Smalltalk Tidbits, Industry Rants

Looks like someone is building a Smalltalk implementation native to .NET - have a look at the Vista Smalltalk blog. This is a great explanation of the value, to my mind:

As internet connectivity improves, we will increasingly be building ad-hoc, highly connected applications. Think of how online games or workgroups might evolve as Peer-to-Peer networking becomes commonplace, or think of how applications aggregating data simultaneously from dozens of webservices might evolve.

We will need a more powerful way of doing programming to build ”instant” applications robustly and quickly.

Smalltalk originated in the powerful biological concept of “protected universal cells interacting only through messages that could mimic any desired behavior (Alan Kay)”.

With its simple messaging paradigm and minimal syntax, Smalltalk is probably the best language yet invented for harnessing the increasing potential of the Internet.

Scoble talks about syndication technology, referencing this post by DeWitt Clinton:

Users don’t care about specs, or arguments about formats. When you understand that you’ll understand how RSS got so big in the first place. Dave Winer evangelized RSS by building a publishing tool (Manila and later Radio UserLand) and an aggregator (Radio UserLand and later Share Your OPML.

Where’s the Atom publishing tool and aggregator that demonstrates Atom’s superiority?

Makes me wonder whether Robert got through more than the first paragraph of the post. Two thirds of the way down, DeWitt says:

Put it this way -- I couldn’t be doing half of the work that I’m doing right now on search syndication without Atom. Sending back search results snippets over RSS is one thing. Syndicating rich search content is an entirely different thing, and that requires a non-lossy syndication format.

My recommendation to application developers today is to use Atom 1.0, not RSS, as the basis for your content syndication.

The tools for Atom that demonstrate it's superiority are exactly what DeWitt said: they're all the tools and services being built up around micro-formats. Now, it didn't need to be this way - RSS could have been that spec. Sadly, Dave Winer wouldn't allow for that. For reasons understood only by Dave, he thinks that the lossy nature of RSS is a feature. When people on the RSS Advisory Board disagreed with him, he called their employer (note the resignations). When that wasn't an option, he tried threatening someone else with a lawsuit. Meanwhile, his enablers - like Scoble - say nothing. RSS could have been the unitary spec had Dave not been a complete jerk, and people like Scoble bear some responsibility for that by never, ever calling him on his BS.

James McGovern suggests that Microsoft should OSS their products:

Now if he would only do the same thing to his operating system, Microsoft SQL Server and Microsoft Exchange. Seems like an opportunity to one up the folks over at Oracle. I wonder if this will get any industry analyst attention?

Yeah, right. How about the firm you work at, James - would it work out well for the employees if they started giving their core products away for free?

Boris Popov has some small, but very cool mods that might help you trick out your VisualWorks image:

Here’s another package that we use quite a bit here, NewSystemIcon. It adds two new sub-menus to the System menu that allow one to pick a non-standard icon to be used throughout the image as well as a menu to change a global background color. Some may find this useful when running multiple images at the same time or simply when they need to differentiate special ones.

Head on over to his blog for some screenshots.

Technorati Tags: smalltalk

I posted on Ted Neward's failed analogy (in a post about O/R mapping) a week or so ago. Yesterday, at the party we had, a few of us were talking about this and that, and James McGovern's blog came up. One of my friends made an excellent point:

How can you take him seriously, when he posts the kind of silly, unrelated pictures he does?

Exactly. That's a large problem for his blog, and it's the same one Neward had in his post about O/R mapping. When an otherwise ancillary point overwhelms your message, you've failed in the basic task of communication. On McGovern's blog, most of the images he posts are political, and they are bound to irritate roughly half of his potential audience. A lot of the others are just pure nonsense images. In general, none of them have anything to do with the content of his posts.

Those images are like annoying popups - they detract from his message, and make it far less likely that his thoughts will be taken seriously.

Happy Fourth of July!

James McGovern:

Since industry analysts tend to focus on features at the expense of security, I figured I would use several tools to determine of what quality Ruby is relative to both Java and .NET. I wanted to also include a version of SmallTalk, more specifically the version that James Robertson evangelizes but wasn't sure of if benchmarking information could be published.

