14 August 1999: Well, as usual I've been quite tardy about updating this site. Part of the reason for this is that when I got back from my trip I was immediately thrown into a maelstrom of activities at work.

Recent news of interest: I canceled my slip.net DSL service, and am now going with PacBell ADSL, which is about half the cost. (One disadvantage though is that the upload bandwidth is slower.) I simply couldn't deal with slip.net's poor service any more, especially since they got bought by another company.

I'm currently working on a new hobby project, a civilization-style game called Strattica.

12 May 1999: I'm back from Europe. Lots to do...

5 March 1999: Today I leave for a 10-week vacation in Europe! I'll be updating this site on a regular basis with my travel journal, including pictures.

19 February 1999: Well, for those of you who have been wondering why you haven't been able to access this site for the last week, blame Slip.net. Apparently, when my account got created, somehow they accidentally created two duplicate accounts. Later, when the billing department attempted to "clean up" the error, it resulted in my getting my DSL connection shut off. Fortunately, I only lost about two days worth of mail - on Sunday I was able to go to the ACM.org site and redirect my forwarder to send the mail to a different mailbox.

Last weekend I went to Pantheacon, a San Francisco-based annual convention for people interested in neopaganism. I had a great deal of fun - for example, late Sunday night I got to play in a drumming circle, something I've never done before. (Not that I'm an expert drummer by any means, although I've been tapping my fingers on desktops in interesting rythyms and thereby annoying people for as long as I can remember.) I also bought a lovely hand-painted silk shirt, and went out to lunch with a  gorgeous, intelligent redheaded lass. Definately a lucky day for me...!

There was a very interesting panel on "Paganism and Science". I think that most people in the pagan community believe that pagan beliefs do not, and should not, conflict with our scientific knowledge. Some feel that "magic" is merely science which hasn't been discovered yet. Others believe that the rituals associated with paganism are in fact psychological tools, and the "gods" and "godessess" are in fact roles or archetypes of our own creation. This does not make them any less "real" however.

For myself, I've always found the ethical and mythological elements of paganism to be something I strongly resonate with, but I don't know what to make of some of the more "New Age"-ish aspects of it. Certainly, there are some aspects of what pagans call "magic" that are undeniable - we know that affirmation, role-playing, and psychological mood-manipulation can cause humans to occasionally transcend their ordinary capabilities. We know that symbols have the power to cause long lasting changes in both individuals and society. No one has yet managed to turn this into a science, so I suppose it's reasonable if some folks claim it as an art. But a few of the claims (not many) go beyond the limits of what we understand to be physically possible, and as a result I'm skeptical. I find some of the "quackery", especially some of the products I saw being sold in the convention dealers room (aura photographs and magnetic clothing for example), to be a little disconcerting, especially when the majority of the people at the convention were clearly intelligent, sensible, and more mentally healthy than most of the people I meet in the mundane world.

Another panel which I found interested was on "pagan philosophy". Lots of discussion of free will, good vs. evil, the nature of deity, etc. One thing that I got out of this is that nature is both beautiful and savage, and that both sides are reflected in us, given that we are part of nature. I'm not satisfied simply to reduce the universe into some sort of simple-minded dualism ("Light...good! Black...bad! Ugh!"), but at the same time we have to adhere to some sort of absolute principles if we want to function as a civilization at all.

Trying to pretend that one side or the other of our multiplex nature doesn't exist leads to sickness. David Brin once said about having a large ego: "You can either let it master you, in which case you become an annoying bore, or you can master it and let it become a useful tool." Trying to pretend you don't have it at all merely leads to frustration and an eventual breakdown. I would say the same holds true for other aspects of the human psyche, such as our predatory instincts.

The day after the convention, I decided to go to Lark in the Morning (a wonderful music store near Fisherman's Wharf, they sell every kind of traditional musical instrument you could think of) and buy a drum. I now have a Tibetan-style drum (I forget what it's called), as well as some videotapes on drumming. The only problem is that I can't practice at home because my landlord will freak. Sigh.

