OpenCD -- Free Programs for Windows

OK, some of you would rather not follow all the notes I've left on Working With Fedora Core, and prefer to stick with Windows. I understand that. However, there are ways to make your Windows computer run better and cheaper. To help you out with that, the OpenCD project makes a bunch of Open Source programs available for Windows on a single CD. There are a large number of open source programs, including Firefox, Thunderbird (email), OpenOffice.org (Word Processing, Spreadsheets, etc.), games, screen savers, and more.

Plus, if you just pop the CD in the drive and reboot your computer, you get a LiveCD version of Ubuntu Linux ("Linux for Human Beings") to give you an idea of how Linux looks.

The programs are great, I really like Celestia, which is a space simulation which lets you look at the Universe from any vantage point in the Universe. But on my generic Dell computer, Ubuntu came up with a 640x480 screen, so the graphics were not good, and there was not option to change it. A better CD for playing around with Linux is probably Knoppix, which doesn't have trouble detecting my graphics card and giving me a properly sized screen.

But ignore that, and download or order the OpenCD disk. Or just download the programs you want.

PostScript Viewer

Sometimes it's a pain to read PostScript (TM) files on screen, even though it's easy to print them out. On my WorkMac I can't even look at a PostScript file unless I use a postscript viewer of some kind, even though the Mac Preview utility will read PostScript's somewhat less featured offspring, PDF.

Fedora comes with at least one PostScript viewer, gsview, available in the Extras directory. However, the viewer I like best is gv, a front end to the standard ghostscript utility. gv will display both PostScript and PDF files, and lets you zoom in and move around the page. It's a very old program, the source was last updated in 1997.

I could have installed directly from the gv source, but I thought I'd find an RPM file to work with. Unfortunately, it's in none of the standard repositories. However, rpm.pbone.net found RPMs which claimed to be compiled for FC4. Rather than trust that whole-heartedly, I got a copy of the Source RPM and compiled that. The install was straightforward, and now I can view PostScript and PDF files in the old-fashioned way.

Notes:

• Both the Gnome and KDE desktops have PostScript viewers, I just don't remember what they are called and find gv simpler to use.
• At some point I found on the web a suggestion that you partially protect yourself from malicious SRPM files by setting up an account which is only used to compile RPM packages. This isn't a total panacea, but it's a good start, so I set up just such an account. (What this does is keep the SRPM's makefile from strolling through your own directories. Of course, it might choose to scroll through everyone's directories, in which case you're still hosed, but that would take a lot longer time and you might notice that something weird is going on.)

The Last Post

The Hindmost remained in footstool mode, needs explaining, I suppose, since no one picked up on it. When I ran across it the other day, I thought of it as a perfect quote. It could only be written by one person, and anyone who knows that person's work would instantly recognize the author and the story setting, if not the exact work involved.

If you don't know what's going on, the Hindmost is (well, was) the leader of the Pearson's Puppeteers in Larry Niven's Known Space series. The quote is from Ringworld's Children, but to understand that book you'll need to read Ringworld, at least, and probably The Ringworld Engineers and The Ringworld Throne. (Ringworld is Niven's best novel, the other two are, hmm, somewhat less well-done. I can't comment on Ringworld's Children yet because I haven't finished it.)

Anyway, a Puppeteer is an intellegent herbivore who believes that discretion is the only part of valor. When in danger with no place to run (always Plan A), a Puppeteer tucks its two heads (which look like Cecil the Seasick Sea Serpent puppets under its body, along with its three legs, creating something that looks like an ottoman (Plan B).

At All Costs Avoid Plan C: Under no circumstances approach the ottoman from the side nearest the hind leg.

The Hindmost remained in footstool mode.

Well, Maybe a Cookie

OK, it appears that cats lack the gene to taste sugar. [Washington Post, registration required, if you don't know about bugmenot.com it's not like I didn't tell you.] That's right, house cats, lions and cheetahs (Oh! My!) all lack the gene, or, at least 247 "letters" of it.

Now, one of our cats eats roses -- I would have thought because they were sweat, but apparently not.

Bad News, Bears

Since there's a new version of the movie coming out, it's appropriate that we look at where the Bad News Bears are today.

No, not Jodie Foster, the characters in the movie.

Scotty's Beamed Home

As you probably know, James Doohan passed away yesterday. The BBC has put up this RealPlayer clip about Mr. Scott.

