Mark's 20 Percent Time

I'm nearly 3 weeks done with NaNoWriMo and still haven't picked up another project yet.

I've gotten the parts for my new desktop, the one on which I hope to install Xen, Win7, Ubuntu 9.10 and possibly someday MacOS. I've played a bit with Xen and haven't gotten to the point where I can easily build a DomU with full-screen (read: usable as a primary desktop.)

I'm officially at the point where I need to just make the machine workable and start migrating stuff from the old box. So I think I'm going to dedicate one drive to Win7 and one drive to Ubuntu and dual-boot for now. Maybe someday down the road I'll throw a third small drive in to run Xen and a Dom0 and can just attach the physical drives to their respective DomUs.

So, that's what I'm going to do tonight. The Goal: Have a working Win7 (see below) and Ubuntu 9.10 install tonight.

I have installed Win7 on this new system already, and it needed new network drivers for my motherboard. I also tried to get it to do software RAID on the OS drive and I think it didn't like that. So I plan to just start over. I also plan to download a new network driver on the old system and put it on a USB stick. :-)

Time to get working.

Yup, you read that right, I've hit my word count goal of 10k. I've never actually been a writer, so I set a more realistic goal for myself, one that gave me a chance of actually succeeding.

And there you have it, I did.

I didn't do so good on my other goal of writing at least a little bit every day. I probably hit about 2/3 to 3/4 of the days, but there were several days where my schedule just did not permit it, and at least one or two where my energy level had a thing or two to say on the matter. But there were more days where I didn't feel like it, but I did it anyway, and that's what I was going for.

I considered cut'n'pasting my manuscript into the NaNoWriMo validator 5 times to compensate for the fact that my personal goal was 1/5 that of their goal. I hit MY goal and I felt I should get something for that. But, I didn't hit THEIR goal, so I decided against it.

Now. The question becomes, what happens next? I'm only about half way through the outline I already have, which is only about half to a third done. And, the 10k words I've written now are filled with TODOs. Read: There's a crap-ton more work to do on this story if its ever to see the light of day.

NaNoWriMo was an interesting experiment. I'm glad I did it. I might even try it again next year with, say, a 20k word count goal. But a lot of other things have been put on hold for this that I really should get back to. I'm about a year late for some Between The Lines editing, I have about four new computer projects in progress that I really need to get back to, and my Seebeck Clock is collecting dust and staring me in the face every time I sit down at my desk. I'd really like to see this story finished, but it will probably have to wait a while.

I'm ahead of my word count goal so far at 6855 after tonight's writing session.

Unfortunately, the last two days I didn't write at all, but I blame work. I "worked" from 7am to 11pm in our Scotts Valley office (200 miles away) on Monday, and went to bed early on Tuesday after driving home. So I'm down by 4 days total for the month, with good excuses though. It's not like I'm getting lazy.

...that's my story, and I'm sticking to it...

Having said, that, I just closed out chapter 3 and my brain is mush. This writing thing certainly doesn't come easy for me. Though, I suppose if it did, I'd have started a lot earlier in life. I think I'm going to watch an episode of FarScape or two and go to bed.

Let me start off by saying that since this is my first serious attempt at writing ANYTHING, my personal goals for NaNoWriMo are: 10k words by the end of the month, and to write at least a little bit every day.

Open Office and the NaNoWriMo page disagree on my word count by 119 words, but in either case, I'm at about 5500 words right now, a bit ahead of schedule, which is good.

However, I am already down by two days toward my goal of writing a little bit every day. The first day was caused by other commitments (see below) which I made up for by writing twice the next day. The seconday (yesterday) I threw under the bus of my personal well being. Between work pages and family waking me up whether I want them to or not (the dogs too; not just Zoe), I haven't been sleeping much lately. So last night I went to bed at 8:30pm and woke up on my own this morning at 5am, an hour before my alarm. 8.5 glorious hours of sleep: precisely what I needed.

So, I've spent the last hour writing before anyone else got it, and got in nearly 800 words, and developed back stories for 7 characters. I'm very pleased with how this mornings writing went.

