John

First seven (work) days at the new job...

2013-05-16T04:47:20Z

So, there's good news and bad news.

The good news is that I'm working for SQL Support again. The old skill set is coming back quickly, and I feel really good.

The bad news: Microsoft is hoping to run a V- organization here. That is: they want to hire a company to provide services. Which means that I might end up being a star for a contracting firm, *not* for Microsoft, and did I mention that the contracting firm, like all contracting firms has awful benefits compared to full time employment? And I have a major concern about the medical plan, which is grandfathered in under the Affordable Care Act so it can still have annual limits on the amount they'll pay. The limits are relatively high, but I could see a medical crisis burning through it in a scary-short period of time.

Still, it's a chance to dig back into a job I loved doing and was damn good at. It's a steady paycheck, and I'll get by for now. What happens next will be whatever happens next.

The first few weeks are intended to be training. I'm re-reading a lot of old stuff and learning a bit that's new, which is always good. Then, we'll have to see what they want us to do. Typically, they start all engineers in "core" - basic installation and configuration. (You would not believe how complicated setup can be for a major product like SQL Server.) I wouldn't mind getting my feet wet again, but I hope to jump into the advanced team quickly. I'd like to be a great onsite resource for the core engineers at my location but work the advanced cases.

Life has been good in general, but I'm adjusting to new schedules and having the normal "ugh, I have to get up on time, and go to work!" adjustment period :-).

Pondering on emotional ownership/choice

2013-05-09T22:22:56Z

I was reading a book, and I saw a passage about how people "make a choice" about how to feel.

That view is a decent model, I think. Like most models, it's good for what it does, but it's not capital-t Truth. And it doesn't have to be - it's a model!

What's good about it?

It encourages ownership of emotional states. If someone does something, and I get angry, if I view anger as a "choice" then I can recognize one fundamental fact: I'm angry because of stuff that's part of me. And I'm only going to become non-angry because of stuff that's a part of me.

In some ways, this seems ridiculous. If someone punches me in the face, didn't they *make* me angry? Well, no. I might have become scared, or I might have become proud (if I was teaching fighting skills!). If I was a perfect Buddhist, I might have viewed it with detachment - "wow, my face hurts, and this person seems to want to hurt me - and boy, do I feel a surge of adrenaline now!"

Thinking about other people's actions might help me decide if the emotional state I'm in is helpful or productive, but it doesn't change that my reaction occurred because of stuff that's within me. So if I see someone attack another, I might become angry, and might think that's the appropriate response. But the attacker didn't make me angry.

The model breaks down in one important way. "Making a choice" would, in most people's minds, mean making a conscious choice. "That person hit me in the face, I'm going to be - uh - oh, hell with it: angry!" More importantly, one can't "choose" to be not-angry, exactly. A person can realize "oh, right, you *did* tell me you were going to be late, and I forgot, so there's no need to be angry" and just stop being angry. Once a person has had a strong feeling, it alters body chemistry a bit, and it takes time for the chemistry to catch up to changes in emotional state.

What is true is that there's always a choice about what to do, right now, about how one is feeling. "I feel angry. I should..."
"...be angry - there's an injustice being perpetrated!" or
"...try to let it go; it's not worth being angry about," or
"...try to be not-angry, because no one did anything wrong in this case."

Those are all choices one can make, and they may be good or bad choices, depending on the situation, and the actions taken in response to those choices may work out well, or might not... life is like that.

But making choices is usually better than making the default choice of "...do nothing, and just let my emotional responses and mental state push me around, because I'm not aware I can make choices."

That is the ultimate usefulness of the "you choose how to feel" model.

Good things are meant to be...

2013-04-24T19:24:12Z

So, when I had a two week wait on my start date, I immediately thought of one thing: kittens!

I don't like leaving kittens alone all day. I like to have contact with them a lot during the first couple weeks. I want to make sure they know exactly where the scratching post and litter box are, and I want to be sure I'm aware of potential problems (like computer cords! I hadn't thought of that!).

That Friday, when everything was set, there were no kittens to be found.

That Saturday, there was one six month old - I'd almost chosen him, and would have if there'd been another. But for kittens especially, I want a pair.

But they said there were more coming in the next day - they were getting neutered and spayed Saturday. Sunday morning, I found two adorable grey tabbies... littermates, so a lot of socialization issues are taken care of (they're not as likely to decide they hate each other). And the volunteers told me that two boy kitties were typically okay, as were boy and girl, but two girls was possibly asking for trouble.

True? Not true? I know the odds are better for littermates in any case... and even if toms are normally in need of settling dominance, they were fixed at just over two months - no tom behavior was likely.

I brought them home to a single room. I'd never thought of this until the volunteers brought it up, but it made sense. I'd had lots of friends have vanishing kitty syndrome, where a kitten left alone in a big house finds hiding places. So I brought them to a single room so there's no big, scary huge house for them to panic about... and they soon started playing. I left them alone some, and played with them some, but kept them in that room.

That night, I moved them, food, and one litter box to my bedroom. This was a mixture of "good idea/bad idea". Sleeping with new kittens is awesome... except for the "sleeping" part!

Still, I did get some sleep, and woke up once to find one kitten nestled on top of me, and the other curled up at my hand. I took that as preliminary approval of me as an appropriate Cat Person.

Next day, I took them back to "their" room, but after another hour, I opened the door. They started exploring, and now, well, you know cats... now they're declaring the house theirs, but figuring there's *plenty* of house to share with that strange beast with peculiar habits, too few feet, and not enough fur - but a warm body, cuddly hands (when they don't turn into the Evil Hand Monster), and opposable thumbs for operating can openers (or can pulltabs), laser pointers, and other wonderful things.

They're learning to deal with linoleum - I'm pretty impressed! - as well as carpet, deciding which cat toys meet with approval, and which are less fun than boxes and clutter. And I'm learning to be a bit more carefully organized, because cats don't know the difference between "desk" and "folder/book overhanging the edge of the desk".

I've realized I want to rearrange my computer stuff - one poor kitty got hooked by a speaker wire and went into panic mode - she didn't *want* my help, but once I got her untangled, she accepted a cuddle. But I'm really glad I was there - I don't know if she could have hurt herself or not, but she could have been royally freaked out for an extended time if I hadn't.

My house has been catless for far too long; I'm glad that's changing.

Why I've been quiet Sunday and Monday...

2013-04-23T18:48:18Z

So, I was thinking of posting more frequently, but Sunday and yesterday my life got a bit busier....(ED: Pictures aren't working. Sigh. Cut for link, and for pictures if I ever figure 'em out.)

More on (I hope)https://plus.google.com/118200625154501628667/posts.

I can stop any time I want to....

2013-04-13T02:12:58Z

So, I bought a Freshman Physics book. And I started working problems. And I was appalled.

I didn't have a scientific calculator! Well, there are Android apps for that - so, that's fixed.

Then I found how hard it is to sketch vector diagrams on lined paper. Wait - how did I not have large quantities of graph paper around? (Any old D&D enthusiast could ask the same question. I still have my polyhedral dice, though!) And I couldn't lay my hands on my scale! (Um - those triangle-cylindrical rulers that let you measure to different scales.) That's not *critical* - but a scale can really help one make a more accurate sketch. Technically, you could use a ruler, "this line is kinda like 3 times that one" but a scale just makes that easier.)

There - I now have them, and some extra mechanical pencils to boot.

But I did not get triangles, a compass, etc.. Yet. (Yes, yes, I also haven't had time to paint it or build it to scale....)