Making Transparent PNG Images

August 14, 2011: Many years after moving from Comcast to Verizon, and thus losing a big chunk of free webspace, I've finally moved the pictures in this article to Picasa, so that you can see them again.

Also note that if you know the exact color of the background, you can do the conversion from the command line.

One of the nice things about old-style GIF images is that it was easy to set the background to be transparent, so that it would show up on a web page without a funky border. New-style PNG images can also do this, but there doesn't seem to be a reliable stand-alone way to do this. Fortunately, I found directions for making a transparent background using the GIMP. The directions seem to be for an older version of the GIMP (I'm using 2.2.7), so let me write down the current directions here. They should work on a Windows machine as well as under Linux.

1. Bring up the GIMP, and load in your picture. For our purposes we'll use this bit of modern art:
   
2. Now right-click on the image and select Layer > Transparency > Add Alpha Channel. (Alpha Channel is what gets made transparent.)
3. Right-click again, Select > By Color.
4. Left-click the image on the color you want to make transparent.
5. Right-click once more, Edit > Clear. The color you selected should now be replaced with a checkerboard pattern.
6. Save the picture, which should now appear like this:
   
7. I'm told that Internet Explorer doesn't do transparent PNGs well, so for youse guys, try a GIF version:
   

And that's all I know about it.

The Thirty-Eight Minute Wormhole

If you don't know what the title refers to, or if you think that the biggest news last night was the release of Harry Potter and the Half-Blood Prince, feel free to read on, if you wish.

If you do know the title reference, then be warned, there are possible spoilers below. I'm going to explain things to the mundane, then discuss the shows. If you haven't seen them yet, you might not want to read any further. Though, really, there isn't much here you couldn't figure out on your own.

---

Back in my youth, there was a rule that at most one Science Fiction story could be shown on television at one time. Thus The Twilight Zone never overlapped with Star Trek. OK, there were a few shows which tried to cash in, like Lost in Space, Men into Space, and even My Favorite Martian, but they weren't really science fiction. (Men into Space was more like "Blue Angels With Rockets", a techno-thriller kind of film. I loved it, but it wasn't SF.)

Post Star Trek, we got a bit more SF on TV: the dreck Space: 1999, the really-cheesy but great back-story Battlestar Galactica, and, when the local PBS station could afford it, Dr. Who. Then came Star Trek: Next Generation, followed by Deep Space Nine, Babylon V, Buffy, etc.

But science fiction on TV really arrived with the SciFi channel. I know, they run a bunch of grade C movie stuff, but they also run SciFi Friday, which features Stargate SG-1, Stargate Atlantis, and Battlestar Galactica (Mark II, with glowing-backbone female Cylons). These shows are generally as well written as an average Star Trek episode, usually pay more attention to continuity, and have a whole lot bigger budget.

So last night they kicked off the new season for all three shows. SciFi has a strange release schedule: the shows will go on for a couple of months, then they'll show reruns until January, when they'll show the rest of the season. In Britain I understand that the series will start in September and show straight through, so though the UK won't see these shows for a few weeks, by November they'll be ahead of us. Hence the spoiler warnings.

For the past few years a friend of mine and I have been discussing the shows by email. Generally I give a few thoughts on each episode, crack a few jokes, and go on, while he gives a reasoned review of the episode, including how it fits in with his worldview.

Pity you won't get that here, you just get me cracking the jokes. But OK, if he wants to respond, he can, in the comments below.

For what it's worth, last year I liked SG-1 best, followed by Battlestar, and then Atlantis. I can go into the reasons if you like, but that's my prejudice as I look at these shows. This year, though, SG-1 has major cast changes, so my opinions might change.

OK, if you're looking for the possible spoilers, they're just below the line. Don't say I didn't warn you.

