Category: Technical

Recently my DNS provider began experiencing a DNS amplification denial of service attack. This resulted in all of my websites being down for a period of about two days. In an attempt to prevent this type of problem from impacting my websites in the future I have now added two backup DNS providers. When a DNS server fails, the query is attempted on other servers on the list until one works. I am hopeful that this will reduce the impact of any future attacks.

As a general rule domains may utilize up to thirteenth servers, and as I was only using five there was plenty of room for improvement. The root zone uses twelve separate providers for its thirteen servers, and while my domains are not as critical there is technical advantage to removing all possible single points of failure. Perhaps having two backups is paranoid, but it will allow me to use all thirteenth name server spaces on my domains.

As some of my friends are aware, I have recently switched cell phone and service provider, Ting.  Tomorrow evening will be the end of my first month of service with Ting and I thought I would post a few notes of my experience.

I assume that for the most part no one reading this will have ever heard of Ting before, because Ting is a small company and their service just launched in February and therefore have not had the level of publicity that existing carriers have.

The first thing to note, Ting does not have their own network, they are a virtual provider (MVNO) which utilizes the Sprint network. Due to restrictions in this arrangement, to use Ting you must buy a device directly from Ting. Ting is only allowed to activate ESNs that are in their database. This is especially limiting when you consider that they currently only offer Android-powered devices. I hope for an iPhone from them, and have been told that it will come, but just not right now.

The key attribute that encouraged me to change to Ting from Verizon is price. I was on a cheap plan with Verizon, a classic 125 minutes for $25 plan. Oddly enough, the plan had more minutes than I needed. I am not a heavy cell phone user at all. I don’t call very often, I only send a few texts and until Ting I didn’t even have a data plan.  I like having a cell phone because I like the security of having access to it if I ever need to make a call, either out of convenience or mortal need. Since this required having a device with me that I barely use anyway, I eventually got to the point where I wanted a smart phone. Only problem is with Verizon, while the phone would be subsidized by the carrier, the bill for my phone would go to over $70 per month – for something I barely ever use. Ting is interesting because I have my device and if I don’t use it at all during a month my bill is $6. If I use 100 voice minutes, it costs me $9. So just comparing service and pricing, Ting is a better deal, and I get to have a smart phone, which makes the idea of carrying the device with me all the time useful. So the big selling point with Ting – I pay for what I use.

The other thing with Ting that drew me in was the community. Ting support uses a forum for general questions and a helpdesk ticket system for specific questions. The forum is filled with other customers and customer service people. Ting is not reliant on peer support, but it does seem to help things work faster because it is present. Most phone companies look negatively upon “power users” who modify their phones. Ting supports “hacking”, “rooting” and other modifications to the phones. After all the phones are owned by the customers, not Ting, so the approach taken matches accordingly. I have never had a reason to actually call customer service, but I’ve heard there is never a wait and that the help is quick and polite. My contacts with support have all been through the web form (I prefer to type rather than tell my issues), which get responded to fairly quickly. I do get the feeling from the tone of the responses and the commentary of one of their top support people, Ben Lucier, that they are prepared to do whatever is necessary to ensure that the customer is happy. You might think that with the low rates, excellent customer service and roaming agreements that the company might operate at a loss. According to one of the sales staff, Ting currently has exactly 1 customer that they are loosing any money on at all.

I dove in to this “unknown” with a good bit of confidence because the company that owns Ting, Tucows, also owns the domain registry that I have been transitioning to for the past 9 months, OpenSRS. I have been pleased with the level of service and customer focused dedication that I have seen from both Tucows companies. After 9 months on OpenSRS and a month on Ting, I am satisfied with both.

To read more about Ting (and save $50 off of a device) use this link: http://cuku.tk/TingPhone2

Why Ting?: https://ting.com/why-ting/

As I always find a route of avoidance for getting actual work done when it comes to writing papers I have put a little work into my papers website. Take a look at the two images below. One is the prototype and the other is the real beta website that I’ve developed over the past week.

ANSWER: The real website screenshot is the first image. The prototype is the 2nd. My crappy bars and buttons gave it away, right?

I have decided to redesign my academic paper website. Shockingly, I have not actually written a single line of code for the new website. Instead, I have prototyped it in Adobe Illustrator. I guess you could say that I am procrastinating on working on real classwork, and you would be right. I’m not sure what to think of going to this new process of prototyping stuff before I write it. I guess I can blame my Human-Computer Interaction class for this new method. I am hoping to code the new site sometime during the summer and have it ready by fall. The coding and designing will not take me very long, but getting the abstracts, program affiliations and such assembled will be time consuming.

The current prototype appears below.