Heh. Unlike some of the big vendors, who get their panties in a twist over the idea of independent benchmarks, we don't care. If they don't look good, hey - they don't look good, and either we have work to do, or there's a problem with the test. Either way, we'll learn something.

Btw, it's Smalltalk, not SmallTalk :)

James Governor points to the curtain behind which lies the smallish pile of data upon which Gartner build magic quadrants.

Scoble on Hi-Def DVD:

I was hanging out on the AVSForum the other day and saw several posts from people who said that in their comparisons HD-DVD is far superior to BlueRay tests.

Today those posts are getting reported in CentreDaily.

See how the grassroots could be changing popular opinion?

Maybe, but I rather suspect that the biggest driver of opinion for the two formats will be this. For the HD-DVD:

prices start at $500

For the Blu-Ray:

prices start at $1,000

That differential will hit everyone, including those that pay no attention to the online forums. I'm sure that a set of influencers touting performance differences will have an impact - but the price differential will have a bigger one.

We are having our 4th of July bbq a day early - we expect quite a crowd. Things are quiet right now, before they start arriving:

That's the patio we put in by hand a few years ago. It's a nice place now that I've forgotten the pain of moving all that dirt :)

Wouldn't be the fourth without the flag. Soon, this food will be all gone, and more will be arriving to keep it company - not to mention the burgers and dogs!

Happy 4th!

An interesting question came up on the Smalltalk IRC channel. Say I create a class that I want to use as a model for something in a database. Say the database columns start with capital letters, and I want my accessors and mutators to match those names. If you define the class using the class creation tool, it will force the first letter of each of those methods to lower case - not what you want. You don't want to create all those methods by hand, so what do you do?

Well, out of the mists of old muscle memory I recalled class CodingAssistant. Before the RB was integrated, it was a tool that I had added into the browser's menu, so that I could easily generate those methods. It comes in with the UIPainter parcel, so just load that, and then open the tool this way:

Tools.CodingAssistant open

Then enter your class name, select the variables to generate code for, and you're done.

Technorati Tags: smalltalk

I've made progress on freeing two packages from their dependency on BottomFeeder - PatchFileDelivery and SyndicationHandling. The former is a simple toolset for setting up HTTP based updates to an application. To use it, you have to set up an XML based config file on a server, and have HTTP access to that file. Setting up the file works like this:

defs := OrderedCollection new.
list := #('ParcelFileName/pcl').
names := #('My Application Parcel').
descripts := #('Description of this component').
sizes := list collect: [:each | each asFilename fileSize]. 
allows := #(true).
list do: [:each | | version nm timestamp properties index |
	properties := [CodeReader new readInfoFromFileNamed: each] 
				on: OsError, CodeReader fileFormatSignal
				do: [:ex | ex return: Dictionary new].
	version := properties at: #version.
	nm := properties at: #parcel.
	timestamp := properties at: #timestamp.
	index := list indexOf: each.
	comp := ComponentDefinition
		parcelName: nm
		parcelFilename: (each asFilename tail asString)
		version: version
		releaseDate: timestamp
		descriptiveName: (names at: index).
	comp description: (descripts at: index).
	comp fileSize: (sizes at: index).
	comp allowDynamicLoad: (allows at: index).
	defs add: comp].

That sets up a simple configuration file. You place that in an accessible location (along with the parcel, obviously), and you're ready to go. On the client end, you need to have the UpgradeManager class check for updates. You do that with the #grabRemoteUpgradeDefinitions message. That looks like this:

grabRemoteUpgradeDefinitions
	"try to connect to remote server and get the available updates"

	|  url  defs |
	url := self settings upgradeURL last = $/
		ifTrue: [self settings upgradeURL, self settings upgradeFilename]
		ifFalse: [self settings upgradeURL, '/', self settings upgradeFilename].
	defs := (XMLConfigFile loadFromURL: url).
	(defs isNil or: [defs isEmpty])
		ifTrue: [self patches: #()]
		ifFalse: [| tmps | 
				tmps := defs first.
				tmps do: [:each |
					each oldVersion: each getCurrentlyLoadedVersion].
				self patches: (tmps select: [:each | each couldReplaceLoadedComponent])]

That uses the XMLConfigFile class (in a separate package) to load the update definitions. Those are then checked agains what's actually loaded, and we come up with a list of potential updates. In Bf, that's presented to the user - you could automate it from there. In the UpgradeManager, there's an #upgrade: and #upgradeAll API for doing that.