---

• Stargate SG-1:
  • Beau Bridges: (who looks much worse than his Dad did at the same age) tries to be a tough guy. But he's not, so far, he's just moaning about how good the SGC people are, which means he can't yell at them. That never stopped Hammond. This guy needs some character development. How come the President answers HIS phone calls? And who is #1 on his speed-dial?
  • Claudia Black: (sorry, I don't know the new character names, yet, so we'll have to go with real names): I liked her character in the tryout last year, and she'll be OK as Daniel's love interest and gadfly. However, she should get a new dress designer. The current outfit just emphasizes the fact that there's not all that much up there. Also (and more importantly?), as portrayed this week, her character is basically stupid. She actually believes what people tell her about devices they've stolen, and she can't think more than 30 seconds into the future.
  • Other Farscape refugee: This is the star of the show? We'll see when they give us a longer-than-10-second opening. Looks too much like Daniel, as Claudia pointed out, and tries to act too much like Jack, without the essential post-MacGyver cool factor. Also, I'm not sure from the flashbacks why he got the Congressional Medal of Honor, and not any of the other X-302/303 pilots that were holding off Anubis. Wasn't he just doing his job, like all the rest?
  • Daniel: Thank God he lost the beard. His rather intimate relationship with Claudia could be fun. It was last year, anyway.
  • Teal'c: Hopefully he'll get more to do than shoot bullets at walls and try to hold up collapsing ceilings. He's supposed to have some conflict with
  • Lou Gossett Jr.: Why is he slumming here? OK, Isaac Hayes did it last year, but that was only for a couple of shows and he's basically got no career anyway. What possesses a Movie Star to appear on a cable SF series? They'd better do something good with his character.
  • Jack: He plays chess? He's good at chess?
  • Carter: Tapping is on post-maternity leave, supposed to be back full-time in a few weeks.
  • Villains: Don't have any yet, so it's hard to tell.
    <Wholesale Speculation>
    It is interesting that Merlin is an Ancient. This means that the Ancients were backing the Celts at about the same time that the Asgard were backing the Norse. This didn't lead to conflicts? Could this explain the rise of the Goa'uld? Battles among the four races. Come to think of it, this could explain why we never see the Furlings. The big four fought amongst themselves, and the Furlings were wiped out.
    And why do the Nox hide? Are they hiding from the Goa'uld? Maybe they're hiding from the Asgard.
    And who created that plague that wiped out the corporal Ancients? Could it be, ..., I don't know, ... maybe ... the Asgard? You'll notice the Asgard are the only ones of the big four still running around the universe.
    OK, probably not.
    </Wholesale Speculation>
    
• Stargate Atlantis:
  
  • Speaking of Asgard, we're corrupting one here. By the time this season is through, Hermiod (bad choice of name, unless they're going to get Preparation-H as a sponsor) is going to be chucking nukes into stars just to watch the pretty light-show. And why does he curse under his breath? Does anyone on the Daedalus understand ancient Norse? Come on, it's not like Daniel Jackson is there.
  • It would have been interesting to have Dueling Colonels ("I get to draw and quarter Sheppard." "No, I get to draw and quarter Sheppard.") but I guess that would get old fast so they had to get rid of one. Interestingly they kept the one who seems easier to get along with. Maybe he's a pal of Beau Bridges.
  • Plot: Did the Wraith really believe that explosion? They didn't even come down to check. And does this mean that the Atlantis team has to stay hidden for the rest of the season? After all, once they blow up another Wraith ship all of the hive will be back.
• Battlestar Galactica:
  • I don't know. But then, I never know about this series. I'm guessing that Adama is going to be out of commission for a few more weeks, so we'll have some conflict between Tigh's sense that he's unfit for command and his wife's pushing him to take control. We've also got to reunite the cast, which is now spread over half the galaxy.
  • I see how the Cylons were able to win the first time: Our heroes cable together a network of three computers, on a presumably shielded spacecraft, and a Cylon virus is able to break it down in minutes. The virus must have been left in the computer's OS, just waiting for a chance to do mischief. And no ones gone through the code for the last 20 years. Or else, all the programmers are secret Cylons.
  • I keep thinking about writing an SG-1/Galactica crossover. Except that Sam (or even Rodney) would figure it all out in about 5 minutes, and Jack (or Farscape Guy) would blow the Cylons to smithereens a few seconds later.

Stay tuned next week, boys and girls

2,000

This blog is rapidly approaching 2,000 hits on the sitemeter in the right-hand bar. If you are the 2,000^th hit, send me a picture of the meter and I'll send you a genuine emailed no-prize, just like the ones Marvel Comics used to give out. You'll also get recognition in the Blog, if you want it.

And thanks to the folks on SABR-L for driving up the count.

