Archive for January, 2008

One thing that makes me a little sad, though..

Friday, January 25th, 2008

All the tracks that are getting all the downloads were not major production numbers.. the best of them were recorded in a few hours, the worst of them were just captures of jam sessions. The Mischief Committee album, which had hundreds of hours of work put into it.. the most downloaded track on it got 30 downloads last month.

Please, if you haven’t listened to it yet, download the Mischief Committee album.

hey now, you’re a rock star..

Friday, January 25th, 2008

As I recall, at some point about this time a few years ago I was convinced that earth was a radio star – after all, we do emit a buncha electromagnetic radiation.

Come to think of it, I may still think that ;-) But, anyway, I also had some explanation for why all humans were rock stars, which I don’t remember at all. That’s *not* what I mean here.

Also, in all fairness, if one had a sense of proportion about the numbers I’m about to recite, one would admit that they don’t even stand up to the coffeehouse circuit, much less make me Bono.

However, a track I did with the lovely and talanted Esen had 811 downloads last month. I have more than ten tracks on my web site that received more than 100 downloads each in December, and more than five that received more than 200 downloads. Mischief Committee has several jam sessions which received more than 100 downloads. And, the trend in my web traffic has consistantly been upwards the entire time I’ve been maintaining sheer.us.

To me, these are rock star numbers. This is a larger audience than I ever dreamed of reaching.. it inspires me to want to write and release more tracks, to say the very least.

Changing history

Thursday, January 17th, 2008

Speaking of books, my friend How Kuff wrote one that I’ve been slowly working my way through – it’s not a traditional novel, but it has a lot of good ideas. A lot of the same things that bother me bother How, and he has a little more distance from them so he expresses them more artfully. His characters are fighting with problems like what to do about the moral repercussions of your day job (if you make bombs, are you responsible for those they kill?), what the costs of endless commercialization might be, and what it might cost you to try and fight the Powers that Be – and what it might gain you.

You can buy a copy of his book at www.changing-history.com. I suspect that it’s one of those works that will take me many passes to fully appreciate, but there’s definitely some good content there.

Microcontroller-controlled commutator

Wednesday, January 16th, 2008

I should really post about this on the EVDL or evtech, but I lack the energy at the moment and besides, I’d like to get the idea kicked apart on my journal first.

It seems to me that you could build a very cheap, very powerful drive for a EV that had a lot of the advantages of a DC drive combined with a lot of the advantages of a AC drive by having a standard DC series motor which did not have the commutator mechanically connected to the shaft – instead, the comm would be driven by a stepper motor controlled by a microcontroller which would have a rotary encoder to tell it the current position of the main drive shaft (the output from the ‘power stage’ of the motor), the current position of the commutator, whether any arcing was occuring, etc.

Maybe I misunderstand some of the theory, but I think that such a motor would be able to do regen braking fairly easily, would not fail in a full-throttle on mode (because the only way it produces motive power is if the comm is staying synchronous with the main output shaft, which requires that the microprocesser be doing its job), and that you would be able to get considerably higher power drives in considerably smaller spaces because you would only need a single H-bridge of power silicon to control power to the motor instead of the three H-bridges that a AC drive requires. However, you would still be able to get AC’s flat torque curve because you would in essence have the same amount of control over the magnetic state of the motor that a AC drive has.

Anyone want to kick a hole in it? Yes, it has brushes, which are mechanical and therefore frowned upon. However, I think it could be made simpler and cheaper than a ‘real’ AC drive, and aside from needing to swap the brushes every few hundred thousand miles – which a sensor could tell the micro about – it would have all the advantages of ‘real’ AC.

My next challenge on this vein of thinking is to mentally walk through all the failure modes – i.e. what happens when the comm gets energized in the wrong position -out of sync – with the current shaft position of the rotor..

Tuesday, January 15th, 2008

You all are wonderful ;-) I appear to have more than enough work for the next few months.. thank you to one and all, and sorry it didn’t work out to the people who I had to turn down.