I'm done with vectors - trig is easy, after all.

(Ouch. I may have made some enemies with that - remember, I majored in math!)

I even figured out my problem on the one question that had the rounding error - one answer came out to something like 7.83 and the other came out to 7.858 - so, yes, rounding to 2 significant figures would introduce a difference in the answers of .1, which makes perfect sense - the actual difference was less than .03, so both answers were correct, within the margin of error.

Time to "move on". (Hah! Get it? Because the next section is *motion*! I'm *moving on*... okay, I guess I'm taking high school nerd too far now, aren't I?)

Seriously, this is good and therapeutic... I wish I'd taken physics after I started acing math classes, because I just barely remember it, and it was never quite as interesting as it seems to be when I'm following the math in my head.

Question for the hive mind...

2013-04-12T20:08:33Z

Okay. Here's the deal.

I have a friend who used to have a cat (probably brain damaged, as in, probably suffering from head trauma) who peed everywhere. That led to another cat adopting the behavior, too. They both died, and a while later, she got two new cats, and one is peeing in inappropriate places.

I own a house, now. I don't want a peeing cat in an apartment, either, but - I definitely don't want a peeing cat in my house. (Um. I hope no one goes too literal on me - obviously, I want my cats peeing - in the litter box!)

Here's what I've found on the 'net. Most peeing is due to either marking behavior (get your Tom fixed before he's a Tom), or unhappiness or poor health.

Could my friend's new peeing cat just be unhappy? Yeah... yeah, she could.

And I've personally never had a peeing cat, except Skitty, when she had a head tumor. She peed in her bed and against the walls. She couldn't make it to the litter box, I think.)

Does anyone know if there's anything I have to do (other than get happy cats and keep 'em healthy) to avoid having them possibly scent the old urine scents, and thinking "this is where cats pee"?

See, I'm a softie, and I'm horrified by the thought of having a peeing cat who I can't bear to get rid of (since it's hard to get a cat adopted with a known peeing problem - so I could be condemning a pet to an early death if I go to a shelter), and yet ends up doing costly damage to the floors to the point that I hate my kitty.

Does anyone have any information for me? Especially of the "Oh, I had an old cat that did some inappropriate peeing, but the next cat was just fine. They're not like dogs who feel they *must* mark where other dogs marked," variety? Or, of the "Yes, I had a peeing cat, but I did X, Y, and Z to clean up the scents and that fixed the problem" variety?

Doing something I should have done a long time ago...

2013-04-04T19:37:48Z

I've picked up a Freshman Physics text book and am working through the chapters.

Startling self-revelation: I've somehow reached a stage in my life where I simply did not have a scientific calculator.

How did that happen? In high school, I spent a fair amount of money on a good scientific calculator, and afterward, thought it was cool that I'd never be without one again. But then, life happened, my last one got broken, but I found a calculator (left over "take this if you want it" office supplies), and put it away, feeling proud that I had one once again, and never realized it was a *financial calculator*. (Obviously, I use calc.exe if I just need to check my figures - or Excel if it's complicated.)

Thankfully, with an Android tablet, one can quickly find a free or low priced calculator app, and I have one once again.

And now I'm getting frustrated by significant figures....

No, seriously. I thought I knew how they worked, but I think I'm messing something up, because somewhere along the line, I'm getting answers that are just barely different when I work a vector problem two ways (graphically - adding head to tail - and component-wise, converting everything to unit vectors). There's no way that I'm doing something seriously wrong when I get one answer to be 90.3 and another to be 90.1, and the angles match to the correct number of figures, but that difference screams "rounding error" and I'm not sure where it's creeping in. Sigh. Why does it matter if my technique is probably good? Because I want to do these *right*. Like, if I tell you your SQL Server's disks are overloaded, I can explain precisely why, and any SQL expert will agree with me. I want to do these problems so any physics major would agree with me.

This kind of perfectionism can be dangerous - it sometimes leads me to stop doing something because I know I can't do it "right". (If I was learning to play an instrument, or to knit, that would be easier to deal with - those things take practice to learn. But this isn't that kind of thing. Either the numbers are right, or they're not.) Worse, it can sometimes lead me to scorn the attitude that it matters. "Eh. I *got* 90.3; I must have done the problem *right*."

But I also know that if I can understand the "why" of something, it'll take root in my brain and I'll find myself drawing conclusions without even being aware of it.

So I'm momentarily frustrated, but for its frustration, it *is* a bit of fun. I'm doing something that requires some thinking, some re-learning, and where I don't know exactly how things are going to turn out. I think that's helping with depression.

Not that I'd recommend Freshman Physics for antidepressant qualities if you didn't major in mathematics, so that you have absolutely no fear of calculus, trig, or geometry.

Because I know a lot of you care...

2013-03-28T16:42:26Z

http://www.explainxkcd.com/wiki/index.php?title=1190:_Time

(As I joked to kightp, this is one time I could probably just say)

"You're welcome."

The Iraq war - a retrospective

2013-03-27T20:12:44Z

There's a lot of history being rewritten now, so I'd like to put in my perspective.

First: a lot of people are acting shocked, shocked! to find that there were no WMDs found in Iraq.

Folks, before the invasion, we had no solid evidence that Iraq had WMDs.

We had our intelligence agencies estimate that he did - this was specifically noted to be based upon our best assumptions of Saddam Hussein's behavior, given his circumstances and past behavior, and emphatically not based on actual intelligence. We'd had no on-the-ground "category 1" intelligence since the weapons inspectors left.

(I've vaguely heard of categorizations of intel - "category 2" was akin to "reasonable information from a usually trustworthy source" whereas "category 1" was "accepted by this station/agent as true". Are these actual real classifications, or a sign that I've read too many - or too few - spy novels? In any event - we hadn't any any "based on reality" intel since the inspectors left.)

So, we had people whose job it is to look for threats, who said that we should assume they existed. Face it, they should have said that, barring proof that he'd suspended all illegal munitions production; the safest way to fail is to assume he had 'em. But they didn't know.

Ah, but Saddam Hussein wasn't allowing inspections!

The inspectors were in Iraq, and were getting acceptable cooperation. They were destroying weapons they found that were outside of the restrictions imposed on him. Yes, there were negotiations and stubbornness, but the inspectors felt that they were making progress.

The atomic inspectors found him to be clean - these were the people who blew the whistle on him during the Clinton years.

Ah, but we had strong evidence! Like, we had sigint catching a transmission to "clean it all out".

That's evidence for an indictment - not for a conviction. Later, we found caches of dual use chemicals - chemicals Iraq was allowed to possess, but can be used for chemical weapons manufacture. "Clean it all out" now looks like Saddam Hussein was making sure that those dual use chemicals were gone, out of fear that their presence would be used to condemn him.

Given the clear lack of a nuclear threat, and given how little threat Saddam Husein and Iraq posed to us if he had non-nuclear WMDs, there was no excuse to kill a bunch of people, can cause large amounts of suffering to a bunch more. Not unless we had no other choice. And we did. We could have waited, and allowed more inspections.

I've seen people point out that we only had a cease-fire from the first Gulf War, and Iraq was under cease-fire obligations to show they'd dismantled all their biological, chemical, and nuclear research and manufacturing projects. But isn't this a case of never let the law prevent you from doing what's right? You can't hold ordinary people accountable for the responsibilities of an entity that has such power over them. While there was no question that an entity known as Iraq should be pressed into compliance, there was, and should have been, questions about whether the people deserved to bear that burden, unless there was no choice.