Yankees Lose!!! The, THE, THEYankees Lose!!*

It's the All Star break, which makes it a good time to take stock of the season so far. Here are the standings after Sunday night's games:

2005 American League Standings
EAST	W	L	PCT	GB	RS	RA
Boston	49	38	0.56322	0.0	473	429
Baltimore	47	40	0.54023	2.0	431	409
NY Yankees	46	40	0.53488	2.5	478	431
Toronto	44	44	0.50000	5.5	428	381
Tampa Bay	28	61	0.31461	22.0	399	553
CENTRAL	W	L	PCT	GB	RS	RA
Chicago Sox	57	29	0.66279	0.0	413	339
Minnesota	48	38	0.55814	9.0	396	360
Cleveland	47	41	0.53409	11.0	406	365
Detroit	42	44	0.48837	15.0	387	375
Kansas City	30	57	0.34483	27.5	376	485
WEST	W	L	PCT	GB	RS	RA
LA Angels	52	36	0.59091	0.0	420	355
Texas	46	40	0.53488	5.0	476	430
Oakland	44	43	0.50575	7.5	400	386
Seattle	39	48	0.44828	12.5	377	388
2005 National League Standings
EAST	W	L	PCT	GB	RS	RA
Washington	52	36	0.59091	0.0	357	361
Atlanta	50	39	0.56180	2.5	428	348
Florida	44	42	0.51163	7.0	383	368
Philadelphia	45	44	0.50562	7.5	410	417
NY Mets	44	44	0.50000	8.0	387	381
CENTRAL	W	L	PCT	GB	RS	RA
St. Louis	56	32	0.63636	0.0	447	340
Houston	44	43	0.50575	11.5	365	362
Chicago Cubs	43	44	0.49425	12.5	394	394
Milwaukee	42	46	0.47727	14.0	392	374
Pittsburgh	39	48	0.44828	16.5	365	403
Cincinnati	35	53	0.39773	21.0	434	518
WEST	W	L	PCT	GB	RS	RA
San Diego	48	41	0.53933	0.0	406	385
Arizona	43	47	0.47778	5.5	394	479
LA Dodgers	40	48	0.45455	7.5	384	422
San Francisco	37	50	0.42529	10.0	393	457
Colorado	31	56	0.35632	16.0	389	493

In the above table, "RS" and "RA" stand for "Runs Scored" and "Runs Allowed," respectively. More on that later.

From the standings, we see that there isn't any contest in either the AL or NL Central divisions, that the AL and NL West are, if not wrapped up, at least uninteresting for the moment, and that the most exciting baseball is being played in the AL East, closely followed by the NL East, where the Nationals are surprisingly in first place.

And "if the playoffs started tomorrow" we'd have Minnesota and Atlanta as the Wild Cards, and the Yankees would be out of the playoffs. (Yankees delenda est!)

Now one of the rules of baseball is that to win you have to score more runs than your opponent. If we look at the above table, you'll note that the Washington Nationals, though in first place in the NL East, have been outscored 361-357, even though they are 16 games over 0.500. What this means is that the Nationals are winning close games (except this last week) and losing blowouts. This is Not a Good Thing for Nats fans, because the winner of a one-run game is determined mostly by luck.

This is quantified by what is known as the the Pythagorean Method, which in its simplest form says that the fraction of games a team should win is related to RS and RA by the formula:

Pct. = RA² /(RA² + RS²)

So what would the Pythagorean rule say about this season? Let's calculate how the standings would look if each team won and lost according the to above equation:

