Wow, got a bit of an anger issue with academia do ya? Give the guy a break, he was given a task and he turned to a self professed community of ubergeeks to see if they had an answer.
Reading through some of the early responses it looks like they gave him at least one path to follow with moodle which looks like a pretty nice solution. I do feel that he threw out a bit of unintentional flame bait when he stated that they wanted a tool that would allow non-programmers to program when what it seems like he meant was he wanted a system where the users could add their own content without involving programmers.
For an alternative solution he might also want to take a look at http://ocw.mit.edu/index.html which is a solution heavily grounded in a commercial product by ab initio, after all, if it's good enough for MIT it might just fit his needs for flexability and scalability. Not only that, but I believe (and I could very well be mistaken) that OCW itself is available in an open source license format.
Good for SOE, it takes a lot to swallow your pride and admit that your game is not up to snuff. Then to suck it up and put the time, effort and money into improving it shows some backbone and character. I haven't played galaxies for about six months now, but I may very well give it another shot.
When I wrote the original post I was unaware that L3 was blocking all Cogent originated traffic. This is entirely another issue, although related, from the original post. Yes, Cogent does have to pay someone transit, but, under normal circumstances, loosing a bgp session would not cause you to lose all connection to an IP space once the route was adjusted to transit another network in order to gain access to, in this case L3's network.
Please be clear on that fact that if L3 cuts their peering point with Cogent it does not stop anyone on either network from reaching the other as the article suggests. Peering simply gives each providers clients shortcuts to the others IP space. All tier one providers have multiple peering points because the whole point of routing a packet that is not destined for your network is to get that packet to the edge and off your network.
To try for some clarity, just as an example: L3 peers with Cogent, XO, and 4 other providers. Cogent peers with L3, Internap, and 6 other providers. So, if L3 turns off their peering to Cogent, packets that originate from Cogent that are destined for the L3 IP space will have to transit another common peer before they get to L3's network.
This is like attaching a cellphone to your asterisk. It allows you to make calls across the GSM network and to receive calls to the number programmed on the SIM card and pass them to the asterisk box.
There have been some attempts to do what the parent is asking about, but I do not know of any that have been rolled out for public consumption.
Well I guess the interface may change a bit, for example adding a name to my address book:
Phone# configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
Phone(config)# alias exec callbob dial phone.device.213.555.1212
Phone(config)# end
Phone# callbob?
*callbob="dial phone.device.213.555.1212"
It's not as though the telco's haven't been subtly anticompetitive up to this point. I work for a regional DSL provider who leases Bell lines (specifically BellSouth). There is a never ending stream of problems with the bell systems. For example, when we switch a customer it can take up to two weeks to take effect, but when a customer switches back to Bell, very often, our DSL service is cut within 24 hours of their order being placed even though bell service will not be in place for a few more days.
Do we think this will pass? Absolutely. That's wyh we are taking this opportunity to look at other technologies like wireless, so really, in the end this is probably a good thing for competition.
Happy to. Because unless you are a local exchange carrier (a certified phone company) you do not have access to the 911 database. Most 911 centers are staffed by the local municipality, but the architecture is provided by the telephone company. We are going through this right now. We have spoken with the local 911 center and they were happy to let us insert data into the database but the regional bell says there is no mechanism for us to do that. So here I am stuck in the middle, wanting to comply but having no way to do so.
Actually I believe this has already been asked and sufficiently answered. There have been no issues with the United States stewardship of the root servers on a technical level.
In this day and age these servers are truly globally important. Unless there is a problem with their management or function I can think of nothing more disruptive to the internet globally than relegating their administration to a global committee.
Actually, MPLS would fit the bill here quite nicely. It allows you to specify ToS in an external label so that the routers do not need to peek at content to make routing deciscions. This makes for very fast routing deciscions. Keep in mind that you will have to support QoS edge to edge and once a packet hits an egress point to the public internet, boom, the QoS and ToS bits are stripped away.
You might want to look at products from Riverstone, Alcatel, Seimens, and last but certainly not least expensive, Cisco.
Well, you are certainly entitled to your opinion. If only it were as easy as walking in to the server room and sacrificing hundred dollar bills then yes, the problems would have been gone by now.
In the grand scheme of things five months is a relatively short period of time to work the bugs out of a large system like the WoW server farm.
Most of the servers that have been affected by the worst cases of downtime are the highest pop servers and they do need to resolve those problems and satisfy those customers even if that means closing those servers for new characters.
I don't know what industry you are in, but I am in the IT world and every time you have to patch a system, no matter how many times you have done it on the test bench, you hold your breath and cross your fingers. Multiply that by the number of inter-related systems that Blizzard is running and you begin to see that it is actually pretty amazing how quickly they are able to rectify those issues without having to resort to a rollback.
As to the end game, man, just chill for a sec. Did you even stop to look at the in game world while you were rushing headlong to 60? They are working on the content as is plainly evident by a cruise over to http://www.worldofwarcraft.com/. In fact, the patch yesterday introduced some of that content.
