August 5, 2010

Random Musings

Filed under: Geek Stuff — Adam @ 8:35 pm

- Sometimes the mouse bugs me, and I want to just stare at the monitor and move the cursor with my brain. (actually with this sort of control we wouldn’t need a cursor). It would also make for easier dealing with highlighted/unhighlighted windows; I often start typing in the wrong window if I’m focused on the idea, not the computer.

- The task bar being typically aligned to run lengthwise is poor design now that monitors are wide screen. There’s less vertical space.

- Speaking of, the Gnome Shell screenshots I have seen thus far make me want to kiss Fluxbox.

- No, seriously. That is way too much fucking space. Here’s what I need: A system tray and a clock. The system tray is mostly for those damn programs that will minimize to it. I get that it’s “overview mode” or whatever, but still.

- Desktop icons are stupid.

March 26, 2010

Whereupon nVidia potentially loses me forever

Filed under: Linux — Adam @ 7:04 pm

While AMD/ATI are making strides toward improving their open source drivers for video cards, nvidia has decided to drop theirs. Thus, what they are saying is “to use our cards really neat features we want to install code on your system you can’t see that will run at the kernel level.” No, thanks though.

Even though I use them, I have a problem with proprietary drivers (namely, that I use the broken fuckers). I have a problem with them mainly because an issue with a (nvidia…) closed-source video driver is the only time I’ve ever had Linux lock up to the point I had to restart. It crashed and flailed madly and took X down with it so hard X didn’t restart until the computer did.

AMD is making more of an effort; nvidia is making less. That’s a pretty plain message to me.

Also, the thing about IP being secretive is a bunch of crap. If they have a patented technology it doesn’t matter if it becomes public — they still own the patent. It’s not like they’re the 800lbs gorilla of video cards anyway.

I don’t want another nvidia card. The alternatives, however, suck. AMD may be making effort, but a reasonable Linux driver is about five years away for their products. Intel is great, but they’re on-board only. (Dear Intel, please release a standalone card…) So I may have to suck it up and buy an nvidia card if I want a decent video card. It all depends on where the state of video cards is when I build the next desktop.

Maybe the Nouveau driver will improve…

Either way, fuck you very much Nvidia.

March 21, 2010

Ubuntu: Problems with not so benevolent dictators

Filed under: Linux — Adam @ 9:41 am

The other day, I read about fewt leaving Ubuntu, and yesterday I saw this about Jeremy being done with Ubuntu. Their issues are similar, and striking, and also good examples of why I stopped using Gnome and, eventually, Ubuntu. Certain players are making decisions to remove some of the choice from users, or at least make it harder for novices. To keep it easy by keeping it stupid.

The first problem I ever had with a choice made by a free software programmer was when Gnome completely and totally neutered their screensaver. They made it nearly impossible to select which screensavers to have come on in random mode (so, for example, you couldn’t disable webcollage, which has a nasty habit of putting hardcore pornography on your screen), and they also took away configuration options (so, for example again, you couldn’t select which directory to use for pictures, or which font size some screensavers should use). The argument was that any screensaver requiring configuration is broken.

This is simply not the case: In many cases this is a feature. I want phosphor, for example, to display stuff I’ve written at random, and I want the text to be smaller than a billboard. I think it’s funny to randomly see my computer “typing” what I’ve written on the few occasions I have a screensaver on. (I typically just turn the monitor off, and use the screensaver to lock the screen). Or I want to only use pictures from a trip Bendy and I took to Florida, or Christmas pictures. I don’t want to keep my pictures in Gnome’s “home/username/pictures” folder any more than I want to keep them in My Documents\My Pictures.

I am more technically savvy than many, so my solution to this was to eradicate gnome-screensaver on every Ubuntu install/upgrade and then install the unmolested xscreensaver. Further issues with configuration being hidden from me or outright complicated led me to the stop using Gnome all together in favor of fluxbox. If I’m going to have to dick around with text files, I might as well have the full options fluxbox provides me. I also put William Jon McCann on my eternal shitlist for his general attitude about screensavers.

