Lock down the APIs so verdors and closed source projects can link to the kernel. Things like FUSE, NDISwrapper, etc. should be "the way" ALL layers are done. The kernel is robust enough now to lock down the interface and allow the hardware and internals to grow independently of each other.
CS = software development = creation of value = creation of money IT = operational lubricant = stop money waste = preservation of money
From a financial perspective, value creation trumps loss management. If you like both, go CS. If you really like IT, it is worth it to do what makes you happy.
1. Many "Protocols" are in fact determined by source or address. Have you noticed that most P2P apps run on port 80? Does this mean it is HTTP?
2. For many FTTH-type installations, or even certain DSL/Cable bundles, a router is provided by the ISP. Most people greatly benefit from the NAT on these routers as they are in the "request/response" world more than the daemon world and simply don't want unsolicited packets.
3. Virus scanning at a network level is possible and effective. Many customers want this. Also, many customers want HTTP and other content filtering proxies to protect children from predators and inapproproate material. This may also be illegal.
4. Spam filtering and pop-up blocking are mainly done out of band, eg via mail servers, HTTP content filters, etc. but still may be at risk of being illegal.
5. Traffic shaping: completely at-risk. The whole notion of "type of content" is completely arbitrary and will be highly subject to opinion.
6. Port blocking may come in handy to protect network resources from DDoS attacks, infected PCs, etc.
By no means am I an RBOC fanboy, but my point is this is a very slipppery slope. People can vote with their wallets if they feel they are getting screwed.
The whole "advantage" of packet-based networks over "circuit switched" networks is oversubscription. Most, if not ALL, major ISPs perform traffic shaping to make sure this oversubscription does not result in all the bandwidth going to the filesharing app or zombied PC down the street.
If someone on your broadcast segment gets zombied, you WANT the ISP to be able to shut that down, which would be illegal w/o dropping traffic.
Also, since ISPs provide e-mail services, those data services may be considered part of the "data" provided by the contract, and as such, removing viruses and spam would be unlawful.
The point is a non-neutral net is a double edged sword, and many services critically depend on either:
1. them getting a priority boost (eg, SIP, H.323. IGMP, RTSP,...), or 2. some other app not clogging things up (bit torrent, spambots, zombies, etc.)
Just something to think about before championing legislation.
It is trivial to maintain multiple TCP sessions and multiplex in the library. It is a tuning parameter to determine at what point an additional link would be beneficial, but for the most part this only benefits things like HTTP that require pipelined responses to be returned in the same order they were requested. If you have ad hoc messages and the ability to pre-empt message serialization, you can pretty much max out the available bandwidth between 2 points and still guarantee some level of fairness.
Any reasonable app can wrap up similar functions in a well-written messaging library. There may be some minimal enhancements by performing certain operations in the kernel (and hopefully in hardware), but the main stuff can be handled in any reasonable message-oriented transmission kit (bit torrent, jabber, and libraries I have written). The key here: messages good, streams bad. If you take that concept to heart you can avoid a lot of headaches (scaling, repairing, multi-pathing, multiplexing, guaranteed delivery, etc.).
Things that get FULLY documented become so obtuse, esoteric and bureaucratic that they can never actually work without huge teams of people. Look at CORBA, X.500, iCAL, and government defense contracts.
Things that are over simplified (HTTP,SOAP,SMTP) take off quickly and people are left to their own devices to fill in missing protocol gaps, causing huge future headaches.
The best way to split the difference is to make LOTS of simple design docs that allow for multiple iterations, where no single document becomes a monster (the CORBA spec is currently over 1000 pages).
traceroute uses ICMP and not UDP. It is HIGHLY unlikely you'd see any noticable QoS difference (which basically only kicks in with congestions) after seeing the 600ms added by the satellite link clearly dominates the timescale.
As the author of this article states OVER and OVER, it is REALLY EASY to mess up your filters, and it is very tedious (with lots of permutations) to properly build your corpus.
For a centralized spam filtering solution, the goals are:
1. Insulate the users from spam
2. Insulate the users from "administration"
3. Do no harm (no false positives)
For these goals, I would take a "dumb" filter, set it conservatively, and hope for 80% catch rate and zero false positives.
DSpam has a complicated workflow that requires EACH AND EVERY end user to complete a feedback loop. This is WAY to much to expect from people who are barely capable of finding Google. Unless the ONLY access to the mail is web-based, with a VERY clear "This is Spam" button, Bayes is a sysadmin's nightmare.