Keep in mind that WoW is the new kid on the block. What has happened to them is virtually the same as what we see every day here. They were/. According to every interview I read they were not expecting anything like the response they recieved. That kind of slammage is gonna requre a serious rescaling of the backend which will not happen overnight. Imagine if you had 1.5 million users banging away on your project, no matter how careful you were or how much you tested, things happen in the real world that don't happen in the dev sandbox.
As far as endgame content is concerned, again, it's a new game. To compare it to games like EQ with its X number of expansions is insane. Give them time to stabalize their platform and I am sure you will see more content.
Well, as a local market VoIP provider, we offer 911 service. We work with the 911 authority to update their database when we assign a client a number. Of course, the problem here is if the client moves locations and doesn't inform us. The lack of an elegant 911 system is definitely the biggest stumbling block to wide spread VoIP deployment IMHO
The reason this is a big deal (and in the interest of full disclosure I work for a VoIP provider) is that it is one less hurdle. Sure, you can get phone numbers from Vonage and the like now, but for them to get those numbers they have to go to the LEC in that NPANXX area and negotiate with them. In some areas the LEC just tells them to kiss off, we are not going to sell you numbers (this is the case in most rural telecom areas). This ruling will allow providers to go directly to the NANPA directly for numbers.
You are ignoring the fact that the major vendors on top of having to pay for advertising, sales weenies, and a fancy database also swing the stick of major purchasing power themselves. You do not become, and maintain, your position as a major server vendor by turning out junk. Yes, there are exceptions (the IBM deathstar incedent still makes me wince) but on the whole, I would trust the power supply speced by a group of highly trained engineers over the one I would pick out myself.
Well, here goes, (flame on everyone). I work in a hosting environment where the demands on CPU are low and storage is high. I don't need the whizziest new processor in my NOC to do a great job for my clients. I stick with one platform (Dell Poweredges in my case) and I began purchasing on ebay. I now have a very good relationship with one vendor, I pay about one quarter of what a new server would cost, I get a 30 day warranty from the vendor (and we torture test the machines for about a week upon receipt) and I have a NOC full of quad processor, Raid5, redundant servers that I am very happy with and which do the job quite well. Not only that, but replacemnt parts are a snap to buy direct from Dell.
Slashdot Mirror
Wow, got a bit of an anger issue with academia do ya? Give the guy a break, he was given a task and he turned to a self professed community of ubergeeks to see if they had an answer.
Reading through some of the early responses it looks like they gave him at least one path to follow with moodle which looks like a pretty nice solution. I do feel that he threw out a bit of unintentional flame bait when he stated that they wanted a tool that would allow non-programmers to program when what it seems like he meant was he wanted a system where the users could add their own content without involving programmers.
For an alternative solution he might also want to take a look at http://ocw.mit.edu/index.html which is a solution heavily grounded in a commercial product by ab initio, after all, if it's good enough for MIT it might just fit his needs for flexability and scalability. Not only that, but I believe (and I could very well be mistaken) that OCW itself is available in an open source license format.
Someone send one of these to Stallman
Good for SOE, it takes a lot to swallow your pride and admit that your game is not up to snuff. Then to suck it up and put the time, effort and money into improving it shows some backbone and character. I haven't played galaxies for about six months now, but I may very well give it another shot.
By all means, dive into Oracle, it will be a pleasure to see the Ellison flames replace the Gates flames.
Wow, there is a lot of anger in that reply.
When I wrote the original post I was unaware that L3 was blocking all Cogent originated traffic. This is entirely another issue, although related, from the original post. Yes, Cogent does have to pay someone transit, but, under normal circumstances, loosing a bgp session would not cause you to lose all connection to an IP space once the route was adjusted to transit another network in order to gain access to, in this case L3's network.
Please be clear on that fact that if L3 cuts their peering point with Cogent it does not stop anyone on either network from reaching the other as the article suggests. Peering simply gives each providers clients shortcuts to the others IP space. All tier one providers have multiple peering points because the whole point of routing a packet that is not destined for your network is to get that packet to the edge and off your network.
n /0,,sid7_gci212768,00.html
To try for some clarity, just as an example: L3 peers with Cogent, XO, and 4 other providers. Cogent peers with L3, Internap, and 6 other providers. So, if L3 turns off their peering to Cogent, packets that originate from Cogent that are destined for the L3 IP space will have to transit another common peer before they get to L3's network.
For an interesting look at how peering works in the real world take a look at http://searchnetworking.techtarget.com/sDefinitio
I feel rather dense here, but what do fps over say 60 buy you? I use 60 because that is the fps of the average motion picture.
I am not trolling I am actually curious, can the eye see the difference between a card pumping 90 fps and one that is pumping 110 fps?
This is like attaching a cellphone to your asterisk. It allows you to make calls across the GSM network and to receive calls to the number programmed on the SIM card and pass them to the asterisk box.
There have been some attempts to do what the parent is asking about, but I do not know of any that have been rolled out for public consumption.
I have to agree, we have been using IPCop for over a year and are very happy with it.
I suggest that you implement load balancing and failover in your router.
