A beautiful, yummy smelling surprise that appeared in my room...
thank you, you know who you are :)
The first week is, as expected, very hectic. Between moving in, getting to know the campus, and navigating through the CS building, time really flies by. Then all types of accounts, accesses, paperwork, permissions have to be granted, signed, and approved. The project that I am working on partly lives at the University of Massachusetts, so on top of gaining access and logins to machines and code here at the University of Texas, we have to obtain similar access at the U. of Mass. as well. Still, since everyone in CS checks their email regularly, setting everything up ends up being a lot more efficient than expected.
The second week still finds us in the midst of setup - we are attempting to use CVS in Eclipse to get Scale, which consists of over 700 classes (in Java). This requires many specific settings and environment variables to be set, and somehow some package or another is always giving us problems. Still, even when this is completed, it would only allow us to look at the files, and run things through the debugger using TRIPS, rather than PowerPC, to get an idea of how things work. We still have to setup something similar once we get access to PowerPC machines, which is ultimately what we would need to test our code.
On a more fun note, I think Katie and I are settling in and beginning to enjoy our environment. The grad students here have been amazing with making us feel welcome and busy! I think without them this would not be the experience it is turning out to be. I have also started playing racquetball here, and find it really fun. The gym here is also great, although there was a little run-around about all the paperwork since we are not actually students here. There seems to be a lot of that at first, but I think it is finally beginning to wind down.
This week I am going through the files and attempting to get to know the code. Since there are over 700 classes, I clearly do not need to acquaint myself with all of them; however, part of the hard work right now and pretty much from now on is exactly getting to know what we do have to know, and that which we don't. It certainly helps that one of the other already existing backends, the SPARC one, is somewhat similar to the one that we are to write. However, since we didn't write it, we have to get to know that one as well, for it to be any help.
I am also excited because this weekend is Father's Day, and I am going to be going home for it. My father doesn't know, so it is certainly going to be a surprise. I can't wait to go.
I had a great time last weekend going home - I really missed it a lot and felt like I would have liked to be there longer. Unfortunately though, I sprained my ankle playing tennis on Saturday, which meant no gym or racquetball this whole week. At first I couldn't walk on it at all, but what's left now is swelling and only some pain. I definitely feel it put a damper on things.
I found a new author I like, which for me is great news. I bought a book before getting on the plane for home last weekend, by Jodi Picoult, called My Sister's Keeper, and I really enjoyed it. A couple of days after I got back, I went to the local Barnes and Noble here and bought two more books by her - I hope they are just as good.
Project-wise I ran into another problem. It seems that I cannot run Scale even when logging on to a PowerPC machine for several reasons, all of which seem to come down to the fact that the machines are PowerPC/Linux. Katie has been really lucky in this respect - she brought her Mac laptop here, which turns out to be precisely what's needed to avoid all these problems. It's definitely a little frustrating, and I am more than ready to start coding, but unfortunately I cannot do that just yet.
One of the things that I am really enjoying here is going to the weekly meetings of different groups. One of the meeting is for TRIPS, and the other is Speedway. In the TRIPS meetings everybody discussing the progress that they have made that week on their particular task, and then everyone discusses which problems still remain, how to solve them, which ones are priority, etc. The past few Speedway meetings have also been really interesting because they have been critiquing papers that are going to be or are attempted to be published. We get these papers several days before the meeting, along with certain reviews that this paper has already gotten. Then during the meeting everybody discusses different aspects of it, what they saw as a potential problem, what the strengths and weaknesses were. These experiences are great.
Also, Katie and I are each have lunch with Kathryn once every week (alternating). I have just had my first such lunch, and I really enjoyed it. I got some insight into how Kathryn manages being a Professor along with the mother of three young boys, each in itself sounding like quite an accomplishment! I am definitely looking forward to spending some more one-on-one time with her.
This week was just another episode in the getting-things-to-run saga. Turns out that the ppc machines have to be moved to the Debian OS. In the process of moving the first ppc machine, something weird seemed to happen, and we thought that the hardware crashed and needed to be sent away for repair. However, for the first time, luck seemed to be on our side, and everything seemed to be installed and ready. With some additional searching for the required Java 1.4, and then its install, that machine seemed to be ready to build Eclipse and set up Scale on it. Unfortunately, it seems that someone is running timing experiments on that machine that have shut down the network interface (in order to minimize variations in the timing). So I am again machineless. Looks like having that one machine ready will not be enough, and that we have to grab another ppc machine and upgrade it as well.
