Smalltalk Tidbits, Industry Rants

Doc Searls got blasted by Dave Winer today over a number of things, including podcasting:

Doc, how about looking at your words from our perspective. Talking to us through you ain't going to cut it. Your friends who want to earn the respect of the podcasters should explain in the medium, in their own voices, in their own words -- produce a podcast and tell us what the f*** they're doing, instead of leaving us guessing. Then you might see the hostility ease, because that's where it comes from.

Whereupon Doc says that he won't say a word about podcasting until he does one. Ok - this is a really stupid meme that ticks me off in more ways than I can count. What specifically, you ask - the one Winer leads with (but he's hardly the only one making this sort of argument) - summarized, it's: "You can't criticize/comment on that - you've never done it!"

That's a completely lame, stupid argument, and the people who make that sort of argument deserve to beaten with a cluestick. And yes, I've been guilty of the same lame argument with respect to Smalltalk, so I'll keep that in mind. There are tons of things that it's impractical to have first hand knowledge of - it's highly unlikely that I'll be writing for a newspaper anytime soon, or reporting on TV, or researching drugs, or fighting a war. Never mind the things I couldn't do even if I wanted to (give birth, say). In the constricted universe of the "no experience, no input" crowd, that means I don't get to have an opinion on any of those things.

Well heck - how do any of us get to have an opinion of any sort, then? No one can try everything - do we really want to limit debate (pick a topic, any topic) to those with direct experience? Bah.

It looks like Librarian Gorman isn't the only one upset that the little people have voices now - just look at Tina Brown moan:

We are in the Eggshell Era, in which everyone has to tiptoe around because there's a world of busybodies out there who are being paid to catch you out -- and a public that is slowly being trained to accept a culture of finks. We're always under surveillance; cameras watch us wherever we go; paparazzi make small fortunes snapping glamour goddesses picking their noses; everything is on tape, with transcripts available. No matter who you are, someone is ready and willing to rat you out. Even the rats themselves have to look over their shoulders, because some smaller rat is always waiting in the wings. Bloggers are the new Stasi. All the timidity this engenders, all this watching your mouth has started to feel positively un-American.

That's right - the media actually has to pay attention to facts, or they'll get caught out. Oh, the horror. Time to put on ashes and sackcloth for sure - it's just too hard! The condescension in that piece is amazing - she's not even trying to obscure how little she thinks of the great unwashed masses. Here's a tip Tina - if you can't handle the heat, get out of the kitchen. Unlike most of use bloggers, you get paid for the stuff you write - so spend some actual time on it.

Oh, here's a tip for you - I found a definition of the word "fact". It looks like you aren't really sure what a fact is, so I thought I'd help.

There's a new dev build of BottomFeeder online as well - go to the download page, scroll down to dev builds, and grab it. I'm getting pretty close to a 3.9 release - Rich is working on documenting the Blog posting tool (now that it really works with the Blogger and MetaWebLog APIs properly).

I've just added two new features to the Silt server:

• Per category RSS feeds
• Potential per-item RSS feeds

The former are self explanatory - each category now has an associated feed - the XML links are in the sidebar. The per item feeds need a bit more explanation. As an anti-spam move, I started closing down comments to items that had aged out of the RSS feed - I figured they were effectively invisible (to me) once they left the feed, and I wouldn't see any spam that came in. I recently added a feature that allows bloggers to mark comments open for individual posts. So, any such post now has an associated per-item feed (there will be a feed link right on the page for such items).

This guy got my (negative) interest in a couple of ways. First off - Smalltalk is hardly dead - our business is growing year on year - both from existing customers and new ones. Secondly, a .NET Passport required to add comments? Good gosh, why not require a signature in blood too?

One thing that I haven't made that clear about BottomFeeder is that it's been designed for extensibility via plugins - I've created a number myself - the Posting Tool and the Enclosure Handler. I've packaged TypeLess as a plugin, as well as a number of games that Steve Kelly fixed up for later releases of VisualWorks. I've also done some additional work on Bob Westergaard's scripting plugin. All of this stuff either ships with BottomFeeder, or is available for download via the update tool.

