S.KRAUSE

Random Thoughts

Part 2

Design issues: While taking some time to work on a computer today (that is, trying to trouble-shoot problems on someone else’s computer), I for some reason began to think about web-site design issues again. Naturally, this is a large topic - too large to deal with here. I guess it’s a question I have to think about again because I am in the process of re-working the German Dept. site. Before undertaking any changes, I decided to visit other German Departments across the country—online, that is. Of course, I did this before when looking at programs for grad school. In summary, Minnesota’s is pretty, but heavy on the graphics (of course, a lot of college students enjoy T-1 connections); Berkeley’s is ugly in many regards (check out their background colors graphics, for example), but it contains much of the same information our site and Minnesota’s do. Princeton’s is rather minimalistic in some regards, being mostly text-only, black text on white, with a Princeton logo at the top of the pages. I like their list of links, but the rest of their site is not particularly intriguing. Washington’s, while nice-looking at first glance, is rather lacking, being only two-pages in size; they have to have more information than that. As much as I love Pomona, the German dept’s site is quite lacking (since no one has updated in so long). I guess that’s part of it—someone (a faculty member or more likely a graduate student) creates a site/page, but after that, the site just sits there forever, rarely if ever updated. At best, many of these sites become online versions of the dept. brochure. Reflecting upon my attempts to get information from Yale and Harvard senior year, I realize that many of the people who put up these sites never once put themselves in the shoes of someone who might actually want to visit such a site. Who do you contact if you are prospective graduate student? Who do you contact if you are a high school teacher? Who do you contact if you are fellow academic? These questions were rarely answered. Often only the web-master left an e-mail address available, and sometimes the chair’s was listed. But as a prospective student, neither the chair nor the web-master were the people I should have been contacting.

The same is true in many of the commercial sites I visit on occasion. Only the better zines, certain (rare) corporate sites, and a few news sites are organized for content. Many other sites are interested only in their own version of push technology—they want to push advertising on you, assuming this will make you more likely to visit the advertiser’s site. I can’t argue with this in some regards; companies are smart. They have marketing data. Research shows that their methods work, at least given a large, statistically relevant sample. <sarcasm> The web is the next incarnation of the television after all. </sarcasm> The most annoying examples are, of course, the so-called online-communities (see my complaints about GeoCities above), most of which have simply become ad-whores. No wonder people become bored with the Internet so quickly; for many, the web is the Internet (or worse, AOL is), and the web is to a great extent composed of really boring, poorly-done, brainless drivel. That is not to say that there isn’t a lot of great stuff out there—news groups and message boards are one of the success stories of the Internet, and there are many worthwhile personal pages (a small minority, at least), zines, and such.

Back to design questions. Personal pages are by virtue of their purpose self/author-centered. For people to return to them, they have to be more than that, of course. The tricky ones, the puzzling ones, the multi-layered ones are my favorites (pretty much the opposite of this site, actually) The German Dept. site, however, is not a personal page. It must be visitor-centered. Luckily, Kerry’s version of it already is to a great extent—my work, therefore, won’t be too difficult. HTML is content-oriented, not lay-out oriented. I like that. However, there are times when, despite the merits of HTML, the bare-bones nature of “pure” HTML may not fulfill the visitor-centric needs of a site. And this is not, I believe, an inherent problem of HTML, but merely a question of design, structure, the and creative limits of what good hyper-text can do. Enough rambling for now.

—July 8, 1998

Religious Nuts: Pomona was home to quite a few active Christian groups (Christian Fellowship, Pathfinders, etc.). Luckily they weren’t too annoying and pushy. Now, I am here in Madison (and have been for almost a year), and I have, for some reason, encountered a relatively agressive Asian/Asian-american Christian community. Pomona had its wacko who would (and perhaps still does) stand on the corner of 6th and College, passing out brochures and such until Camp Sec. came along to move him on (the ability to do that was both the blessing and curse of a privately-owned college), but why is it that I am regularly approached by Bible-thumpers on my way down Bascom Hill? Often it is a male-female pair (always Asian/Asian-american, mostly Korean), who just stop me (and others), asking if we have a moment to spare. Then they ask if I would like to talk about Jesus Christ (answer: Christ no! sorry ... I mean, Fuck no!), or if I would like to come to Bible study. Or ... (I thought I had left the missionaries behind in Idaho) A serious and associated question is: can anyone give a satisfactory explanation as to the seemingly large, devout, and evangelism-oriented percentage of Korean-americans? I have several hypotheses (mostly socio-historical in nature), but no definite answers.

