The Twilight of Roger Ebert

I had known from vague things that I’d read on Roger Ebert’s journal that he was ill and declining, but had not known exactly why. Today PZ Myers posted a pointer to an Esquire article that explains it all.

I had always respected Ebert’s movie reviews, even when I didn’t agree with them, but had never realized what a good writer he is until recently. His recent post on the demise of Jermyn Street in London is as good as it gets, seems to me. There’s nothing particular fancy or self-consciously artistic/literary about his writing — it’s just good solid prose with a wealth of images written in a natural, conversational style that is eminently readable and entertaining.

Ebert, like me, is a non-believer, and he writes this, quoted in the Esquire piece:

I believe that if, at the end of it all, according to our abilities, we have done something to make others a little happier, and something to make ourselves a little happier, that is about the best we can do. To make others less happy is a crime. To make ourselves unhappy is where all crime starts. We must try to contribute joy to the world. That is true no matter what our problems, our health, our circumstances. We must try. I didn’t always know this, and am happy I lived long enough to find it out.

There’s nothing fancy about the language there, but the structure and flow of the thoughts is quite powerful.

Ebert is not dying imminently, but according to the Esquire piece he’s fading very gradually.

When his last day passes, I, for one, will miss his writing terribly, but for now, I am grateful for each new post to his journal that comes along. He’s not short-winded, and for that I’m very thankful, as it means there’s more to savor. And the conversations that ensue in the comments, with Ebert participating, are worth almost as much as Ebert’s journal entries.

Addendum: Ebert himself responds to the Esquire piece and points out that the article’s “Ebert is dying in increments” line is true of all of us. Indeed, it certainly is.

Why Do the News Outlets Think We Care About Tiger Woods?

I’ve been listening all day to progressive talk radio and I’ve heard a report on Tiger Woods at the top of every hour. And then the host of one of the shows launched into the subject, and his audience emailed him en masse and asked him to SHUT UP.

I don’t give a frigging rat’s ass about Tiger Woods. And nobody who is not a member of his family or one of his corporate sponsors or a big golf fan really cares, either.

I just do not understand the way the news media think.

Bush’s Moon/Mars Plan is Dead

It’s been a while now since Obama’s NASA budget came out, but it’s pretty clear that manned spaceflight has been put on the back burner. This means that Bush’s Moon/Mars plan is probably dead.

Of course, I said that a long time ago, here and here.

My Impressions of Google Wave

Google Wave was in the news the last week, Sept. 30th was the date that platform was opened to the wider developer community. The actual big news event was last May when they presented a demo at their developers conference. The video of that demo is an hour and 20 minutes long, but quite illuminating. I sat down over the weekend and watched it, expecting it to be a slog, but it was quite entertaining.

While watching it, I took some notes, and what follows is a digested version of my immediate responses to the demo.

First off is that all of these web services have on main flaw, a single point of failure, i.e., network connectivity (either yours or Google’s). If, for instance, you use Google hosted services for your Wave conversations, if your Internet is down, you’re dead in the water. Of course, if your Internet is down, you also can’t receive email, so perhaps that’s not so big a deal, but the advantages of Wave come in the real-time and near-real-time collaboration, whereas email suffers very little from the latency problem that a local internet connection failure imposes.

Google certainly has big pipes to and from their servers and lots of redundancy, but they occastionally do have failures in some of their apps that cause them to be unreachable or slow. But consider if you decide to run your own Wave server — likely you would do so on a commercial hosting service rather than on your local office’s servers, but either way, you’re again in the situation where a crucial app is network-based, and only as reliable as the networks you depend on.

Embedding the Wave in a Blog:
Is just anyone allowed to participate in a WAVE embedded in the blog, or do you have to have user authentication in place?

Brilliant application of their Google search spelling algorithms, but how often does it actually fuck up?

The default is edits are immediately visible, indeed, they haven’t even built the feature to hide immediate updates. If you have a 10% rate of comments that you don’t want others to see, how many times will you end up accidentally sharing something you don’t want others to see until you’ve finished it? It seems to me, this makes it necessary to be very aware of the nature of the communication before you initiate it, and with certain people you’d want the default to be HIDE updates until SEND, while with others, you’d you’d want it to default to immediate.

A complicated problem, and one that will, I think, cause endless problems for end users — how many people pay for the email services that allow you to undo SEND in an email?

The presentation has confused me. I thought I understood the difference, but now I’m confused. They mentioned a distinction that a robot was server-side and an extension client-side, but the demo of Polly the Pollster seemed to obscure this — I feel less distinction now than I did before going into the example. Perhaps this because they’ve successfully abstracted the underlying technology so that to the user the difference is undetectable.

At the conclusion of the robot demo, Lars says “OK, that’s EXTENSIONS for you” (1:04), which just goes to show that I’m not the only one who is confused.

The difference is clearly server-side vs. client-side:

  1. an extension runs in the client. Updates to it get passed to the server as part of the wave and distributed through the normal wave distribution process.
  2. a robot is code running on the server that waits to see something pass by it in the wave that triggers its server-side behavior, whatever that may happen to be.
  3. robots have client-side UI elements that dump relevent XML into the wave that will then trigger the server-side action from the robot.
  4. thus, robots are client-side extensions that modify the wave plus a server-side process that reacts to those changes to the wave stream.

