alot. Many 18-24 year olds are happy to get a chance to work in a dot-com company...of which there are not many left still hiring. Right now we have about a 1/50 ratio on admins to servers. That will get raised a bit because NOCSTER will not have phone support, which takes up alot of an admins time for regular BURSTNET service. We also differentiate between admins, level 1 & 2 support techs, assemblers, and sales reps...of which I am not couting in that ratio.

<< without calculating in any of the expenses of running a datacenter like electricity, power, a/c, security etc. >>

Already have the generator, ups, space, a/c, security, router, etc....all paid for and in place...and with a ton of capacity untouched...only costs about $4 per month average for power to a server.


We have worked the numbers, and NOCSTER is profitable. Our company has been profitable for almost 11 years, and has never filed a LOSS.
NOCSTER will survive, and it will survive while regular BurstNET customers do not get any service degradation. That was the main thing we were concerned about. We spent the last 6 months straightening out our support dept, and we plan to keep it that way. You'll notice the fact that there has been very little support complaints about BurstNET in the past 6 months...even though we have about 5x as many servers now then we did back then! Much of this is due to our new Customer Service Manager, Matthew McCormick, whom we hired from Dialtone a few months back...who I must say is doing a remarkable job for us. With our new 12000 series router coming in this late week, and with the new OC3 capacity, our network problems of past 2-3 weeks will clear, and we are ready to roll just in time for the NOCSTER release date.


Sean R.
BurstNET

http://bassboy.net/frostelves.jpg

They are scary.

and if that's not "farm" enough...

http://bassboy.net/snakes.jpg

Originally posted by BurstNET
You'll notice the fact that there has been very little support complaints about BurstNET in the past 6 months...even though we have about 5x as many servers now then we did back then!

Good job on that - I have noticed that... we've been working on the same thing right now ourselves.

Originally posted by UmBillyCord
allera, thanks to this, I finally understand what you mean. Holy crap, it took me a few minutes to stop laughing after catching this.

Hehe, yea, he misspells every other word in that post and only questions one. I guess if he would have correctly spelled Jose it would have been more ironic.

By the way, I can spell pretty well and I can't grow a weed if I tried. Thanks to this thread, I now know why. :)

I suffer from ITFTIT (I Type Faster Then I Think). Its a very sad disorder, although ive started running my paragraphs through microsoft word :) hehe, so hopefully some improvement will be noticed...

So burst according to your own figures your making 2400.00 per server but to provide service for that server your spending:

1000.00 for a server, yes i know you say it doesn't cost you that much but for 5 years of my life i built computers in mass quantity so i know alot about hardware and alot about quantity purchases, so you don't look so bad we shall say 1000.00 because even if you are buying in quantity your getting crappy chassis and ram if your not spending close to that.

600.00 for bandwidth

1040.00 for a admin

4.00 for power, i gave you the benefit of the doubt on this one but if your running a carrier grade data center i can guarantee it costs you much more then this.

Your total costs for this server are: 2644.00 and your income is 2400.00 YOUR STILL making a loss on every server you sell.

Now how about we go one step further and i hook you up, we take off a couple of hundred for a cheap low end server you build in house, that brings your expenses down to say oh 2400.00 even, HEY great your breaking even now. Thats what every ones in the business of doing.

But hey ill even give you one more, at 2400.00 (break even) you can still make money because you can oversell (not every ones going to use all their bandwidth) and like you said some people will buy a bit more expensive server. BUT your still screwed you forgot to factor in:

1. Sales people, someone has to sell the servers.
2. Billing, if you do the volume rackshack does or even close your going to need more billing personnel.
3. Same goes for customer service.
4. In 2 years your going to have some servers go bad, a expense that will come out of your pocket.
5. Your going to need new equipment, new racks, your buying a new router, all other expenses out of your pocket.
6. You'll have to hire installers to build the boxes, (you gonna have one guy support, and build them) i hope not.
7. You forgot all about cpanel, you cant tell me your getting these for free. This alone is a large chunk out of your monthly revenue.
8. Your going to have to hire additional level 1 and 2 techs to handle support for these boxes, again i don't think your going to have one guy supporting all of them, building them, etc
9. And heres the kicker that screw up your whole formula, you based your figures on a 2 year life span of the server, what you forget is that in one year you'll have new and better servers for the same price. Your customers will cancel their current server and order a brand new one for the same price, at that point you'll either have to scrap their hardware or resell it as a used older model for less then what you have listed on the web site today.

I can go on and on, even if you get personnel for cheap, either way you look at it your prob loosing money on the whole deal in the end, or breaking even. Even if you were making a little money. I don't see how so much work and effort is worth a few dollars.

sorry to branch out again, but, spelling was mentioned.


http://www.iespell.com/download.htm

is an interesting tool to use, especially for the folk posting alot on forums.

I am not going into details to the penny about our NOCSTER business model, just suffice to to say you're off by about $1000 per server Charles. You didn't fully listen to everything I said in my last posts regarding operating costs.

NOCSTER is just added sales over what we actually need to operate.


