wonkablog

I’m still getting overdue stuff taking care of, this time in Planet Larry. I’ve finally cleaned up the scripts and the configs to make it much, much simpler to debug and run the update process. Not only that, now it runs about 15 times faster (under one minute, actually).

Something interesting I realized the other day is that Universe gets about 5% of the traffic that Planet does. As a result, I changed it from updating twice an hour to once every 2 hours. Planet, on the other hand, I changed it from twice an hour to every fifteen minutes, and I might make it update even more frequently than that.

Anyway, all that’s changed is a bunch of boring backend stuff, which was some much needed cleanup that I’ve been putting off for a very, very long time. The fact that planet has been a maintenance nightmare is what has mostly kept me from dragging my feet on getting anything done, but now that that headache is gone, I feel hugely relieved.

There’s still more stuff to do, but I think I’m over the worst of it. What I’d like most is to work on next is getting the actual frontend cleaned up, and get a new design. That’s not really my thing, so if someone can help out, that’d be much appreciated.

I took a few minutes today and cleaned up a few small bugs on Planet Larry and friends.

One thing I get asked for every now and then is if I have archives of past posts. Well, I do now. I just copy the HTML file of the last post to $date one each run. A simple and unelegant solution. I’ll be doing the same thing for Planet Gentoo soon.

Also, fixed the FeedBurner link on the main page — I didn’t even realize it was broken. While I was at it, I created one for Universe as well.

On the packages website, I finally fixed it so you can search against just packages again. That’s been annoying me for a while. By default the search is way too wide, I think. It will search the full atom, the package description and the package name. I have to do the package name twice because of regular expressions (starting with, ending with, exact matches, etc.). And there’s still no simple way to search for packages containing multiple words, which is also an annoying little bug. Advanced searches for GPNL and Packages has been something I’ve wanted to do for a long time, but have been putting off since I started the projects. Sheesh. Every time I sit down and start to poke at it, though, I realize just how big a beast it is, based on what I’d like to accomplish. I really need something for the interim, though.

Anyway, I better quit before this post gets any more boring. One last thing — we can use more users who are Gentoo users and have a blog on Planet Larry. Just drop me an e-mail and I’ll get you setup.

I’ve made an awkward and difficult decision that I hope doesn’t make anyone feel bad: effective immediately I’ve removed any feeds from Planet Larry / Larry the Universe that were from developers who have retired from the Gentoo project.

I setup the planet feeds mainly so that regular users of Gentoo Linux could talk about and share their experiences, and I want to get back to keeping it with them primarily in mind. I tend to think that ex-developers have more weight with their posts, which tends to cause some unbalance that I was never comfortable with.

Speaking of users, I am always looking for new bloggers to get added, so if you are a Gentoo user who blogs about computer experiences, be sure to sign up.

That’s all for now.

Edit: I originally took this post down, and am now restoring it. I still haven’t decided if this is a permanent change or not. I might create a planet just for ex-developers, or reintegrate them somehow. I dunno. Comments and suggestions are welcome.

I’ve just updated my apache config, and larrythecow.org and planetlarry.org now have a permanent redirect to http://planet.larrythecow.org/

There’s no need to update your RSS readers if it handles redirects okay, but eventually the content for the main domain name will change — not anytime soon, though, that’s for sure, I have zero plans right now. Even when I do change www.larrythecow.org to something else, I’ll still have the RSS feeds redirect so as not to break anyone’s feed.

Anyway, you’ll see why I’m moving to a subdomain here in a day or so.

Edit: More spring cleaning, now Universe has its own subdomain as well: http://universe.larrythecow.org/

The time has come to make some major changes to Planet Larry, to keep things simplified and make life easier on Alex and I as well. Lots of stuff has changed, and here’s the summary:

There are now only two planet feeds, similiar to Planet Gentoo / Gentoo Universe, we have Planet Larry and a Universe feed as well. I’m using Planet Venus’ filter option with a simple regular expression to filter out only certain topics to the planet. That means that the main planet will be mostly computer related stuff. The Universe feed remains the same as before, and pulls in everything.

