Older blog entries for brg (starting at number 11)

I released WebNews 1.2.2. I fixed the race conditions, and threw in a few speed enhancements. By way of fixing News::Newsrc, I learned to write perl modules and use perl's version of class method inheritance, which turned out to be a big win here because I could just override the offending methods without having to hack the module.

Guile has become a much better language for extension than it used to be. Mostly this is because of the manual they have finally written. I really wish they would post it on their web page, even if it's an incomplete version. Right now you have to check it out from CVS. But it's already more than enough to get you started, which makes Guile much better than it used to be. It still has some noticeable startup overhead though.

I've become disenchanted with Tcl lately. I can't understand why arrays (hashtables, for those whose native language is Perl) are not first class objects in standard Tcl. Not only is it the Right Thing to Do, there doesn't seem to be any obvious technical reason why it should not be possible to make that change to the language.

I think I need to learn to use SWIG, as well. Extension languages are a maze of twisty little passages all alike that you can get lost in forever if you're not careful, and SWIG looks like it just might provide a reasonable abstraction over them. Here's hoping.

I released vmips 1.0, finally. Sometimes you have to kick the code out the door, long after it should have left of its own free will.

I have been working on a race condition in WebNews which results in corrupted .newsrc files. I think I have a fix. It's kind of lame that the News::Newsrc perl module doesn't give you an option to set locks on newsrc files. I am contemplating whether to send in a patch.

I have made WebNews work with mod_perl. To the extreme credit of the authors of mod_perl, it didn't take any modifications to WebNews. I added my notes to the WebNews documentation. I created an entry for WebNews on Freshmeat.

A couple more features have squeaked by into vmips, and I am building up a build status page. vmips is now essentially code-complete for its 1.0 release; I will probably not add anything more, but I will quite possibly take things out. I'm not quite ready to branch yet.

I GPLed one of my research codes (vsimII), but I have yet to think of why this would help, because it depends on non-free software. Hopefully I will have an opportunity to GPL other more self-contained codes in the future. I wish academics were granted more freedom to do as they pleased with their intellectual property (sounds silly, doesn't it?) not to mention with their minds in general.

My XML-based procmailrc generator has evolved to a point where it might prove useful, but it still languishes in limbo due to other coworkers' projects that have yet to see the light of day.

Today, my salary goes to the EFF in the form of a membership fee. Joining the EFF is quick and easy, and they take plastic. Click here (SSL) to join.

On the free software front, I have been working on fixing bugs in vmips and its manual with a view to an eventual release.

I released a new version of WebNews, and I worked on the vmips manual. WebNews's speed problems have begun to get serious. I need to develop a benchmark for it so that I can start to do some serious performance tuning. I would be happy to hear suggestions, if anyone out there has any experience benchmarking CGI scripts.

I benchmarked GCC 3.0 prerelease against its predecessors 2.91.66 and 2.95.2 using the Stepanov Abstraction Penalty Benchmark, and the results made me pretty unhappy. I hope it gets better, because some of my code has performance problems that seem to track the slowdown on this benchmark pretty well. At least I can assume that it will get better, because good performance on this benchmark is one of the release criteria for 3.0.

I have begun developing an XML-based procmailrc generator with a web-based editor, for use in secure IMAP server installations.

My lone experience with the SourceForge support system resulted in approximately one month turn-around time for my bug report. (Happily, once someone got to looking at it, they immediately did the right thing, so I can't exactly say I'm disappointed, but neither can I say that I am pleasantly surprised.)

One thing I've been noticing lately about C++ is that you really have to have a recent compiler to get support for anything resembling the standard.

I set up a SourceForge project for vmips, but unfortunately, it doesn't look like SourceForge has much to offer beyond what I have set up at home, with the possible exception of a bug tracking system (and it's not like those are hard to set up) and more bandwidth (which might be useful later, but not now.) So I am just mirroring my web pages there for the moment.

In terms of actual free software work, the only thing I've done lately is work on the embryonic vmips manual. I also made a new release of WebNews on March 15th, with better article sorting support, e-mail carbon copy of posted articles, and more bug fixes. I will probably mirror WebNews's site onto sourceforge eventually as well.

The more I use C++, the more I find I like it. And, speaking of C++, I still like GCC 3.0, which I am actually using on a daily basis now (especially because it has libstdc++-v3 inside(tm)). I wrote a little web page about how to play with GCC 3.0. I highly suggest you try it.

I have been working on the IRAM project, doing benchmark hacking. My free software projects have received a lot less of my time lately, but I'm learning a lot more. After learning about IRAM, going home to my Intel and AMD boxes makes me kind of sad.

I have become very interested in automatic vectorization in C compilers. Someday, it would be nice to see a free software vectorizing C compiler.

The Instructional Software CD was released, without Java.

I have been working on getting GCC, Glibc, GNU Binutils, and the Linux kernel to work together in concert to produce a free-software cross-compilation environment for vmips -- this has meant that I have to use prerelease versions of both GCC and Binutils. I abandoned my earlier idea to go with a NetBSD-based environment. Besides, compiler hacking is fun. I'm pretty happy with the new GCC, and excited about its eventual release as GCC 3.0.

Speaking of reasons not to use non-free software, the Instructional Software CD has hit major speedbumps due to Berkeley instructors' use of the Sun JDK, which we can't redistribute.

I made a conscious decision to spend a lot less time chatting online these days. I'm finding that it's actually rather nice.

I added project pages for some of the projects I am interested in: vmips, WebNews, mutt, and STk. Sometimes I wish there were an "Evangelist" or "Proponent" relation between people and projects. (An "Adapter" or "Customizer" relation might be useful too.)

In many people's experience, romantic relationships are so powerful that they have the capacity to (inter alia) utterly destroy many wonderful things.

You might enjoy reading Fear and Trembling by Søren Kierkegaard, on the power of relationships. I am currently taking an excellent class at Berkeley (from a remarkable teacher, I must say) which includes this book.