The plugin mechanism is pretty simple. There are a number of events that BottomFeeder reports that make it easy for applications to deal with things happening:

• #bfOnline - reported when the application goes online
• #bfOffline - reported when the application goes offline
• #bottomFeederStarting: - reported when BottomFeeder is starting. The RSSFeedViewer instance is sent as an argument
• #bottomFeederQuitting: - reported when BottomFeeder is quitting. This happens both in development and in deployment - the RssFeedViewer instance is sent as an argument
• #quitting - reported in deployment only, when BottomFeeder is quitting. BottomFeeder will not quit until all handlers have returned
• #addedFeedList: - reported when a feed list has been added. The FeedList is sent as an argument
• #addedFeed: - reported when a feed has been added. The Feed is sent as an argument
• #newItemsFor: - reported when there are new items in a feed. The Feed is sent as an argument
• #removeFeed: - reported when a feed is removed. The Feed url is sent as an argument
• #removeFeedList: - reported when a feedlist is removed. The Feedlist url is sent as an argument

The other thing you need to do is register your plugin with BottomFeeder. The way you do this is to add a small message to the #postLoad action of your plugin parcel (the plugin should be packaged as a parcel). That snippet looks like this:

#{RSS.RSSFeedViewer} ifDefinedDo: [:cls | 
	cls registerPluginClass: BlogTools.PostingTool 
	startupMessage: #openWith: 
	label: 'CST Blog Tool'].

That's the message send that's used by the Posting Tool to register itself. There are actually two forms of registration possible, depending on what your plugin is going to do:

• registerPluginClass: aClass startupMessage: aSymbol label: aString
• registerPluginClass: aClass startupMessage: aSymbol label: aString isForUI: aBoolean

You can send the second to specify that there's really no UI if necessary; the first will assume one and add an item to the 'plugins' menu. Finally, there's a way to unregister a plugin as well. This code should be put into the parcel preUnload: block:

unregisterPluginClass: aClass

With that, you can properly unhook the menu addition. That's pretty much it. You should probably be familiar with the codebase for BottomFeeder to work with this, but that's easy enough - it's all in the public store.

I've been meaning to pick up a book I just saw referenced on the Samizdata blog - "Armageddon - The Battle for Germany, 1944-1945". So far, the only thing stopping me is the really, really melancholy nature of the subject - I'm both interested and repelled. I'll get to that book eventually, but there's another WWI related tome I intend to pick up first - "Paris 1919: Six Months that Changed the World". I've been trying to get a handle on the first world war, and I really have to go through the post-armistice conference to get a fuller picture. Maybe I'll be able to turn my attentions to a different topic after that; at the moment, I'm kind of immersed...

Via Daver Winer comes news that the term "Podcast" is being trademarked - I found more on it over here. This post on the Trademark blog has the details - Apparently the shameless folks at Shae Spencer Management are the ones doing this - relying on the general cluelessness of the USPTO, I guess.

Dave Thomas (the OTI/ENVY guy, not Prag Dave) spoke at the AOSD conference. I have a lot of respect for Dave, but there's one thing being reported that I have to take exception to:

The later adopters have a long list of questions they just need to have answered. Often have very different skills and much less software engineering experience than any AOSD attendee can imagine. Want standards, customer references, case studies and insurance of major vendors/ISVs. Make sure you always operate with honesty, integrity, and modest claims. Dave : we actualy persuaded ourselves that people could program in Smalltalk - they can't! Gurus can, but the man on the street can't.

Oh? Well heck, they sure can't code in Java or C# then, because those are vastly more complicated. Or Fortran. or C++. Sheesh. People think Smalltalk is hard because it's different, not because it's actually hard.

This column by Allen Holub of SDTimes magazine has to be the best thing I've read about development in eons. He points out that software engineering isn't engineering, and that mathematics simply isn't that relevant to most developers. I couldn't agree with him more:

In fact, computer science is neither a science nor an engineering discipline. Science concerns itself with the formulation and proof of hypotheses. Programmers just don’t do that. Similarly, all engineering disciplines except software engineering concern themselves primarily with the mathematical analysis of structures, be they physical structures or electronic circuits. Programmers don’t do that, either.

Indeed, “software engineering” books don’t talk about engineering at all, at least not in the way that mechanical or electrical engineering books cover their subjects. Software engineering is about process, not structural analysis. The closest thing that software engineering has to real engineering is the study of design patterns, but even these are nebulous. There is no single “correct” structure for the realization of a design pattern. Even calling the creation of software “engineering” is a misnomer in my mind.