Shuttleworth is, apparently, just as bad. But that kind of figures. He’s made great strides for Linux and F/OSS. It looks like he’s lost sight of the core principles of Ubuntu — humanity to others. Telling people, especially many people, that they can’t second-guess someone is a bunch of crap. The community is all about having discussions, and no one is higher up the chain than anyone else. We don’t have room for elitists.

I’ve used Macs enough to be irritated with the placement of the title bar buttons. I don’t think we want to clone Macs. Fortunately, at this point, I’m using Debian on all but one computer in the house (haven’t seen cause for replacing Bendy’s Ubuntu with Debian — yet), so it’s a non-issue for me. My next machine will probably run Debian, as well, with perhaps some Slackware or Arch as a secondary, testing operating system (though I may just install them in KVM or get VMWare).

The big problem is that this sort of attitude drives away people who are drawn to the Linux community for reasons similar to my own. I like the control it allows me to have over my system. When people take that control, or try to, I don’t like it. Especially for arbitrary reasons. We seek logic; don’t move a button just because and call it an unquestionable design decision. Don’t say screensavers shouldn’t need configuration and then offer up some half-ass, back way around when the original configuration screen was more than adequate. Above all, don’t pretend we’re idiots. We left Windows behind because of that.

I changed to Ubuntu when, inexplicably, my printer stopped working in Slackware and nothing I could do would make it work. I tried everything and spent about four hours on it one Sunday. Google, reading documentation, reinstalling, nothing worked. I finally had enough — it had taken a long time to get the thing running in the first place, and I hadn’t changed any settings. Ubuntu was supposed to be easy. This was just after the release of Dapper Drake. I was there right up to Gutsy. I had problems with every single upgrade, though. Not once did I follow the instructions for upgrading and not have a huge problem. Most of them broke and I ended up just installing over the old OS. Finally, when I had an issue with screen resolution, and my Debian laptop did not with the same monitor, I said a few expletives (“Expletive it, I’ll expletive-ing install expletive-ing Debian then!”) and installed over the old OS.

Debian isn’t perfect, either. No distribution is. Fluxbox isn’t perfect, xscreensaver isn’t perfect. These are all tools made by flawed humans and therefore flawed themselves. The problems come when we allow more of our flaws to influence decisions, even against the voices of people who may well know better. Perhaps Mr. Shuttleworth should take a step back and wonder why the peasants are revolting.

March 13, 2010

Price Per Gigabyte

Filed under: Geek Stuff — Adam @ 5:11 pm

There was a time, not too long ago, when the dollar/gigabyte barrier was crossed. That was when 100GB drives were finally available for $100 or less; $99.99 at first, of course. When that happened it was as though some sort of magic line had been crossed, and it seems like the price per gigabyte has dropped like a stone since. Today, one can buy a great brand’s 1TB drive for $100 or less. A thousand times the space, give or take for how much scratch you save.

So now a gigabyte is a tenth of a cent.

That isn’t the same for USB thumb drives or (Micro | Mini) SD(HC) cards, of course — those haven’t quite crossed the magic dollar/gigabyte barrier. But I have a few SD cards of varying size (both physical and spacial) around the house. It seems like everything uses them now — both my and Bendy’s Sansa has a card added in to it, my phone takes a MicroSD card, and my laptop came with an SD port. (I also have a collection of Micro/Mini to regular SD size adapters and (Micro | Mini) SD (HC) to USB adapters… some are even in the shape of regular USB thumb drives). RAM seems to have crossed a magic barrier, too — 4GB of high end RAM is now under $100. (And it’s actually 4GB in the way a computer looks at it, unless it’s a 32 bit Windows system, in which case it’s 3.2GB.)

Space is cheap. At work there are many TBs of drives. I don’t even want to think about how painful that would be before the wall came down.

Space and RAM aren’t the only things that are cheap — LCD monitors are cheap. I’m sitting in front of a 24″ 1080p ViewSonic that is easily the best monitor I’ve owned (an upgrade from my first flat LCD monitor, a 17″ beauty that cost the same as my 24″ did just 3.8 years later), and Bendy has a 21″ LCD (also a Viewsonic, an upgrade from a 15″ CRT that was probably a hand me down from Moses). Keyboards (except the beautiful clicky kind), mice, webcams, digital cameras — all cheaper and better. Of course, all of this is obvious to everyone that buys computers or computer accessories. Duh, right?