Meanwhile, I borrowed Katie's laptop for a few hours, and got to run things on ppc for the first time! I think we are now going to try to pair program, although at this point that will not be so easy. Other than the fact that Katie expressed her dislike of doing that from day one, now we are no longer on equal footing with our knowledge/awareness of the code.
All in all, I am desperately, desperately awaiting to be able to work on my own machine. I will then have to spend many hours to catch up, and then Katie and I will have four weeks to accomplish that which we should have had ten for.Week 7
Another ppc machine has been upgraded to Debian. Following that was many hours and days of trying to build Eclipse from source. Everything seemed ready to go, but for the bug in Eclipse 3.0 for building on ppc. At this point I am no longer surprised, disappointed, or even anything, whenever something new goes wrong. After many more hours, I finally build Eclipse 3.0 from source!! I am very happy about that; it seems like this has really been the first hurdle in pretty much all these seven weeks that I have been able to cross. It is yet to be seen that everything in it works successfully, but I have my fingers crossed.
Of course, if only I was now set up, I would sound a lot happier. Unfortunately, there is yet another piece to this puzzle. Scale currently does not have its own parser, which I think is currently in the works. It therefore uses EDG, which is what I still need to be set up for a ppc machine in order to proceed. I am really hoping that next week's entry will find me set up and working. :)
This week a miracle happened - I was given a Mac laptop to work on!!!! This happened after several days' work on trying to get EDG all ready to go, rather unsuccessfully. I do not know why if one is accessible to me it took so long for me to get it (in the beginning I was told that there were no mac laptops that were accessible), but I am thankful that I have one now! The funny thing is that when I was waiting for a wired connection to be set up by my desk, they brought me a second one (laptop)!! Turns out they had their signals crossed or something, but either way it was really funny - from none to many!
Katie is helping me with getting everything set up on it, since she has already gone through this before on her own laptop. I am going to be using Xcode, which is free for all Mac owners and part of Developer's Tools (or at least, it used to be, I think they separated it from the bundle now, but it is still available for free). I've never used it before, but I am looking forward to it, because from my short experience with the mac I am already beginning to like it a lot more than anything else I've used.
This has been one of the best weeks for me here. Katie has come up with something vital that I could work on which should not interfere with anything that she is doing, so I am finally working on something other than setup, and all the things that come along with that. Also, Katie and I gave a presentation to the Speedway group (the group that we usually read/critique papers in, or presentations get practiced on). We talked about many things in it: the DMP program - the whole reason while we are here, its intentions and goals, how those goals compared with the actual reality of our experience here, how everybody contributed to this same experience, and of course our assigned project and its progress. I feel like the presentation went well.
Something else made this week different from the others. I volunteered for First Bytes, which is a camp for girls in high school with the same goal as the program I am in, except on one level lower. The program tries to introduce girls to computers, programming, and overall idea that women and computers can - and do! - mix. I feel that if nothing else, the program made the girls not afraid of a computer - we'll take what we can get.
Back to what I have been working on this week. So it turns out that doing a "Hello World" is easier said than done on Mac OsX (very easy on linux, however). The reason is that the way that Mac OsX handles functions is with position independent code (IPC). My task is to implement this, so that we could test things on the Mac with printf's.
This has been my final week here in Austin. I will be leaving soon, but not without taking some things I learned with me. As I look back on the past 10 weeks, I realize even more that I was very lucky to be placed among the group of people here; that same group whose faces I have grown accustomed to seeing every day. They are an amazing bunch, and I wish all of them luck and happiness.
As for the project, I am happy to report that I have successfully implemented PIC, and also had the chance to work on some more issues. Although I have expressed the fact that I would have loved to continue putting the PowerPC backend together from home to Kathryn (my mentor), I do not think that I will have that opportunity, since I am about to hand in the laptop. Kathryn suggested perhaps making it possible to log in to that laptop, but then I will be back to the idea of working command line, and not having something like Eclipse or Xcode at my disposal (by the way, I really liked Xcode).
All in all, thank you everybody for this experience - you know who you are.
Questions? Feel free to email me at email@example.com