The other package, Syndication-Handling, is not as cleanly separated yet. There are a few ugly #isDefined checks, but it does load and work independently of Bf. Once you load it, you can see how it works by doing this:

doc := Constructor  
	documentFromURL: 'http://www.cincomsmalltalk.com/rssBlog/rssBlogView.xml' 
	forceUpdate: true 
	useMaskedAgent: false.
cls := Constructor determineClassToHandle: doc content.
target := cls objectForData.
feed := cls 
	processDocument: doc content
	from: 'http://www.cincomsmalltalk.com/rssBlog/rssBlogView.xml' 
	into: target.

You'll get back this if you inspect to feed object:

That code should work against any version of RSS or Atom out there, and should give you back a FeedList (a collection of feeds) if you hand it OPML or OCS.

My daughter's friends are all out (they had a sleepover last night), and now it's time to get ready for a holiday barbecue. We're having it today instead of on the 4th itself - this way, no one has to head out early for Fireworks. We've made mighty preparations - I have 12 pounds of beef ready to lay across the grill. Happy Fourth of July weekend to all!

I mentioned in passing that I was subscribed to 320 feeds the other day. This evening I did some weeding - it turns out that some of the feeds had been inactive for a long time (some for over a year). Pulling those out dropped me down to 284. Still a lot, but it's under the 300 level :)

With the Summer Release out the door, it's time to take a look at what's new in the upcoming Winter Release. Here's the list of what's planned:

ObjectStudio

• ObjectStudio 8 initial support (Probably Preview)
• ObjectStudio running on the VW VM
• Bug fixes for ObjectStudio 7.1.2 (7.x existing VM path)

We will have more details on ObjectStudio 8 as initial beta testing happens

VisualWorks

• VM
• 64 bit support for HP (PA Risc) and PPC (PPC Mac, AIX)
• Support for intel based Mac OS X (should be available after the summer release, and before the winter release)
• New, more stable PPC Mac OS X VM (should be available after the summer release, and before the winter release)
• Shared Perm space implemented on 64 bit platforms
• Loadable GUI on PPC/AIX & HPPA/HPUX
• 64 bit DLLCC on all supported 64 bit platforms
• Initial support for deploying a Smalltalk app as a DLL/shared lib with callable APIs
• Base System
• Non-blocking DNS lookups
• Fix for font lookup issues that impact some Linux distributions
• Threaded COM Support
• Win CE support folded into the base system
• Continued work on simplifying the deployment of Smalltalk applications
• Delivery of a smaller base.im
• Continued Unicode improvements, including VM/image level support on the Mac
• More Locales - Chinese, possibly others
• Store
• Maintain Override loading order
• Atomic Source Loading
• Configuration Management with tools
• Improvements to the Merge Tool
• Tools
• Preview support for Splash, the Pollock based GUI Builder
• Continued work on unifying the tools sub-strata for inspectors, debugger, and browsers
• XSchema moved to supported state
• Security
• Ciphers - CTR Mode, RSA-PKCS1v2 padding
• OpenSSL wrapper - add RSA, DSA, DH (preview)
• X.509
• certificate creation APIs
• more extensions
• SSL
• TLS v 1.1/1.2
• Pluggable APIs for the OpenSSL wrapper
• ASN.1 - further improvements
• Net Clients
• MIME enhancements
• HTTPS proxying
• POP3/IMAP/SMTP
• Authentication
• SSL
• Smalltalk to Smalltalk (Opentalk)
• bi-directional connections
• firewall/NAT traversal
• secure connections (SSL)
• Web Services
• Header support in WS Tools
• X2O binding editor
• WSDL binding editor
• Pollock - see the Roadmap for details
• Deliver Feature Set 3
• Web Toolkit
• Scripting (startup) support
• Better headless operation
• Browser (Web) Plugin
• Full support for Windows IE
• CAB installer support (Windows)
• Possible support for Linux/Mozilla and Safari
• GLORP
• Move to supported state
• Database
• Sybase 15 support
• SQL Server 2005 Support
• Oracle timestamp data type support
• ODBC Connection Pooling
• MySQL Support
• Connects for 64 bit platforms

I guess it was a good idea to just turn trackbacks off - since about 4 AM on the 29th, there have been 14,201 attempted trackbacks (and most likely a handful were actual, non-spam trackbacks).