That's exactly right. Think about Unit tests, for instance. A structural engineer building a bridge doesn't need tests - what he needs is a blueprint and the materials. Armed with those, he can tell you whether the bridge will stay up or not. In software, the basic design tells us nothing of the kind. If it did, we wouldn't need tests. So given that, what is software development? Back to Allen:

So if programming isn’t science or engineering, what is it? It’s a liberal art. Modern programming bears more similarity to creative writing than to engineering or physics. The design process that you go through (or at least should go through) to create a program is almost identical to the process that you use to write a book: research, formulating a thesis (or problem definition), an orderly exposition of the thesis. These steps are central to both expository writing and object-oriented analysis and design.

In other words, it's not really a "hard" science at all - it's one of the soft, squishy ones, where the answers are subjective and we have to puzzle out the best answers from among many possible ones. Consider what we actually have in the popular languages in this light - we have primitive data types in Java (and in the CLR as well) - because someone decided that premature optimization of low level math stuff was important. Yes, I know that some people need that speed. But for the vast majority of developers on the vast majority of projects - you aren't one of those people. The problem for most of us is developer productivity and time to market - not how fast we can add up the numbers that show how many users of IE hit the web page today. Carrying on from that point:

We should really give up on the notion of math being the foundation of computer science, drop hard-core mathematics from the curriculum, and replace it with English composition (which teaches you how to write large, complex, documents like computer programs) and Latin (which teaches you how to analyze complex linguistic systems).

Bear in mind that logic—the one “mathematical” subject that is inarguably part of programming—is traditionally taught by the philosophy, not the math, department. As an added benefit, if programmers were trained as writers, they’d be able to write coherent documentation and put meaningful comments in their code.

The mathematics that is actually relevant—a bit of set theory and the like—is easily covered in a one-semester class on the order of the Math-for-English-Majors classes offered by most universities.

There are a lot of developers who will have a really hard time letting go of the idea that all inbound CS people should take a compiler class. Ask yourself though - how many people are doing that kind of work, and how many are creating what amounts to glorified reporting interfaces?

In the last few days, I've had two things that I took to calling "Heisenbugs" - one of the blogs here had a never-changing referer list, and the server, after having been restarted, stopped sending blog pings. These were weird bugs that I just couldn't understand - I kept querying the server, and things looked ok - but wouldn't work correctly.

What I discovered is that they both had (of course) logical explanations. For the referer problem, there was an old log file that was being read from - and sure enough, it had the "never changing" list of referers. After I cleaned that up, the problem went away.

The other one was stranger. On posts, I send blog pings to a variety of places - blog.gs, Technorati, etc. These had worked before, but they weren't working now - and the failure seemed inexplicable, a MessageNotUnderstood deep in the bowels of the Http library. The issue was simple, once I figured it out. There's a cached dictionary of header fields mapped to classes that handle them. What happened was this - when the server started, that cache got filled before all the code was loaded into the system - leaving it partially filled. The solution was simple - clear the cache, and force it to reload.

So I didn't have Heisenbugs - I had bugs I hadn't figured out yet.

Patrick Logan notes that Microsoft has hired a new VP with a suitable background :)

Chris Petrilli may actually think less of IT industry analysts than I do. I wouldn't have thought such a thing was possible :)

"If you’re evaluating tools, and you don’t bother to install and test them, then you’re in paper comparisons and someone will always look better. Feature wars are pointless exercises. You can lead an analyst to water, but you can’t make him any less of a myopic moron. "

"Solve real-world problems in simple to use ways and you will have the best advertising in the world… word-of-mouth. I’ve already told over a dozen other groups about Trac, all of which have adopted it. None of which would have even looked at it were it not for a recommendation. "

I stand by the comment, because I think most “analysts” are little more than overpaid sock puppet shills.

If you are a Smalltalker - or interested in learning about Smalltalk - in the greater Washingtin DC area, then check out the new DC Metro Smalltalk page. We are investigating places to meet now.

Patrick Logan says that Java and C# (et. al.) are not only dead ends, but doomed:

Not only did Lisp and Smalltalk languages and tools support evolution as well as backward compatibility, the culture and organization encouraged even drastic swings in features.

J/J and C/C do not have those benefits, neither technical nor organizational. They are ultimately doomed, so the word to the wise would seem obviously to get off those platforms ASAP. Invest in systems that have proven themselves already for decades.