Well I guess the interface may change a bit, for example adding a name to my address book: Phone# configure terminal Enter configuration commands, one per line. End with CNTL/Z. Phone(config)# alias exec callbob dial phone.device.213.555.1212 Phone(config)# end Phone# callbob? *callbob="dial phone.device.213.555.1212"
It's not as though the telco's haven't been subtly anticompetitive up to this point. I work for a regional DSL provider who leases Bell lines (specifically BellSouth). There is a never ending stream of problems with the bell systems. For example, when we switch a customer it can take up to two weeks to take effect, but when a customer switches back to Bell, very often, our DSL service is cut within 24 hours of their order being placed even though bell service will not be in place for a few more days.
Do we think this will pass? Absolutely. That's wyh we are taking this opportunity to look at other technologies like wireless, so really, in the end this is probably a good thing for competition.
Happy to. Because unless you are a local exchange carrier (a certified phone company) you do not have access to the 911 database. Most 911 centers are staffed by the local municipality, but the architecture is provided by the telephone company. We are going through this right now. We have spoken with the local 911 center and they were happy to let us insert data into the database but the regional bell says there is no mechanism for us to do that. So here I am stuck in the middle, wanting to comply but having no way to do so.
Actually I believe this has already been asked and sufficiently answered. There have been no issues with the United States stewardship of the root servers on a technical level.
In this day and age these servers are truly globally important. Unless there is a problem with their management or function I can think of nothing more disruptive to the internet globally than relegating their administration to a global committee.
Actually, MPLS would fit the bill here quite nicely. It allows you to specify ToS in an external label so that the routers do not need to peek at content to make routing deciscions. This makes for very fast routing deciscions. Keep in mind that you will have to support QoS edge to edge and once a packet hits an egress point to the public internet, boom, the QoS and ToS bits are stripped away. You might want to look at products from Riverstone, Alcatel, Seimens, and last but certainly not least expensive, Cisco.
Well, you are certainly entitled to your opinion. If only it were as easy as walking in to the server room and sacrificing hundred dollar bills then yes, the problems would have been gone by now.
In the grand scheme of things five months is a relatively short period of time to work the bugs out of a large system like the WoW server farm.
Most of the servers that have been affected by the worst cases of downtime are the highest pop servers and they do need to resolve those problems and satisfy those customers even if that means closing those servers for new characters.
I don't know what industry you are in, but I am in the IT world and every time you have to patch a system, no matter how many times you have done it on the test bench, you hold your breath and cross your fingers. Multiply that by the number of inter-related systems that Blizzard is running and you begin to see that it is actually pretty amazing how quickly they are able to rectify those issues without having to resort to a rollback.
As to the end game, man, just chill for a sec. Did you even stop to look at the in game world while you were rushing headlong to 60? They are working on the content as is plainly evident by a cruise over to http://www.worldofwarcraft.com/. In fact, the patch yesterday introduced some of that content.
Keep in mind that WoW is the new kid on the block. What has happened to them is virtually the same as what we see every day here. They were /. According to every interview I read they were not expecting anything like the response they recieved. That kind of slammage is gonna requre a serious rescaling of the backend which will not happen overnight. Imagine if you had 1.5 million users banging away on your project, no matter how careful you were or how much you tested, things happen in the real world that don't happen in the dev sandbox.
As far as endgame content is concerned, again, it's a new game. To compare it to games like EQ with its X number of expansions is insane. Give them time to stabalize their platform and I am sure you will see more content.
Well, as a local market VoIP provider, we offer 911 service. We work with the 911 authority to update their database when we assign a client a number. Of course, the problem here is if the client moves locations and doesn't inform us. The lack of an elegant 911 system is definitely the biggest stumbling block to wide spread VoIP deployment IMHO
The reason this is a big deal (and in the interest of full disclosure I work for a VoIP provider) is that it is one less hurdle. Sure, you can get phone numbers from Vonage and the like now, but for them to get those numbers they have to go to the LEC in that NPANXX area and negotiate with them. In some areas the LEC just tells them to kiss off, we are not going to sell you numbers (this is the case in most rural telecom areas). This ruling will allow providers to go directly to the NANPA directly for numbers.
Guess you'll make it to the resaurant at the end of the galaxy before me. Give me a shout on the rap rod when you get there.
You are ignoring the fact that the major vendors on top of having to pay for advertising, sales weenies, and a fancy database also swing the stick of major purchasing power themselves. You do not become, and maintain, your position as a major server vendor by turning out junk. Yes, there are exceptions (the IBM deathstar incedent still makes me wince) but on the whole, I would trust the power supply speced by a group of highly trained engineers over the one I would pick out myself.
Well, here goes, (flame on everyone). I work in a hosting environment where the demands on CPU are low and storage is high. I don't need the whizziest new processor in my NOC to do a great job for my clients. I stick with one platform (Dell Poweredges in my case) and I began purchasing on ebay. I now have a very good relationship with one vendor, I pay about one quarter of what a new server would cost, I get a 30 day warranty from the vendor (and we torture test the machines for about a week upon receipt) and I have a NOC full of quad processor, Raid5, redundant servers that I am very happy with and which do the job quite well. Not only that, but replacemnt parts are a snap to buy direct from Dell.