Stone Soup 0.6.0

That last post was way too negative, and since I have a two-day hiatus coming up and won’t be updating this site, I would rather leave something at the top of the page that’s a little more fun. This is the newest version of Dungeon Crawl Stone Soup, which recently got bumped to version 0.6.0:

That’s via telnet crawl.akrasiac.org, on a 120Mhz Pentium machine running Crux. As you might be able see in the lower corner, the memory and processor load are negligible, probably because the bulk of the responsibility is shifted off this machine and out into the ether. I have installed crawl directly on to this computer, but the work involved seems a bit overkill when I can just use telnet to jump into a game remotely.

And I know some other people have used Internet Explorer to telnet into the game and play, or so they told me. I have tried that in Firefox in Linux and come up empty-handed.

It’s worth reading the changes between version 0.5.2 and this one, since there are new gods and monsters, the attribute system has changed slightly, and an entire spell school is gone. If you’re used to playing with one particular set of spells or rules, things might be different now. Take a look at the wiki while you’re at it.

It’s also worth noting that the developers attribute the new release to a recent upsurge in interest in the game, and the number of people willing to contribute. Which is how the system should work, really.

Enjoy!

P.S.: There is a tiled version too, if you can’t stomach the thought of playing a game at the console. …

What Linux needs

Learning about Linux has been as much an education about computers as it has been about human behavior. Take, for example, this somewhat elderly thread complaining about photo management.

Linux Needs a GOOD photo management app … Does anyone else agree that great photo management software for linux does not exist? Something along the lines of google picasa would be excellent. (Yeah I know it can be run under wine, but I really would rather run a native program).

It is slightly irritating (only slightly these days … I’ve been watching that thread pop up again and again for years now) that, because nothing fits in the “great” category for this one particular person, that such an animal doesn’t exist. Or rather, “Can we all agree that this animal doesn’t exist?”

Here‘s another one.

No REAL PIMs for Linux? … If the definition of a PIM is where all your notes, appointments, tasks and contacts are stored but more than that, linked then I have not found any PIM available in Linux. I have to use Windows XP under Virtual box and that is slow.

Perhaps it is my understanding of English, but I find that even more acid. Define your animal, and then insist the animal is impossible to find.

One more:

Is there a REAL replacement for MS Project? … I have been looking at several open source PM programs, but all of them have serious issues. … Does anyone know of any program that is a true replacement for MS Project? In other words, something that can calculate task logic and critical paths correctly?

Probably not as crass as some of the others, but still somehow arrogant. Here, there are animals that are similar to the one I describe, but they are not exact duplicates. Therefore, I seek the “true” animal.

I hope this doesn’t sound like a personal attack against any of these people, because really, what amazes me is the method of argument, not the person behind it. I have no formal training in logic or dialectics, and added to that I have a rather unusual grasp of English at times, but in each case the tendency toward generalization is interesting. Each one seems to be saying, “Because the options didn’t work for me, it doesn’t exist.” Because I didn’t find the grail, the grail is a myth.

I wonder what makes humans do that? Why are we all so sure that our case is the only case, and that our experience is replicated endlessly, six billion times over and over again, over the surface of our teeny blue planet?

For me, the oddball part of all three of these threads is that in each case, the original poster knows exactly what they want, and asserts — more or less — that nothing short of a mirror image of that program will suffice. Let me be equally blunt, then: You should be using Windows.

In fact, I can answer all three questions in one fell swoop: If the only application that will fit the bill is mired in Windows, then you should be using Windows. If the game you need to play doesn’t behave as cleanly or run as quickly in Linux, then you should be using Windows. If the icon you want for the button you like on the application you prefer is only available in Windows … then stop telling us what Linux needs.

Because what Linux needs is people who want to use Linux. I said a long time ago that if Linux was Windows, nobody would need Linux. By corollary, if what you need is Windows, don’t tell me Linux has shortcomings. Linux doesn’t need application X, it needs someone who is so prepossessed with the idea of creating application X that they spearhead an independent movement to duplicate it, gather a few like-minded and talented peers, and as a group they roll up their sleeves and get to work making application X.

Aha. Now that is what Linux needs.

Or maybe I should just learn to avoid threads with words in all capital letters in the titles.

P.S.: Yes, I know, there is another level of irony at work here: Windows XP Classic makeover for IceWM, anyone?

Greetings from Slitaz 3.0

It’s been a year since the last “stable” release of Slitaz, and whether or not the yearlong break is part of the development plan, I have been chomping at the bit, waiting for this new version.

Slitaz is very impressive, no matter how you slice it. Distributions like Ubuntu or OpenSuse or Fedora shudder to their foundations when someone points out that it’s possible to run an entire desktop on a measly 30Mb disk image. And the fact that it will install and boot and run a graphical desktop on machines as old as this one only underscores the kind of bloat that everyone talks about, but nobody challenges.