You can get off the dead end here

I've been making my way through "The Myth of the Great War : A New Military History of World War I", which goes through a military history of the great war. The details are fairly astounding - it looks like the Allies spent the entire war in a parallel universe, convinced that their huge casualties were not so bad, because Germany was getting worse. In fact, as Niall Ferguson detailed in "The Pity of War", nothing of the sort was happening - and by 1917, both the BEF and the French army had ceased to have significant offensive capabilities - they had been bled out by follies like the Somme campaign and Nivelle's 1916 disasters. It's not a matter of cowardice in any way shape or form - based on the accounts I'm reading, the French and the British soldiers (not to mention the Canadians and Australians) fought bravely - it wasn't their fault that they were being led by fools.

If you take a look at the other side of the ledger, Germany (and Austria-Hungary) had knocked Russia, Serbia, Rumania, and Italy out of the war prior to the entry of the US - had the US not come in, it looks to me like the Germans would have turned West with a vengeance (witness the nearly successful 1918 spring/summer offensive) and taken the UK and France out.

Of course, there are issues with Mosier's (the first book above) narrative - he ascribes too many faults to the Allied powers, and too much brilliance to the Central powers. Even given that, it's a worthwhile book - combined with Ferguson's book, it's given me a fresh look at the mostly forgotten conflict that the pre-WWII generation called The Great War. By all means, read the book yourself, and draw your own conclusions. The more I read about that war, the more I realize that it is the central tragedy of the 20th century. Yes, WWII resulted in many things that were a whole lot worse - but WWII was one of the things set in motion by the endgame of WWI. You can look at the 1919 settlement talks and trace down many of the current issues in the middle east and the Balkans. Some wars have results that you can look back on later and be somewhat happy about. There's really nothing about WWI that looks that way.

If you've tried downloading the screencast in the (rather large) zipped AVI file, you might have gotten a server error. If so, try this link for the file. Sorry about that.

It looks like I'll need to upload the AVI file along with the WMV file each week - nearly 30% of the accesses to the screencast I pushed up earlier this week are for the compressed AVI file. Who knew?

It's time to take a look at the log files again. I've walked through the Apache logs for the period between March 6 and March 17, and filtered on the XML accesses - here's a look at the tools that are being used to access the various RSS/Atom feeds on the Cincom Smalltalk site:

I've omitted anything that fell under 1% - so the table above represents 83.7% of the accesses - just over 16% of the accesses to the feeds come from other tools. One thing I've noticed is that the Mozilla numbers keep going up - which tells me that more people are using Sage and similar things. While there are a few IE RSS plugins, not many people are hitting this site with them. Anyway, I'm going to start filing these log reports in their own category so that they're easy to find - I might even be able to spot trends that way :)

I guess we need to make it simple for Scoble and the IE team:

No, I'm not saying that. But, I am saying that what'll get more people to switch browsers in the future is things like having better security (that's why most Firefox users tell me they switched from IE to Firefox). Having better RSS support. Having better tagging support. Having better tabs support. And so on.

I'm not saying that standards don't matter. They do.

But what about Dave Winer's request for a better editor?

Or my request for better inking support? Or, or, or, or.

I'll spell it out - your CSS support sucks eggs. What does that mean? It means that pages that look normal in Firefox look like crap in IE - and it's not because Firefox is adapting for crap, it's because the IE team has apparently decided that CSS2 is "flawed". Well.

You know what? As far as I'm concerned, MetaWebLog API sucks, Atom is pointless, and SOAP is a huge duplication of CORBA on port 80. But you know what else? It doesn't matter what I think, because I have to support these things anyway. Whoever runs Product Management/Marketing for IE - it's time to fire their sorry butt, get someone new in there, and slap them around with a cluestick until they get it. I'm tired of paying the IE penalty for CSS, and so is everyone else.

Let me know if the IE team doesn't get it - I'll use smaller words next time.

Via Learning Seaside comes a pointer to a PDF presentation on Seaside - and I'm quoted on the first page :)

I'm of two minds on this article by eWeek's Scot Peterson. On the one hand, the kinds of security issues that face IT departments are awful - phishing attacks, spam, viruses, worms, etc. - it's a full time job just keeping up. On the other hand, this recommendation - while mostly valuable - has a huge downside:

Make systems that are locked down from the beginning and that restrict what users can do, where they can go and what they can download, or significantly cut the number of power users at your company. It will be hard for users to accept—IT administrators need to be jerks, meanies, hard-asses or whatever you want to call them. This may not be "new" thinking, but until technology can be made simple enough, then perhaps its uses need to be simplified.

