Posts for March 2003

2003-03-15: Initial Entry

No one can, as yet, see this, unless you play games with HTTP requests. But sooner or later everyone will be able to. *grin*

I have now convinced MT that it really doesn't like .cgi extensions, but I've not yet convinced it to use adequate-sized fonts. I have the font size increased by two notches in Mozilla to just be able to see what I'm writing. I'm sure that this will be very customizable in the entry view, but I'm not sure how customizable the editing screen is. But I expect to enter most all of my entries via XEmacs anyway, so I may not care.

I'm clearly going to have to do a lot of work on the presentation style of this software, though, because it's all way too small and not at all the style that I like.

2003-03-15: XEmacs

This is my first update made from inside XEmacs. Let's see if this works correctly. *grin*. I'm using the stuff at to do it. It's mostly just an entry poster and editor, but it looks like it can also do some template editing. The only problem is that it can't assign titles, so I have to go back to the MT interface to fix the title. I'll have to play with that some more.

I think I now have a layout for the journal that I like, although I chose a bunch of colors and I'm making extensive use of style sheets, and of course I've only checked any of it in Mozilla. So it may look horrible in IE. I'll have to check that or get someone else to. I also haven't touched a lot of the templates that are for things I'm not really using yet, and I'd still like to rewrite the XHTML in all of the templates to not be quite so ugly. But that can all wait.

I'm staying up tonight doing this in part because I've finally come down fully with the cold that's been threatening to attack for some time now. Of course, it would do that just in time for the weekend rather than waiting until the week so that I could get some time off work. And I had a lot of things I wanted to get done this weekend. Well, hopefully it won't be too bad and I can mostly ignore it.

2003-03-15: Done fiddling for a while

I found a new, even better Emacs client for Movable Type that can actually handle categories and other useful things, so I'm using that now. We'll see how this works. I've also now fiddled with the layout and style sheets quite a bit and things should now look fairly well in both IE and Opera. Thanks, piranha!

The lack of standardized support for CSS really annoys me every time I run into it. I love writing CSS and playing with page layout right up to the point where I have to try it or get it to work in something other than Mozilla and then I just get infuriated at how poorly other browsers implement the standard. The viewing bug in Opera 6 just isn't going to get fixed since it's minor and I can't find a good way of correcting it.

The cold is much better today, although I didn't get very much sleep. I'm currently contemplating whether I'm going to try to get things done or just relax. The latter would probably be the best idea, but there are several things I was really hoping to get done today.

2003-03-16: New day, new layout fixes

More minor this time, though. Just making more things consistent with the fixes that I put in last time, getting rid of a failed experiment in changing link colors, and fixing the comments preview.

On the agenda for today: verifying some n.a.n software changes that I made yesterday and then reviewing NNTP working group drafts (belatedly; in the future, I need to get on that quicker).

Let's see if the XEmacs code to set entry topics works better this time, now that I've updated my xml-rpc.el library.

2003-03-16: Reviewed NNTP drafts

I finished a thorough read of all of the current NNTP protocol draft and sent all of my comments to the mailing list. Now that I'm the working group co-chair, I really need to get to these sorts of tasks sooner than I did; that draft was published two weeks ago.

I think we're in really good shape. There are a few nits to clean up, and we have to decide whether to standardize LIST OVERVIEW.FMT, but other than that and a few more things to clean up, I think we're ready to publish. We should be able to make our working group deadline to finish next month, and then implementors can work from an accurate NNTP standard. Not complete yet, since we've not yet standardized streaming or authentication (those are the really big extensions to tackle next), but it will be a considerable improvement over RFC 977.

I also reviewed the first TLS draft and the discussion of it so far, and I think we've pretty much reached consensus on the issues identified so far. INN will need a little work to be compliant with it, but not too much, and it looks like the document will be tight and small and pretty straightforward. Those are the best kinds of standards. Jeffrey Vinocur has done an excellent job carrying that draft forward.

2003-03-18: Meetings and teeth

Yesterday was sort of a null day, during which I accomplished very little beyond worry about lingering cold sensitivity from having my upper wisdom teeth removed. I finally called that morning to check with the surgeon and they didn't think that was normal, so I was worrying there was decay or something more seriously wrong.