—July 8, 1998

I’m on a roll: (and Tom Petty in the background is helping) Okay, third entry today. This isn’t a journal—it’s just random thoughts—so I can do as I like with it (hell, even if it were an online journal, I could do whatever wanted with it). In any case, my current thought centers around webrings. They were a fad recently—perhaps they still are. Online journals are/were a fad. So were animated GIF and blinking text, and “awards” still are, but I digress. I currently belong to one webring—the Exchange Students Webring—and for rather good reasons. Plus, I have the satisfaction of being one of the earlier members (number 6, that is). There are a lot of great webrings out there, and I find them a useful means to finding interesting sites related to a particular topic. In addition, they are a great means for increasing the traffic to your site, and a much more acceptible means than that damn GeoGuide at GeoCities (have you see the mess it makes when the java-script version is used on a page, and java-script is turned off? For all their marketing attempts, you would at least expect GeoCities to be able to implement something in a decent fashion ... or not ... but again, I digress). On the other hand, it’s amazing how many people simply join a webring for the added traffic, making a number of such rings virtually useless. A number of rings have simply begun to become much more self-selecting and closed. In any case, I’ve come across a few rather nice rings the past few days (most are relatively closed, and I doubt my site would cut it), and through them I’ve found some rather nice personal pages. Definitely a cut above the rest. On a side note, it would be nice to have more people visit this site—I know that basically no one does. I took down my counter a while back. It was basically useless as it was, as well as depressing. On the other hand, I’m not sure how willing I am to go publicize this site (I refuse to go the GeoCities route and use that heinous GeoGuide) simply for the sake of getting visitors. I don’t go around signing everyone’s guestbook, begging them to visit my page. To get more traffic, I would probably have to come up with some interesting content, so for now I guess I’ll just have to live with a truly “personal” site—one which only I visit.

—July 8, 1998

Trouble-shooting: Actually, this little adventure began yesterday, but today was the slightly more annoying part. All in all, however, it was all quite fun, plus, I got paid for my time.

Anyway, it starts as follows...

Yesterday I went in to the department (something I haven’t done in a while), primarily to look for some old hardware (ethernet cards and such) left behind by the ex-webmaster, who left last week for Utah. Anyway, nothing there. One secretary asked me to help with a few Quake codes (which I did gladly), and then the main secretary told me that her computer was essentially dead.

Last month she had sent her P100 (and a nearly identical machine belonging to a Prof.) in to be upgraded. P233MMX processor and new motherboard; 64MB RAM upgrade; new, 2.5GB hard drive. The prices paid were far too high, but that’s another issue, and one I don’t care to discuss now, since it is off the main topic. The computers came back from the shop, and the Prof.’s computer worked fine. It still works fine (I had installed Windows 95 on it last semester. Kerry—the ex-webmaster—put a CD-ROM drive in it while I was back in Idaho). They (Kerry and the main secretary) decided to upgrade to Windows 95 on her machine, and to set up the new hard drive as the main hard drive. It crashed. It crashed during installation. And it didn’t want to work. Kerry reformatted the hard drive(s) and tried again. Not much luck. Since he (Kerry) was leaving for Utah in a day, he had to run, and recommended the computer be shipped back to the people who did the upgrade; he figured the motherboard may have been part of the problem.

So, the computer was sent back, the drives were set up correctly, and Windows 95 was installed and “tested” (as for the latter part, I doubt the quality of the “diagnostics”). The machine returned to the office, and it was still crashing on a very regular basis. It was impossible to install any applications: it would crash during the installation procedure(s). They did get the machine hooked up to the network, though (since that required practically no effort).