The Times catches up with the fact that every customer interaction is now part of marketing - they have an analysis up of the "please cancel my AOL" thing that made the rounds a few weeks ago (as well as the "sleeping Comcast technician" incident). Here's their take:

How should Mr. Finkelstein have responded? By writing a letter of complaint to some distant regulatory authority that will require years before it acts? Far more effective means are now at hand. He recorded, then uploaded the video clip with some humorous asides about missed appointments and unfulfilled promises, and got immediate satisfaction in the act of sharing. More than 500,000 viewers have watched Mr. Finkelstein's video "thank you" note to Comcast.

AOL and Comcast executives in charge of customer service may long for the good old days when they had to deal only with a finite number of federal regulators and state attorneys general, not a universe of millions of Web-savvy customers.

Maybe those execs should buy Glenn Reynold's book :) The fact is, every customer interaction is now a potential marketing incident - and the more Kafkaesque ones can create huge blowback. To wit - AOL has been flagged by regulators for this sort of thing before - but it didn't take. This sort of thing has a far better chance of succeeding, because it puts the negative experience of the people affected right on the front burner:

AOL internally boasts to its employees that third-party verification is an "industry-first initiative to guarantee quality," but isn't this like a parolee showing off his electronic ankle bracelet as proof of how trustworthy he is? The public embarrassment of the settlement faded with time, but then Mr. Ferrari's five-minute recording undid 10 months of public relations repair work.

Seems that even the slow learners at AOL have finally gotten the message:

On the Monday after the public debut of Mr. Ferrari's call to AOL, Scott Falconer, an AOL executive vice president, sent an e-mail message to company employees alerting them to Mr. Ferrari's blog post and warned, "On any interaction, you should assume that it could be posted on the Web."

That's only been obvious for a few years now. You would think that a supposed tech company would get that, but their failure to adapt to the broadband world has led them down a really stupid path. Reading the rest of the article, it sounds like they'll need a few more object lessons before they really get it.

James McGovern:

I suspect that if I refer to the original creators of the agile manifesto as a club they may go into defense mode. Maybe they could tell us why none of the original members worked for large enterprises, the federal government, large consulting firms and so on. While the work was good, was this in rebellion to something else?

Well, the creation of XP itself happened on the C3 project at Chrysler. Which was (and is), a pretty big enterprise shop. In fact, the C3 project was an attempt to rescue an enterprisey project that was failing.

Scoble explains in a Wired interview why RSS will remain important - but will also stay niche:

In an RSS aggregator it shows you any new things, so it makes it so much more productive. So the first thing I ask is, do you read more than a couple of sites? Do you keep going back to the same sites?

And that's why it's not going to go mainstream, because if you're reading only one or two, there's not enough of a productivity gain for you to care.

That's why journalists and newsfreaks care, people who are passionate about keeping up with lots of things. It's the passionate ones, and they're the ones getting everybody to pay attention anyway.

That's about the size of it. People like me (I'm subscribed to 311 feeds right now) are outliers - most people just aren't interested in that much news and information. If you visit a handful of sites once or twice a day, a browser works great, and bookmarks are sufficient. It's only when you start trying to follow tens (or, like me, hundreds) of news sources a day that an aggregator becomes critical.

I've posted a new dev build - the one I posted yesterday had some issues with the update tool. THis one seems fine that way - go to the download page, and scroll down to the Dev section. Bear in mind that this is a dev build, so it would be prudent to back stuff up :)

Update: If you do grab this dev build, then grab the update for the main (BottomFeeder) component - there was a missing code problem in that piece, and the update addresses that.