30 January 1999: Fixed a few minor bugs in ScanDoc, but I haven't submitted a new version yet (it's in CVS, but not in FTP). Also, some additional bug reports have come in, which I haven't gotten to yet.

I realized that the image map on my main page has never worked with Apache; I just fixed it.

So far I've made $1.98 in Amazon referrals!!!! Waaaahooooo!!!!

I decided to re-organize the main page, mainly because I was bored with it. I may change it some more later.

15 January 1999: Gosh, is it 1999 already? I've submitted an announcement about ScanDoc to freshmeat.net. Let's see how much feedback I get on this. (Several hours later: Wow! Downloads from 45 different countries!) No feedback yet, however...I suspect I'm going to be busy for a while supporting ScanDoc, however I'm hoping that others will want to help.

31 December 1998: Well, the installation of Linux on the Sony Vaio went swimmingly. I have a more extensive report which I've emailed to Jerry Pournelle to post on his site; If he doesn't post it, then maybe I'll put it up here.

Stig came over and helped me tune up my servers. One side effect of this is I now am able to set up so that the outside world can access my server via ftp without poking a zillion holes in my firewall. The hostname is ftp.sylvantech.com. I've already posted the sources for ScanDoc, and the CGI generator programs used on this site. If there are any problems, send mail to me.

28 December 1998: I've been spending pretty much all of my spare time playing with Linux. I just installed the latest version of Gnome, as well as icewm (I tried Enlightenment, but it was too slow.) I've also been reading the various gnome mailing lists, and I've actually been posting on the GUI topic. (Mostly I just lurk.)

I just bought yet another computer to run Linux. (This is becoming like an addiction.) This one's a laptop -- a Sony Vaio 505 FX. It's so small, I simply couldn't resist it. A co-worker had bought one, and I played with it for about half an hour to see if I could actually type on it, with fairly good results. Hopefully today we'll see if Linux can be installed on it.

Christmas was mostly uneventful. I did go to one party, but other than that I haven't done much.

I've also been spending some time attempting to make this site publically available. I can't believe that after all of this time I'm still having DNS problems! Getting the secondary DNS working is the last remaining obstacle to having my servers be available to the world. I managed to figure out how to do anonymous CVS (you have to hack the source - Fred Fish gave me the clue.)

I have added my current Linux and BeOS projects to the current projects page.

15 December 1998: Joe bought me the new Zelda-64 cartridge for Christmas! I see that my spare time is going to be allocated for the next week or so ("week" was optimistic, it turns out.)

2 Dec 1998: Just got back from LosCon (The Los Angeles Science Fiction Convention). I spoke on a couple of panels, which was fun. Met some new friends.

24 November 1998: Well, it seems I have a new job. Sort of. I'm working for a different company, although I still go to the same building and sit at the same desk. The company I was working for, Postlinear Entertainment, really doesn't have much for me to do right now. The project I was working on, 10six, sort of ran off the end of it's contract. As a result, my employers have asked me to do some work for their sister company, Transactor Networks Inc.  I know nothing about e-commerce, howeer they've offered to give me a relatively free hand and so I'm going to give it a try for a few months at least. I'm currently digging into JavaScript, something I haven't looked into up to this point.

I went to the "Open Source and Business" forum at Stanford last week. This was very interesting, and I managed to pick up a copy of Oracle for Linux. (What I'm going to do with this I don't know.) One of the more interesting discussions was on what the relationship between open source and proprietary software should be. One idea which was espoused by Bruce Perens is that we should distinguish between the basic infrastructure which allows for interoperability between programs, and end-user applications. If you can imagine the conceptual space of software programs laid out like a tree, then you can imagine that the infrastructure (in other words, software that connects to many other kinds of software) would be in the center of the tree, whereas applications to perform specialized tasks, and which need only limited and specialized interoperability would be near the periphery. Bruce felt that the infrastructure at the center should always be free and open, whereas it was all right if some of the "leaf nodes" at the edges of the tree were proprietary. (I realized that this model could be taken further, by considering that any successful application eventually turns into infrastructure; For example, Netscape, which was once an application, is now a platform. As leaf-nodes are surrounded by other leaves, they gradually become infrastructure, and should at that point become open source.)