Skip ahead to yesterday, and enter Steve. I booted the machine, and it seemed to work fine. I tried to install some internet software (in a package known as “WiscWorld”)—crash. It crashed Windows, and I ran scan-disk (in DOS-mode and Windows). In DOS-mode, several corrupted files were found (including *.dll files in the windows\system folder), and that corruption could very well have been causing the problems. Luckily, the hard disk itself was just fine. I needed to reinstall Windows (due to too many corrupted and truncated files). So, I nuked the hard drives (fdisk-ed and formatted them), and got set to do a clean reinstall of Windows (a procedure that usually takes me all of 15 minutes).

Ah, but no! The available Windows95 floppy disks (the “emergency reinstall disks”) didn’t have any CD-ROM drivers on them. There was a disk-only version of Windows95 sitting around somewhere, but I didn’t want to install the first version of 95. I played around, trying to find a CD-ROM driver (I checked out the files on the floppies—these floppies were made for the OEM, not the end user, and the OEM had failed to setup the floppies correctly. Morons). I decided I would go home and come back the next day (today) with my own copy of Windows—I’ve used it, and it’s never given me a problem.

Move ahead one more day. I went in relatively early. I put a game or two on another secretary’s computer, as well as a nice transparency utility. Then I moved on to my main target. I had no problem installing Windows. Got to the final part of the installation, and then things started crashing. Got Windows back up, but the network wouldn’t install correctly. Couldn’t install any programs. I left for lunch.

I came back from lunch and reinstalled Windows again. No real problems this time (had one small crash towards the end of the install...but it was a sign of what was to come). I got the Plus! pack to install, and after fixing an incorrectly installed ethernet card, I got the network login to work. For a moment, I thought I was out of the woods. Then, another crash.

I rebooted Windows. Logged into the network. And ... crash. Repeat. Reboot. Login. Crash. Actually, by this time, the nework login wasn’t even finishing before the system would crash. Annoying? A bit.

I had a hypothesis.

I decided to test things a bit—I had brought my RedHat CD with me that morning. I popped in a CD and floppy, and I was off to the races. The RedHat installation is a breeze. I figured it would take a few minutes to install everything via CD-ROM, so I went to the other secretary and put Plus! on his computer (not too useful, but a few nice, cosmetic changes). A bit later I returned to the computer, only to find a screen filled with garbage. I tried to parse the on-screen text; it seemed that Linux had encountered a problem and had tried to gracefully shut itself down: it failed, and died.

By this point, I had reached a conclusion. Windows 95 was not at fault. This, despite the fact that one of the techs who had worked to upgrade the machine earlier had claimed that the problem was the use of an early version of Windows 95. It sounded like a bogus explanation, though. The question was: was our problem hardware or software in nature? Since Linux was killed, too, the software trail was a dead end. As for hardware, however, the hard drive was out. I had scanned it. The motherboard as such was out, too, since it was also used in the Prof.’s computer, and he had had no problems. Of course, that didn’t (and doesn’t) preclude a problem on the particular motherboard—I’ve encountered dead motherboards before, and they can be a pain to diagnose. So, the most reasonable answer, and one supported by both the Linux crash and the error messages Windows had give me was that some bad SIMMs were installed.

The computers had been sent in to get 64MB new RAM each (for a total of 80 - of for 64, if they simply wanted to return the old 8MB SIMMs to us to use in other machines). That is: 2x8 (old) + 2x32 (new). However, the company working on the computers (and the people who had built them in the first pace—they generally do good work, they’re local, and they get a lot of business from the University) had fucked up. The machines came back with 48MB each (obviously 2x8 (old) + 2x16 (new) = 48 as the only possible configuration), even though we got billed for 128MB total (4x32). No one had bothered to check—I did, since I had decided to check out the BIOS setup to see if there were any problems there. In any case, both machines will be getting sent back for the correct amount of memory in the near future (I would do the installation myself ... an explantion as to why I won’t do it is at the end of this tale).