It just feels strange. I could, if I had the urge, purchase VMWare Workstation and install a variety of OS’s and run them within my current OS. Dual-boot? Why?

February 8, 2010

OpenOffice.org is a turd.

Filed under: Geek Stuff, Linux — Adam @ 5:49 pm

I’m a huge advocate of free/open source software. I like to use it, and I think it makes the world of computing better. Some of it is that the open source software I’ve been using is easier to customize and more fits my particular mindset — I like the way Fluxbox and vim work, for example. Usually, it comes down to open source stuff pissing me off less than closed source.

OpenOffice.org is an exception. I’ve had to use it a lot for work lately, building spreadsheets and whatnot, because there is no Linux version of Office. I come into this as someone who has spent more time in OO.o than in Microsoft Office by about a 10:1 margin, 1000:1 when it comes to the 2007 variety.

For one, we ran into some compatibility issues between OO.o and two versions of Excel. Of course, we also ran into problems between the two versions of Excel — I had a file saved in Pre-2007 Office, opened it in 2007… and it crashed the program. This should never happen. I finally tracked it down to an issue with trendlines, deleted those graphs, and it works. So that’s actually a point in OO.o’s favor. No file saved by one Excel should ever crash a NEWER version.

But the real problem, and the real reason that OO.o is a dog, came running large spreadsheets. It is obvious that no one uses OO.o’s Calc to do any real work, because when I had four sheets filled with data — the spreadsheet is around 400kb — the thing lagged my system down to unbearable speeds. My machine is a dual-core, 3ghz processor monster with 4GB of RAM running CentOS 5 and Fluxbox. The underlying OS uses almost no memory. But a few worksheets and hundred rows of data and OO.o Calc takes sixty seconds to highligh 100 lines of data (one column, 100 rows).

Excel in a virtual machine worked 100x faster, no exaggeration. Point to Microsoft Office.

Then when actually trying to make graphs, the ability of OpenOffice.org Calc to do polynomial trendlines is crippled. I spent at least an hour on the Google trying to figure out how to do it and there was some macro I could download and use (imagine that makes it much slower…), and that is a deal breaker. When someone has to go to a conference and give a report in front of people they don’t want to rely on some wonk macro to get the R^2, they want to push a button and get it. Point to Excel.

One more point in OO.o’s favor, though this may be fixed in 2007: In Calc I could label data points with text, while Excel (pre-2007) only wanted to use their value. What the hell? So point for OO.o — tied game and all.

Excel, like Word, is just more polished. I know that OO.o is probably safer to use, but this is a big point in the favor of MS Office. It’s just prettier and easier to look at for long periods of time, it makes better looking graphs (OO.o is just above the command line graph program), and it has, in my opinion, a better interface (even the ribbon). Options are easier to find. Point for MS Office.

You may say, well, it’s a close game here, and OO.o is free! Win for OO.o!

But that’s not how this thing called reality works. That trendline thing truly is a dealbreaker, as they are a requirement for this project. So is the general sluggishness — sure, OO.o is free, and Excel crashed, but there is only so much waiting I can do on a project that is going to be do. You can’t go to a client and say “hey our project you paid for is going to take 50% longer because our software lags.”

I love Linux, and I love open source stuff. But OO.o is NOT an acceptable replacement for Microsoft Office. Open Source proponents are going to have to do better. Some applications — vim, fluxbox, the gnu tools, ImageMagick, xscreensaver (or even xlock!) — are svelte and do what I want, following a perfect sort of logic. But OO.o is not one of them. At least not when it comes to Calc or the Word clone (Writer?). Fortunately, at home, I don’t need any sort of complicated spreadsheet, and Calc is passable.

As a side note, I tried to do some of the work in Gnumeric, and it also wasn’t up to snuff, but it didn’t irritate me quite like OO.o did. I didn’t try KOffice. But Excel won hands down against the spreadsheet programs I tried. (Personally, I’m starting to think “awk and a perl script for the math that runs gnu plotutils” for my personal spreadsheets…)

January 27, 2010

Constellation Nixed

Filed under: Final Frontier — Adam @ 8:19 pm

Apparently confused by what exactly the words Aeronautics and Space mean, Obama wants to push to have NASA study climate change and forget all that astronaut nonsense:

When the White House releases his budget proposal Monday, there will be no money for the Constellation program that was supposed to return humans to the moon by 2020. The troubled and expensive Ares I rocket that was to replace the space shuttle to ferry humans to space will be gone, along with money for its bigger brother, the Ares V cargo rocket that was to launch the fuel and supplies needed to take humans back to the moon.