It's time for my weekly look at the logs - first up, BottomFeeder downloads, which ran at a rate of 161 per day last week:

Platform	BottomFeeder Downloads
Windows	409
Update	201
Mac X	117
Linux x86	112
CE ARM	94
Mac 8/9	54
Solaris	34
AIX	29
HPUX	28
Windows98/ME	23
Sources	14
Linux Sparc	9
SGI	3
Linux PPC	3
CE x86	2
ADUX	1

Those look about the same as always. Next, the HTML page accesses:

Tool	Percentage of Accesses
Mozilla	64.1%
Internet Explorer	25.1%
MSN Bot	3.2%
Other	3.8%
Opera	2.8%
Megite	1%

That looks like the normal distribution for the site. Interestingly, the absolute number of pageviews doesn't seem to be rising much, while the number of unique IP addresses in that mix is rising. Good trend, I'd say. Finally, the RSS accesses:

Tool	Percentage of Accesses
BottomFeeder	21.1%
Mozilla	20.2%
Other	13.8%
BlogLines	10%
Net News Wire	8%
Safari RSS	4.4%
Google Feed Fetcher	3.6%
NewsGator	3.2%
Internet Explorer	2.4%
Planet Smalltalk	1.7%
BlogSearch	1.7%
RSS Bandit	1.5%
RSS 2 Email	1.3%
Opera	1.3%
SharpReader	1.1%
JetBrains	1%
Feed Reader	1%
Java	1%
MSN Bot	1%
Liferea	1%
Lilina	1%

Slight increase in share for Bf this week, but about normal otherwise.

Doc Searls links to Mike Arrington, who's taken note of PayPerPost.com. Here's Mike's description:

Ted Murphy, CEO of advertising firm Mindcomet, has launched a new service called PayPerPost.com. You guessed it, it’s a marketplace for companies to connect with bloggers who are willing to blog about a product - for a price. The companies can set guidelines for their requests such as whether a picture must be included and whether they will only pay for positive blog coverage. There does not appear to be any requirement that the payment for coverage be disclosed. There is a requirement that PayPerPost.com must approve your post before you are paid. Wow.

This sounds like a blogosphere version of product placement. You know - you watch a movie, and when a laptop gets opened, it's a MacBook. When a soda gets drunk it's a coke. And so on. I'm not sure I'd call this the end of the blogosphere as we know it, but it sure will make separating the wheat from the chaff a lot harder.

Of course, you already need to pay attention to the messenger. I do Smalltalk advocacy here, and you should take note of the fact that I'm the product manager for Cincom Smalltalk. The relevant bit here is that I disclose that - and as Mike notes above, there's no apparent disclosure in the PayPerPost system. What this boils down to is that anyone using this service had better hope that they aren't found out - their stealth marketing will turn into an anti-campaign immediately.

I've just posted a new development build of BottomFeeder on the site - this is based on the newly released VW 7.4.1, and should support NTLM proxy servers. As well, the spell check code for the editor is now pure Smalltalk - which means that it's faster, and works on all supported platforms. The build scripts are updated as well - you should be able to build from the Public Store or from parcels.

The download site has been updated - follow the link that was sent to you in email, or register for the NC here now. VW 7.4.1 and OST 7.1.1 are ready for download. Enjoy!

Update: All the configuration file issues have been dealt with now; the new bits are ready for download

Microsoft shuffled their exec team again - it's starting to look a lot like IBM back in the bad days of the early 90's....

This is a funny story, but at the same time it illustrates a problem not addressed by the humor:

This recently hired pilot fish is trying to improve security by installing security cables to lock each laptop to a desk. That's simple enough, right?

So IT locks each laptop to a desk to prevent theft. Ok... but that also completely destroys the rationale for getting a laptop in the first place. The point is mobility, both inside the office and outside of it. If your plan is to lock them down, just get desktop machines and flatscreen monitors - it will be less expensive, and easier to upgrade.

It's things like this that lead the user community to just shake their heads at the ideas that flow down from management and IT...

Put Lessig on the list of utopians who believe that we can get neutrality legislation for the internet, but not get any of the content restrictions that have appeared on TV and Radio with it. You want the "7 bad words" banned? Then advocate for neutrality. You want the political free for all of the current blogosphere to fall under campaign finance laws? Then advocate for reform.

While you're at it, check the pile of manure - there might be a pony in there too.

It looks like OSS might get hurt as a side effect of this patent suit - even though the patent itself is blatantly stupid. FireStar claims to have invented O/R mapping in this patent. Hmm.