And there were choices.

There are a lot of people who are pretending otherwise. But I was there - I was reading the newspapers, and the news websites. And I remember feeling like I was going crazy, where one story would say that Saddam Hussein was being stubborn, and reluctant, but negotiating to allow inspections, and allowing them - and another that said he was "acting like" he had WMDs, this latter story never explaining how one "acts like" one has WMDs. Is this a part of the method?

I will grant this: it was frightfully easy to have fallen into the trap. I myself believed that the US knew that Saddam Hussein had WMDs. No one would ever risk the bare-assed embarrassment of being so sure, so rock solid certain, and end up being wrong, when being wrong means the pointless death of thousands.

It never even occurred to me that "the pointless death of thousands" could be so easily shrugged off. Live and learn, I suppose.

I'm lousy at blogging, but...

2013-03-21T04:58:51Z

... I'm sure other people are interested, too.

You know that song in Horton Hatches The Egg?

It really exists. (And yes, the link to the lyrics works.)

I'm sure there are plenty of people who could have told me this. This is hardly news; this is something well known, I'm sure, and many people will say that it was silly to post this given the breadth of knowledge of this fact on the part of so many people.

If so - I have to angrily demand to know why they didn't tell *me*, hm?

Not exactly musings on Physical Assisted Suicide... (but close)

2013-02-28T00:31:04Z

In thinking about, and debating, phsyician assisted suicide, I had an interesting revelation, one that I think is very important.

I'm not exactly in favor of physician assisted suicide. Is anyone actually "in favor" of it? I reckon some people consider it better than the alternative, but it's a bit mean spirited to make it sound like they approve!

Okay, but my feelings on suicide are complicated. I've seen such strong moralizing about suicide in general, where people who commit suicide are loathed and reviled, that my initial feeling was that the only opposition to physician assisted suicide was this same kind of moralizing. Oh, sure, there are other arguments made, but lots of times people make arguments that they think will sell, when their real motivation is completely different. I just assumed that, deep down, they were hateful of suicidal people, or too unwilling to let go and admit that people die, and I didn't have respect for either viewpoint.

I have a friend who helped me realize that this was not the case. I won't try to make her argument, but I hope I can attempt a rough paraphrase. It goes like this:

At any stage of life, if a person feels suicidal, we consider that a medical emergency, and seek to preserve their life, and return them to a state in which they are no longer suicidal. It doesn't matter why they feel that way, or what intractable-seeming problems they're having, we always fight to protect their lives, and help them to find answers to continue to live out their alloted time. Why would we change that, simply because that alloted time is smaller than it is for most?

As for the problems with end of life, if they are in pain, we provide better pain control. If they are despairing, we provide counseling and other supportive measures. If their symptoms are otherwise horrible, we try to find ways to relieve or ameliorate them. And we do this because this is what we do for living people, even if they are not going to continue living for long. (Note that this does not include forcing life-extending treatments on a person who does not wish them. This is all about letting a person die in their own time, in as much comfort as we can provide.)

That argument is compelling for me; I agree with it completely. At the same time, I see some exceptions. We may not be able to relieve their physical suffering, be it pain, or something else. And there's also just a sense I have that, if I knew I was dying in the next week or so, I might find some empowerment in deciding that I've lived as long as I need, and Saturday night would be a fine night to check out.

And that's where the more important revelation came in.

Who is pushing for the greatest help for those who are dying? The person who says "if we can't control the pain, we need better pain control methods!" or the person who says "it's intractable pain; but we can provide a lethal dose of drugs."

Who is more compassionate? The person who wants to help people break through a natural stage of end-of-life grief, or the person who agrees they might want to die in the same situation, so why not help the patient do so?

Who will push us to find other methods of treating suffering? The person who insists we must figure out how to help people die naturally and without suffering, or the person who might think it's impossible, and besides, there's assisted suicide if the suffering is too much?

We need people who push back. Even if there's some way in which I am "right" about whether some people should be given help in dying near the end of their lives, we need people to push back against that notion so we continue to test our compassion and stretch our understanding so we make it less and less necessary.

More importantly, we need such people to be totally committed, to be angry and unreasonable. If I were ever in a position where I had to make a call regarding assisted suicide (and I don't imagine I'll ever be, unless it's for myself), I'd need to feel that I was letting this friend down. Because otherwise, I might let my own feelings, about how I can understand why a person might want to "check out early", cloud my judgment and not realize that there are other, better options in that specific case.

It's not easy to accept that good people can take all of our carefully constructed, well thought out, compassionately reasoned ideas, and say "you're wrong!" and attack our beliefs and struggle against us as if we were enemies. But it would be worse if we could see that they were coming from a place of wisdom and compassion, and nevertheless wish that they were silent, so we went unchallenged, even when wrong.

Ever read a good instructional manual?

2013-02-22T00:48:49Z

Have you ever read a book to learn about something complicated?

Did it help?

If so, mentally thank the author again, if you would, please. I'm learning just how hard that is.

I was laid off a few weeks back, and that's got me trying to blog about SQL Server. I would have sworn that explaining how database engines use transaction log files, and how that ties in to backup and restore and recovery models would be an afternoon's work. It's been two weeks and counting!
(updated: and about 6200 words - to be honest, that's *not* bad output. I feel a bit better now.)

I'm also working in a way that seems to work for me... but it's not what I expected.

Here's what I hoped I would do:

1) think about the topic I need to write about.
2) make a *very* rough outline. Something like this:

Simple - no tlog, full diff only
Full - tlog, lots of goodies, tail of the log
Bulk-Logged - tlog, "full, BUT!!!!"
Changing between them - NOTE THAT SIMPLE TO FULL IS REALLY FUNKY!

And then, I figured, I'd fill in the outline. I thought of this as kind of like "engineering" the writing... putting up a rough structure, then doing the rest of the "building".

And so far, I haven't done that. Ever. I've been doing what I think of as "beginners writing" - sitting down, starting to write on a topic, and then finishing. And then maybe deciding I have to split it in two, and doing that, and so forth.

I'm curious: do any of my friends out there write the first way (putting together a kind of outline, and then filling it in - even if the outline is purely in your head)?

Did you do that from the beginning? Or did you kind of need to learn how you wrote first, so you started to realize what kind of outline you'd follow?

Or do you do like I do (or "know some fine writers who...") and just get an idea, and start writing, figuring you can re-do later?

"Victory through air power"

2013-02-16T04:08:00Z

Lawyers, Guns and Money posted this - here's the link they used:

http://www.lawyersgunsmoneyblog.com/2013/01/victory-through-air-power

(I agree strongly with the caption to the video :-) )

Bill Gawne, I'd love to know your opinion of this (but I'm not asking you to actually watch it unless you're interested).

Second SQL Server blogging

2013-02-06T23:11:03Z

More SQL blogging. This one should be a lot "crunchier". If any computer-geek friends would like to review it from the perspective of "wow, I don't know SQL Server, but, reading this, I (feel I learned important stuff/was bored to tears)", I'd appreciate it. Here, especially, I'd love to have someone say "I felt like I could have understood X, but I don't, can you explain better?" so I know if I'm going over the head of an informed layperson.

Now, I just said I'd appreciate this - but this is more like "here's a story of mine, if you want to read it." You don't have to, I won't feel you're a bad friend if you don't. But I'd like to learn if I've got the right touch for blogging about this. If I can't even write a couple of quick articles on fundamentals, SQL blogging is probably not for me.