2005 American League Standings
EAST	RS	RA	PW	PL	Pct.	PGB
Toronto	428	381	49.0953	38.9047	0.55790	0.0000
NY Yankees	478	431	47.4348	38.5652	0.55157	0.6605
Boston	473	429	47.7338	39.2662	0.54866	0.8615
Baltimore	431	409	45.7770	41.2230	0.52617	2.8183
Tampa Bay	399	553	30.4701	58.5299	0.34236	19.1252
CENTRAL	RS	RA	PW	PL	Pct.	PGB
Chicago Sox	413	339	51.3816	34.6184	0.59746	0.0000
Cleveland	406	365	48.6664	39.3336	0.55303	3.7152
Minnesota	396	360	47.0860	38.9140	0.54751	4.2956
Detroit	387	375	44.3540	41.6460	0.51574	7.0276
Kansas City	376	485	32.6598	54.3402	0.37540	19.2218
WEST	RS	RA	PW	PL	Pct.	PGB
LA Angels	420	355	51.3291	36.6709	0.58329	0.0000
Texas	476	430	47.3552	38.6448	0.55064	2.9739
Oakland	400	386	45.0491	41.9509	0.51781	5.7800
Seattle	377	388	42.2493	44.7507	0.48562	8.5798
2005 National League Standings
EAST	RS	RA	PW	PL	Pct.	PGB
Atlanta	428	348	53.5788	35.4212	0.60201	0.0000
Florida	383	368	44.7170	41.2830	0.51997	7.3617
NY Mets	387	381	44.6875	43.3125	0.50781	8.3913
Washington	357	361	43.5098	44.4902	0.49443	9.5690
Philadelphia	410	417	43.7467	45.2533	0.49154	9.8320
CENTRAL	RS	RA	PW	PL	Pct.	PGB
St. Louis	447	340	55.7473	32.2527	0.63349	0.0000
Milwaukee	392	374	46.0667	41.9333	0.52349	9.6805
Houston	365	362	43.8590	43.1410	0.50413	11.3883
Chicago Cubs	394	394	43.5000	43.5000	0.50000	11.7473
Pittsburgh	365	403	39.2058	47.7942	0.45064	16.0414
Cincinnati	434	518	36.2953	51.7047	0.41245	19.4520
WEST	RS	RA	PW	PL	Pct.	PGB
San Diego	406	385	46.8612	42.1388	0.52653	0.0000
LA Dodgers	384	422	39.8603	48.1397	0.45296	6.5008
San Francisco	393	457	36.9863	50.0137	0.42513	8.8748
Arizona	394	479	36.3194	53.6806	0.40355	11.0418
Colorado	389	493	33.3822	53.6178	0.38370	12.4790

Here "PW" and "PL" are the wins and losses as predicted by they Pythagorean rule, and yes, I've kept way to many decimal places. (You'll note that the total number of wins in this table isn't equal to the total number of losses. That's because the Pythagorean rule is applied on a per-team basis, so a win for one team isn't necessarily a loss for another team.)

For the Central and Western divisions of both leagues these results are essentially the same as the actual records. In the Easts, however, things are drastically different. For one thing, Toronto is actually a very good ballclub, ten games above 0.500. The Yankees aren't doing as badly as Mr. Steinbrenner thinks, they're in contention for the Wild Card. And in the NL, Atlanta is in its accustomed place, and Washington is at about 0.500.

Of course, what this tells us is that the Pythagorean rule isn't exact, and will have some discrepancies, especially over only a portion of a season. But it does suggest that the Nationals have been playing over their heads, as anyone who watched the games with the Mets and the Phillies will attest, and that the AL East is going to be very interesting.

OK, now for the fun part: Let's assume that for the rest of the year teams will win at their current Pythagorean rate, but, of course, keep the wins they already have. Then we get the following standings:

2005 American League Standings
EAST	W	L	Pct.	GB
Boston	90.1499	71.8501	0.55648	0.0000
NY Yankees	87.9191	74.0809	0.54271	2.2307
Baltimore	86.4629	75.5371	0.53372	3.6869
Toronto	85.2847	76.7153	0.52645	4.8652
Tampa Bay	52.9923	109.0077	0.32711	37.1575
CENTRAL	W	L	Pct.	GB
Chicago Sox	102.4070	59.5930	0.63214	0.0000
Minnesota	89.6109	72.3891	0.55315	12.7961
Cleveland	87.9241	74.0759	0.54274	14.4829
Detroit	81.1966	80.8034	0.50121	21.2104
Kansas City	58.1550	103.8450	0.35898	44.2520
WEST	W	L	Pct.	GB
LA Angels	95.1631	66.8369	0.58743	0.0000
Texas	87.8488	74.1512	0.54228	7.3143
Oakland	82.8355	79.1645	0.51133	12.3276
Seattle	75.4218	86.5782	0.46557	19.7413
2005 National League Standings
EAST	W	L	Pct.	GB
Atlanta	93.9466	68.0534	0.57992	0.0000
Washington	88.5878	73.4122	0.54684	5.3589
Florida	83.5174	78.4826	0.51554	10.4293
NY Mets	81.5781	80.4219	0.50357	12.3685
Philadelphia	80.8821	81.1179	0.49927	13.0645
CENTRAL	W	L	Pct.	GB
St. Louis	102.8784	59.1216	0.63505	0.0000
Houston	81.8095	80.1905	0.50500	21.0689
Milwaukee	80.7379	81.2621	0.49838	22.1404
Chicago Cubs	80.5000	81.5000	0.49691	22.3784
Pittsburgh	72.7981	89.2019	0.44937	30.0803
Cincinnati	65.5210	96.4790	0.40445	37.3574
WEST	W	L	Pct.	GB
San Diego	86.4367	75.5633	0.53356	0.0000
LA Dodgers	73.5189	88.4811	0.45382	12.9178
Arizona	72.0555	89.9445	0.44479	14.3812
San Francisco	68.8848	93.1152	0.42521	17.5519
Colorado	59.7777	102.2223	0.36900	26.6590