As for the other commitments I mentioned: I've been helping a friend of mine work on a hardware project involving a tiny surface mount camera, and interfacing said camera to an FPGA. I won't go into the details here yet (I need to get ready for work), but suffice it to say that I finished the circuit and board layout on Tuesday night this week (the first day I missed NaNoWriMo.)

Right. To the day job!

I said before that I was crazy enough to attempt NaNoWriMo and so it appears that I am. It starts today, but I've not been able to get started yet (it's family time right now," he says while writing a blog post on his phone.)

Maybe a little later today.

In other news, MovableType works, but not exceptionally well, in Safari on the iPhone.

I wasn't really planning on banging out this project so quickly, but it turns out there is a triathlon this weekend that I volunteer for every year (Scott Tinley's Not-Dirty-But-I-Still-Call-It-Dirty Aventure) that's at Lopez Lake, spitting distance from my house. I thought about how far the radio integration project was and realized that if I took the iPhone part of it out of scope, it was nearly done. So I banged nearly all of it out last night, and finished up just a little bit tonight (had to solder 4 more wires).

I just got back from my test ride, and the audio reports are pretty universally positive. There's a little bit of wind noise at freeway speeds, but all things considered, very readable. The GPS is getting signal and broadcasting updates every few minutes on APRS, as KR6ZY-2.

A few things I'd like to fix as time allows, but not blockers for this weekend:

• The PTT is a little flaky. It's spent a lot of time in the elements and can stand to be replaced.
  
• Audio is only in one ear. I know why and its easy to fix. I just need to bridge two pairs together.
  
• I still need a UHF antenna (Ok, I lied. This IS a blocker for this weekend, but I think there's one waiting for me at the shack at Cal Poly)
  
• The audio is a bit tinny, always has been. I think there some caps in the audio path that are too small, making a high-pass filter. I'd like to increase these some to drop the frequency response a bit.
  
• Make the whole radio system switched with accessory power. Right now, it pulls directly from the battery; if I forget to flip the switch in the luggage, it'll kill the battery pretty quickly.
  
• Integrate iPhone audio. This is the much larger project, involving mixing and audio amps, and microcontrollers so I can hit the big stop/start button and what-not. Being able to charge the iPhone while riding is a plus.
  

So, yeah. Just a few things... ;-)

Thanks to Nick, KF6UZB, for helping me out with the audio checks this evening.

In further Mark-Doesn't-Have-Enough-Time-For-His-Current-Projects-What-The-Fuck-Is-He-Thinking-Now news, I've signed up for NaNoWriMo. I have no real clear understanding WHY I've done this, but I have. I kinda have a story idea in mind, but not anything closely resembling a plot or outline or anything.

...NO idea what I was thinking...

Strictly speaking, today wasn't a 20% day, but since Cindy is itching to have some of this work done, I was able to spend a bit of time working on Koi today.

Some more research suggested I need to get a different driver installed for this chip set. Further research suggests that the driver that's being loaded is compiled into the kernel statically by RedHat, making it neigh on impossible to load another module in its stead. So I've just kicked off something akin to madness: I'm starting a CentOS kernel RPM compile, after hacking the config file to remove the "bad" driver and compile the "good" driver as a module... ...on an 800MHz C7... ...with a RAID array that's in PIO mode... Yeah, this is going to take a while...

So, I'll come back some time in 2011 with a status report. (In all seriousness, I suspect it'll still be going when I head to bed in (hopefully) not too long.)

While waiting for things to happen with Koi, I put together my Atom board, 1GB of RAM, a 20GB 2.5" laptop PATA drive (quiet and low power), and a DVD-ROM drive. This is likely to become my Boxee system. I can't find my XP install CD, so I installed CentOS on it as a proof of concept, to make sure the system works. This is the same system that was giving me fits with the SATA drives earlier. Still haven't figured that one out...

Anyway. Once I find my XP install, I'm going to put Boxee on it and hope its fast enough. I may try Ubuntu/Boxee in the mean time and just accept that I won't be able to watch Hulu or NetFlix.

Been keeping busy... :-) Go bears!

Oops, forgot to write this yesterday when I was done.