The second thing that’s happened is I’m dropping support of any languages other than English. Sorry to all the non-English blogging participants, but the thinking was this — I only speak English (and Spanish), so for me to support other languages is not a reasonable expectation. Also, it’s incredibly easy to setup your own planet feed, so if some Gentoo users want to get together and make their own feed, then go right ahead. Even the most popular non-English feed was getting about 5% of the hits that the main feeds were, so it’s not worth the effort to keep it up. I’m also going to stop including feeds that are *only* non-English (again, sorry), since frankly, I have no idea what they are saying, and again it’s gonna be easier to start your own. If you speak English and another language, that of course is fine. The filters for the Planet feed will probably kill most non-English posts anyway. If someone is really really interested in keeping theirs alive, then contact me and we can work out a way to manage it yourself.

Now, as far as the filtering on Planet / Universe, right now it’s really simple. In fact, the exact expression is this: “filter = [Ll]inux|[Gg]entoo|[Cc]omputer|emerge|portage”. It sucks, I know, because I’m not even doing case insenstive matches. I just need to read the docs some more or figure out how to hack the source code a bit to figure out how to get it working so I can add that. Also, that filter isn’t set in stone … I’ll be adding more keywords as I see them commonly cropping up. And of course you can subscribe to the Universe feed if you like instead, and get everything unfiltered.

Alternatively, and this is really cool, I can change the regexp filter for each planet feed. So if you want your filter to be more permissive (or restrictive), then just let me know.

That’s it for now. I hope everyone enjoys the new changes. I already like the Planet feed’s “on-topic” style already, myself. Feedback is always welcome though. I get about 3 comments a year, so don’t be shy.

 Update: I’ve expanded the regexp quite a bit, and at the same time cleared out a lot of dead blogs and links.  So you should be seeing more content and less deadbeats.

I’ve been having problems with the cache of the RSS 2 feeds, and in an attempt to clean up config files, broke it completely. Whoops. As a matter of record, though, I think that I would recommend the Atom feed above the RSS ones, after doing a little reading on the two. Personally, I’d like to ditch RSS 2 completely, especially if the issues persist (which I can’t really explain too well, and it might just be a problem with my reader — anyone else having problems?), but we’ll see what happens.

I should probably elaborate what’s happening.  Using liferea, with the RSS 2.0 feed for Planet Larry, the posts from ~5 days ago always refresh as “new” posts.  It’s really odd,  especially since it’s the only one of all my feeds that does it, and of all the ones on Planet Larry, it’s just the RSS 2 one that is doing it.  The Atom one is fine.  Who knows.  It could just be a bug.

After a few bumps in the road, Planet Larry has finally migrated over to using Planet Venus on the backend. Right now, only the front page (the universe feed as I call it) is switched over. The per-language ones will continue to use the original Planet until I have time to switch them over.

Also, I added a link to the Feedburner version of the RSS feed. If you wanna use that instead, have at it.

The old universal planet feed is still available, but it will eventually be going away.  You can get it at http://planetlarry.org/planet/ but this one will only update on a daily basis instead of every 30 minutes like the main feed.
More changes coming on the way, though they are probably going to be trickling in for a while. One thing I’m thinking of doing is kicking off any subscribers who only post partial RSS feeds. That’s a part of my general “get rid of annoying stuff” directive which you’ll see more as I get my backend admin cleaned up. Venus’ ability to use regular expressions to filter posts is going to also play a big part in that.

As always, let me know if general problems with the feed / page, and we’ll try to get it taken care of. And send in some hackergotchis / avatars.

I’ve cleaned up the Planet Larry feed powered by Planet Venus quite a bit, so it should look exactly as it is when we switch over to that backend instead.  I plan on doing that pretty soon in fact, like, tomorrow.  So please check it out, let me know if there are any bugs or what not that I should know about.