Based on these predictions, Atlanta will keep its accustomed first place in the NL East, Washington is a good shot for the Wild Card, just because they've won so many games in the first half of the season. And Boston and Minnesota will hold one to the AL East and the Wild Card to (yeah!) keep the Yankees out of the postseason.

Of course, these projections are just that, projections, and aren't a guarantee of anything. Don't even think about using these to place bets. At the end of the season we'll see how well these predictions stack up.

---

^* I could not love baseball so much, did I not hate the Yankees more.

Where Have All the B-Movies Gone?

Gone to Sci-Fi, every one.

MP3 Rips With Sound-Juicer

Well, things work well when you get help:

A bit ago, I said that it seemed hard to get sound-juicer to rip CDs to MP3 format. But then, I got feedback from Ross Burton, the developer of sound-juicer, who told me to look for the gstreamer-lame plugin to handle MP3 files.

This took a bit of detective work: Mandrake has the plugin under that name in the Penguin Liberation Front archives, but the dependencies to install it weren't compatible with Fedora's shipped version of gstreamer. Eventually, I thought to look in the the list of available RPMs for Fedora Core 4. Doing
$ yum list > yum_files
and then looking at the output, I found the line:
gstreamer-plugins-mp3.i386  0.8.8-0.lvn.1.4  livna
which looked promising. (The program is in the Livna repository, which is I installed with help from Fedora Core Tips & Tricks.) So install the package via:
$ sudo yum install gstreamer-plugins-mp3
and off to the sound-juicer manual to learn how to add MP3 ripping capabilities:

1. Run the program gnome-audio-profiles-properties
   1. Press "New" to get a create a new profile, then "Edit"
   2. Name it MP3
   3. In the box labeled "GStreamer Pipeline" enter:
      audio/x-raw-int,rate=44100,channels=2 ! lame name=enc
   4. Fill in the "File Extension" box with "mp3"
   5. Click the "active" check box
   6. Exit
2. Start up sound-juicer
3. Under "Edit=>Preferences" go to Output: and select "MP3"
4. Rip normally

And this works: except that I get error messages of the type
Couldn't find matching gstreamer tag for track-count
Couldn't find matching gstreamer tag for encoder
Couldn't find matching gstreamer tag for encoder-version
for every track, and at the end sound-juicer hung up and I had to kill it manually. We'll see if this problem persists.

Anyway, thanks again to Ross Burton for helping out.

What Do You Do With a 873-pound Tuna?

Apparently, you explain to your wife that the house is going to stink for a few -- years.[*] And prepare your kids to eat tuna steak forever.

This fish was caught in the mouth of the Delaware River. I'm not sure I'd be eating it at all, but, OK, if you want to live dangerously ...

[*] If you didn't know about BugMeNot.com, now you do.

Linux, Firefox, and rtsp Revisited

Without a doubt, the most-accessed entry in this blog is one I did on getting the rtsp aka RealPlayer protocol to work with Firefox. Don't believe me? Go to Google, enter "firefox rtsp" in the search bar, and then hit "I'm feeling lucky."

Anyway, as a result of feedback I requested on that post, I've got some new information. You don't have to edit configuration files buried three levels deep to fix this. No, you just have to do a little bit of mouse clicking to change the configuration file from your browser. This post describes the procedure for Mozilla, but the same trick works in Firefox, to wit:

1. In the address bar of your browser type the string
   about:config
   and hit return. You'll get a long confusing list of options. Some are in bold, some aren't, but don't worry about that.
2. Right-click anywhere in the display area. A box will appear. Click on the New option.
3. Another box will appear. In it, type
   network.protocol-handler.app.rtsp
4. Yet another box will appear. In it, type
   /usr/bin/realplay
   or the location of whichever program you want to work.