Sean R.
BurstNET

Originally posted by BurstNET
I am not going into details to the penny about our NOCSTER business model, just suffice to to say you're off by about $1000 per server Charles. You didn't fully listen to everything I said in my last posts regarding operating costs.

NOCSTER is just added sales over what we actually need to operate.

Sean R.
BurstNET

Sean,

I'm sure if you guys are building these boxes in house, your covering the whole cost with the setup fee. Prices from Pricewatch:

Intel Celeron™ 1Ghz ($71)
MSI M6368 Socket 370 MATX Motherboard ($67)
512MB PC133/DDR SDRAM ($59)
60GB IDE Hard Drive ($75)
ATX Case w/Power ($30)
-------------------------------------
Total: $302

Factor in bulk buying power and I'm sure you'll save the few $$ you need to break even with the $299 setup fee - at least for the hardware...

Take care,

Brian

Sean,

Here's an important question regarding Nocster.net. I have a burst.net server now. Can I transition it to a Nocster.net server? I really don't need the managed support anyways. I'd rather have the lower price and more bandwidth.

Is this something your going to allow? I hope so. :-)

Thanks - Brian

Originally posted by bteeter
Sean,

Here's an important question regarding Nocster.net. I have a burst.net server now. Can I transition it to a Nocster.net server? I really don't need the managed support anyways. I'd rather have the lower price and more bandwidth.

Is this something your going to allow? I hope so. :-)

Thanks - Brian

Check out their FAQ, they've got the question covered :D
http://www.nocster.com/faq.shtml#42

Try buying all that on pricewatch see how much it really comes to after shipping and do you really want your server running on a msi motherboard with non ecc memory, and a cheap atx case with a cheap power supply, and a 5400 rpm drive, i guess its fine if what your doing isnt mission critical.

The setup fee will def. help pay for the case, if they didnt charge a setup fee i dont ever see them even breaking even.

Brian,

You're right on target :-)
Factor in wholesale pricing, and bulk discounts, and it becomes economical.

And let me stress once again, we are using NAME BRAND PARTS:

Asus Motherboards
Intel & AMD CPU
Samsung, Micron, NEC, Crucial, Toshiba, RAM
Maxtor & Seagate Hard Drives
Adaptec SCSI Controllers
IBM, HP, 3Com switches
etc....


Sean R.
BurstNET

No matter how you look at it you cant get servers for 309 with those specs

Originally posted by JG


Check out their FAQ, they've got the question covered :D
http://www.nocster.com/faq.shtml#42

Hmmm... So I'd have to order a Nocster server, pay another high setup fee, then move my clients. Well, I think I'll stick with Plan A, which was to move them to servers in other datacenters...

Thanks for the help,

Brian

You're right on target :-) Scary, a server < $350.00. I am sure this is why RS moved to Compaq servers.

Originally posted by RackMy.com
Scary, a server < $350.00. I am sure this is why RS moved to Compaq servers.

That, and I'm sure that Compaq (or is it HP now) gave them a hell of a good deal on their servers.

In fact, I'd wouldn't be surprised if they are leasing them. Cheaper pricing that way, and they don't have to worry about what to do with old machines three years from now...

Hardware obsolesence is a problem Rackshack will encounter in another year or so when all of the Cobalt boxes they have are approaching or more than 3 years old. Who will want to host on a 300 mhz cobalt when the "low end" server at that time will be a 2ghz box with way more memory and hard disk space? And then another year after that with 1ghz duron boxes.

It seems to me that any server more than three years old is going to look pathetic when compared to current offerings...

Take care,

Brian

Originally posted by bteeter
MSI M6368 Socket 370 MATX Motherboard ($67)
60GB IDE Hard Drive ($75)
ATX Case w/Power ($30)

*shudder*
Nothing against cheap hardware, I have a PC with the same specs, but the PC is used only as a surf station - I never thought about such a garbage PC as a server...

That, and I'm sure that Compaq (or is it HP now) gave them a hell of a good deal on their servers. Well, I am sure they did not get them for less that $1000.00.

Originally posted by Walter


*shudder*
Nothing against cheap hardware, I have a PC with the same specs, but the PC is used only as a surf station - I never thought about such a garbage PC as a server...

So do I, well, similar specs anyways. (My home workstation is an eMachines T1100, 1ghz Celeron, 320MB RAM, 20GB HD, GeForce 4 MX 420.) Its fine for working from home and games, but the hardware is ultra-cheap, I wouldn't want to keep it under load for long. That chinsy 200 Watt (I think) power supply would probably fry. :-)

But, that isn't to say that a 1ghz Celeron, 60GB drive and a decent MB wouldn't be a bad combination for a small server. I don't think I'd use the MSI board I quoted, but a lot of pretty decent Celeron boards can be had for only about $20 more.

So really, if you added another $20 for a good MB, another $30 for a nice case & power supply, another $20 for extra fans, another $20 for name brand RAM and a few extra bucks for a cheap Video Card, your looking at a decent web server built all with name brand parts.

Sure its not SCSI, sure its not a Pentium 3, and sure, its not a multi processor box. But, it'll do a pretty decent job for all but the most demanding tasks.

Take care,

Brian