There's a problem here - not with Scot's prescription, but with the all too common implementation by IT shops. In my experience, IT shops simply aren't that good at distinguishing classes of users - they like coming up with standards, and then enforcing them across the board. It can be nearly impossible to convince them that you really need a different set of access rules. This is a difficult problem - all the more so because it's not really a technical problem. It's an interaction issue between business units and IT.

Since Steve Kelly decided to start using Silt, both Silt and the client posting tool have been improving a lot. Bugs that I had forgotten about (or simply not run into) got noticed, and then either fixed or pointed out. An extra pair of eyeballs - even one 6 timezones away - is simply invaluable

When IBM did this sort of thing, it was called FUD. In this case, I think it's just an MS marketing guy with no grasp of technology speaking out of his... well, just imagine the orifice he's using. In a CNET article discussing JavaScript and DHTML (now called AJAX, because we needed a new acronym), Charles Fitzgerald says:

The software giant, which pioneered several of the technologies developers are now re-evaluating, dismissed any threat to its plans for XAML. "It's a little depressing that developers are just now wrapping their heads around these things we shipped in the late 20th century," said Charles Fitzgerald, Microsoft's general manager for platform technologies. "But XAML is in a whole other class. This other stuff is very kludgy, very hard to debug. We've seen some pretty impressive hacks, but if you look at what XAML starts to solve, it's a major, major step up."

Yes, a UI description language is a real step up from something completely different. Maybe someone can get Fitzgerald an Apple and an Orange, and see what happens.

Scoble asks a generally good question about the browser business:

Finally, I'm gonna ask a provocative business question of Opera (and other browser manufacturers): What's your business again? If all the browsers have the same underlying features, and they should only add things that are standards, what differentiation are you offering your customers and investors? Are you saying Firefox's developers can't propose anything new that'd push the Web forward? Hey, how about some linking technologies like Greasemonkey? Is Firefox not allowed to add anything like that that the W3C didn't propose and that the WaSP didn't approve of?

I'd take it a lot more seriously if IE didn't screw up my css fairly regularly - and if the IE team weren't already backing off of supporting css properly in IE7.

Blogging Roller thinks that Dare and I are being too hard on Atom. Hmm. Perhaps, but I'm not convinced. A few points - in response to Dare's assertion about category support:

Atom's support for categories is unclear, of course, because that section of the specification has not been written yet. When Atom Protocol is complete, it will offer more, not less, functionality than the MetaWeblog API. Contrast that with the MetaWeblog API, where the spec is "complete" yet categories are still unclear.

Suffice to say that this does not fill me with confidence. Categories are not complicated. A post can have N of them (never mind that this server only supports one category per post :) ). A client has to allow some way to attach one or more categories. While I don't much care for the woefully underspecified MetaWebLog API, it at least takes a stab at it. They've been at Atom with the hammer and tongs for a few years now - if something as basic as categories causes confusion, then I have rather large doubts.

I expressed concern over the base64 inline attachments that Atom allows - to which the answer comes back:

No, that is not the Atom Protocol solution for Podcasting. To reference a Podcast from an Atom entry, you'd use the Atom Link element.

Seems to me that by having both, some people will do one, others will do the other. The inlining is a disaster waiting to happen. I can see it now - I'm traveling, and I've found a hotel that only has dialup. I start my aggregator off getting updates, only to find out that some puzzlewit has inlined a 10 mb file. Oh joy - I get to enjoy downloading that until it clears his feed, because - even with a smart HTTP library, the entire thing will be "new" each time a new post gets added.

Having said that, the Atom posting API looks mostly rational. The trouble there isn't with the spec, but with Blogger (Google) having rushed out with support for the 0.3 feed spec - and now, apparently, support for gosh knows what intermediate version of the posting API. This doesn't simplify my life :)

Here's a problem I ran into with VisualWorks parcels just now - code that "shouldn't" be there. What do I mean by that? Well, take BottomFeeder as an example application. I ship an image with the various code components pre-loaded (there are now 31 potentially upgradable pieces). During the lifetime of a version, I release updates as new versions of the parcels, which BottomFeeder can download and load (either immediately or at startup).