What is logging, and why do you care?

Okay, so, you want to learn more about SQL Server. Good for you. Let's delve into one of the most fundamental pieces: logging. I don't mean the writing of important messages to the SQL Server log (by default, named "ERRORLOG" - but it includes non-error messages, too). I mean the logging of database activity, called "transactions". These are logged in what's called the "transaction log" of each database.

What's the use of a transaction log? One of the biggest uses is data protection. If you log a proposed change, and make sure the log is "hardened" (physically written to disk) and then make the change, if anything goes wrong during the change, you know what was supposed to happen. So you can check, and see if it actually happened that way - and if not, you can fix it. So if power goes out, or the OS crashes, and the database server is restarted, it knows what changes might have been in progress, and can fix the data so it's in the appropriate state.

Another use for logging transactions is allowing you to group a set of commands into a single transaction, where everything will happen, or nothing will happen. So, if a bank customer pays for an item with a check, money is paid to the check holder, and removed from the customer's account - unless both pieces succeed, the transaction throws an error. This way, you never end up with only the payment made, or only the money removed - you either get both, or neither, leaving the accounting in a consistent state.

This also allows you to make provisional changes. Imagine an online retailer that has many warehouses. A customer places an order. This might be several sales recorded, several inventory adjustments, maybe a warehouse selection (or multiple selections - perhaps the warehouse will change based upon the specific items in the order, to minimize the number of necessary shipments), maybe a special discount loaded from the customer's account record... oops! The customer decides to cancel the transaction and empty the shopping cart. No biggie - just roll back everything, and you know you're all cleaned up.

Or, if you're a DBA like me, when you're making configuration changes, or the rare, but occasionally necessary, back end data changes, you'll wrap everything up in a transaction and make doubly-sure that the changes are correct before committing[1] them.

There's another subtle bit to the transaction log. Every thing that's logged has a LSN - Log Sequence Number. This means that there's a kind of running total to all logged database changes. (Most changes are logged, but some are not. From now on, I'll talk about "changes" and let you assume they're logged, unless I mention that they're not.) This lets you track database history. This becomes important in backups, restores, and in certain high availability and disaster recovery scenarios. For example, if you have a copy of your database, SQL can tell if it's synchronized to the original by checking the LSNs of completed and active transactions - if they match, then you can roll back all active transactions on both, and they're perfect copies. (Okay: they're perfect copies, unless one has data corruption, or unless someone has shut down SQL, made modifications to the data via hex editing of the data file(s), and then restarted SQL.)

These LSNs are intimately tied to SQL Server's backup and recovery models, so let's talk about that for now.

SQL Server has three "recovery models". These determine two things: first, they determine how the transaction log is used and re-used, and second, they determine how you can restore a backup of a SQL Server database.

The transaction log is like a whiteboard. Imagine scribbling notes about data changes on whiteboard before you make the changes. Once a change is absolutely committed, written to disk and done with, you no longer need the notes you scribbled, right? You can't roll the change back, and you know the disk changes have been made permanent, so you don't need to save the changes. So what do you do with them?

The simplest recovery model is called "Simple". When you're using the Simple recovery model, once a transaction has been committed, SQL Server marks the transaction as complete. And then, later, it will essentially erase the whiteboard section holding it. This is called "truncating" the transaction log. That's a really bad name - people expect the transaction log to become smaller, but it doesn't. It is, in fact, exactly like a white board, in this respect - when you "truncate" the transaction log, you've just erased the unnecessary sections of the transaction log, so they can be re-used. There's no change to the size of the transaction log file, there's just a change to the amount in use.

This wiping of old transactions happens at what's called a "checkpoint". I'll explain what that is in a bit - it's important, and is intimately involved with the transaction log.

Now: the problem with the simple recovery model is that every time a transaction is over and done with, the transaction log data gets thrown away. That means that the only way to restore anything is by restoring a full backup, or a differential backup.

(A full backup is a complete dump of all data in the database. A differential backup is a dump of all changes since the last full backup.)

This isn't horrible, in some applications. Imagine a database holds an inventory of your office supply cabinet. Well, if that database crashes, and you can't restore it to the point in time when it crashed, you can rebuild the data by taking a new physical inventory of the cabinet. That's a pain, of course, but on a properly maintained server, it's rare to lose data, so you can probably afford to take the chance. You can probably just back up the database with a full backup once a week (or even once a month!) and take differentials every day that you don't take a full. Because, really, you probably don't care about when you used a particular ream of paper... you only care about whether the ream was still there as of 6pm one particular night.

However, there are other applications that require better data tracking. A sales database is a prime example. If your server crashes at 12:37am, you might need to be able to restore it to 12:37am, exactly, and if you can only restore it to 12:00:01am[2], you might find it necessary to update your resume.

Here's where the transaction log and use of the LSNs can save your job. There are two other recovery models: there's "FULL", and "BULK-LOGGED".

These recoery models are identical except under certain circumstances. For most transactions, SQL Server logs the transactions in the transaction log, and then, when they're completed and written to disk, SQL marks the transaction as complete - but it doesn't truncate the log. It just marks the completed sections as complete.

When you take a transaction log backup, then SQL Server truncates the transaction log - any transactions that were committed, and backed up, are erased so that the space can be re-used[3].

Let's think about this for a moment, and let's go back to the idea of a white board.

With a Simple recovery model, you need a transaction log/"white board" big enough to hold all of the transactions that are in progress at one time[4]. As soon as each one is completed and written to disk, you can count on SQL erasing that section of the white board.

With the Full and Bulk-logged recovery model, you need a transaction log big enough to hold all transactions that occur, or are in progress, between transaction log backups. So if you run transaction log backups every hour, you need a transaction log file big enough to hold an hours worth of transactions for that database. And you need to take backups often enough to keep your transaction log from filling up.

Wait, you say. Isn't that the same thing, said two different ways? No. See, let's assume you have a big, busy database that will perform two gigabytes of transactions in an hour.

Well, if you have a four gigabyte transaction log that can grow to 6 gigabytes maximum, you need to make sure that if no transaction log backup has occurred in two hours, you know you can find and fix the problem in under an hour.

Simultaneously, if you take transaction log backups every hour, you need to have a transaction log big enough to handle anything that might happen in that hour... and you should remember that it's possible for some actions (a massive data load, or a huge index rebuild) to take multiple hours to complete. So you might need a larger transaction log than you might think. (This also holds true for Simple recovery model - any large transaction can require a lot of transaction log space. But with full, you need to wait for the next backup to occur, after the large transaction finishes. With Simple, you just have to wait for the next checkpoint - and SQL Server is likely to run a checkpoint after a large transaction completes.)

Okay: back to discussions of recovery models. As I was saying, you need to back up the transaction log when using Full and Bulk-Logged. (In the Simple recovery model, you can't back up the transaction log. Even if you could, it would be meaningless, because there'd be no guarantee that the information you collected was useful.)

A transaction log backup typically contains all data changes recorded in the transaction log, plus the LSNs involved.

Because of this, if you had the first full backup of a database, and all transaction log backups taken since then, you could completely restore the database to its current state, no matter how much time has passed.

With Full recovery model, you can also restore the database to a point in the middle of a transaction log. You can designate either a specific LSN (if the LSN was marked - don't worry about what this means right now, just remember that you can't just stop at any arbitrary LSN), or a specific time, to stop, since the transaction log backup will know when each transaction completed - so it can restore all transactions that committed before a certain time, and it can restore all transactions that committed before a particular transaction completed.