The Trust Center stuff in MS Office has been bugging me for a while now. I get sick of having to “enable editing” on documents that I created on another system, but may now be on a network drive location or floating about elsewhere. The thing that makes this more annoying is the part where I am a domain user in an Active Directory environment, so theoretically, there is a unique code that each installation of Office could use to validate me as the creator of the work. I know that my user-id is stored (as KULARSKI\curtis I believe) in the document’s meta data, but that isn’t reliable, so I understand why that isn’t validated to determine if the document can be assumed to be safe (because it originates from me). Each object (including users) in Active Directory has a unique ID, referred to as either an SID (cause they are like S-1-####), or a UID (for unique ID). It seems reasonable to me that perhaps Microsoft could have written Office to pick up on that ID (from a network OS model they designed) to help with eliminating redundant prompts for authorization to edit documents created by the user at a different location. Another thought that has occurred to me on this line is the idea that perhaps future versions of Windows should by default have the ability to carry a user certificate, without the aid of AD Certificate Services, for basic in-domain authentication tasks, such as automatically assigning a digital signature to documents or allowing a user a temporary login to a part-time connected domain client system utilizing a file on a removable storage device to authenticate. These implementations are things that I would consider to be “low security” situations. Obviously not a good practice for financial institutions, military or other high-security situations, but for situations where physical security is pretty well established or there is a low risk of falsification of the credentials, it is a workable solution that could make things a lot easier and avoid certain problems, such as verifying a document’s source (when you wrote it) or not being able to log into a domain system just because it can’t connect to a domain controller/global catalog server to authorize a login.

This is probably an irrelevant rant, but in some ways I feel like Windows security could be aided by adding some simple measures of authentication. It may be basic but it may help. If a low-security user-check stops users from disabling the security features all together, then maybe there is an overall increase in security?

By this point, I suspect even the least techy of my readers has seen the little square barcodes… like everywhere. They are square for a reason, they are 2 dimensional barcodes, there is information stored both vertically and horizontally, whereas traditional barcodes only scan horizontal measures.

Below are 2 barcodes that related to resources in GITI. The first is a QR code, the 2nd, a standard code 39 barcode. While the initial reaction may be “wow, the QR barcode is huge!”, you would be correct. It is rather large as barcodes go, but it also contains the exact location inside GITI for me to retrieve the data it is pointing at, whereas the 2nd one only provides me a barcode for the assignment number. While initially the idea of using barcodes inside GITI was merely to facilitate end-of-semester grade updates and the like, it was designed with the intention of using a standard barcode reader, which in reality is quite annoying, and saves very little time, especially with numbers so small. QR codes on the other hand, will allow me to scan the code using a basic smart phone camera, or other such device and have direct access to the data I am looking for. On such a small device it is quite a handy thing to not have to type the URL and then the number to access what I’m looking for.

On the other hand, there are serious culturally implications to the use of QR codes. The widespread use of QR codes is more infectious than the wide use of barcodes because of who they target. Barcodes target only “information personnel”, who serve a function in a data collecting system, such as sales clerks, librarians and hospital staff. QR codes affect a much larger number of people.  The integration between human and machine has perhaps become too close. People already are very attached to their smartphones, whether staring mindlessly at the screen, letting it control their schedule or being directly connected to it by a Bluetooth headset. This takes it a step further and makes technology pervasive in all sorts of printed materials, advertising and product packaging. Barcodes at least have a human readable component, QR codes do not. So why are they presented so prominently in human view? To allow the human that is thoroughly integrated into their device to retrieve information more quickly, perhaps almost with machine-like precision.

So why would I want to use the evil little buggers? I am an information worker, and I enjoy geeky little things like this. I like to experiment with things like this. While I may include the QR codes occasionally on things, I don’t see it being something I take on with any seriousness. I have for the moment added the ability to use them to GITI for the view assignments action only. If they prove useful they may stay, if not, then [POOF]. My primary use for them, as I am considering their usefulness at the moment would be on printed materials that I produce (such as final papers), giving them a QR code for locating a PDF copy or some such craziness.

In January 1985, Dr. Jonathan B. Postel, the Internet’s first IANA (Internet Assigned Numbers/Names Authority), set forth a new era in the Internet’s progress. He brought the DNS (domain name system) into existence, essentially giving form to what we today consider the very essence of the Internet. That first glimmer of light on a shapeless form.

In the beginning, there was .com, .net and .org. That was the public sphere of the DNS naming system, all that existed, all that could be outside of restricted space. Of course there were other TLDs, .gov, and .mil primarily. In February of 1985, in coordination with IEEE, Dr. Postel provisioned a ccTLD for each legally recognized country in the world. After this point there were no more additions to the Root Zone for over 15 years, give or take a few changes to the ccTLD space caused by trouble in the Russian territories and the UN requesting the creation of .INT.