And 30Mb is packing an enormous amount of stuff. Openbox, composite effects, Midori, PCManFM, mtpaint, games, an entire X-driven desktop, Leafpad, PDF viewers, Osmo, HardInfo, a slew of customizing packages and scripts, plus … plus … plus. …

If I could build an entire system — and I have tried — I would want it to turn out just like Slitaz. Even my lightest, fastest efforts with outdated software in custom-built configurations can’t stand up to what Slitaz gives you for nothing. It’s fantastic stuff.

Slitaz isn’t the only ultra-lightweight distribution out there — Tiny Core Linux is excellent, someone will no doubt mention DSL, and Kolibri was mind-blowing — but it’s definitely one of the best. I heartily recommend keeping this on hand, even if you don’t use it on a daily basis. A gold smilie for the Slitaz crew:

P.S.: LinuxTracker has a torrent here.

dockapps: Everything old is new again

I could spend most of the day just sifting through dockapps.org. I’ll admit that a lot of them look a bit outdated, but some of them are quite smooth. I mentioned two a lo-o-ong time ago that are still very attractive — wmhdplop and wmforkplop.

 

Both of those are available in AUR and in the Ubuntu repos, but you’re not restricted to what you see there. One nice thing about (most all of) the dockapps is that they require very little to build. And of course, if you’re using Arch, a good number of them have corresponding PKGBUILDs ready and waiting.

Just don’t go overboard with them.

(Because I know someone will ask, top to bottom that’s wmforkplop, wmhdplop, wmlenovo, wmc2d, wmix and wmauda. And that’s barely scratching the surface. …)

elinks and zlib 1.2.4

In between everything else that has been going on around here, I’ve been trying to pin down some strange elinks behavior this machine running Crux. For a week or so, I had gotten bizarre broken pages and error messages when trying to access what seemed like otherwise normal pages — GMail, for example, or LinuxTracker.org. Messages like, “Error reading from socket,” or even just a dead black page.

I had cloned my system early in March and performing a systemwide update didn’t seem to correct the error, which to me meant there was some sort of package update that was silently killing some Web access. Needle in a haystack is the expression, I think.

The obvious culprits were things like openssl and openssh and even iptables, which all were bumped in Crux in March. Strangely though, the errors still appeared when I excluded those packages from the update. In the end, after about four or five days of recompiling in the background — which didn’t really hamstring the machine at all, since it is more than capable of compiling at full bore while playing music or showing a movie — the answer was … zlib.

Yeah, I know. That’s what I said too. “zlib? Huh?”

I don’t know how or why, but the 1.2.4 update to zlib causes some Web pages in elinks — at least on my contorted little home-grown system — to return error messages. I’m apparently not the only one with the problem though — both Arch and Gentoo users seem to have had the same problems.

I’ll see if it’s possible to fix it, or if I should just stick with zlib 1.2.3, which is what I’m doing right now. I don’t mind holding back a package until I can be sure it won’t taint another program; I sat on an old version of mplayer for about six months, because the updated one wouldn’t compile without errors.

Still, the satisfaction in this little adventure was tracking down the offending program, and seeing that it wasn’t my problem after all. Even if it did mean restoring the drive twice over USB1.1, and then recompiling packages one by one at 550Mhz. …

To compile or not to compile

As always, I’m late to the party. But I’m still glad someone has come along and done a comprehensive benchmark of Gentoo’s optimizations, including -Os, -O2 and -O3 flags for gcc. And it’s probably good to throw Ubuntu into that test too, for another level of comparison.

There were very few surprises there for me — a lot of what the author reports and concludes is similar to my own experiences. For my older systems I usually build Crux with the -O2 flag, although while I was using machines with a mere 16Mb of memory, I relied on -Os, and I still try that from time to time. My logic is, if -Os builds smaller binaries, they should load faster. But maybe I’m overthinking it.

Where you sit on the issue depends a lot on what you use your computer for, and how much performance you require and at what cost. Like so many things with Linux, you have the freedom to sculpt your computer to your liking — or not at all. If you’ve never built a system from source code, it’s worth trying once or twice. If it appeals to you, you won’t mind the time spent doing it. On the other hand, if it doesn’t, you can rely on binary distributions, and there’s no shame in that.

As a final note it’s worth pursuing if you’re working with underpowered or out-of-date hardware. Speaking from experience, a customized system might be what determines whether or not the machine is usable at all.

Something fun, for a change

I’ve been far too serious in recent days, and so for a nice change, here’s a screenshot for you.

If the seconds in the screenshot look a little corrupted, it’s not a jpg artifact. The numbers are melting between seconds, as the display changes. This is xdaliclock, which a reader mentioned in an e-mail a while back and I am finally getting around to.

If I read the PKGBUILD correctly, this comes from the stable of Jamie Zawinsky, whom you probably owe a lot of thanks to, after looking over this “resume.” xdaliclock is just one of many many interesting things listed there.

For the clock however, it’s fun to see so many machines from so far back, running it at once. Openbox users should see the potential in this at once — set up the configuration files to spawn this at start, paste it to the bottom layer and run undecorated. Instant desktop flair.

As opposed to something like conky, this animates continually and is fun to watch too. Ubuntu has a package for it; Arch users can compile out of AUR. Resourcewise it is barely noticeable, and I imagine you can run this on anything that will run X. Otherwise, may I suggest … ?

