Sunday, 14 September 2014

Some background information

I was told by my PI to graph the number of physical memory addresses that are passed through our simulator, and compare these to the structural addresses that these are mapped to. Many of the simulations didn't complete because the files' simpoints needed to be recompiled, so we recompiled the simpoints. To do all of this, we are using a software called Intel PIN tool that is a program through which one can run C++ code with a great deal of control over how the code is run. Simpoints are sets of instructions that are run. So, for example, if we compile 3 simpoints per benchmark, we are given 3 sets of output per benchmarks, each set being the result of ~25,000,000 instructions. We are using SPECfp as a benchmark, which is a standard floating-point algorithm benchmark that works by running state-of-the-art algorithms such as the Einstein evolution algorithm and quantum mechanical simulations.

I apologize that I can't give more accurate details as to the actual code that I'm working with, but I'm not really allowed to talk about it because the research is going towards an unreleased paper.

Tuesday, 2 September 2014

Good morning world

I'm splitting my time between the lasauil project that Neil, Ryan, and Evan are working on, and an internship with UT. For my internship I need to ssh into the UTCS computers, as well as a TACC (Texas Advanced Computing Center) allocations. Unfortunately, AISD blocks all SSH requests, so I've been trying to find ways around the block, with the help of Sam Grayson- I'm also contacting AISD directly about an exception. The work I am doing is related to cache architecture, specifically caching algorithms.

Friday, 16 May 2014

Graphing

Most libraries built for graphing data on web pages are large and clunky, so I opted to write my own client code to generate graphs of user data for the uil website. The data is given as two groups of time stamps and question ids, one for questions answered correctly and one for questions answered incorrectly. First, the data is normalized so that the time stamps are between 0 and 1, inclusive, so that the question with time stamp 0 is your first question answered and 1 is your last question answered. The data is then sorted and iterated through and graphed, keeping track of the ratio between your number of questions answered correctly and incorrectly, generating a graph of your general performance over time.

This style of data crunching is briefly very computationally intensive, so if users are opening a lot of pages it could chew up some bandwidth. Additionally, I'm not sure how well this will work if users have a very long history, but there won't be more data points than there are questions in the whole database, so that should be fairly easy for us to control. 

Tuesday, 29 April 2014

Let's try this again

Code is very dry, and for the most part, uninteresting to read in large quantities on a blog. For the most part, blog posts consisting of a lot of code just end up being blindly copied and pasted by someone scouring the web looking for the specific css they need. How do I know this? I do it- all the time. Yet, when I was looking for a guide to making columns in css I found the amazing blog A List Apart, whose quirky sketches and almost philosophical writing immediately caught my eye.

A List Apart presents practical, thematic css guides, giving the user a general set of rules and inspiration to use in their web pages, rather than a simple index of terms. Their guide to columns and page layouts describes how the proportions of the divs on a page define the page's tone and give it character.

It's also clear that they understand what is key in a website, because their website is easily maneuverable and manages to avoid being overwhelming, despite consisting mostly of giant masses of text.

This fluid and satisfying design philosophy is something I hope to bring into not only my projects, but also my code. As our UIL database project becomes larger and larger, it is crucial to effectively comment and manage the code- avoiding things such as 1000 line long route files and enormous blocks of CSS. But this is not as easy as splitting as splitting files into more javascript files with mildly descriptive filenames- for this only pushes the problem to the app.js file, forcing us to balance many more route files. So what is the solution? I have no idea, really. For now I'll consult The Pragmatic Programmer and maybe Code Complete. 

Monday, 28 April 2014

manifesto.json

For years we have been limited to two dimensions. We have been squashed flat by our languages and have been unable to flourish in the third dimension as we have so long desired- but no more! We shall thrive and reproduce, gaily prancing around in multiple planes. The Z-Axis gives us the ability, nay, the god-given right to express our web pages in a multitude of additional ways. For how long have we seen boxy, single-dimension pages? No longer, for we have been given a gift of vertices and CSS3. Z indices and transforms will revolutionize the way we experience the internet, and will gift unto us a new era of html.

Web workers of the world, unite. You have nothing to lose but your chains!

Friday, 25 May 2012

...

First of all, the class certainly prepared us for the AP, without question. (Not that that's a very big accomplishment, I certainly hope the AP does not accurately represent the content covered in a college class.) Second of all, despite any problems I had with the class, I very much enjoyed it as a whole.


This class should not be any simpler. In the description for the course it says that computer science requires prior knowledge of java/another language. If anybody requires a longer introduction than the one given, then they should take computer programming first. I understand that oop and programming in general can be hard to teach, but if somebody is having difficulty learning about them they should instead take an introductory course. If the course were any simpler, it would seriously deter those who did understand the concepts. 


Although self-motivated projects are inherently engaging, teacher guided projects will always be faster and ultimately more informative. Self guided projects are usually guided through online resources, and thus there is no guarantee I will be able to finish said projects. The guidance of a teacher, or at least a single well-written walkthrough, will both increase the productivity of a student on a project and what a student learns from a project. This leads into my next idea,


Have more projects. Pong was fun to do; having a clear set goal for a project, and the knowledge that I have to finish said project, is extremely fun. More so, everybody was working on the same project, and so getting advice on an issue was extremely easy. Pong itself was relatively simple, more complex projects would certainly be welcome, especially if it required a lot of thinking about how to efficiently code the program. There was little effort actually put into the general method I would use to make pong work, instead the challenge was simply working out bugs.


Not enough focus on efficiency and general coding practices. In the first coding class I took, I was told that the question is not whether you can make a program do something, it's whether you can me it efficient. This isn't really something that was ever mentioned in this class. The use of profilers, for example, to find computationally heavy spots in code should definitely be talked about in class.



Saturday, 19 May 2012

Last 3 weeks

Two weeks ago I was primarily studying for the AP exam, and after taking the AP I have to say that I did not need to. The practice materials had some content I needed to look up: ternary operators, parts of the string class that I don't normally use (a lot of this), and some other miscellaneous code. However, the AP itself had very little even remotely complex content. A specific question I remember quite well had a list of 3 search algorithms which, in theory, found the largest value in an array. One of these, however, used -1 has a default that would be replaced by the next value in the array if it was stated to be the maximum value. Because of this, if -1 was the largest value in the array (there was no precondition that would exclude this as a possibility), then the array wouldn't work. This was the only question that had this small of a problem and thus I thought it might be designer oversight rather than a hard question. Other than this, the AP was fairly basic. This isn't to say that it was badly written, it covered the content it was meant to cover and did it quite well, just that they don't seem to be grading high school APs on too high of a bar. (Watch as I get a 4...)

Outside of class, I've been working on a quadrillateral-based 3D engine. To map the textures I am using affine transformations, a very low-cost method of perspective transformation. The reason this is not used normally is that it isn't entirely accurate and thus must be modified, but this required more processing power. The affine transform is more acceptable on quadrilaterals, thus why I decided to make a quadrilateral-limited engine. A while back I derived a series of equations to run this off of, using perspective matrices. However, I've been experiencing some weird image warping with this, so I need to re-derive the equations. Going off on a bit of a tangent, but learning about the actual application of matrices in computer science has shown me something about how idiotic it is to solve matrices by hand, as taught in algebra 2.

I've also been studying for UIL on my own time... when I read a question which required you to figure out the sorting algorithm a program was using based on the time it took to execute it hit me just how much there was study. Big O notation and sorting algorithms in general seem to be a huge part of the higher up UIL exams, so I'm putting a large portion of the time I spend studying into studying these.