After the meeting, I went to a pizza place with Stig, Mark Stone from O'Reilly and Associates, and a bunch of other Linux hackers and had a really interesting and enjoyable conversation. Mark mentioned one thing that really stuck in my mind: There are now people buying computers specifically so that they can use Amazon.com.

30 October 1998: Well, it seems that I only manage to update this about every six months or so. However, I believe that this may change. For one thing, I now have a full-time internet connection at 384K (a DSL connection from Slip.Net). This will make it very easy to update and maintain the page. For one thing, I plan to use CVS to control access to the pages.

I haven't decided what HTML editing tools I'm going to use. Up to this point I've been maintaining the web as raw text files, using the Alpha editor on my Mac PowerBook. I think that the level of work involved in hand-editing HTML is part of the reason why I've been so lax in updating this site. At the moment, I'm using Netscape Composer (V 4.05), running under Red Hat Linux. It's not the best possible solution. I've used FrontPage, and it's definately easier to use although the HTML it produces is a mess. But I'm resolved that I'm not going to be using any Windows-based software at home if I can possibly help it. I get enough of that at work.

The DSL connection has been much longer process than I had ever expected. I ordered it about 4 months ago. At that time, I was told that there would be a two month wait. Apparently the ISP and their DSL provider, Covad, needed to work things out with the phone company and check out my phone line to see if it could support the connection. After two months had passed without hearing from them, they said that a guy from Covad would come out next week. The guy came, and set up the connection, or at least most of it. Apparently, there was something going on with my phone line where Pac Bell needed to do some work on it before it could go through. The Covad guy also left behind a Diamond Lane ADSL router. A week later, the Pac Bell people came out (I wasn't here but the landlord let them in) and did some work, and said that the Covad people needed to do more work. This went back and forth for about three or four weeks. I, of course, was calling Slip.Net several times a week, most of it playing phone tag, trying to find out what was going on. Eventually they got it straightened out and the line started working as of one week ago.

My friend Ash, who's a professional network consultant, helped me set up a firewall box using Red Hat and a spare Gateway P90 we had (Joe and I tend to have lots of old hardware lying around.) Now all I need is for Slip.Net to supply DNS services. I finally got a call from them today, it seems that the person I was talking to before no longer works there.

My plan is to set up a server and use them in a number of creative projects. For one thing, I'd like to host one or more open-source projects. I've been really excited about open source development ever since I read Eric Raymond's paper, and I also met a lot of cool people at Open Source developer day down in San Jose. (I should remember to post the trip report here, although you can read it on Jerry Pournelle's web site.)

I also plan to experiment with these pages. Having access to Apache and being able to install modules such as PHP, as well as set up my own database, will give me a lot of options in terms of making these pages more interesting.

Work-related: My current project at work is 106: The Million Player Game. You can see some screenshots on the Postlinear web site. The game is kind of a first-person Command and Conquer in a massively-multiplayer persistent world.

Personally, I feel that the choice of the name is rather unfortunate. If, for some reason, we find that we cannot actually support a million players (which seems likely), it will be far too late to go back and change the name of the game to be more accurate.

However, if we ignore the hype for a moment, things are actually going rather well. We have tested the game with up to 40 real users (which is all we have), and up to 400 simulated users. This is on two servers sharing a common Oracle database.

Now, you might ask "what do you mean by 'simulated' users?" Well, the fact is that in order to load test the server, we need to have lots of simultaneous connections going on. However, we also need to be able to control the amount of load at any given moment. For example, suppose the server crashes - we want to be able to disconnect all users, fix the problem, and then slowly ramp up the number of logins, while at the same time monitoring the performance characteristics of the server. Unforunately, there's no way we can do this with 400 real people, we simply can't coordinate that many people, and we couldn't afford to have that many people available on call.

