Pleasing your patrons in academia

Wow, it's been a few days since I posted, but it was a busy weekend for me, I guess I have to try harder next time. Ok now

http://chronicle.com/article/Public-Research-Universities/126464/

The link I just pasted basically tells university researchers to spend more time on projects that bring money to the state. It gave me the sense that many people in the private sector feel that university research is pointless unless it's helping the local community (or the state). I think many people forgot that good research is a slow process that takes a lot of thought, time, and retries. You are basically searching the truth space and sometimes it's not always obvious. These days as we have new data, and faster supercomputers, we are able to find correlations and patterns that gives us some insight on the underlying truths of the world. But people are not patient enough to wait for these truths. I read an article the other day saying that the President wants to spend more on engineering projects instead of "scientific" projects. Engineering is all about implementation, and not really about discovery or really advancing the field. Trust me, I am a software engineering, so all I do is implementation, but my implementations depend on scientists (way smarter than I) to come up with truths that give me the assurance that AES encryption is really hard crack or that a small world graph will give me logarithmic hop-count to any destination (on average). Most times engineers are to bogged down with the hardships of implementation to worry about the scientific (or theoritical) stuff.

Now back to the article, I see it as an omen, because of this new wave of austerity measures sweeping the nation, scientific research funding will definitely take a hit. The future is going to be tough, since there will be less NSF funds and an increasing number of PhDs fighting for it. I guess over time researchers will look more towards companies for funding and maybe that's the future of research. Companies will drive research in universities as a result, they save money in R&D and universities make more money without depending on tax dollars and everybody wins. To reiterate, departments won't have to close because researchers are able to get money for industry, state saves money because they can allocate less to universities, industry saves money because they get a bigger bang for their bug by outsourcing their R&D to academia. For now, I'll just leave it at that. (I actually proofread this one, I guess I'm a sucka today).

Staying connected

We all know how popular Facebook and MySpace have been in the past few years. In the past, people mainly stayed connected through email. But now, with social networking site, you can know a whole person's history from the classes that they are taking to what parties they plan to attend (if they let you). These social networking sites, realizing each other's need to constantly stay connected, began providing frameworks that allow application developers access to that connection link between peers. Developers can use that information to create applications that connect friends in unprecedented ways.  Some applications automatically update user profile's based on some criteria. For example, smartphones that can automatically upload pictures to Flickr or Facebook. With this connection information, applications can be much starter about who, when, how, and where to connect to each other.

• Imagine a peer-to-peer file sharing application that only connects to your facebook friends and allows you to share only with the people you want.


• Imagine being able to run your own website privately and allow only your friends access to it and not have its content show up on Google.


• Imagine graduate students at different universities connected together and allow their spare CPU cycles to be shared when their CPU is idle.


• Imagine friends being able to share their files with using email.


• Imagine friends being able to stream their video files to each other without having to download the entire file.


• Imagine being able to do all of this in 5 minutes or less.

Well, if you can dream it, then you can acheive it. Check out this link below, it may not be the answer to all of these dreams, but it's a start.

http://socialvpn.wordpress.com

How to turn your Windows PC lab into a Linux-based Condor Pool

How can you get the most use out of a Windows PC lab?

Most Windows PC labs around the country are severely underutilized. Despite the high cost of power consumption and management, average PC labs only uses about 10% of their total CPU utilization in a particular day. Meaning that 90% of the time, the CPU is just waiting and running some idle process that does nothing. Well, there is a solution at

http://www.grid-appliance.org

This Grid Appliance project you to turn a Windows PC lab into a Condor pool without interfering with normal lab use. Through the use of virtualization, this can be acheived without requiring any change to the infrastructure of the lab. And all this can be done effortlessly. Here is a list of projects/software that have made this possible.

• 
  
  VMware (www.vmware.com)
  
  


• 
  
  Condor (www.cs.wisc.edu/condor)
  
  


• 
  
  IP-over P2P (www.ipop-project.org)
  
  

If any of the links above are not clear enough on how this is possible, simple reply to this post and I will be glad to provide more details.

 Enjoy

ICMP or No ICMP

Lately, I've been looking a lot into ICMP packets. Most people only pay attention to TCP and UDP packets, while most just know that ICMP packets are used for pings, traceroutes and others. Well, I've been looking into various other uses for ICMP packets, but one question came to mind? Should ICMP packets be used for other purposes besides their original intent? Would the makers of the Internet get upset if people started using ICMP packets to communicate data, the same way that they use UDP packets, if so, why would anyone want to even consider that approach?