When you put it that way …

These days, when I see posts or threads lambasting the command line, my reaction is almost amusement. In the old days I found them obtuse, but now that I am cresting a fifth year with Linux, I see them more as lack of education than willful exacerbation.

Still, threads asking when the command line will quietly submit are abrasive. It’s unfortunate that so many people see the command line as some sort of throwback, when in actuality it is often what’s doing the work — they just don’t know it. Perhaps the most diplomatic way to bridge the two sides is with this kind of answer: that the appeal might lie with not needing to use the command line.

But as I’ve mentioned in the past, for some people (like me) that idea is exactly the opposite: The appeal lies in not needing to use a graphical interface. Wishing for an end to the command line is just as ludicrous and just as asinine to some people (like me) as it would be for me to openly pine for a day when graphical interfaces are finally abolished.

When I put it that way, it sounds quite silly, doesn’t it?

Its beauty is in its potential

I spent most of this week in various shades of the new Ubuntu, with everything from pure command-line installations to full-blown Gnome desktops, and just about anything in between. I don’t pretend to know all the ins and outs, but it was nice to get back to the system that started me out, so to speak.

It hasn’t all been daisies and happy puppies though. To be perfectly honest, the newest Gnome edition of Ubuntu just doesn’t do it for me.

I have any number of reasons to say that. Aesthetically I think it’s a bit clumsy — yes, putting the buttons on the left is a hazard, the purple-black-orange color scheme is bizarre, and the wallpaper is amateur at best — but I say that as a founding member of the Your Default Desktop Doesn’t Matter Club. Three clicks and it’s gone.

It’s more than just looks though — I get sluggish performance in Lucid, a slow mouse, Nautilus takes seconds to open the home directory, system updates take forever, the hard drive is constantly churning for something, ext4 seems slow compared to ext2 … and so forth. If I was still using my old 1Ghz machine I would expect that kind of drag, but on 3-year-old core duo … ?

In the old days I probably would have gone on a quest to trim back all the things that make it slow and unwieldy, but these days I have a different philosophy. Ubuntu is slow because Ubuntu is slow, and me wasting my time trying to perk it up is just that: me wasting my time. There are dozens — no, hundreds of distros out there all designed and intended to run fast or light or speedy or flexible, and trying to cram Ubuntu into that category is like trying to cram an elephant into a Volkswagen Beetle.

The funny part is, if you strip down Ubuntu’s Gnome — or even better, start without it in the first place — the underlying skeleton is actually quite efficient. Not lightning-fast, but at least workable. Aside from minor points of core design — like the fact that aptitude is a slug when compared to other package managers — you gain an immense amount of speed and function just by whacking off the tumor and grafting on something new.

And Ubuntu’s repositories are immense, so your selection of prepared software is vast. That fact coupled with the previous says to me that, if your hardware is willing and your mind is creative, you can come up with some impressive home-grown systems that run circles around anything the Ubuntu overlords dictate. If you want a system that’s very easy to put together the way you want it, Ubuntu would be my first suggestion.

A long time ago I made a wish that Ubuntu would concentrate more on performance and reliability over looks and flash. This was back when Beryl and Compiz were the rage, and every cool kid had a 3D spinning cube on their desktop, and the lure of accelerated graphics by default was powerful.

I didn’t get my wish, or at not from where I stand now. The trend turned toward razzle-dazzle, and from my perspective the result is an orange-purple-black splash that only looks kinda good, and eats up a lot of my system’s resources even when idle. I could have asked for a lot more from Ubuntu, given the two or three years I have watched it grow.

But the bright side of that — and the reason that I will still use it on occasion — is that it is so very easy to peel off the outermost layers of that onion, and find something useful to start with inside. That’s what makes it worth keeping, in my opinion.

Turning off desktop effects to boost fps rate

One of the reasons I went through all the rigamarole of installing Neverwinter Nights — aside from a desire to play it again — was an offhand comment I saw on the Ubuntu forums a few days ago, about getting a slight bump in framerates in fullscreen games by turning off desktop effects.

I wanted to try it for myself, and to be honest yes, I got the same result. In NWN you can display a frame rate counter by entering ##trace fps in the Talk bar (but the counter doesn’t appear on screenshots, which means you’ll have to take my word for it ).

But what I was seeing was a 4-6 frame-per-second increase (on average) when Appearance > Preferences > Visual Effects were set to “None,” and when they were set to “Normal.” Jacking up the visual effects to “Extra” didn’t seem to have any worse an effect than just “Normal.”

Again, I have no way of proving the slight bump, and I have almost no education on why or how these things work. But I do know enough to realize that there are a lot of factors at work here, not least of which is my relatively mediocre video subsystem. It could also depend on the game, or even just the graphic demands. And of course it’s possible that on a more powerful system that frame rate difference is negligible, or unnoticeable.

Be that as it may, if people ask (and sometimes they do), I’ll probably suggest they turn off their desktop effects if they plan on playing fullscreen games that require video acceleration. At least, it can’t do any harm.