Just off the top of my head, The Object People created TopLink back in the early 90's. Around 1992 ParcPlace shipped the ObjectLens. There are likely hundred of others out there; those are just two Smalltalk specific implementations that come immediately to mind. Reading through the patent, it looks like the ObjectLens should certainly be prior art.

This is yet another example of how patents and software don't mix. Copyright ought to be enough.

Explain to me again how class action suits help people?

Boris Popov has extended the RBStoreExtensions package to add color to the graph of version history. Check out his extensions in the public repository

I should mention that Boris has moved his blog there - you can subscribe to him here.

The Summer Release of Cincom Smalltalk is officially out - upgrades are being sent to existing customers now. I'll be updating the NC download application today.

Marten Feldtmann has joined the blogging lineup here - go check out his thoughts on Smalltalk and C# over here.

James McGovern throws the gauntlet down to the Ruby crowd again, challenging them to get something published:

Folks reading this blog entry need to consider adding him to their blogroll... Awhile back I threw out the challenge to the Ruby community that if within thirty days, they could get a single Fortune 100 enterprise whose primary business isn't technology to tell a story in a public forum (conference or magazine) about how they used Ruby to develop an enterprise application (aka system of record) that I would make a sizable donation to a mutually agreed upon charity. I still have my money in my pocket.

He repeats the challenge further down in his post. The thing is, non-technology companies have no real motivation to do that. If problems are being solved, they tend not to care how it was done - at least in the corporate marketing and executive suites. In the standards/architect groups, and in IT groups, on the other hand, they tend to be overly concerned with following the herd, due to the perceived safety - see my post earlier on that.

The people who most need to wonder about potentially better ways of doing things are the architects and IT managers. Asking for case studies - which will have to flow from the Marketing group - isn't going to move the ball forward.

David van Couvering has two issues with DabbleDB:

I did look at the privacy policy, and I have to admit after all the news of data getting stolen and compromised, it didn't give me a warm fuzzy. No discussion of security auditing; no discussion of how they are securing your data; no clear ability to ensure that your data has been removed completely when you want it removed. Hmm...

I had thought of maybe using it to quickly create some apps for my personal data, but found myself backing off from putting my data on their server. It appears that their focus is on SMBs, but would SMBs want to put their business data online like this? I also found myself not wanting to pay the monthly fee; I do hope they can find another business model that doesn't require the consumer to pay. Maybe they could make it free up to 10MB of data or something. Otherwise I'm concerned that they just won't get the uptick I think they deserve.

I understand the first concern - but looking at the privacy policy, it seems reasonable to me. As to getting data out - not having done more than just look at Dabble, I can't really say much about that.

The second concern is one of those things I find fascinating - the expectation that stuff should be free. Richard Stallman may be happy with that theory, but here's the thing - it costs real money to pay a mortgage and put food on the table. Where is this expectation of free coming from? Maybe developers should stop getting paid - that may be the only way to insert some reality back into the software space.

If I doubted the need for a better deployment answer for Cincom Smalltalk, I have been being reminded all afternoon. The process of working through Runtime Packager issues on a new release (7.4.1 in this case) are never fun. It's always fairly straightforward once I get the packaging script saved with the right parameter set, but until then - utter pain. This remains at the top of my agenda for things that need to be made better in the product.

I'll be getting an initial build of BottomFeeder on VW 7.4.1 out this afternoon - the major new stuff is plumbing (we now have NTLM auth support, for instance). I'll have a summary of what's new soon - I'm hoping to release Bf 4.2 on top of 7.4.1 in the very near future.

I'll be at a customer location today, completely without network access (I know, the horror :) ). I should be back online later this afternoon.

My wife and daughter wanted to see "The Lake House" this afternoon - I went in with no expectations, but I was pleasantly surprised. The movie hung together pretty well, once you bought into the "two years apart" plot device. I did periodically think "why doesn't one of them Google the other?", but that would have destroyed the plot.

It's definitely in the "chick flick" area, but it was pretty well done. On balance, I recommend it.