I wrote out so many wonderful ideas to work on with koi (the name of the new server.) I was expecting the OS install to be the easy part, right?

...not so much... First of all, I thought I had an ISO for CentOS 5.3 i386 already. Turns out I had a couple aborted downloads of i386 and a good disk for x86_64. So, I spent the first 3.5 hours of the day downloading a DVD ISO and researching software.

Then when I finally got started with the OS, it took about 2 hours to format the 1TB RAID array, and that is without having to re-build the RAID array itself (was already built from the last time I tried this with CentOS 5.0) Something is seriously wrong with these drives.

After the format, the install was relatively uneventful. Started playing around, and came to the conclusion that I don't want the OS on the CF card; I'm too afraid of write wearing on a non-redundant file system. So I took out the CF card and re-installed the system (again) on the RAID pair. This of course necessitated a repartition and therefore a RAID rebuild, which is still going on now.

So, yeah. These are brand new 1TB SATA drives. They should _NOT_ be going this slowly. Something is seriously wrong. hdparm claims that they're not using DMA; I'm not even sure I know what PIO mode they're using. When I tried to enable DMA using hdparm, it claimed that function was not allowed. The BIOS doesn't have any useful settings on the SATA controller other than "enabled" or "disabled."

There is a newer BIOS that I'm trying to get installed, but my DOS boot USB stick doesn't work on this system, so I'm having to wrangle some other tricks to get FreeDOS installed on a partition I left for it on the hard drive. I just haven't gotten to this part yet, so I don't know whether it works. But the goal is to have a DOS environment on the hard drive that is bootable from the normal grub install that I can use for BIOS upgrades and what-not.

I'm _HOPING_ that a new BIOS will fix a lot of the problems I'm having (possibly even the inability to boot from USB!). I'll report on that when I've tried it.

But if I can't fix the SATA, this motherboard, as cool as it is with its 4 physical serial ports, is not going to work out. :-(

In the many hours I was waiting for koi to do thing, I started researching media PC options. I looked at MythTV, The Roku Box, and Boxee.

I have a Myth front-end on my neighbor's back-end which is cool, but it a) is illegal, b) only kinda sometimes works, and c) likely has a limited time-line once Charter decides to ditch their analog channels. We're looking for alternatives.

The Roku box looks pretty cool as a hardware platform, but it only supports NetFlix, Amazon Video On Demand and Major League Baseball (which I really couldn't care less about). Those two are cool, but I really want a better selection, and Boxee seems to be the best available. Roku claims that they'll add more as time goes on, but I'm not willing to bet $100 on that just now.

So I'm really looking for something that will get legitimate content, and a wide array of it, over IP. Of the three, Boxee running in Windows seems to be the most functional option. Windows is required for Silverlight and therefore Netflix. This is a bit of a bummer because it means I likely won't use the system for anything else. But if I can build a small, low power, quiet system that'll do Boxee, I'll be happy with that. So that's what I'm playing with. I'll let y'all know how that turns out. :-D

Today's project is to keep working on the rebuild of our home file/DNS/DHCP/everything else server. I have a new low-powered motherboard and a pair of Western Digital Black 1TB drives and a quiet power supply. I just need to put it all together and setup the software. Having lost my old 1TB drive, I don't have anything to port over, so if there's a silver lining to be had with the loss of 1TB of data, it's that I get to start clean now.

The problem is, the CentOS 5.0 install DVD I have doesn't support my network chip, which means I can't get on the network to install updates which support my network chip. *sigh* So I'm now downloading the newer CentOS 5.3 install DVD, but that's going to take another 3 hours. I _THOUGHT_ I already had this downloaded somewhere, but for the life of me, I can't find it.

In the mean time, I'm putting together a list of all the functions I want this box to perform and investigating software options. So far, the list is:

• The obligatory DNS, DHCP, Samba, and print server


• iTunes streaming server (DAAP) of the files contained on the server


• Graphing data from my solar electric system on the house


• X10 controller for lights and stuff (specifically, lights on my fish tank)


• APRS data


• Remote control of my Ham Radio HF station


• Possibly a media player, connected to the TV in the other room.

Off to investigate software options for the above.