(I hate turning people down, because I feel like I’ve lost the opportunity to work for them forever… and I really do want to help people do things, beyond the money aspect)

I think I’m maybe a bit overly afraid that I’ll be unable to find work. Thus far I’ve never had to look very hard to find someone who wanted me to do something for them – I just feel like surely there are a million people who are better, faster, smarter, and more creative than I am – but what I forget is that there are probably billions of things that want to be built – far more things that need fixed or created than people to fix or create them – and that I do genuinely produce – eventually – good products. I’m sure that there’s still much to be learned, but I’m also sure, at least on some level, that I’m not completely worthless.

Of course…

Friday, January 11th, 2008

as usual, sysinternals.net had a tool. ;-)

What’s really terrifying is what I found out *using* that tool. I was having trouble starting IIS on my win2k machine, because something else was already bound to port 80. Guess what it was…

…. skype.exe

Yes, boys and girls, *skype* is a web server. I guess that’s one way to get around those pesky firewalls – almost *everyone* lets people open a connection to port 80…

Ah, how I hate them, let me count the ways…

Friday, January 11th, 2008

this microsoft support article in essence says M$ is holding hostage a hotfix that will let you see what processes on a windows 2000 computer are opening what network ports. They could have easily just included a link to the hotfix, but this article seems to me to be saying, ‘we are going to make you pay $200 for a support incident to get a copy of this tool’.

Looking for work..

Thursday, January 10th, 2008

My boss decided to cancel the software development project that I’ve been working on for the past year. My guess is it would have taken me about a month and a half to fix the bugs in the user interface code – which was being written by a offshore development firm. However, he didn’t want to continue working on it.

This leaves me in the position of looking for work. My resume is at http://www.sheer.us/resume/SheerResume-html-ca.html, although I’m working on a nifty database-driven searchable resume which I hope to post sometime next week. If any of you know of a position that would be a fit for me, I’d love to hear about it.

Housekeeping..

Tuesday, January 8th, 2008

So, my spam folder – just the spam automatically weeded out by SpamAssassin – was 1.3G this year. I’m tar/bzipping it up and then deleting it, and I’m really thinking someone needs to do something a lot more aggressive about spam.

I’ll post a followup to this with how much you can compress spam, because I’m curious. Does it compress better or worse than regular mail? ;-)

Time estimation..

Tuesday, January 8th, 2008

Well, I have yet again blown a time estimation. I guessed a job at 80 hours that took 200 – and because I wasn’t keeping track, I didn’t really notice until my boss totalled up the bills.

I don’t know what to do – I feel awful about it. Most of the time was spent getting audio playback working in a java application on a sprint phone – sprint requires the midlet to be signed in order to have access to the audio playback API (someone, please tell me there’s a reason for those?) and it took me forever to find out how to get the midlet signed – especially since the directions on sprint’s web site were *wrong* – and to figure out how to test to see if I had developer root enabled on the phone – and then I had to fix some threading issues that came up on the phone but not on the emulator – and a hundred other little things. The rest of the overrun was a set of fixes to a windows CE application that I thought would take a few hours .. but which instead took 50.

I suck at time estimation.  And I try and warn people that I suck at time estimation, and I try to never take jobs where I get paid a flat rate because I know how badly I can get screwed by those because of my lack of time estimation skills..

I don’t even know what the fallout from this latest screwup is going to be. Will they refuse to pay me? I’m already sure they’re upset about me not keeping track of my estimate – the thing is, I did tell them when things weren’t going well, but I didn’t say ‘I’ve reached the number of hours I guessed at and I’m nowhere near having things working’ because I wasn’t even keeping track of how long it was taking..

Compared with the problems of my friends.. one of whom has a friend in the hospital because he was attacked by a crackhead with a 2×4, another of which has had a good friend suicide.. a third of which has just had their grandfather die.. I have nothing at all to complain about, and I really shouldn’t feel so awful..