That's it, RealPlayer should now work. You can check your work here.

Windows Users! I note that some of you have been searching for a solution to this problem as well. The problem seems to be in RealPlayer, not Firefox. If you go to Word_Whore, scroll past the politics, and look at the bottom of the post, you'll see that you should enable Real-Time Streaming Protocol by clicking on "Preferences=>Advanced=>Other Media" in RealPlayer and enable RTSP.

Hope this works, I haven't tried it myself on a Windows machine.

!!THUD!!

It looks as though Terry Pratchett's next Discworld book, Thud, will come out in October. You can already preorder on both Amazon and Amazon UK.

The US version is cheaper over here, so I'm going to buy it, but I was tempted to get the UK version. Why? Well, look:

UK Edition

US Edition

There's no comparison.

Ripping Direct to MP3

Of course, the main purpose of previous entry was to convert ogg formatted sound files to mp3 so that I could burn them to a CD and play them in my car. Why did I need the conversion? Because I was using sound-juicer to rip files from CDs, the juice defaults to ogg for compressed files, and anyway in FC4 there is no support for MP3s in there.

I thought that maybe, just maybe, I could use the same trick on sound-juicer as we did on SoX, but that turns out to be rather difficult. So the best thing to do is follow the advice in that link and use Grip to burn CDs. As long as the lame package is present, Grip will rip CDs to MP3 format if you ask. There is a small bit of configuring required, but it works very well.

Now I just have to re-rip all those CDs.

Putting MP3s into SoX

I mentioned previously that one of the things I wanted to do with Fedora Core 4 was to find an MP3-enabled version of SoX, "the Swiss Army Knife" of sound processing programs. In FC3 I found a SoX+MP3 RPM file online. I still haven't found one for FC4. Fortunately, however, I found a hint for creating such an RPM online (search for "sox").

The procedure is rather simple, if you're willing to rebuild RPM files:

1. Set up your account so that you can build RPM files.
2. Find a copy of sox-*.src.rpm, where the "*" is the version number. The current version is 12.17.7-3, so you'll look for sox-12.17.7-3.src.rpm. This is available in any mirror of Fedora Core 4, I found it in
   ftp://ftp.linux.ncsu.edu/pub/fedora/linux/core/4/SRPMS/
3. Copy sox-12.17.7-3.src.rpm to ~/rpmbuild/SRPMS.
4. Make sure the packages lame, lame-devel, libmad, and libmad-devel are installed. If not, you'll need to install them. Assuming you followed Fedora Core Tips & Tricks to enable the freshrpms repository, you can do:
   sudo yum install libmad libmad-devel lame lame-devel
   where sudo allows selected users to run root commands. If you these packages are already installed, yum will let you know about it.
5. Now create the RPM. From a terminal window:
   $ cd ~/rpmbuild/SRPMS
   $ sudo rpmbuild --rebuild sox*
   and wait awhile.
6. When the program is finished,
   $ cd ~/rpmbuild/RPMS/i386
   $ ls sox*
   There should be three files with RPM extensions: sox*, sox-devel*, and sox-debuginfo*, where "*" hides all the version numbers and stuff. Since these RPMs were compiled with the lame and mad libraries, they contain an MP3 enabled version of SoX.
7. Now for the tricky part: we have to get the non-MP3 version of SoX off the system and replace it with the MP3-enabled version. This is difficult because a) the version of SoX on your system is probably the same as the one you just complied, so the rpm database won't recognize your version as an upgrade; and b) some programs depend on SoX to run, so the dependency checker in rpm will generate an error. To get rid of the old SoX, then, we must force the uninstall:
   $ rpm -e --nodeps sox sox-devel
   assuming that you had sox-devel on your system.
8. Now we can install the brand-new version of SoX:
   $ sudo rpm -i sox-12.17.7-3.i386.rpm sox-devel-12.17.7-3.i386.rpm
9. Now test the program. Run
   $ sox -h
   At the bottom of the output is a list of supported file formats. If you see "mp3", then SoX works.
10. Use SoX as before.

Note that if freshrpms ever updates SoX, you'll lose the MP3 capability and will have to repeat all of this stuff again. :-(

Windows: Another Hole In IE

If you're still running Internet Explorer under Windows (and 46% of you reading this are), then you need to take note of yet another security hole in IE.

