[ozgaming]

Isn't it nice that someone is offering a free service for the little guys out of the goodness of his heart. Just see it as that. And good luck to him and his project.

[Arrow768]

Looks like a very interesting project,

but how are you going to found it once it is getting more popular
(You might need a dedicated server with a lot of bandwidth, once it comes to 10Gb/s its getting very expensive)

There might be a storage problem too:
As a rough calculation: there are 27000 Maps(/Map packs) for CS:S on gamebanana lets assume they have a average size of 30 MB, then you need 810GB Space

Another Calculation:
Lets assume there are 1000 Servers using your Service.
300 of them are changing the map at the same time (to a 30mb map)
Each Server has 20 Users
You have to provide 180 GB within a minute (so its a fast dl) (300 ser * 20 us/ser * 30 mb/us
This would result in a bandwidth of 24 Gbit/s ((180GB / 60s)* 8 bit/byte)

Ok, this is a worst case scenario, but if your service is getting successful you might have to think about the founding.

I think this is also the reason why gamebanana doesn't provide a fastdl.

I think your project is a great idea, to help people for the first months.
Maybe you have to think about limiting the usage duration of your service (until they are successful enough to afford a own fastdl / have time to set one up)

As I said, your service is great for a new community, but as they are getting bigger, your service is going to become inflexible because they have to wait for the approval of their maps/materials/sounds/...

[Chefe]

Originally Posted by Arrow768 Looks like a very interesting project, but how are you going to found it once it is getting more popular (You might need a dedicated server with a lot of bandwidth, once it comes to 10Gb/s its getting very expensive) There might be a storage problem too: As a rough calculation: there are 27000 Maps(/Map packs) for CS:S on gamebanana lets assume they have a average size of 30 MB, then you need 810GB Space

I have actually 2 tb available for this project.

Another Calculation: Lets assume there are 1000 Servers using your Service. 300 of them are changing the map at the same time (to a 30mb map) Each Server has 20 Users You have to provide 180 GB within a minute (so its a fast dl) (300 ser * 20 us/ser * 30 mb/us This would result in a bandwidth of 24 Gbit/s ((180GB / 60s)* 8 bit/byte) Ok, this is a worst case scenario, but if your service is getting successful you might have to think about the founding.

That isn't a calculation. Why 180 gb / 60 sec? The reason to use a fastdl is not that everyone gets his map in a time windows of 1 minute. The reason to use a fastdl is to give the clients the possibility to use more bandwidth from their connection to download the map. And im confident that I can provide that. Sure, if the project gets such big, there are plans for bandwidths limit per connection/server. But not yet.

I think this is also the reason why gamebanana doesn't provide a fastdl.

I think that has many other reasons, including gamebananas everlasting storage problem.

I think your project is a great idea, to help people for the first months. Maybe you have to think about limiting the usage duration of your service (until they are successful enough to afford a own fastdl / have time to set one up) As I said, your service is great for a new community, but as they are getting bigger, your service is going to become inflexible because they have to wait for the approval of their maps/materials/sounds/...

Thanks. It isn't intended to replace all fastdl servers in the world ;) Its like you said: a possibility for young clans/group/community's to get there servers up with fastdl without pay for whats already there. That its maybe not such fast as a webspace you may host for one single or few servers is self-explaining.

[Ade]

dont think that high of a number of servers will change map at same time, and since it's zipped it shud take seconds to dl 1 map
the first step would be to have at least 2 servers, in US and EU so it is indeed 'fast' dl id rather pay to have fastdl on my side of the pond at least, than have free slow dl on the other, it breaks the purpose

[YamiKaitou]

Originally Posted by Ade the first step would be to have at least 2 servers, in US and EU so it is indeed 'fast' dl id rather pay to have fastdl on my side of the pond at least, than have free slow dl on the other, it breaks the purpose

Not really. FastDL is desgined to allow the clients to download over HTTP rather than in-game as Steam imposes caps for downloading content during the game. Downloading over HTTP will be faster than in-game regardless where the FastDL server is located.

[Powerlord]

Originally Posted by YamiKaitou Not really. FastDL is desgined to allow the clients to download over HTTP rather than in-game as Steam imposes caps for downloading content during the game. Downloading over HTTP will be faster than in-game regardless where the FastDL server is located.

Only if the server's sv_maxrate and client's rate is slower than the FastDL is.

[Pitbull3]

Time to use akamai when you become well known. Then lets hope you can shell out $2,500-$10,000/month for their CDN services.

[Chefe]

Originally Posted by Pitbull3 Time to use akamai when you become well known. Then lets hope you can shell out $2,500-$10,000/month for their CDN services.

Not really. Therefore I have many other analog structures I can use. But maybe someone who can put this so clumsy in here can shell that out?

[MPQC]

I love this idea, though I can't see it going very far unfortunately. Currently I don't see any way to raise money -> this will soon cost quite a lot if people decide to adopt it.

Best of luck with it though.

[eGOTheJudge]

Very nice.