Bulk-logged is almost the same. Bulk-logged is a special mode intended to be used during bulk operations. Bulk-logged works just like Full if there are no bulk operations (like BULK INSERT, or the use of BCP - these are tools designed for moving lots of data into (or out of) a database). If there are bulk operations, the database will employ a technique called minimal logging.Instead of logging all data changes in the bulk operation, it only logs the data pages (data in a database is organized into 8kb pages) that were touched by the operation. So, during the bulk operation, instead of having to log all data changes (before and after values), it just has to say "and we did a bulk operation here." This means that a lot less data is written to the log file. This means that the task can complete faster (remember, all changes have to be logged - less logging means less time spent waiting for the log to be hardened), and requires less log file space.

(Requiring less log file space may not seem like much, and it isn't much - if your log file is large enough to hold the logging for the bulk operation. But if it's not, then the log file would have to grow, and log file growth is a very expensive operation.)

There's a trade-off for this minimal logging. If there are any bulk operations performed during the time covered by a transaction log backup, you can't restore to a point in time. You need to restore the entirety of the log backup - it's all, or nothing. Why?

Well, it's pretty obvious once you understand. Remember, when doing bulk operations in a Bulk-logged recovery model, SQL just notes which pages were touched by the bulk operation. So, when you back up the transaction log, it doesn't have enough information to rebuild the data in those pages. Instead, it takes a snapshot of the current contents of those pages, as of the time the log backup was taken. So you can't stop in the middle, because only have good data for how it looked at the very end.

Now, I just ran a little test. I created a Bulk-logged database, performed no bulk operations, and used a transaction log backup to restore to a point in time. It worked. It should have worked. But should you trust it to work? No - if you trust it to work, Murphy's Law will dictate that someone will end up doing some tiny operation that gets recorded as a bulk operation, and you'll lose the ability to restore to a point in time at the worst possible moment.

The best practice is to plan your bulk operations in advance, and switch to being in bulk-logged recovery model just until the operations are complete, and stay in full recovery model at all other times. It's a bit of a pain - but it's much less painful than finding you can't perform a desperately needed restore.

One final interesting bit of trivia: all databases - even those marked as Full and Bulk-logged - use the Simple recovery model until their first full backup is taken. This is because you can't restore a transaction log backup unless you've first restored a full backup, so there's no point in saving transaction log information until there's a full backup taken. So: always plan your first full backup of any new database with this information in mind.

Wow, this has gotten really long, but I did promise I'd discuss checkpoints, and that requires going over recovery, so let's cover those two topics.

Remember, the point of logging transactions is to log what happens before you make any changes to the data. That way, if something goes wrong while writing the data to disk, you still have a copy of what was supposed to happen. There's another aspect to this. SQL Server doesn't make any changes to any data on disk. It always makes them in memory. Now, if you made a change in memory, and then immediately wrote it to disk, you'd waste a lot of time writing to the disk. Some data might change a million times in an hour (think about inventory levels on a busy internet vendor!) Since the changes have all been written to the transaction log, there's no need to copy things to disk immediately. Even if you lost power, you could redo all of the committed changes, right?

In fact, that's what SQL does when a database comes online. It runs a process called "recovery". It scans the transaction log, and sees what was active when the database was last online. It then checks and undoes everything that was written to disk and shouldn't have been. And then it checks and redoes everything that had been changed and committed, but hadn't been written to disk.

If the server was shut down cleanly, this is very fast - the undo/redo was already done during shutdown.

But while we always hope our servers will be shut down cleanly, we know they won't always be. There are crashes, there are power failures, there are people who press the wrong server's power button, etc.. Since nothing can happen to a database until the undo operation is complete, SQL Server wants to keep that as short as it reasonably can. So it has a setting that determines how often it will run a checkpoint.

A checkpoint is when SQL Server flushes all changed data to disk - whether the changes are committed or not! - so that this recovery process should occur in a reasonable period of time, if possible. Now, you may wonder why SQL Server doesn't write only the committed changes to disk, instead. Wouldn't that be safer, and eliminate the "undo" period? It might - but there are multiple reasons it doesn't work well. First, it's very expensive to determine if all changes to a data page are committed (we'd have to look at the transaction log to see if the last LSN is committed or not - for each page!). Also, if a page is really busy, it might always have uncommitted changes. Second, remember SQL has to make all changes in memory - if we didn't write uncommitted changes to disk, we couldn't work on data sets bigger than memory. So SQL writes both committed and uncommitted changes to disk.

There are two big side effects of this. First, if you have to undo a large transaction, it can take just as long to undo it as to do it - because it might have to make just as many changes to the disk[5]. Second, if the server crashes during a huge transaction before it gets committed, you might need to wait a long time before that database comes online - the incomplete transaction must roll back before the database can be used.

SQL Server has its ways of estimating how long that will take, and when it decides it is past a certain threshold, it runs the checkpoint process, to flush data to disk, and (for Simple recovery model) to truncate the transaction log. Unfortunately, if there's a huge transaction that might require a massive undo, there's no way for SQL Server to cut the recovery time short.

[1]Finalizing changes is called "committing"; cancelling them is "rolling back".
[2] 12:00 is never am or pm - it's noon or midnight. Yes, I'm nitpicky.
[3] There are other reasons SQL Server might delay truncating the transaction log - Replication, and Mirroring, and some other issues, may prevent the transaction log from being truncated. But all databases in Full or Bulk-logged recovery model need transaction log backups to allow truncation - only some of them need to have other tasks complete.
[4] Technically, you need enough space to hold all transactions that occur between the times that SQL chooses to run a checkpoint. The log is only truncated at each checkpoint.
[5] Disk IO is the slowest operation on a server, and usually takes the vast majority of the time on large operations. So, the same amount of disk IO often (but not always) means the same amount of time.

SQL Server blogging, and test of cut tags on Dreamwidth

2013-02-06T22:43:18Z

I'm hoping to do some SQL Server blogging. *Please* feel free to skip the stuff after the cut tag.

If you know enough about computers that this doesn't go over your head, and can give some friendly advice about the writing, I'd like to know how it comes across. (If it's dead boring, and no, it's *not* that you don't understand it, I'd like to know that. I mean, I'd rather know that it's fascinating reading, but only if it's true :-) )

So - this is a blog dedicated to teaching folks about SQL Server, sharing my knowledge with others.

Why? Well - I'd like to learn to write better. And I'd like to learn to teach better. And I'd like to "pay forward" some of the people who've helped me learn SQL Server. And, of course, I like to show off, and love to hear myself talk :-).

Let's start with the very basics. What is SQL Server? If you're a Windows Sysadmin, and you've just been told you need to support a bit of SQL Server, what are you supporting? SQL Server is a relational database management system. So, what's that?

Okay: one of the things a great many programs will need to do is store information. Even a great many games need to save information - what stage of the game have you reached? What have you earned in the game so far? What is the game state (are there tasks you've started but haven't finished? Are there tasks you can no longer perform? What's different between now, and a brand new game?) These types of storage are usually easy to manage; you might need one or a few files, and with proper care and testing, a programmer can create the routines to manage that storage without any difficulties.