So, robots and extensions are *not* two different things. A robot is just an extension paired with server-side actions/

While it might seem tempting to integrate a Wave into a website, the problem is the same with the Wave as with SideWiki — you don’t have control. It’s collaborative, so you have all the problems that come with collaboration, where everyone is equal and nobody gets veto power. That is, unless they’ve actually engineered it for superusers who have the privilege of editing the way, i.e., removing from the playback those things that they don’t want part of the permanent record.

Interesting that the conclusion of the talk is Google’s realization of the importance of developers in making a platform successful. This is something MS always understood, something that Apple has only imperfectly understood, and something that this video shows Google is obviously coming to understand.

Blogging Pachelbel Note

I’ve received some really helpful information about the history of recordings of the Canon from Robert Fink and will need to absorb and research some of the information he’s provided me. That means going back and changing some of the entries already posted (e.g., I now know that the Paillard I listened to is not the 1968 version but a 1989 recording). I won’t re-order the posts, but it does mean that I’ve placed some of the recordings in the wrong chronological order.

More to come when I have time to get back to this — the most interesting recordings are still to come!

Blogging Pachelbel #10 — New Bach Collegium

Based on a close listening, I’m guessing this recording is misplaced in my chronology. The New Bach Collegium of Leipzig was founded by Max Pommer in 1979, and based on this recording, I’d assume they play modern instruments (it’s at 440), but in a historically-informed style. Certainly this performance is quite lovely, in fact — a breath of fresh air after the hackwork of the previous five performances discussed here.

This group is not afraid of playing notes legato when they have no slurs over them. The bass is also refreshingly grouped metrically, so that the 1st and 3rd beats of the measure are stronger than the 2nd and 4th — imagine that!!!

I strongly suspect that this recording post-dates the Musica Antiqua Koeln and Hogwood recordings that come next in this “chronological” review, simply because the earliest date I’ve come up with for this recording is 1989. But I don’t have an exact date.

Also, I’ve classed this as “orchestral” because it’s clear that in some sections at least, more than one violin is playing on a part. But it also seems that there’s some nice use of solo/tutti throughout, in order to make dynamic contrasts. While I don’t think there’s anything in the original that suggests any need for that, if the decision is made to use multiple players on a part, I think it’s a good idea to vary their number. It certainly broadens the potential contrast in terms of both dynamics and tone color.

In short, I’d conclude that this performance is well worth hearing.

Blogging Pachelbel #9 — Vienna Baroque Ensemble

This appears to me to be a recording in search of a reason for being. That is, the results seem to me to suggest that the performers said to themselves “There are so many recordings of this piece already — how can we make ours stand out?” They appear to have concluded that the appropriate response was to completely forego the use of legato.

The bass is pizzicato. There seem to be no slurred notes. Every beat is banged hard, but the notes are all short.

The recording as it exists in the downloaded MP3 appears to be truncated, as it picks up with m. 9.

There is really no reason for this recording to exist.

Blogging Pachelbel #8 — Gerhardt/National Philharmonic

Another workaday performance — they play the notes and rhythms on the page within the style they’ve been trained to follow.

The era of this recording is not entirely clear — the continuo bass is very heavy, with an unimaginative tinkly harpsichord playing nothing but unidiomatic unrolled chords on the beats of the bass line, and a hint of organ at the beginning. The overall approach to the canon is to just play it and let it come out as it comes out, except for a tendency that’s quite obvious in a lot of recordings at the beginning, to delay the foregrounding of the fifth couplet (m. 9-10 in v. 1) until the second half of the measure. This is likely as much an acoustical phenomenon as anything else (the 1st violins are below the 2nds here), but it has bothered me in all the recordings so far, given that what is presented seems to me to be an exaggeration of what happens naturally acoustically. It’s as though the first violins hold back in the first half of m. 9 in order that when they leap up to the C# they can become VERY, VERY PROMINENT. The wave form seen in my audio editing program confirms this — the amplitude jumps noticeably at this point.

Other than that, this is yet another performance where no one is imaginative enough to figure out that notes without slurs over them can still be played legato.

In other words, more of the same.

Blogging Pachelbel #7 — 101 Strings

I’ve already discussed this performance in the post on the Paillard recording, where I was attempting to come up with a justification for the cuts seen in the Baumbartner and Muenchinger/Stuttgart recordings, and the apparent need in the Paillard to de-emphasize the same passages that got cut in the other two recordings. My surmise was that performance practice of the time led to unpleasant results when confronted with the notation of that passage.

This recording, like the previous one, seems to be your typical sight-reading session for session musicians brought in for the purpose of laying down some tracks that will be cheap to produce and engineer and, thus, profitable to sell.

101 Strings was certainly a fictional orchestra, not a resident musical group but a group of musicians who were hired for particular commercial recording jobs and engineered to have a certain sound. I’m not sure I hold this recording against these players as much as I do the awful Royal Philharmonic arrangement recording. But the same disregard for musicality found there and the grade-school-orchestra level of subtlety found in the London Philharmonic sight-reading sessions are evident here, though with an admirable variety of nonetheless inappropriate agression lacking in the previously-discussed recording.

Blogging Pachelbel #6 — London Philharmonic

This is yet another “professional” performance, and sounds like nothing more than a sight-reading session from unedited parts. All non-slurred notes are aggressively detached, and the 8/16 octaves of the bass line are aggressively out of tune (as though it was no accident). There is no dynamic contrast — they just play the notes. There is no change of affect anywhere, and no subtlety at all — a moderately-talented grade-school orchestra could be more musical than these jokers.

If the only recording of the Pachelbel canon that you’ve ever heard is this one, you’ve simply never heard the Pachelbel Canon.