RBL, finally

After years of resisting it (too many legit messages dropped) I have finally implemented the RBL on my mail server. It’s gotten absolutely absurd. I’m dropping almost 90% of all incoming email as spam, and there’s still a significant number of spam messages that make it trough every day.

So, that was the first thing I did. Then I added a rule to drop all incoming email from .ru and .kr domains. And finally I added the list at http://www.securitysage.com/files.html of hosts and email addresses that are known spammers.

This is all stuff that postfix will do before it even makes it to SpamAssassin. We’ll see how this affects things. I suppose I don’t have particularly high expectations. But I can already see in the logs that it is dropping some of the connections that I expect it to drop, even before it considers the message itself. So maybe it will actually make a difference.

Oh, yeah, and I discovered one of the reasons that my filters weren’t catching the FLOOD of viagra spam. Turns out that the messages contain HTML comments in the middle of words, causing SpamAssassin to not see the keywords. In fact, the majority of the message was composed of HTML comments. Stuff like vi<!–jksdfoew–>agra for example. So I added a rule to SpamAssassin to look for “viagra” with, optionally, HTML comments between any pair of letters. I haven’t seen viagra spam since then.

It amazes me that these people will go to such lengths to send me email when it is obvious by their actions that they *know* I don’t want to get it. That is, they are intentionally and aggressively violating my wishes.

Software and community

Someone asked at OSCon why Open Source people are so obsessed about community. I answered with my standard answer – that it’s really that community-oriented people are attracted to Open Source – that software is really just a vehicle to get to community, and it is really much more about the community than about the software.

Andy disagreed with me, but then we got distracted, and the conversation died there. I’m really interested in continuing the conversation. So here’s my bit of it.

I joined the Apache project for the software. I stayed for the community. Likewise Perl. The software is interesting, but the people are more interesting. So now that I’m really not even writing much Perl, I’m still involved with the community, to some degree, because they are cool people.

Is there a PhotoShop Community? Well, sort of, but they don’t really have any ownership of their community. They are at the mercy of some large organization of which they are not a part. The Apache Community, on the other hand, has their hands in the thing that drew them together, and can remake it into something cool.

It’s very cool that I’ve had lunch with Larry Wall on a couple different occasions – not that he’d remember me particularly. It is very cool that I’ve chatted with Bradley Kuhn, Eric Raymond, Eric Allman, Tim O’Reilly, and a variety of other people – again, not that any of them would particularly remember me. But all of this is to say that it’s the opportunity to meet interesting people that has been one of the most valuable things about getting involved with Open Source Software. But what’s way cooler than that is the people that I’ve actually gotten to know in all of this – people that I would have *never* run into without OSS, because they live a world away. Folks like Mads Toftum, Greg Stein, Kevin Hemenway, Jesse Vincent, and the list goes on and on. There is no chance I would have ever met these folks in “the real world.”

So it appears, as I think about this, that I’ve often expressed these thoughts, but never really unpacked them completely, as Brother Bourbon would say. I’m very interested in hearing other views, or whatever.

So, why am I involved in Apache? Well, there’s a few reasons. I’m good at something (which happens to be writing about things in terms that beginners can understand) and this makes peoples’ lives easier. Doing something that potentially millions of people will benefit from is very cool, both from the perspective of helping people, but also from the sheer hubris of all of those people thinking I’m cool. And then there’s the community of people that it puts me in touch with. I care a great deal about the Apache Software Foundation and what it does, but I’m not so sure that the Apache HTTP Server is the primary part of what I care about. Presumably, if we had to, we could create that from scratch again. But the community is less replaceable. (Tirade about people destroying community ommitted for your reading pleasure.)

Thoughts?

ssh hostname tab completion in ssh

Because I can never remember, here’s how I did it.

In the file /usr/share/zsh/4.0.4/functions/_host, I removed the line:

hosts=( “$_cache_hosts[@]” )

And added:
hosts=( `perl -pe ‘s/(.*?)[, ].*/$1 /;’ ~/.ssh/known_hosts2 ~/.ssh/known_hosts | egrep -v ‘^[0-9]’ | sort | uniq` )

This allows you to tab complete to all of the hosts in ~/.ssh/known_hosts* – ie, any host that you have ever ssh’ed to before. Ssh’ing to a new host makes it get added to that list, so that it will then be in your tab completion list automagically – nothing to update.