So instead, Joe Pearce has written a Java version of the 106 client program called JavaBot. This allows a single PC workstation to simulate up to 30 simultaneous users. The 'bot currently simulates a player logging in, moving around, teleporting to different regions, etc. Whenever we want to do a load test, we fire up about 20 copies of JavaBot on various machines around the office (of which there are plenty, now that Vigilance has shipped).

When we first tried the 'bot, the server hit a wall at around 56 users. After a little head-scratching, we realized we had forgotten to raise the limit on the number of filehandles for the server, which was set to 64. (Remember, on Unix a network socket counts as a file handle.) Then we hit another wall at 120, which was quickly solved. Then 152. Then 240. Then 300. And finally, after a few days, 403, which is the current record.

This technique has really saved our butts. Apparently, there were many who didn't believe that we could adequately tackle the server scalability issues. I admit that I had doubts myself, until recently.

3 June 1998: Well, the role-playing game I mentioned earlier collapsed, although the reasons for this had nothing to do with the game itself. It had more to do with the inability of some of the players to attend.

Recent books: I recently read Gael Baudino's "Strands of Starlight" series. I had read the original book about eight years ago, and I was affected by it quite deeply -- so much, in fact, that later when the sequels came out I let them sit on the shelf for five years before I could work up the nerve to read them! However, while I enjoyed the sequels, they did not move me as powefully as did the original experience, although the last book "Strands of Sunlight" I liked quite a bit.

17 October 1997: Last week, I game mastered a role-playing game with my friends Mark, Leo and Rebecca. This was like the first time in 15 years that I've run a game, and I had a great time doing it. The rules system I used was FUDGE, a freely-available rules system.

I've also been working on MeV (my music sequencer for the BeOS, which I'll tell you about later).

I recently gave a talk for the CGDA (Computer Game Developers Association, local chapter) on "Aspects of Community Design". I'm becoming very interested in the anthropological aspects of online communities. More on this later too.

I've recently been re-reading a lot of my older favorite science fiction and fantasy novels. I just finished re-reading Marion Zimmer Bradley's The Mists of Avalon and I'm glad I did. It's been at least 10 years since I read it, and I've changed enough that the re-reading was a substantially different experience for me than the original reading. I recognized a number of interesting points that I would not have been able to detect at an earlier age. There are very few books that can do this trick of being a different and deeper experience each time they are read. The Lord of the Rings is the only other one I know of.

For example, the first time I read the book I was totally sympathetic with the pagan protagonist, Morgaine, in her struggle against the intolerant Christians. However, since then I've been doing a lot of studying of early religious beliefs, both Pagan and Christian, and my views about people don't fall into such nice, simple categories as they used to. I have come to the conclusion that while the Christian characters in the book were quite bigoted, Morgaine's attitudes were somewhat unfair as well. The Christians in the story believe that all women carry the seed of evil because Eve brought sin into the world; Morgaine never seems to tire of using this against them, to show how anti-nature and anti-woman they are. But the truth is that the story of Eve implies absolutely nothing about the nature of women in general, and only implies something about the nature of one particular woman. And while the bigots on either side don't seem to realize this, there are some Pagans as well as Christians who don't try to use their doctrine as a club against the other.

In the end of the book, Morgaine discovers that the Goddess's purpose was not to fight Christianity, but to transform it. We see signs of the Goddess's presence in the statue of the Virgin Mary, just like ol' Joseph Campbell has been telling us all along.

There is another way in which this work of Bradley's reminds me of Tolkien. I've seen Bradley at conventions, and read her other books, and I've read Tolkien's letters, and I would say that in both cases these two works transcend their authors. It's almost as if they were channeling some greater, wiser spirit, one that was not so limited in viewpoint or narrow in prejudices. (Not that I believe in channeling or spirits, you understand...or rather, I think that one can adjust one's perception of the world where such things do exist and are useful, but that's a different kind of reality than a bullet.)

 How can you seek knowledge of the spirit...
...when you don't believe in spirits?

 How can you seek guidance...
...when you don't trust the guides?

 How can you engage in ritual...
...when you abhor repetition?

 How can you mediate...
..when you have an itch in the middle of your back?

 Have a nice day!