This morning, I got in to see the dentist, and found that there was nothing wrong at all, just more of the nearby teeth exposed. Apparently, teeth undergo a process very similar to oxidation in metals and develop a protective layer after having been exposed to food and saliva, and the newly exposed tooth area hasn't developed that layer yet. So it's nothing, and the sensitivity will fade with time as that layer develops and the gum fills in again.

Given how difficult it's been to find information about wisdom tooth extraction that isn't for teenagers and that only involves the upper teeth (which is a completely different matter than the lower ones), I'm tempted to put up a web page about my experiences. Might help someone else who wants as much information as they can get their hands on, like I do.

Today was the every-other-Tuesday solid meeting day, after the morning dentist appointment, so I'm completely exhausted. I was pretty much in meetings or talking to people about things related to the meetings from 10:30 to 18:15 with just a few breaks to read some of my e-mail. But it looks like my various deliverables for spring break are proceeding well, as are the rest of what my group is working on, so that's a good thing. Now I just have to get through a bunch more meetings tomorrow and the rest of the week will look much nicer.

Maybe I should bow out of the meeting tomorrow on choosing a Linux distribution. Walking all the way across campus and then back for another meeting burns a lot of time that I could be using to try to figure out what the current status of the WebAuth LDAP support is.

2003-03-18: USEFOR progress

Finally. Andrew Gierth and Pete Resnick are taking over as chairs of the working group and posted their road map for going forward, which involves splitting the current monster draft into four documents.

One will specify the article format, akin to RFC 2822. Another will specify the interpretation of the messages on the wire, the roles of various agents in the process, and all the other stuff that goes into Usenet like relaying posts to moderators and processing control messages. Both of those documents will be standards-track.

The third document will be an informational RFC about best practices for software, moving all the "Ought" stuff out of the current draft. This is an excellent move.

The fourth document will be an Experimental protocol for supporting internationalized newsgroup names, thereby getting the single most controversial proposal out of the critical path into its own document where proponents of UTF-8 can propose their solution without it necessarily being the only permissible solution. This is excellent.

I'm very, very happy. I really feel like the first two documents, which are the ones that I'm really interested in as an implementor, have a real chance of getting done now without including a bunch of unnecessary, untested cruft or getting bogged down in endless arguments about new work.

The timeframe is very aggressive, and I want to find as much time as I can to help. We might actually get both a new NNTP standard and a new article format standard by the end of the year.

2003-03-19: TeX and War

Today was devoted to tracking down the latest versions of the various fonts, macros, and other packages that were installed in Stanford's previous TeX installation and re-installing them in the new teTeX installation that I've been working on. I made excellent progress, too; I'm almost done with everything except CJKV (Asian language support) and some of the utility programs. We definitely should be able to make the spring break upgrade.

In other news, the new directory server may be available for mail routing on Monday (or shortly thereafter), but it won't be available for WebAuth support for a little while yet. WebAuth 3.1 is therefore delayed until it is, so I'm not working on its documentation and instead will spend more time working on TeX.

I have all the pieces for a new INN STABLE release and just need to drop them on the FTP site and send out the announcement. Maybe tomorrow.

War has sort of begun but not really yet. CNN is a joke, as usual, but still occasionally entertaining. The funniest moment of the day was watching CNNfn, which had turned into CNN Europe or some such thing. CNN Europe was following the regular CNN feed, which switched to a press conference in Iraq. Then they cut away from it, so CNN Europe cut away from CNN USA to go back to the news conference. Then they rejoined CNN USA when CNN USA picked up the press conference again... and then the whole thing happened again.

Only CNN would cut away from CNN for breaking news. It verges on being a parody of itself.

I really wish I could watch the CBC. I probably could watch the BBC, but it's a high-numbered digital cable channel, so it would disturb my TiVo. Which is frankly a lot more important than watching a bunch of reporters say that we don't know anything new for hours on end.