You have to dig deep down in this advisory to figure out what the problem is. Clicking on the various buttons that hide text, it seems that the bug triggers the Java Virtual Machine to crash IE and "gain the same user rights as the local user," which in Windows usually means administrative privileges.

If you have to use IE, Microsoft has a few suggestions, but the best one is in this Washington Post article: switch browsers.

Of course, Linux & Things recommends Firefox, but there is a wide variety to choose from.

July 6: Microsoft has released a patch for this bug

A Measure of our Ignorance

Science, second only to Nature in prestige, is celebrating its 125^th aniversary by publishing a list of the Top 125 Things We Don't Know. Actually, that should be The Top 125 Things We Don't Know But Which We Might Figure Out Soon, but it's still a good list, including things like:

• Are we alone in the Universe?
• What is the Universe made of?
• Why do humans have so few genes?
• How, and where, did life here arise?
• How hot will the greenhouse effect heat up Earth?
• Is Quantum Mechanics all there is?

and much more. Read it.

Using the Backspace Key in vim

The Fedora Core 3 version of the vi-clone text editor vim let you use the backspace key to delete the character you just typed, just like an old-style keyboard/typewriter.

In Fedora Core 4 the default is that the backspace key is the same as Ctrl-H, like on an old style teletype. To fix this, it's necessary to edit the ~/.vimrc file, adding the line:

inoremap ^? ^H

Actually, to type that using vim, you press the keys:

inoremap Ctrl-V Ctrl-Shift-? Ctrl-V Ctrl-h

as Ctrl-V means: insert the next key I press literally, dummy.

This makes vim behave the way it should in insert mode: Pressing the backspace key (or Ctrl-H) deletes the character you just typed.

Getting A Text Terminal

Way back in the past, even before the X window system for Unix-like operating systems was in widespread use, it was realized that sometimes a simple terminal screen wasn't enough.

So it was decided that there should be multiple terminals available to the average user. These days the default number is set at 7. So if, for example, you're logged into terminal #1, and you want to do something completely different, you can hit Alt-F2, and suddenly you're in terminal #2. You can even log into terminal #2 with a different account, so two people can work on the same terminal screen, provided that they don't beat each other silly fighting to press Alt-F1 or Alt-F2.

When X windows was added, the extra terminals were left intact. A good decision, as anyone who has watched a Windows graphic freeze up with no solution except to reboot can attest. In fact, in a default Linux system the X-window graphics module runs in terminal #7. The only difference between systems with X and systems without X is that you need an additional keypress to get to terminal #3, say, pressing Ctrl-Alt-F3 to get to terminal #3 and Ctrl-Alt-F7 to get back to X.

So if a program does hang up your X-windows system, all you have to do is hit, say Ctrl-Alt-F1, log in there, do ps xawu to find the offending process's ID number, and kill it. Most times X will come back by itself, but if it doesn't, you can log in as root and kill X. It will restart when you do Ctrl-Alt-F7.

Except with Fedora Core 4 and my Intel 82845G/GL "integrated" graphics device, it didn't work. Hitting Ctrl-Alt-F1 gave you a blank screen. I think that you could log in and Do Stuff, but I couldn't verify anything, and logging in seemed to freeze the Ctrl-Alt-F7 mechanism to get back to X.

This bug was often reported, but not fixed. Now, thanks to Fedora Weekly News, I find that there is a workaround (see Bug#2 in this link). It seems that the offending software is part of the new X distribution, specifically the library libvgahw.a. Apparently there is something in the new gcc 4 compiler that doesn't like the source code for this library. So the solution is:

Duh, use the library file from Fedora Core 3. The procedure is:

1. Download the latest and greatest version of this file from
   ftp://people.redhat.com/mharris/libvgahw.a
2. Log in as root.
3. Copy the downloaded version of the file to
   /usr/X11R6/lib/modules/libvgahw.a
   you might want to rename the old file first, just in case you need it around.
4. Restart X and see what happens. For me, it works.

This isn't the optimal way of doing things, of course. Hopefully a real fix will arrive soon.

Working With Fedora Core 4

I've collected all of the Fedora Core 4 posts from this blog and put them, in chronological order, in (duh):

Working With Fedora Core 4

As in the past, I'll keep adding posts to the file as long as I'm using FC4.