Keith Ray notices something:

Why do companies that hire only "the best" employees, and want to provide "market-leading" products or services (again, their goal is to be the "best") choose to go with average technology in implementing these products and services? And then defend that technology choice by saying that they want a bigger pool of developers from which to hire -- more people that already know that technology.

If you want the best, you won't find it in the center of the herd.

Net Neutrality went down, and I for one am glad to see that:

A U.S. Senate panel narrowly rejected strict Net neutrality rules on Wednesday, dealing a grave setback to companies like eBay, Google and Amazon.com that had made enacting them a top political priority this year.

For those of you who think this is a bad thing - Recall the FCC's actions after the Superbowl "wardrobe malfunction". If you think the US government is going to lay down neutrality rules and then keep a hands off attitude beyond that, you probably also think you'll find a pony under every large pile of manure...

DabbleDB launches tomorrow, and there's more press around that, here and here. The latter story notes that DabbleDB is in the space (end user development) that Microsoft is only talking about.

I was listening to the latest Gillmor Gang, and the end of the conversation caught my attention. Someone (can't recall who) suggested that Gillmore start including transcripts of the gang podcast, as a way of increasing listenership - apparently, the opportunity to scan the transcript drives more people to listen. Gillmor was having none of that, but it got me thinking.

I listened to the podcast while I was out jogging (have to take advantage of the dry air here). If I want to take something in while jogging, audio is my only option - I can't read while I'm plodding along, nor can I watch video. While I'm in my office during the day, I can watch video, but it sucks down time - for that matter, so does Audio. If it's much longer than the average RocketBoom, I probably don't have time for it - I'd much rather read text.

During the evening, I don't want audio or video. I'll take my laptop into the living room, but my wife finds it offputting if I slap headphones on while I'm sitting with her - so again, text works best then. Now, if I drove more, that would fit in more time for podcasts - and admittedly, my low rate of driving is an anomaly compared to most people. Even so - video doesn't work for driving.

To my mind, that limits the potential reach of video blogging - it takes as much time to get through as audio, and requires a lot more attention. For me, at least (and I suspect, for lots of other people), the potential time for audio is much bigger: exercise time, working in the yard/garden, driving. Text works anytime I'm in front of a PC, and chews far less of my time than audio. Video? Beyond training and education, I'm thinking that the time allotment for professional use is pretty slim.

At the same time, I wonder about the advertising model for audio. I was just listening to the GillMor Gang, and Steve slaps a 6 minute ad at the front of the show. Zip - using fast forward on my iPod, I zoomed right past that. If he made the ad shorter - say 15-30 seconds - I might not bother. Six minutes though? You've got to be kidding.

And then there's astoundingly big. From Greg Linden:

Google reportedly had an estimated 450k machines two months ago and adds machines at roughly 100k per quarter. In 2004, each of these machines had 2-4G of memory, and, two years later, likely are up to 8G standard.

That means that Google can store roughly 500k * 8G = 4 petabytes of data in memory on their cluster.

I'm guessing that their local power company loves them, a lot.

Hat tip Patrick Logan

Ted Neward analogized the Vietnam War to O/R mapping, and now has to respond to the predictable complaints:

"The Vietnam War is a bad analogy for O/R-M." Vietnam remains, for most Americans, as the quintessential symbol for "bloody, ugly, unresolvable quagmire". And, as some have pointed out in comments on the blog post already, all analogies break down eventually, and this one is no different--as one commenter put it, nobody ever died from a bad O/R-M tool. (Though the day is not far off when such could occur, given the incredible spread of technology into all corners of our lives--it's not too hard to imagine a day when a patient dies because a doctor received incorrect information about a medical allergy from the enterprise system he/she uses to call up patient records.)

Rule number one when making an analogy: Don't pick one that will immediately drive the conversation into a ditch. It's like saying: "whatever you do, don't think of a zebra"...

Doc Searls, in the process of explaining how the non-compliance of low power FM transmitters (for the iPod, etc), gets to why I don't trust governmental regulation:

One is the fact that Congress and the FCC have done more to make "free over-the-air radio broadcast services" unattractive than a million little FM transmitters ever could. Between relaxed ownership rules and increased "indecency" fines by the FCC, the AM and FM bands have become boring beyond endurance. I have my problems with satellite radio too (for all their diversity, all the channels on both services are owned by one company apiece, and the two silo'd systems are entirely incompatible); but my Sirius radio provides infinitely more usefulness than I can even imagine getting from "free" radio today. All that's left for me on "free" radio are our local AM news station and NPR/PRI programs, most of which I get now via Sirius or podcasting.