2003-03-20: INN 2.3.5

INN 2.3.5 has been released. It took a lot longer than I had been intending to get it out, due to getting distracted by random other things going on. This is a small bug-fix release, knocking off a couple of annoying bugs in INN 2.3.4 that cropped up right after we released it.

See the release notes for all the details.

INN 2.4.0 is still pending me finding some time to track down what broke about makehistory. Hopefully that's the last major bug.

2003-03-21: CJK font fun

Now I remember why I hate doing this so much. Every time I try to install Asian fonts for TeX, something doesn't work the way that it worked the last time I did it. Installing the CJK package works well enough, and is quite straightforward, but it tends to expect fonts that are nowhere to be seen on CTAN. At the least, the CTAN:/fonts/CJK directory never gets updated.

I've even lost track now of where I got the TrueType Chinese fonts that I already had, but they're apparently no longer the right ones. I've now finally located the Arphic fonts, which appear to be what everyone wants to use, via a lot of Googling on the fonts that the CJK TeX documentation source required. There are no mention of these fonts at all in the documentation proper, just in the TeX source code for it.

I've already decided to give up on Thai support for right now, since the pointer in the CJK documentation is to a resource that flatly doesn't exist. I found the SRPM (ick) that appeared to be the closest possible match, and I think it has the stuff that I want in it, but I failed to figure out how to properly extract it. Of course, there appears to not be any reasonable upstream source listed; heaven forbid that people who package things mention where they got the stuff that they packaged originally or provide it in some usable form for people who aren't using some Linux distribution.

And as usual, one of the innumerable little utilities that one has to have to get all this to work right has completely changed all of its options, configuration, and calling method. This happens every time I go to upgrade the CJK TeX installation. This time, it's ttf2pk, which has now spun off a separate ttf2tfm program which is supposed to be called with a variety of parameters that I don't know, and which bear no resemblence to the parameters that the old version of ttf2pk used. *mutter*

I'm currently taking apart Debian source packages to figure out how Debian solves these problems, and then reusing those techniques. This seems to be a promising approach, but this is all taking far longer than I was hoping.

So annoying.

2003-03-23: Geneva convention

Free hint to the Bush administration: if you're going to invoke the Geneva convention to complain about Iraqi television showing captured American soldiers, you might want to do something about the fact that US television has been showing tons of footage of Iraqi soldiers surrendering, being captured, and being marched away. Otherwise, to anyone who's paying attention, you just look utterly hypocritical and manipulative. Yes, Iraqi television went a step farther by actually interviewing people, but the Geneva convention is quite clear and what the US has been doing is quite definitely in violation of the exact same section.

There are many areas of tragedy in this war, but one that I've been thinking about quite a bit are the common Iraqi soldiers. A bunch of conscripts who had no choice in fighting for Saddam Hussein, they've been put on the field against a vastly superior force with little food, no pay, and no training. Some of them are probably just scared. Some of them are probably trying to do their best in a doomed cause to defend their country against a foreign invasion, just like many Americans would regardless of their opinions of the current political leadership.

And then, when they do face the choice of fighting to the death or surrendering to overwhelming forces, they're publically humiliated, treated like common criminals, and shown on international satellite television as a propaganda technique by the US government to demoralize their friends and neighbors. Some of the treatment is necessary, since soldiers can and have used surrendering as a ploy to lure the enemy into an ambush, but the international broadcast is completely unnecessary and cruel. Just ask Rumsfeld, who had quite strong opinions about this when the Iraqis responded in kind.

While we're in the middle of a war, it is obviously necessary to treat any Iraqi military as the enemy in an immediate tactical sense. But I wonder what will happen after the war is over. Will we start treating the Iraqi population, including those who were conscripted into the military, with respect and dignity and give them power and control over their own fate and help them feel empowered to rebuild their country along whatever lines that they want, or will we impose a foreign military dictatorship because we know what's better for them and haul unknown numbers of captured troops off to prison camps for arbitrary amounts of time? Will we create yet another generation of people with a deep and justified hatred and resentment of the United States and a need to avenge their humiliation in any way that they can?