There will be no lunar landers, no moon bases, no Constellation program at all.

…

In the meantime, the White House will direct NASA to concentrate on Earth-science projects — principally, researching and monitoring climate change — and on a new technology research and development program that will one day make human exploration of asteroids and the inner solar system possible.

One administration goon said “We certainly don’t need to go back to the moon,” later in the article.

I don’t know how to put this any clearer for all of you anti-space travel fuckers, but this is our future. We have to keep going. There is no alternative. We have to reach higher, go further, get out there. No one is going hungry because of space spending, no one is going uneducated because of it, no one is losing their job because we’re spending money on this. The amount of money given to NASA every year is a pittance.

But hey, at least the guy is finally keeping to a campaign promise.

January 9, 2010

I’m really starting to dislike Monty Widenius.

Filed under: Geek Stuff — Adam @ 2:48 pm

A few weeks ago I had no idea who Monty Widenius is. There are only a handful of Open Source developers whose names I know — Linus Torvalds (Linux kernel, git), Richard Stallman (a lot of GNU and FSF stuff, the GPL), Eric Raymond (the Cathedral and the Bazaar), Theo de Raadt (BSD work), and William Jon McCann (the dickhead who broke Gnome Screensaver). See, a short list. Now, this guy is all over the Oracle/Sun merger because of the fact Oracle will now own MySQL, the database program that a lot of the internet is run on.

Widenius was, as far as I can tell, the main force behind MySQL for a long time. Then he courted investors, and then Sun bought the company. Widenius claims he didn’t know Sun was going to buy them until after it was a done deal — which sounds odd to me if he was a shareholder of any magnitude — and now he wants MySQL back. Basically, he got like a billion dollars and wants his toy back now. I get his argument about Oracle controlling the database, though. It would be a Bad Thing. For about ten minutes:

A lot of people have, however, pointed out that people could just go to PostgreSQL. This is a very real option. In fact, if Oracle made things too difficult on people using MySQL — a lot of which is protected by the GPL — then the free stuff could be forked or worked on with PostgreSQL.

Widenius, who is at least somewhat to blame for the state of affairs, is now saying that, if PostgreSQL actually comes to dominance as a product for free databases, then Oracle would just somehow either buy off the developers or buy the companies supporting it, thus dashing all hope and crushing the internet into the shape of Larry Ellison’s colon. Or something.

Other than the fact he’s a raging hypocrite and projecting his own issues (being bought out) onto others, I have a few problems with his theory/whining:

1. He’s assuming that other companies would just let Oracle stomp around at will. IBM, Microsoft, Google (which really should have bought Sun), even RedHat, all have a vested interest in Oracle not holding all the cards. I don’t know if Widenius is aware of this, but those companies are pretty big. Oracle is one company. They are many. Oracle doesn’t control everything.

2. A lot of software works with non-MySQL databases already, especially PostgreSQL. Once you’ve modified your software to support more than one database, adding another shouldn’t be any big problem — good programmers would make the hooks fit into a module of some sort. PostgreSQL is not the be all end all either: Open source developers are legion. Especially if suddenly Oracle starts handing out checks to them. I would start my own SQL DB if there were a billion dollar check at the end of the road. Dear Larry Ellison, I promise not to make LawsonSQL for a pittance — a million dollars! I am a thousand times over more cheaper than Widenius. Invest in me!

3. No, seriously, invest in me. I guarantee I will not produce a database of any sort.

December 8, 2009

All hail Google!

Filed under: Geek Stuff, Personal — Adam @ 7:53 pm

A coworker has a Droid phone, and the other day started to do a Google Search. It popped up previous searches — from his home computer and his work computer. Obviously, Google is keeping track of what the account searches for. That alone isn’t too scary, but:

“If you have something that you don’t want anyone to know, maybe you shouldn’t be doing it in the first place”

Yeah, that’s not frightening in the least.