Once the Feds define the net as a "public utility", it will go the way of TV and radio - with bozo rules like the ones Doc rightly condemns here.

Ed Bott is worried about where Microsoft is taking the WGA program, and I don't blame him. The non-denials from Redmond are none too encouraging either. The measured walk toward 1980's style IBM bureaucratic stupidity seems to be transitioning to a jog. If they actually start requiring that you have the upgrade system on, then they'll have broken into a full-blown dash.

Can you imagine the hair pulling in IT departments as they try to sort out driver problems after a blown upgrade that they couldn't test first?

We will have the Summer Release of Cincom Smalltalk shipping as of June 30th. As soon as that's done, we'll post the new bits for NC download. The details:

Release Date: June 30, 2006

The summer releases of Cincom Smalltalk are maintenance/bug fix releases. As such, you won't see any new features coming out. You will see enhancements, bug fixes, etc.

The new VMs for Mac OSX (Power PC only and intel Mac VM) will follow after the summer release, but before the Winter release). We will formally ship these new VM's in the winter release, but they will be available via vw-dev (and to other interested parties) before the winter release.

Highlights for VisualWorks

• Security - We are including an implementation of PKCS #8
• COM - The COM Automation Wizard can now save and restore settings to create a VW COM server image.
• Font Matching - on Font matching failures, the system will return the best match it can find instead of raising an exception
• WS* - Further Enhancements to various aspects of our WebServices implementation
• NetClients - We have implemented support for Digest Authentication and for NTLM Authentication

There are various other improvements and bug fixes; the file fixedARs.txt on the CD includes an exhaustive list

Highlights for ObjectStudio

• We are providing Early Access for ObjectStudio 8 by request only - please contact James Robertson if you are interested
• OLE bug fixes and enhancements
• Database bug fixes and enhancements
• Both the XML Parser and the Opentalk framework have been synchronized with the VisualWorks implementations

Looks like the hardware vendors have figured out that a Blu-Ray/HD-DVD format war isn't a great idea: there are plans to start shipping dual mode devices:

Samsung and Toshiba have joined forces to end the format wars for good. They are releasing a hybrid player that plays both Blu-ray and HD-DVD formats. But that’s not all! Sony and NEC are also releasing a dual-format player of Blu-ray and HD-DVD. This is back-up for Sony, since they did loose miserably with Beta.

I'm certainly more likely to buy one of those in the short term. My other option would be to hold off until a clear winner emerged.

Cincom's President, Tom Nies, is speaking at a UN sponsored conference in NY this evening. I posted details on this here last month.

Steve Rubel:

In a terrific reversal of its prior stance towards the site, NBC is set to run promos on YouTube , according to AP.

Maybe they could explain it to the MPAA and the RIAA.

Looks like Avi and Andrew have decided that they need an infusion - Om Malik is reporting that they've taken $2M in venture capital:

Dabble DB, a Vancouver-based online database / hosted application creation company has raised around $2 million dollars in Series A venture funding from Ventures West, a Canadian VC fund. Paul Kedrosky, a venture partner with the fund, is leading the investment in the eighteen-month-old company. The company will also come out of “beta” tomorrow.

That's interesting, given what Avi said about venture money last April:

DabbleDB came out of their experience in consulting - the ad-hoc spread of semi-shared data that really should have been fully shared (eg - emailed spreadsheets). Had they tried this a decade ago, they would have gone the whole VC "take the money" route. That's not the way they went - they believe in a "late binding" approach to business planning. Once you take venture money, a lot of options get closed off - you are committing to a specific set of plans. So Avi's take: Taking Venture $$ is a premature optimization

Sounds to me like they are still in late binding mode though - they held off on venture capital until they needed it. I hope they don't run into any of the classic "business optimization" scenarios that can come with funding.

Update: Mike Arrington covers the story

Update2: Tim Bray has some kind words for DabbleDB, Avi, and Andrew.