Perhaps the Bush administration should be giving some thought to that, rather than attempting to cynically manipulate the Geneva convention for propaganda purposes.

2003-03-24: Domesticity

Spent the day catching up on various domestic chores. I remain not particularly good at taking care of those in little bits on an ongoing basis, and tend to put things off until they bother me and then do a lot of work at once. I suppose an argument could be made that it's more efficient that way, and I could pretend that's why I do it.

Spent much of the day updating The Agonist, which is far and away the best source of war information right now. I donated some money to help pay his bandwidth bill, which must be huge (had to use PayPal to do it -- ick). CNN continues to be worthless for real information. The alternate CNN (CNN Europe, I think) that CNNfn changes to in the evenings is considerably better. I got dragged into a bit of war discussion, but then extricated myself again. I'm really not in the right mood to get into an extended debate over the war; right now, I'm primarily in an information-gathering mode, and wanting to really know what's going on rather than just getting the stuff that CNN thinks makes them look good enough to bother covering.

Other than that, I did little. I did compile a bit more software for quarter break pubsw upgrades, though, and will have to start prepping those tomorrow. Hopefully I can get them released on Wednesday and then take it a bit easy for the rest of the week, working on my own projects.

2003-03-25: cvslog 1.44

Prompted by some e-mail, I finally sat down and figured out what to do about diffstat and diff output from binary files, and did something that's kind of a hack but which should at least cause people not to get errors when they commit. I just put <<Binary file>> in both the diffstat and diff output.

2003-03-26: Term::ANSIColor 1.07

Just documentation changes, updating the URL for the ECMA standard for formatting codes and adding information about Windows telnet, OpenSSH under Cygwin, and PuTTY.

2003-03-26: pubsw upgrades

The quarter-break pubsw upgrades have gone out. That's been my work for today and yesterday. 21 packages in total, including two new packages, 18 upgrades, and one rebuild of the same version (but making it available in our Linux tree and writing a wrap file for it). Now that I have wrap files for most everything, it's really easy to upgrade lots of packages very quickly; I didn't start working on upgrades until last week.

There's still some minor cleanup to do and a few other things I need to work on, but that finishes off most of what I've been working hard on lately, and should mean that the day job will let up noticably.

2003-03-27: INN 2.4 getting closer

Today I finally found time to sit down and really track down the reported bugs in makehistory in the INN CURRENT tree, and with the help of Purify I think I tracked it all down. And found various other bugs in makehistory for tradspool; I wonder how it was working before. Anyway, lots of problems with the recovery tools should now be fixed.

This means that the only thing I have left on my real to-do list before the INN 2.4 release, provided that no more bugs crop up, is to install INN's header files. The interface is unstable still, but we should note that in the documentation rather than not installing the header files as it just makes matters more difficult for packages that need to build against INN's libraries.

I'm hoping to find time this weekend to take care of the last few issues.

2003-03-27: The price of civilization

I've been thinking about this topic a lot lately, with respect to several different things.

I believe that there is an inherent price to civilization, to respect for human rights and the rule of law: letting the other person take the first shot. And even more than that, once the other person has taken the first shot, measuring one's response to respect human rights as much as possible and acting with the minimum amount of force required to resolve the situation.

This is a very real price, because the first shot can injure or kill people. The measured response can give the other person a chance to repeat their crime, or even escalate. And yet, I think that this is a requirement of civilization, that being civilized is actually more important than human life.

It doesn't feel to me like US residents always think of this that way, even if you replace the word "civilization" with "freedom" as is common in the United States (without really that much change in meaning for this discussion). Sure, if phrased as "would you die for freedom," this is a strong part of US culture and the local spin on history, and in that context people are quick to say that freedom is more important than human life. But when faced with the harder question of dealing with a potential threat to freedom that already exists, the answer seems to change.

It is more important to not attack another without provocation than it is to ensure that they don't hurt you. It is more important to follow the laws of civilization than it is to prevent crime or death. We know that, at least in the short term, simply summarily executing anyone arrested for a crime would drastically lower the crime rate. We don't do that because that's not the way that we want to lower the crime rate. Because there are other things more important than less crime.