I already know the posts on there are a bit whacked, it’s because I cleared out the cache completely.  It’ll fix itself right as it pulls in new feeds, so don’t worry about it.

I also cleared out a few users whose blog was getting a 404 page not found error.  So if you want to be back on there, let me know that, too.

I know people have heard me promise repeatedly over a span of quite a few months that I’ll get Planet Larry cleaned up of its various issues, but this time I have some actual news.  I finally started working on an interface to administrate all the bloggers.  Can you believe there are 110 of them?  Unless I missed someone.

Right now it’s nothing more than a small db schema and an ugly front page that I use.  I’m still undecided how to  integrate user interaction so they can do their updates, so I’ll come back to that later.  But, at least I’ve got something.  And believe me, anything will make the entire process go much faster.

Well, cleaning up transcode was relatively easy (though I’m ignoring the minor dvd::rip bugs for now). I bumped the latest rc candidate for 1.0.4 and got a new CVS snapshot of 1.1.0 in the tree, so I feel like I’ve at least gotten something done. Upstream is having a few issues with some bugs, unfortunately, that will delay the next releases. I’m sure they could use with a few contributors.

The next project on my plate is to finally tackle the problems with the planets. Specifically, Planet Gentoo, Gentoo Universe, Gentoo Developer Blogs and Planet Larry. I have been putting this off for a very long time, and there’s a lot to be done. Things have been nagging me, and it’s time to finally clean things up and do them right.

The first two (Planet Gentoo and Gentoo Universe) are actually pretty good to go. The only thing I’m looking at that might change them is the backend we use. I’m experimenting / playing with Planet Venus, which isn’t so much a fork as much as a lot of fixes to Planet. I’m already running a copy of it on Planet Larry to see how well it works, and so far it’s been great. The things I like the most are that it uses less memory, can update more than one feed at a time, and most of all, lets you filter out posts using regular expressions. For those on Gentoo’s planets, that last point could prove incredibly useful for those who don’t have a blog that let’s them setup categories.

The only other changes that would go into those is creating some tools to help me manage and take care of users more easily. It has long been a royal pain in the butt to manage planet’s files because it is essentially one large .ini file where you have dozens of entries. I’m first going to write a frontend that I can use to automatically generate the .ini files for each user, instead of all in one global file, and store them in a small database. That way, making minor changes will be a simple feat.

For Planet Larry, I’m going to take it a step further and let users manage their feeds themselves. I’ll have an entire user authentication system where they can login, set their feed URLs, choose their language, set their location, etc. so that Alex or myself don’t have to do it all. The only thing left to us to administrate would pretty much authorize any changes — which isn’t as bad as it sounds, it’s just that in the past we’ve had people want to get on the planet for nothing more than pure self-promotion, so we can’t just automatically include anyone on there or we’d get spammed.

Something else that I’m going to do when I setup the new user administration system is this — purge all the entries and start from scratch. I’ll let anyone sign back up and get back on, of course, but I want it to be completely voluntary.

Planet Larry has grown quite a bit, and while my original idea of pulling in anything from anybody who randomly mentions Gentoo now and then was fine for when there was a few people, the scenario now is that most of the content is off-topic completely, and makes for some pretty boring reads sometimes. If there’s some interest, I can easily setup a filtering mechanism and / or let users choose which categories to be pulled into the feeds, so we can have a Planet / Universe mix of on and off-topic posts. The real hurdle is just getting a frontend setup to manage all the data, and once that’s done, we’ll be in a position to make a lot of customizations. One idea I really like is to even setup personalized planet feeds, where you can ignore some users if you’d like.

That’s what I’m going to be working on soon. If you have any comments or suggestions, now is the time. Just ping me at my gentoo.org e-mail address (beandog) if you have any ideas.