What happens when it's complicated? For example, what if it's an auto parts inventory and sales program? Now you have to track parts by a variety of factors (one air filter fits several cars, over certain model years, but might have multiple manufacturer identification codes). Here, things get complicated - when you start developing the system, you need to think about all the possible ways that the data might be used, and what people might need to use to search for parts or sales or deliveries. You might need to use the inventory program to interface with the bookkeeping program, and might need to calculate sales commissions or vendor discounts. Each time you think of new factors, you might need to change how you model the data, and how it gets stored. You also need to know a great deal about the computer operating system and the hardware, so you can figure out how to make sure the data gets written in a consistent manner, no matter what happens to the server, or the hardware. And you need consistent performance - you need to know how the program will work under expected, and higher than expected, workloads.

Let's face it; that gets hideously complicated, for any but the simplest of programs. It's easier to have someone else do all of that work for us, and then license that tool for building applications. That's what a relational database management system is.

It's a database, because it stores data.

It's relational, because we can define relationships between data. What kind of relationships? Well, let's go with that auto parts inventory. Let's say we have part number 2356, a zerk fitting. (Yes, that really is a part name - I have no idea what it does, past the Google search results.) It might have an OEM part number, it might have been part of a recall, so it has a new part number (to make sure that you don't use the old, recalled part), it has a description, it has a list of vehicles in which it will work, possibly of different makes and models and years. You purchase it from a supplier, and you have a buddy across town who will sell you one at cost if your mechanics need one now, but you're out.

There's certain information about it that describe that one item - ideally, you make that the main Part database record, and you have a unique identifier to identify that part. That's called a "key". (It's kind of like a key that fits in a lock - the idea being that only one particular key will fit any given lock[1].

What makes a database engine "relational" is that you can use that key elsewhere. You can use it on an invoice, you can use it on a delivery, you can use it in your inventory records, you can search cars that use that part, you can use it on a voucher, or a sales report, and you can do all of this using that same key. Rather than having to store all of the information about the zerk fitting on each of those things, you only need to relate it back to the Parts database record, and we can get that information using the key.

So, that's it: a relational database management system is about storing data in a structured manner, in ways that you can relate items to each other.

Microsoft's SQL Server is named SQL Server after SQL, which stands for Structured Query Language. Per ANSI standards, it's pronounced Ess Cue Ell. I pronounce it like "sequel" because I don't think ANSI has any right to dictate how I act in my personal life. Yes, I'm a rebel - all of my actuary friends wish they could be so bold and reckless!

(I passed two actuarial exams. I'm allowed to make jokes. So, why am I a DBA and not an actuary? T-shirt and jeans, my friend... t-shirt and jeans. None of that suit and tie stuff for me, thank you very much.)

SQL is a standard way of querying a relational database engine. Since it's a standard, people can build a tool (like SQL Server, or Oracle, or MySQL) and people don't have to be locked in to using a particular one. Most (but certainly not all!) of the commands will work on any given SQL-using database engine. If you're going to manage SQL Server, you need to learn the basics of SQL commands[2]. Microsoft includes some excellent management tools with good GUIs, but most of the information about the server and all the various settings is stored in various system tables and views, which are accessed using standard SQL queries. While I hope to include some of my favorite queries on this blog, I don't intend to teach you SQL - there are excellent online resources to do that better than I can!

What else do you need to know about SQL Server? Well, it's intended for the enterprise. "Enterprise" is a bit of a question begging word[3], but it's intended to store data properly, or not at all (look up ACID to understand storing it properly), to have tools to alert you to, and try to deal with, or recover from, hardware and software problems, and to be scalable - meaning you can put it on a machine that's twice as powerful, and get almost twice the performance out of it. It has tools for high availability, and disaster recovery, and lots of bells and whistles that allow you to build very powerful, very sophisticated applications with a reliable engine backing them. It also has a great support team backing it - if you find yourself stumped, or are sure that something isn't working right, you can call Microsoft's CSS team for SQL, and open a new case (a "pro" case, for those without support contracts, is surprisingly cheap compared to pounding one's virtual head against a virtual brick wall), and they'll help you out. (Yes, that really is kind of necessary for the "enterprise" - every program has bugs and its rare for a program to work exactly the way you expect it to, even after you've read the documentation.)

So, that's the basics. SQL Server is an engine that is meant to be able to store and retrieve data quickly, that lets you relate one piece of data to others, and uses the SQL language. It's enterprise-ready - trustworthy storage and retrieval, scalable, and can be used in highly available applications, with good disaster recovery possibilities. It has a rich feature set, and good support. It's a fine choice for a database back end.

[1] Technically, a key can fit multiple locks, and a lock can have multiple keys that will turn it. Don't strain the analogy too hard - the important idea here is that a data "key" means it fits something specific (like a key fits a lock)

[2] SQL commands are called "queries". This makes sense when it's a SELECT - that's the basic command for "tell me what items have the qualities I'm about to request". It makes less sense when it's an INSERT or an UPDATE - that's not a query, that's a command. But "query" is what the Structured Query Language is about.

[3] "Question begging" means we're assuming another question has already been asked and answered first. In this case, that question is "what the heck does 'enterprise' (or 'enterprise ready') mean?"

Interesting thoughts on word choices...

2013-01-22T20:11:25Z

There's a word that reverberates in my head sometimes - "strong".

One of the reasons I find it intriguing is that it pops up in Anime, sometimes referencing something that doesn't mean "able to apply a lot of brute force". "He's a really strong martial artist" for a guy who doesn't split boulders - but is wise and knowledgeable and skilled. (This in an anime where people *do* split boulders... hence the allusion.)

It's also a word that gets used for a lot of people who bear up under a lot of crap. And it's sometimes a painful word - IIRC, a friend of mine once reported an ex- saying that he treated her badly at a certain time because she was strong, and could take it.

Sometimes people who are suffering are told they are "strong" when they feel weak, and they hate it because everything is telling them that they're not.

And me? I've been finding myself angry about the word, because I'm fighting harder and harder for fitness, but I still have physical issues after a heavy workout.

After a year and a half of pushing myself, I can run 30 minutes at a 9 minute mile pace - and that, only because I have a treadmill that forces me to keep the pace up. I don't think I could do it running around a track; I think I'd unconsciously slow down to a more comfortable pace.

And yet, in spite of the difficulties and frustrations, I've managed to go 387 days in a row without skipping a workout. Sometimes it was only 25 minutes of walking, but even when I was sick, I worked out. (Admittedly, I've only had colds or the equivalent, so far.)

Two days ago, I tried running for 35 minutes at a bit over a ten-minute mile pace (6.1mph). It was easy - as easy as 5 miles an hour used to be. So that's good. But I had some tiredness problems afterward. I sure didn't feel strong.

And that's when I realized there was another word that described what I did feel.

I felt tough. (If this was written by Stephen R. Donaldson, I'd feel "an annealment precipitated by a confluence of mental and spiritual inanation".)

I didn't feel strong - a strong person would have done that workout easily, and then been ready to go dancing. But me? I knew that workout might ruin my night... but I needed the workout, and I needed to know how it would affect me, so I did it, knowing I wouldn't be dancing, but willing to suffer through that.

Just as I'm willing to suffer through day after day of exhaustion when I need to.

Just as I'm willing to struggle through depression, and never stop remembering that I deserve, and need, better.

Just as I'm willing to bear up under significant mental, emotional, and physical pain when necessary.

An expression I like to ponder is that "you never know how strong a person truly is until that person is free of their heaviest burdens." A person might be truly mighty - but carrying so much that they appear weak.

One of the burdens I find I was carrying was a need to be "strong" and a frustration that I wasn't. And I can set that down, because I have another word that better describes what I am, and better describes my day to day victories. I may not feel strong when I make it through the day, feeling battered, but I can accept that I was tough enough to make it through.