This country will have to learn to apply this to terrorism as well. There are more important things than preventing terrorism. There are standards to which we should hold ourselves that mean that acts like 9/11 cannot always be prevented. They will happen again. More people will die. And in part that will be true because we want to retain our freedom. Because we want to continue to be civilized, and not descend into a cycle of suspicion and violence that could well prevent attacks, but which would lead to other, even worse things.

The price of civilization is holding ourselves to a higher standard. The price of civilization is to make it more difficult to stop evil because we refuse to be evil. The price of civilization is to not extend self-defense to possible threats, only clear and immediate danger. I hope the United States is willing to pay it.

2003-03-29: spin 1.27

Finally finished a modification that I've been wanting to make to spin for quite some time. It now looks for a .sitemap file at the top of the source tree and uses that to figure out the structure of the web space for navigation links at the bottoms of each page. It doesn't, as yet, generate an HTML sitemap, but that should now be fairly easy to do, and I think I now know how to get it to pick up changes in page structure automatically and regenerate the appropriate pages (although that may require some restructuring to other parts of the code).

This is the first in a series of changes that I want to make to publish some more information on my web pages. The next is to have a file that tracks the version numbers of all of my released packages so that I don't have to edit the web pages by hand when I release a new version, and so that I can easily include the version number next to download links for scripts.

2003-03-30: podlators 1.26

Only change is to add "1;" to the end of the module. Apparently require in Perl 5.8.0 is pickier about modules returning a true status and got confused by Pod::Man under some circumstances. (I think this is a dumb thing for use/require to require and never did understand the point, but ah well. It was an easy enough fix.)

2003-03-30: Beautiful breeze

It's just a beautiful day today. Not too hot, with a lovely breeze. I have nearly all the windows open in my apartment for the first time this year, and left the bedroom windows open last night for the first night this year. I probably could have done that a few days ago, but last night was lovely. This is also the first day that I've worn shorts.

I really like spring at Stanford, even though the campus has some minor insect problems (oak worms, primarily, with a side helping of bees). It stays a nice, reasonable temperature and there's generally a nice breeze.

On the agenda today are such things as catching up on various bits of e-mail that I need to act on, catching up on personal correspondance, and hopefully getting a bit of house cleaning and random chores done.

2003-03-30: War links

Here are a few things that I wish more people would read and really think about.

From the political angle of what's behind the decision to go to war in the first place: Noam Chomsky on the underlying motivations. I've seen independent confirmation of both the proposed new foreign policy doctrine (Richard Pearle is the most frequent public proponent), and of the poll figures cited here giving the success of government and media propaganda at connecting Iraq with 9/11. A majority of Americans believe that "many or most" of the 9/11 hijackers were Iraqi (not a single one of them was; they were almost all Saudi).

From the operational angle of how the war is being conducted: a New Yorker article about Rumsfeld and the way that he's choosing to run this war. The degree of arrogance and self-certainty of the people that Bush has surrounded himself with is frightening, and they're proving to be spectacularly, catastrophically wrong. I have the feeling that Colin Powell attempted to warn them.

Regardless of one's position on the war, Rumsfeld deserves and needs to be thrown out on his ear. He's simply too arrogant, dangerous, and blinded by his own propaganda to be permitted to remain in control of a significant government department. He's getting US soldiers killed, unnecessarily, purely for political reasons, and that's something that both pro-war and anti-war people (and all the people in the middle) should be able to agree is unacceptable.

2003-03-31: NNTP working group catchup

I'm now caught up on the NNTP working group traffic again, and have posted some general information and some questions about how we should handle LIST OVERVIEW.FMT going forward to Hopefully we'll get enough information to decide that question soon. I'm still in favor of not standardizing it, but maybe someone will come forward who actually uses it.

I think the only major unresolved questions left are whether to keep the special language for using <0> to handle messages without a message ID, the LIST OVERVIEW.FMT thing, and how to handle 480 and 483 (for groups that need a privacy layer) in the basic specification. We should be able to make our April target.

Last modified and spun 2017-02-20