My only gripe w/ SpamAssassin is performance. If I could get SPAMD to analyze headers in 25ms instead of 2000ms I'd never look back. As it is, DSPAM's performance has me very jealous.
Actually, if they wanted free replication and caching they could distribute MD5 checksums in DNS so that you could create a ciphertext based on the e-mail address and see if it is in the (DNS) database.
All they need to do is set up a web service that responds YES or NO to whether an address is blocked. There is NO NEED to publish the list itself. In a single line:
This is based on reality and my personal knowledge -- not what is on their website.
PhD in CS for a damn INTERNSHIP!
http://www.google.com/jobs/internship.html
Google is a great company founded by a bunch of academic aristocrats that will not consider employees that we not good students from good schools (Ph.D. at Stanford is par for the course). I bet if they looked at the results of their challenge they would se SEVERAL people who did very well with little or no academic training.
Meritocracy is the way to determine who knows their stuff and academic merit does not hold up well to real world merit in my experience. Of the best coders I have ever met many are college dropouts and others did their college in totally different subjects (chemistry, physics, music).
----
Linux users are cheap
on
Kylix in Limbo
·
· Score: 2, Insightful
Linux users by their nature are averse to paying for software. I would rather roll my own stuff using Java, Tcl, PHP, etc. and then not be dependent on a company like Borland.
I looked at Kylix as it looked cool but now it appears I was correct in avoiding it. I pity companies who try to sell software to people like me who are addicted to free (as in beer) software.
it would be quite difficult to get high performance IO through a PCMCIA bus. I can see its use for large matrix computations but not generally useful as an "OS" CPU.
Unless they have some monster NUMA architecture RAM access will also suffer dramatically. RAM contention on a 64-way system would be AWFUL and there is NO WAY in hell they would access system RAM through PCMCIA. It is certainly an interesting idea but I do not see a way for this technology to be useful without lots of changes to existing software and program design. This is NOT like adding 64 CPUs to your home machine.
If hundreds of volunteers were to write software to increase EFFICIENCY of small operations it would only reduce the human resources needed to complete a job, and would in fact CAUSE unemployment.
The only thing to help mom/pops EMPLOY more people is to increase their revenues, so volunteer marketing / web pages would be better.
Or, if we are simply talking about job creation, how about sending spam? Spam increases marketing for small organizations, sells computers because of the increased loads to process the spam, consumes expensive bandwidth, and creates product opportunities for anti-spam software and services.
Unline faxes, VOIP needs lots of things customers can not do for themselves (soft switching, namely). Simply buying a VOIP phone (the product) without Vonage or some similar service makes it IMPOSSIBLE to swith between PSTN (public phone newtork) and the internet. Sure, one POTS dies and EVEYONE on the planet has their own domain name tied to an IPv6 address (nobody can memorize IPv6 addresses) the "service" element will drop off, but as of today, the service element is absolutely critical.
"The magazine and the article itself are intended for executives, so the technical aspect is at a beginners level."
Is this to say CIO's are not techincal? Any CIO who is worth his / her salt should be able to understand technological issues at a profound level.
Is yours?
you are certainly getting a full t1 to them...
on
How to Test Your T1?
·
· Score: 1
your local loop provides a continuous pipe of 1.5MBit/sec to THEM, which has a max theoretical bandwidth of 192KB/sec. However, even w/ a cisco router doing T1 over a cross connected v.35 cable (to itself), you'll never see TCP traffic over about 150KB/sec. It is physically impossible for them to "steal" this.
Now, just as certain as you get the full pipe to him, you will never get his full pipe out. If people did not oversell the internet would be mathematically impossible. There simply is not enough througput available on the planet to guarantee no bottlenecks between you and anyone else you would actually perform a transfer with.
Look for a large provider that is well cross-connected, like Internap. The more networks a provider has direct access to, and the better the routing, the better your overall performance will be. If you are on a tier-3 provider, you may be blazingly fast to some places, but getting your packet across town may take it though abu dabi.
Slashdot Mirror
Lock down the APIs so verdors and closed source projects can link to the kernel. Things like FUSE, NDISwrapper, etc. should be "the way" ALL layers are done. The kernel is robust enough now to lock down the interface and allow the hardware and internals to grow independently of each other.
CS = software development = creation of value = creation of money
IT = operational lubricant = stop money waste = preservation of money
From a financial perspective, value creation trumps loss management. If you like both, go CS. If you really like IT, it is worth it to do what makes you happy.