Anyone need a 7" tablet?

2013-01-21T23:59:32Z

Okay, so, for my birthday last year, I decided to buy a relatively cheap 7" tablet. Not a super-cheap one, but I wasn't going to buy an iPad Mini (which didn't exist anyway), or the equivalent.

It's not a bad tablet - dual core processor, SD card slot, and it was a decent brand - everyone agreed that if you weren't buying a Nexus or a Kindle, this was about the best you'd find in the price range.

And, about a month later, I bought a 10" tablet, and haven't used the 7" tablet since.

I don't want my old tech to just sit and rot, so - does anyone want it?

Minor preference given to anyone local who can arrange for pick-up.

Comments screened, because it feels like the kind of thing where comments ought to be screened.

Hapy new year, everyone...

2013-01-01T03:51:34Z

So, today, I woke up and did a 25 minute work out on my treadmill, to cap 366 days in a row of workouts. I'm torn between working out again, tomorrow... and deliberately skipping, so I can't ever again feel trapped by not wanting to break a streak :-).

I also meditated for the first time in a long time yesterday, and had the goofiest reaction... but let me explain.

One of the simplest meditations is just paying attention to your breath. That's not easy - your mind will wander, right? And you'll do things like say to yourself "okay, thinking about my breath" - and that's not focusing on your breath, that's *talking* about focusing on your breath, right?

But it's like a balancing exercise. Doing it wrong, feeling ridiculous and wrong, doing the crazy-balance-dance, those are all part of it. You're finding out where you're weak, and noticing yourself getting stronger. Even the worst flub is still an exercise, and you're still learning and growing from it. If you could do it perfectly, you wouldn't be doing it - you'd be doing something else. So, you work at it, for the sake of growing.

So you keep noticing when you're thinking and returning to your breath, as best as you can.

One of the biggest questions I've seen people talk about is "what does it mean to focus on your breath?" I've seen people say to imagine a particular spot, and feel the breath moving past that. I've seen people say to imagine the whole act of breathing, and think about that. But I had an idea... what if I stopped breathing?

Breathing is both a voluntary and involuntary reaction, after all. There I was, breathing in when I kinda-sorta decided to, then breathing back out.

So I stopped. I didn't deliberately stop - I wasn't holding my breath. I just quieted my brain as best as I could, and waited.

And suddenly, I felt my belly expanding, my diaphragm moving downward, air coming into my lungs - I was ready to burst out laughing at the surprising way it felt. It was, in a certain sense, like watching a comic do a deliberate pratfall on a banana peel - HAH! Look at that! - only it's not someone getting hurt, it's surprise and a bit of delight (because you know the comic is a pro, and isn't going to be hurt).

I think I managed to sit there for a full minute just thinking about what a crazy thing this body of mine is, every now and again suddenly moving in funny ways and drawing air in, without my doing *anything* to make it happen.

Of course, now the next time I meditate, I'll probably find my brain wandering to "I wonder if I can feel that again" and it's impossible to *feel* something like that intentionally. But it was fun, and goofy, and spiritual, all in one... so, unless I can think of something better, I'm going to let that be my last post of the year :-).

Trying to update more frequently...

2012-12-31T02:31:24Z

...and not succeeding, but I figured I'd run an update today, now, before I could forget (and it got too late to write, even if I remembered).

So, health updates: Today I ran 35 minutes with the treadmill set to 6.2 miles an hour. I should be able to do a lot more than that, after a year of training, but there have been complications.

My hip issues are nearly resolved (that was one of those complications). I had some incredible tightness in the ileopsoas region - that's the stuff that connects your lower back and your hip (and thigh?). I'm still not well, but today, I realized I was doing stretches that I used to hate, hate, *hate*, and they weren't painful, either while stretching, or while releasing.

I also found that I can do one of my balance exercises (too complicated to describe) 15 times in a row on my left side. When I started, I could do it maybe 4 times in a row if I was lucky, and often ended up doing the balance-dance quite a few times.

Each time I manage to release a bit of my hip issues, I found my stride changing a bit, and this time, I think I feel the insides of my legs doing the work they should, and that's great, but a bit achy :-). I can feel a few tight areas, but I think I've mostly got the same flexibility on my right and left side, now - just, on one side, it feels a bit more stretchy when I use that flexibility.

I'm still battling depression, though I'm at my "winning the fight" stage. The problem is, this is where I usually get stuck - I'm not depressed, but I'm not normal. I do what I have to do, I react to things that must be reacted to, but I don't tend to *act* - and that's not a good way to live. (There are other problems too - no real pleasure or satisfaction in most areas of my life.) This time, I'm making sure I keep up the fight, but it's hard. Meds are one of my primary options right now, and meds take up to 12 weeks *if* you're on the right med. That's not long compared to a lifetime, but it's excruciating when you're fighting to remain present and in the struggle, knowing how easy it is to stop struggling.

See, with most depressed folks, they have someone around who can tell them "you know, I don't think those pills are working" (or, conversely, "you might not have noticed it, but I have - you're doing better"). In my case, I have to notice that. And I have to make sure I don't stop thinking about how I feel so I don't stop short of as much remission as I can hope for. And that means I have to think about how bad I feel, not just whether I'm feeling better. It's... well, it's complicated.

Worst thing is, I'm on a med that shouldn't have a particular side effect - and it seems to be having that side effect, a decidedly unpleasant one. And so the question is, will it pass or not? But regardless of those concerns, I am coping for now.

I'd like to include something more interesting here, but I'm afraid I'm at a loss for that, so I'll leave you with what I have, for now :-).

Hey, everyone...

2012-12-13T06:25:20Z

So, yeah, I'm still reading, and thinking about you, and caring about you, and trying to remember to write in here more often.

But I'm also struggling with depression and other health issues right now. I think I'm on the mend... I found that my generic Wellbutrin sucked, and I'm using a different method of taking it, and I think it's helping, but the generic XL form seemed to make some things worse.

Interestingly, there was a news story that one brand of the generic XL form was recalled because it didn't meet the correct criteria for bioavailability (i.e.: it didn't release the medication steadily over 24 hours). There are two things aggravating about this - I'd been on a generic formulation of the XL form before, and it didn't work for me - I'd just forgotten. The other thing that's aggravating is, this is one of those things that's well known among people who are depressed and follow forums on the internet... but it wasn't enough to trigger a quicker review.

If it had been a pain killer, and people swore up and down it didn't relieve pain as well, well, people would listen. But since it was "just" an antidepressant... well, how do you know if it's the drug, or the progression of the problem? Then again - that's all the more reason to be more careful about these things, right? If you say your generic time release drug is just as good as the brand name, you can't have it release at a completely different rate, precisely because it's that much harder to track any problems the different release rate causes!

Ah well.

I also went to see a physical therapist about my hip. She said I might need a new one.

No, seriously... a guy who's running 3 miles, four or 5 times a week, and walking the remaining days might need a new hip.

See, she asked if it felt like my leg was out of socket. I said well, it's weird - my left leg is so turned out, my left foot comes down so funny, it *does* kind of feel that way, but, of course, if *that* was a problem, I'd have more serious problems than this, right?