I’m starting to rethink my use of gmail and other Google services. While it’s true that they are damn convenient, they’re also starting to become pretty horrifying. They’re keeping too much data and becoming too convenient. At the same time, though, most of my use of the internet is pretty boring; I look up Linux, sports stuff, politics, fantasy/sci-fi stuff.

Still, Google is starting to give me the heebie-jeebies.

November 22, 2009

My horrible idea.

Filed under: Geek Stuff, Personal, whiskeytangofoxtrot — Adam @ 12:33 pm

I told some friends this over the last few weeks, but this is my great, genius, evil idea:

Remember those Magic Eye things, where you stare at them and see a picture? I loved those stupid things when I was a kid. One day, I had an idea, though:

A Magic Eye, but when you look at it? It’s goatse. If you know what goatse is, take a minute to think about how funny that would be. If you don’t, don’t ever look it up and avoid links with those letters in the name. It’s just an awful picture from a shock site. And it will never leave your brain once you see it.

July 15, 2009

The State of Computer Science Education

Filed under: Geek Stuff, Personal — Adam @ 6:03 pm

Now that I’ve been working at a real job with real responsibilities for a little over six months, I have some comments on my education and what I think it should have had and all that good mess.

When I started work at the lab, I had finished all of my computer science education at USM. I only needed 11 hours of credits after that, in any subject matter. I chose some random classes to allow me to spend more time with Bendy and I chose easy classes to let me spend more time at work. It’s a good thing I did.

Because I didn’t learn jack.

Honestly, I learned more in the first month I was at the Lab than I did in a year and a half at USM. I taught myself Perl in the first week — and I’m not trying to pat myself on the back. I just needed to know it. Perl and bash scripting were my first priority, and I learned them quickly. I’ve learned more about C in the months since than I did at both USM and Bendy’s school. The main thing I learned while a student was PHP, and I taught that to myself for a project in the senior capstone. The only other languages I learned well I haven’t used — assembly (for the Motorola 68K) and Java. What I learned of Java was basically an extension of my knowledge of C/C++.

I took five of what I considered to be “programming” classes — Programming with C I and II, Data Structures, Assembly (wasn’t called that), and Java. The way I see it, Data Structures was mostly the third in a series with the first two programming classes. It was also one of the best of the five. If I had to rank them for what they taught me, it would be: Data Structures, Assembly, Programming w/C I, then II, and finally Java.

I think every computer science program should emphasis programming at least to some extent, and I think these five classes are a good example, but I’d expand it thus:

Programming I, II, and III would cover up to the end of a good data structures class, and Programming IV would cover algorithms. These would all be in C/C++. Building them together would give more of a grasp of programming.

Assembly would follow a Boolean Algebra/Discrete Math sort of computer logic class. These could be taken concurrently with the above. Java would come in the form of two classes and be “Advanced Programming with Java” or something. This combination would expose the student to three languages, and that’s a good foundation. Hopefully somewhere along the way someone would mention scripting languages (Perl, bash, Python).

I also took a databases class. It was mostly useless and I had to teach myself because the teacher was obsessed with the visual representation of a database and the logic but never really got down to the actual, y’know, use of the code. That’s only slightly important after all. I think everyone should be exposed to that, as well, and possibly a bit of web programming (PHP, Javascript).

Then comes the more theoretical classes — Operating Systems, Programming Languages, Language Theory, Automata. I wouldn’t so much change these classes from what I took as make them more effective. Programming Languages was a disappointing class but it struck me as pretty important. This adds to the knowledge background above. Automata, at least the class I took, was hard and boring at times but looking back one of the more educational classes I had. But again, it’s mostly theory, unless I develop and Operating System or a Programming Language.

Most of the rest of the classes I took were math (Calculus, Stats) or electives (Artificial Intelligence, Ethics) and those I wouldn’t change,but I would have preferred more suitable electives being offered.

Regardless, my point is, I don’t feel like a CS degree necessarily prepared me for the work force. I’ve looked at the classes for a masters and a doctorate in the field and they don’t look like they’d help all that much either. Other than making the classes teach more I don’t know how I’d fix it. I really would have appreciated a better foundation in the science and less theory, though.

At least I didn’t get a liberal arts McDegree though.