1. Many "Protocols" are in fact determined by source or address. Have you noticed that most P2P apps run on port 80? Does this mean it is HTTP? 2. For many FTTH-type installations, or even certain DSL/Cable bundles, a router is provided by the ISP. Most people greatly benefit from the NAT on these routers as they are in the "request/response" world more than the daemon world and simply don't want unsolicited packets. 3. Virus scanning at a network level is possible and effective. Many customers want this. Also, many customers want HTTP and other content filtering proxies to protect children from predators and inapproproate material. This may also be illegal. 4. Spam filtering and pop-up blocking are mainly done out of band, eg via mail servers, HTTP content filters, etc. but still may be at risk of being illegal. 5. Traffic shaping: completely at-risk. The whole notion of "type of content" is completely arbitrary and will be highly subject to opinion. 6. Port blocking may come in handy to protect network resources from DDoS attacks, infected PCs, etc. By no means am I an RBOC fanboy, but my point is this is a very slipppery slope. People can vote with their wallets if they feel they are getting screwed.
The whole "advantage" of packet-based networks over "circuit switched" networks is oversubscription. Most, if not ALL, major ISPs perform traffic shaping to make sure this oversubscription does not result in all the bandwidth going to the filesharing app or zombied PC down the street.
...), or
If someone on your broadcast segment gets zombied, you WANT the ISP to be able to shut that down, which would be illegal w/o dropping traffic.
Also, since ISPs provide e-mail services, those data services may be considered part of the "data" provided by the contract, and as such, removing viruses and spam would be unlawful.
The point is a non-neutral net is a double edged sword, and many services critically depend on either:
1. them getting a priority boost (eg, SIP, H.323. IGMP, RTSP,
2. some other app not clogging things up (bit torrent, spambots, zombies, etc.)
Just something to think about before championing legislation.
The inability to "block impair, discriminate or interfere with anyone's services or applications or content," makes the following illegal:
- QoS
- NAT
- Virus Scanning
- Spam filtering
- Traffic Shaping
- Pop-up blocking
- Port Blocking
This means the traffic on the Internet will now be even more dominated by malware and scumbags then ever before. This is a good thing?
It is trivial to maintain multiple TCP sessions and multiplex in the library. It is a tuning parameter to determine at what point an additional link would be beneficial, but for the most part this only benefits things like HTTP that require pipelined responses to be returned in the same order they were requested. If you have ad hoc messages and the ability to pre-empt message serialization, you can pretty much max out the available bandwidth between 2 points and still guarantee some level of fairness.
Any reasonable app can wrap up similar functions in a well-written messaging library. There may be some minimal enhancements by performing certain operations in the kernel (and hopefully in hardware), but the main stuff can be handled in any reasonable message-oriented transmission kit (bit torrent, jabber, and libraries I have written). The key here: messages good, streams bad. If you take that concept to heart you can avoid a lot of headaches (scaling, repairing, multi-pathing, multiplexing, guaranteed delivery, etc.).
Things that get FULLY documented become so obtuse, esoteric and bureaucratic that they can never actually work without huge teams of people. Look at CORBA, X.500, iCAL, and government defense contracts.
Things that are over simplified (HTTP,SOAP,SMTP) take off quickly and people are left to their own devices to fill in missing protocol gaps, causing huge future headaches.
The best way to split the difference is to make LOTS of simple design docs that allow for multiple iterations, where no single document becomes a monster (the CORBA spec is currently over 1000 pages).
traceroute uses ICMP and not UDP. It is HIGHLY unlikely you'd see any noticable QoS difference (which basically only kicks in with congestions) after seeing the 600ms added by the satellite link clearly dominates the timescale.
As the author of this article states OVER and OVER, it is REALLY EASY to mess up your filters, and it is very tedious (with lots of permutations) to properly build your corpus. For a centralized spam filtering solution, the goals are: 1. Insulate the users from spam 2. Insulate the users from "administration" 3. Do no harm (no false positives) For these goals, I would take a "dumb" filter, set it conservatively, and hope for 80% catch rate and zero false positives. DSpam has a complicated workflow that requires EACH AND EVERY end user to complete a feedback loop. This is WAY to much to expect from people who are barely capable of finding Google. Unless the ONLY access to the mail is web-based, with a VERY clear "This is Spam" button, Bayes is a sysadmin's nightmare. My only gripe w/ SpamAssassin is performance. If I could get SPAMD to analyze headers in 25ms instead of 2000ms I'd never look back. As it is, DSPAM's performance has me very jealous.