She'd already figured out that my left hip was jutting up and forward; she figured that the reason this was happening was a tear in the gunk inside my hip socket. An MRI would be a good idea, she said. I finally asked the prognosis of this torn-gunk in my hip socket - there's nothing good; the bone's probably worn away, the gunk doesn't grow back (except when it does, and doctors don't study *that* because what is there to study?), blah blah blah, maybe eventual hip replacement. Oh, I don't want an MRI based upon ludicrously idiotic guesses? Well, here are two tiny exercises that won't damage my hideously injured hip, we have to let it *heal*, and *DO NOT RUN*.

I ran every day for the next four days; do I think I have a hip problem? No. I cancelled each of my two appointments, and didn't make another.

kightp had some psoas muscle problems, and I tried some of the stretches she was doing - they helped. I kept doing the other stretches I was doing that seemed to help. I kept doing the balance exercises that illustrate the problem, figuring that if I kept strengthening the right muscles, it'll help.

I'm not well - but I'm a lot better, and I feel like I'm getting well. My gait is normalizing (which means I have achy legs from using different muscles), and I can sit up straighter for longer. My left butt seems to lay on chairs closer to the way my right one does. All the little things that feel wrong seem to be feeling more right.

I think I've had one time where a physical therapist actually helped me - patellar tendonitis. Each and every other time, it seemed like they completely misunderstood the problem, had me doing things that didn't help all that much, and then slapped on a TENS unit and an ice pack that didn't do a damn bit of good. Maybe it's me - maybe I talk too much to them; maybe I geek out too much. But I think I'm one for four when it comes to effective physical therapy, which isn't very heartening.

But I am on the mend, I think my depression is lifting, and I think I have fixes for some of my other things.

So - enough babble for tonight. Tomorrow - tomorrow I'm going to beg for help finding a present for a teenaged girl (my niece).

Here's hoping life is treating you all well!

Tiny victory over "the man"...

2012-11-29T05:23:18Z

Apple makes some fine products, and some awfully sexy ones. But I was running into a frustrating situation. I have a Honda Fit[1], whose factory stereo interfaces with an iPod. Not a Zune, not an Android device, an iPod(/phone/pad, I reckon).

I have on the order of 30 gigs of music. So, my 16 gig iPod Touch isn't good enough to hold all my music, and even if I had purchased the 32 gig Touch, I'd end up having to swap out music someday. If I want to hold all of my new music, I'd need to spend $225 on a new iPod classic.

Then I realized that I was being silly. I don't need a new iPod. I could buy used! I mean, what, the battery is nearly dead? Who cares! The USB plug that feeds the stereo is both a charger and a reader.

And then I realized I was being even more silly. I don't need a used 160 gb iPod classic - an old fashioned 80gb iPod would be just fine. So, now, for half the cost of a new iPod classic, I have all of my music collection ready to be played in my car. And, I didn't pay extra money for a (let's face it) somewhat overpriced MP3 player to a company that has more than enough money already. Plus, it was probably good for the environment, and meant one less iPod through the factory in China....

[1] Depression, (n): A state in which it's not abnormal for one to realize one has bought a new car four months ago and hasn't told one's online friends...

Random thought of the day...

2012-11-27T03:47:26Z

A life focused on *being* something is risky, because there's a cultural meme that completes the becoming as "...and then I'll live happily ever after". And unless the focus is on being someone who lives happily, one might find that one has spent so much time Becoming Something that one has forgotten to learn how to live happily.

How I can tell I'm feeling better...

2012-11-14T22:59:59Z

http://www.msnbc.msn.com/id/49823152/ns/technology_and_science-space/

You know what's really happening? An eerily earthlike planet had its eerily luna-like moon ejected during a nuclear explosion involving magnetic radiation. And this moon (and it's base, named Alpha) traveled through various star systems running into strange alien races and having strange adventures.

But then... then they started to ponder. Why didn't their beloved earth send a rescue mission? "What, you have a fully functioning space program and can't be arsed to figure out which direction your frickin' *moon* has gone off in and try sending a rescue mission?"

So they've been accreting mass, forming this gas-giant, and now, thinking it's still 1999 - or maybe 2001, by now - they're going to maneuver their way here.

What will they do when they think we've created a new moon, to replace them?

Thank heavens...

2012-11-07T06:37:42Z

It looks like Obama won.

There are times when I'd consider Mitt Romney a decent President. These are not those times.

I mean no disrespect to Republicans - there are good people in the Republican Party. But their party has been taken over by people who are too willing to forgo true leadership.

The Republicans refuse to raise taxes - period. It's good to fight for the lowest taxes possible; it's not good to take a tax cut that threw us from surplus to deficit, and refuse to even allow that tax cut to expire, because that would be "higher taxes". That's not fiscal conservativism... sure, allowing a few years of deficits, just in case economic growth wipes 'em out, that's reasonable. Deficits as far as the eye can see? It's time to bite the bullet and admit the tax cuts were a bust.

The Republicans allowed people to question the President's birth certificate and many actively encouraged it. They allowed people to think that maybe the United States was so stupid, so incompetent, that they couldn't even verify the citizenship of a man about to ascend to the highest office of the land. Obama had a passport - that means his birth certificate was verified. He was on the Foreign Relations committee which means he had a full background check. His birth place was known with certainty. To allow people to call that into question, to allow such rampant, insane distrust of the government to fester is not ordinary behavior.

The Republicans talked about "death panels" - the cost saving panel can't refuse or ration care; but they continued to insist there were death panels anyway. There are lies that politicians tell, and really nasty lies that politicians tell - that's part of the game. But come on - death panels? Think about the distrust one is seeding if one wants one's followers to believe such a thing!

The Republicans in Congress put the full faith and credit of the United States at risk. Now, a lot of people have threatened to hold back on voting for debt limit increases in the past, hoping to apply some pressure to do things. That's part of the game too. But for an entire party to drag it down to the last couple of days, insisting that they were willing to "shoot the hostage" if their demands weren't met? That's beyond ordinary politics.

There are a lot of things I don't like about the Republicans. There are a lot of things I would hate them for doing. But those are political differences. Those are things good people of good mind, heart, and spirit can argue about in good faith. But there are too many things they've done that are not just points of disagreement.

I want a Republican Party that will occasionally tell vicious lies, that will occasionally commit awful grandstanding, that will be intransigent on ridiculous things to get their way - but will do so with honor and, well, sanity.

For the sake of balance, I'll grant you that the Democrats occasionally tell ridiculous lies, commit awful grandstanding, and are sometimes intransigent on ridiculous things to get their way. That is part of the process, part of the game. It's not a pretty part of the game, but it's part of the game. But they haven't questioned that maybe the President was illegitimately elected. They haven't spread lies as hideous as "death panels"; they are willing to compromise in reasonable ways; they haven't sown similar amounts of mistrust in the government and their opponents, even at their nastiest.

I hope that Obama's re-election changes things, and starts to move them back to the way they used to be. Just as nasty, just as fiercely opposed, but with an acknowledgement that we're all on the same side, that we all want what's best, even when we disagree on what that might be.

Sigh (/bleagh)

2012-10-30T04:36:34Z

So, I'm dealing with fatigue and depression. About three days after my last treatment that might help the fatigue... I find that I'm sick. Friday I felt like crap; Saturday, I stayed in bed 80+% of the time and learned that the Harry Potter books are good, light reading (hadn't read any of 'em until now). Sunday, the same, but I went out for chinese food to get protein and vegetables in my system. Today, the same, but I went to Target so I could finally brush my teeth with actual *toothpaste* (rather than baking soda - a decent emergency substitute, especially if you follow up with a fluoride rinse), and other necessities.

I'm really (scary thought) hoping I can get to work tomorrow.