My conclusion is/was supported by the following: few things kill Linux. Bad memory is one. The Windows errors were mostly page-faults and such—all associated with memory access. Windows would install, but not much more. Let me explain this: Windows can install on a system with 8MB of RAM, and during the installation, Windows does not (probably) have access to (much in the way of) virtual memory. Once Windows-proper gets running, however, it’s a resource hog, and starts swapping as soon as it gets a chance. It is entirely possible to install Windows and skip the bad memory. However, the process of installing a large software package within Windows is almost guaranteed to use up a lot of memory and lead to the attempted access of bad memory.

There is/was a sure-fire way to test this hypothesis further—simply remove the offending SIMMs and see how the system responds. If it doesn’t crash, then you can be pretty certain about the hypothesis. However, there is one problem: the case. The case to this computer is rather nice—completely tool-less. Just snap off the front and pull of the metal casing. However, the location of the SIMM slots makes installing/removing them very difficult without taking out the motherboard, or unhooking the cables, things which I have no intent of doing. Hence, my explanation will go without confirmation, and the machine will get sent back to the shop.

This is the end of the story. I reported my findings to the main secretary. A side note is that the university’s local tech support—DoIT—refused to even touch the machine ... morons (they claimed that they didn’t want to invalidate any warranties ... more like they were too lazy to take a look at it ... they never even looked at the machine). So much for having on-campus tech-support. All-in-all, not a very interesting story. I should have reached this conclusion earlier, but the fact that the machine had, at some point, been working, seemed to indicate that it was mostly a configuration problem and matter of corrupted files - i.e. software oriented.

On a side note, another Prof. has an 8600/200, and the internal (infernal?) Zip drive isn’t working. It hasn’t been working for months—ever since she got the machine. However, she has to rely on DoIT for tech support, and I’ve already mentioned how worthless they are (for more evidence, check out their “store,” where they sell computers for inflated prices). She brought in her external drive from home, so she could back everything up and send the machine off to be repaired. However, she forgot her drivers for it. No problem—just download the necessary files from Iomega’s site. Ah, but here’s the twist: she uses one of the most recent versions of Netscape. Iomega’s site crashes it—it’s the java script. No problem: turn off the java script. Ah, but there’s yet another problem: the download area uses java script (it’s done in frames, and the whole setup integrates java script—without it, you can’t view shit). Finally, with great care we managed to download the files (sure, we could have looked elsewhere, or looked for an ftp site, but should that be necessary? This is an extreme example of really horrible site design. Any site that uses java script (this extensively) and doesn’t provide an alternative, especially when regular frames or a cgi script would have been better, deserves to be boycotted) In addition, this Prof. had a nice run-in with a virus on her machine. The humorous part is that she needed a software update for the virus-protection software (from their web-site), but enabling the software caused a conflict that wouldn’t let her get online. Of course, disable the software, get online, grab the update, enable software, and there you are, but I still find the process a tad humorous. On a final note, my little “browser torture” (the nested frames) totalled Netscape. However, it lasted longer than any other browser I’ve seen—didn’t get to the first "rest area" however.

Okay, that’s it for now.

—July 9, 1998

More Rambling: Last night and this evening have been devoted to browsing various personal pages—mostly good ones (IMHO). In fact, the inferior nature and design of my own site was driven home while visiting these other sites, and I realized that I definitely do not have a future in writing web-pages. I’ll stick with running sites, instead. One in particular that I like belongs to Bill Osborg (his use of tables, while not necessarily original, is very well done, and provides for a very calm and settled feeling), and—as he points out—no matter how good your site is, there is someone out “there that knows more and can humble you in just a few seconds.” I was a tad bit more awed by (the graphic design aspect of) Karawynn’s site, the design of which I found rather appealing, despite my strong dislike of frames and java-script (on the other hand, I know they can be tastefully done as well). I guess the best thing about these pages—at least in terms of their effect—is that they inspire me to rework my own pages. They also remind me that a personal web-site is definitely a process. Hence, even if web-design borders on the artistic, it is/can be quite different than many “final result” oriented forms. At least that is the case if you treat your site as a form of creative expression. I suppose that as far as personal pages go, you almost have to do it that way (unless your only interest is presenting your resumé). If only I were a more creative person.

—July 9, 1998

[ Part 1 | Part 2 | Part 3 | Part 4 ]