In 2000, there was a movement to add additional General Purpose TLDs. The idea caused quite a stir in the Internet community. There were geeks absolutely thrilled with the idea of new TLDs, and of course the ever-conservative Internet Engineering Task Force had their concerns about the stability of the root zone network due to the additional delegations required. In June 2001, 7 new TLDs were permitted to be added to the root: .BIZ, .INFO, .NAME, .MUSEUM, .COOP, .AERO and .PRO. For a time .INFO and .BIZ (and .NAME to a lesser extent) had quite a bit of interest, until it was time for the registries to actually go live that is. How many websites do you frequent that actually use any of these TLDs? For myself, it is not many. In the 11 years since these TLDs went live, I have seen only a few domains in active use (for legitimate purposes).

We are presently in a 90 day window before the application process will begin for yet more TLDs. On January 27th, ICANN will allow any organization, corporation or individual register for whatever TLD they wish, so we can have TLDs that are as arbitrary as .PEPSI and .COKE. As long as the new registrant can prove that they have the competence to operate the TLD, their assignment of their registered name is guaranteed. With known registry operators such as VerisignGRS and Affilas offering their support and services (for a fee) to new registrants are on stable ground to be accepted.

There are several problems with these new TLDs, first of all, the last round of new TLDs did not shake loose the dependence on .COM, .NET and .ORG, so why should these new TLDs be any different?  I do not believe that these new TLDs will pull any weight away from COM at all, and will simply add another level of complexity to the naming system. I think what we have learned based on the first 7 new TLDs is that the “powers” of the Internet have no idea what Internet users want, or what content providers want. Take for example the result of the opening of the ccTLDs of .ME (Montenegro) and .TV (Tuvalu). Those 2 ccTLDs had more of an impact on the landscape of the Internet than the creation of 7 new gTLDs. There is no way to anticipate reaction to new gTLDs, but there is one thing that seems pretty clear to me, no one seems want anything that is 3 characters or longer, everyone wants 2 character TLDs, especially ones that they can use in conjunction with their own names. The “powers” of the Internet do not seem to understand that in the current economic and political atmosphere, the Internet’s consumerist phase is past its climax, and the Internet is now becoming social (and of course, the Internet is for porn, that one hasn’t changed in 20 years). I also believe that this new round of TLDs will be directed at more consumerist ambitions, and will not add to the overall mission of the Internet. Not keeping the consumer space limited to .COM and .BIZ seriously changes the mission of the Internet.

On a philosophical level there are both benefits and problems with the opening of the TLD system. First, an opening of the TLDs is consistent with critical theory ideals concerning not categorizing things and not restricting the categories or identities. However, because of the large financial cost of registering a TLD (not to mention operation costs), it still is not possible for everyone to select their own identity. If I wanted to have the .kularski TLD for example, I would have to pay $185,000 to apply for the name, and then pay additional money to a registry operator to maintain infrastructure for the TLD. There is most certainly not an openness to the process.  

There have been a few TLDs added through sponsorship applications in the past few years, but the process for their introduction was fairly restricted. .JOBS, .TEL, .MOBI and .TRAVEL were added as specialized TLDs for members of professional organizations. .JOBS and .TRAVEL specifically are restricted to members of those organizations. .TEL is a TLD that is highly restrictive in what content is allowed. The domains have no name servers other than those for the TLD itself, and in essence the TLD functions as an Internet phone book. .MOBI is restricted to mobile-enabled websites, targeting a specific audience (smart phone users). These TLDs are very specific and had an audience already established to receive them, unlike the creation of generic TLDs without specific sponsoring bodies that represent a specific community. Also this year the .XXX TLD will become operational. I have mixed feelings about the TLD, but there are many who believe it is a good TLD. I don’t see it as a move toward a more child-friendly Internet (nor do I care for a child-friendly Internet), but I do see it as an alternative to overloading .COM with more porn domains. Pornography is a big part of the Internet and its culture, so having its own space is not a horrible thing. On the other hand, having a TLD that is used to categorize a certain type of content or a certain group of people is not in the best interest of the culture of the Internet. I believe that in addition to the XXX TLD there should have been some type of ARTS TLD created to balance things and provide a space for more creative forms (including creative erotica) to exist.

At the present time, the Root Zone is still a very manageable thing, if you would like to see what makes the entire named Internet work, take a look at¹: http://images.cmkularski.net/blog/root.zone.txt

¹The root zone file is a constantly changing file, and this copy, which is used for reference purposes should not be accepted as the authoritive Zone, for that, please download a copy at IANA.ORG.