By default, completion is to your /etc/hosts file, which seems moderately useless to me.

DateTime

Thursday night was the Perl DateTime BOF, with Dave Rolsky. It was mostly Dave telling folks about the state of the project, rather than much of a planning session. Dave gave a lot of credit to other folks for the work that has been done, but it is abundantly clear to me that Dave made this thing happen, and has done the bulk of the actual work. Sure, he has taken a lot of existing code into what he has done, but the sheer volume of code that Dave has produced is staggering. Many kudos, Dave.

Meanwhile, if I actually get this book finished any time soon, I should actually have some time to work on DateTime stuff some in the coming months. As usual, I didn’t get as much work done at the conference as I had planned, having been continually distracted by the darned conference. So, you see, gnat, it’s your own fault for putting together such a good conference!

OSCon, day 5

Day 5 was a little bit of a disappointment, really. I think that I had Milton Ngan’s talk built up a little too much in my expectations, so that when he did not show a preview of Return of the King, I felt rather let down. He showed some very cool stuff about Gollum and the Ents, but nothing approaching the amazing stuff that he showed us last year. Apparently he was unable to obtain permission to show us anything of the upcoming movie. I would expec that it would be inthe best interests of the movie makers to have the geek community fired up about their movie, as long as they didn’t show any “spoilers.” Not that you can really show spoilers when the book has been out for 40 years. Sheesh.

George Dyson and John Von Neumann

This talk just rocked. This guy works in the office where Von Neumann used to work (I’m a little fuzzy on the details) on the MANIAC computer. He discovered boxes of notes from that era, and shared then with us. I can’t begin to tell you all about it, but his slides will be on the web RSN. These folks made decisions, out of several options, and we’re still doing everything the way that they decided was the right way. These documents included a substantial number of “let there be light” proclamations, and a great deal of source code for these initial machines.

What is sad, however, is that most of this work was for the purpose of building a bomb, and that Von Neumann died at a very young age of bone cancer from having worked on the bomb.

It’s important to note, however, that Von Neumann did not think of things in these terms, saying something like, “No, I’m not thinking about anything as unimportant as a bomb, I’m thinking of important things. I’m thinking about computers.”

I’ll link to the slides as soon as they become available.

Travel bugs

Ran out between talks and traded some travel bugs, which will now make their way down to Los Angeles. 🙂

Now I’m in the TicketMaster talk. It is somewhat content-free, but I like these guys because they have hired Stas Bekman (a long time ago) and recently hired Geoffrey Young and Ask Bjorn Hansen, under what basically amounts to a patronage. They get to sit at home and work on mod_perl, and not have to worry about things like income and insurance. They also give a lot of money and resources to the Perl foundation and Perl.org.

OSCon, Day 4

So much to say about Thursday. The goal was to work on the book all day, but I ended up in talks for much of that time. Highlights were Geoffrey Young’s talk on Apache-Test, and the Perl DateTime BOF.

I know that if I don’t write about this stuff now, I’ll probably never get back to it – like last year – but I’m very tired and and grumpy and still disappointed about missing my geocaching date this morning, so I think I’ll sit and sulk and work on my book instead.

Wil Wheaton, a geek like me

One of the highlights of yesterday was a trip over to Powell’s technical bookstore, which is the most amazing bookstore I have ever been in. Let’s put it this way: This is the only bookstore that I have *ever* been in which has a better Apache collection than I do. 😉

Anyways, Wil Wheaton was there, reading from, and signing, his new book, “Dancing Barefoot.” I have to admit that I didn’t expect much, but I was very impressed. It’s stories about his time on Star Trek, but, more than that, it’s about being young, and a geek, and unsure of oneself, and trying to deal with the world. He read one story, and an excerpt from another. He’s a very good writer. And he was very good at reading it as well.

I got a signed copy of the book, and have already enjoyed a number of the stories.

See also his remarks about it, and about the conference. He seems like a pretty cool guy.

Geocaching this morning

I feel like a complete idiot, and I’m sure that someone is cursing my existence. I was suposed to meet someone at 5:30 this morning to go caching. I waited for an entire hour for him to show up, and then went to check for his full name, so that I could call his room. I had been waiting in the wrong place. I’m amazingly disappointed, as I was really looking forward to doing some difficult caches. Not to mention having gotten up at an ungodly hour to do it.

The Margin Is Too Narrow