Actually, if they wanted free replication and caching they could distribute MD5 checksums in DNS so that you could create a ciphertext based on the e-mail address and see if it is in the (DNS) database.
All they need to do is set up a web service that responds YES or NO to whether an address is blocked. There is NO NEED to publish the list itself. In a single line:
wget http://nospam.gov?address=some@address
which would return:
Content-Type: text\plain
NO|YES
Why is that so hard?
If he wants only one GUI, why not use KDE only? It beats GNOME hands down in every category IMO.
This is based on reality and my personal knowledge -- not what is on their website. PhD in CS for a damn INTERNSHIP! http://www.google.com/jobs/internship.html
Google is a great company founded by a bunch of academic aristocrats that will not consider employees that we not good students from good schools (Ph.D. at Stanford is par for the course). I bet if they looked at the results of their challenge they would se SEVERAL people who did very well with little or no academic training.
Meritocracy is the way to determine who knows their stuff and academic merit does not hold up well to real world merit in my experience. Of the best coders I have ever met many are college dropouts and others did their college in totally different subjects (chemistry, physics, music).
----
Linux users by their nature are averse to paying for software. I would rather roll my own stuff using Java, Tcl, PHP, etc. and then not be dependent on a company like Borland.
I looked at Kylix as it looked cool but now it appears I was correct in avoiding it. I pity companies who try to sell software to people like me who are addicted to free (as in beer) software.
it would be quite difficult to get high performance IO through a PCMCIA bus. I can see its use for large matrix computations but not generally useful as an "OS" CPU.
Unless they have some monster NUMA architecture RAM access will also suffer dramatically. RAM contention on a 64-way system would be AWFUL and there is NO WAY in hell they would access system RAM through PCMCIA. It is certainly an interesting idea but I do not see a way for this technology to be useful without lots of changes to existing software and program design. This is NOT like adding 64 CPUs to your home machine.
If hundreds of volunteers were to write software to increase EFFICIENCY of small operations it would only reduce the human resources needed to complete a job, and would in fact CAUSE unemployment.
The only thing to help mom/pops EMPLOY more people is to increase their revenues, so volunteer marketing / web pages would be better.Or, if we are simply talking about job creation, how about sending spam? Spam increases marketing for small organizations, sells computers because of the increased loads to process the spam, consumes expensive bandwidth, and creates product opportunities for anti-spam software and services.
Unline faxes, VOIP needs lots of things customers can not do for themselves (soft switching, namely). Simply buying a VOIP phone (the product) without Vonage or some similar service makes it IMPOSSIBLE to swith between PSTN (public phone newtork) and the internet. Sure, one POTS dies and EVEYONE on the planet has their own domain name tied to an IPv6 address (nobody can memorize IPv6 addresses) the "service" element will drop off, but as of today, the service element is absolutely critical.
"The magazine and the article itself are intended for executives, so the technical aspect is at a beginners level."
Is this to say CIO's are not techincal? Any CIO who is worth his / her salt should be able to understand technological issues at a profound level.
Is yours?
your local loop provides a continuous pipe of 1.5MBit/sec to THEM, which has a max theoretical bandwidth of 192KB/sec. However, even w/ a cisco router doing T1 over a cross connected v.35 cable (to itself), you'll never see TCP traffic over about 150KB/sec. It is physically impossible for them to "steal" this. Now, just as certain as you get the full pipe to him, you will never get his full pipe out. If people did not oversell the internet would be mathematically impossible. There simply is not enough througput available on the planet to guarantee no bottlenecks between you and anyone else you would actually perform a transfer with. Look for a large provider that is well cross-connected, like Internap. The more networks a provider has direct access to, and the better the routing, the better your overall performance will be. If you are on a tier-3 provider, you may be blazingly fast to some places, but getting your packet across town may take it though abu dabi.
I guess not since we all know this site it suspect before we get here.
hello
I hade over 40,000 users on a Pentium 667 w/ 256MB of RAM, and this includes LDAP, POP, SMTP, IMAP, webmail, qmail, apache, etc.
If that machine ever hits 1% utilization he did something stupid.
It speaks poorly for that loser that he can be bought out for a measly $19,000. That doesn't even pay an intern!
I would have thought it'd cost a couple million to pass a shite law through congress. I guess in a down market every penny helps.