24 July 1997: Well, as you can see it's been quite a while since I posted anything here. Mainly, that is because the site has not been up for almost six months. The computer which was hosting my web page has been shut down and disconnected from the net (along with the rest of the company where it was located). However, I now have a new site, a new domain name, and a new net connection. The site is now running on a 486 box which I own, and is running under Linux, which is highly confusing at times but seems to do the job. (As annoying as Linux is, it's still better than Win95!)

 Here's what's been happening: I have moved up to San Francisco, where I got a job with PostLinear Entertainment. PostLinear makes internet games, and a technology called Transactor which facilitates digital commerce within internet games. I'm basically functioning as a toolsmith, developing technologies and tools which will be useful to the project teams. I decided to take the job rather than looking for investors for my business plan, mainly because I was running out of money to live on (One of the few times in my life when I've actually been motivated by money.)

 This month I'm proud to announce (i.e. brag) that I have two articles in major magazines. One is in WIRED, called "Conspicuous Conservation" in the Idees Fortes section. The other is in Communications of the ACM and is called "The Power of Negative Thinking." I've been writing a number of other essays as well, I hope to be able to post them here soon.

20 January 1997: I just finished writing my first Photoshop plug-in. It's called CrossFade and it makes cyclic textures, such as those used by web pages. Everything's finished except for the part having to do with shareware registration. I'm not quite sure how to go about this, especially the part having to do with generating registration codes that are relatively unhackable. Once I get that figured out I'm going to post it on the net.

 I've also been starting work on writing a business plan. I have a lot of really cool ideas for projects that I want to build, things that I think will do really well in the marketplace, but I need to get some sort of outside investment. If I just spend the next two years implementing these ideas myself, I think it will be too late. Many times I've come up with what I thought was a great idea, and have been prevented from working on it (either through lack of resources, or prior commitments), and then later find that someone else has done the same thing and made money off of it.

 In the mean time, I'm going to look around for various short-term contracting work. I don't want to get too distracted from my main goal (the universal online game front-end and server) but I need to build up some resources.

12 January 1997: I attended MacWorld expo and made a number of important contacts. I met with Ben Wen of Think Fish, Inc., which makes a very cool real-time renderer called "Picasso". This product can render "cartoon" style animations (with black outlines and everything) in real-time. The images are so lively and personable, they make photorealistic computer graphics look like dead plastic. I need to get him together with Bruce Naylor (the guy that invented BSP trees).

 I also met with some folks from Metrowerks, who might be interested in my documentation processor, ScanDoc. This is similar to Javadoc, but works on C++ files (and produces output that's a lot prettier.)

 After MacWorld, I attended the Be Developers conference, which was great fun, and I met a lot of the people I have been talking with on the BeDevTalk mailing list. They showed a lot of really cool demos and talked about the exciting changes they are making to the Be operating system. I got to talk with a lot of the engineers and discuss ideas with them.

 I also picked up a copy of the Photoshop SDK. I've got a couple of ideas for plug-ins I want to try out.

 Hey, I made some money! My friend, Kee Nethery of Kagi Shareware asked me to write a bar-code scanner for him that would read the fax files from the Global Village FaxModem, and translate the bar codes into ASCII text. I spent about 32 hours on the problem and created an Apple-Script driven application (using PowerPlant for the Applescript stuff, libtiff to read the faxes) which seems to work pretty well. The hardest part about reading the barcode was dealing with all the noise and dropped pixels on the faxes (not to mention the fact that the inkjet printers printing the barcode tend to bleed ink a little bit, causing the black bars to be thicker than the spec), but I came up with a heuristic which seems to work pretty well. Now Kee will be able to directly transfer the information from the faxes into his order processing system without having to print out the faxes and scan them with a hand-scanner.

25 December 1996: Recently discovered Electric Minds which is Howard Rheingold's new web-based BBS (sort of an updated version of The Well). The people there are very supportive and friendly, more so than other conferencing systems that I have been on. It seems like a place where I can comfortably wax philosophical, which is something I don't have much of an audience for in real life.

Last updated: 
Back to Talin's main page