Now, say we have a parcel that currently contains a method called blogBasicsPassword. In the next version of the parcel, I've renamed that method to blogBasicsBPassword (why? In this case, to affect the ordering of settings in the settings tool). When BottomFeeder gets the update, the new version of the parcel is loaded over the old version - but the old method - which does not exist in the new version - is not deleted. This is why, if you grabbed this afternoon's update you saw the username and password field duplicated in the Posting tool settings.

So what can we do about that? Well, it's a bug so far as I'm concerned, so I've reported it to engineering. In the meantime, this is Smalltalk, so we can do a field repair. Parcels have a variety of actions they can take during loading and unloading - in this case, we need an action to take place when the new version is loaded - a bunch of methods need to be deleted. The code below manages that:

BlogTools.BlogSettingsDomain class 
	quietlyRemoveSelector: #blogBasicsPassword. 
BlogTools.BlogSettingsDomain class 
	quietlyRemoveSelector: #blogBasicsUsername. 
BlogTools.BlogSettingsDomain class 
	quietlyRemoveSelector: #blogBasicsCBlogId. 
BlogTools.BlogSettingsDomain class 
	quietlyRemoveSelector: #blogBasicsBlogId

The #quietlyRemoveSelector: method will fail silently if the method doesn't exist - which is what we want in a runtime application. Why are we sending these messages to the class? In this case, it's because they are class methods.

The end result of all this is a new rev of the parcel that cleans up its own mess. I love Smalltalk :)

Julia Lerman points to a list of deprecated APIs in .NET - i.e., ones that will actually leave the next rev of the CLR. This is something that Sun is utterly unwilling to with Java - either at the language level, or at the VM level. It's why I expect that dynamic languages might someday work well on the CLR, while I will be completely astonished if they ever do on the JVM. Microsoft is willing to break a few eggs to make an omelette, while Sun treats each egg as having the worth of gold. That's why the new stuff in Java gets more and more baroque too...

Update: Julia and Rafe correct my interpretation in the comments - apparently MS is not doing what I thought.

Patrick Logan reports on an interesting code sprint event in Portland this April. Sounds interesting - Python and Smalltalk look like they'll be in use.

I've got the screen cast I mentioned earlier uploaded. There are two files:

• A Windows Media Viewer file (13.2 MB)
• A compressed AVI File (69 MB compressed, 132 MB uncompressed)

A few people asked for access to something other than WMV, so there it is. This week's screencast - creating a local RSS feed using a SAX driver script from within BottomFeeder

I'm in the process of preparing a new screencast - this time I'm showing how you can scrape an HTML page into a local (file url) RSS feed using the scripting capabilities of BottomFeeder. None of that is really Bf specific; if you leave the capabilities in your application, any Smalltalk application can do the same kind of thing. I've got the baseline AVI file done, and am preparing the wmv. I thing I'll compress the AVI and upload that to the server as well, for people who can't easily consume wmv. bear in mind, the uncompressed AVI comes out to 123 MB :)

Freeform Goodness makes a number of good points in this post, but I really like this one:

Once it’s in my browser, I will fold, spindle, and mutilate at will. Tools that let me do this, as long as I’m in control of the Big Red Button, are A-OK with me. If you, as a publisher, can’t handle this, I hear there’s this advanced technology, Big-Ass Clay Tablets Engraved With Flaming Fingers, that will let you make sure that everyone sees your content exactly as you laid it down.

He's spot on here too:

If you’re reading this in an aggregator, you may not have seen the icon on the main blog page that makes it clear that I’m very short, brown, male, with a head much larger than the rest of my body, and living in a cold climate. I’m still trying to figure out why this is important, but it apparently matters to some (koff) leading lights in this field.

Hear that Levy and Jenkins? That noise the two of you hear is my subscription expiring.

It's been a busy morning of Silt updates, along with some matching upgrades to the client blog posting tool. The posting tool menus have been re-organized - they should make more sense now - and the options are enabled/disabled properly based on the posting API in use. The server now deals with trackbacks properly - apparently, I turned those off awhile back by accident. Steve Kelly spotted that, and has been busy fixing a number of other issues as well. This server is getting nicer all the time, thanks to Steve's help.

I still plan to put a screencast up sometime today - it's just looking like it'll be later today :)

It looks like Steven Levy of Newsweek isn't the only one capable of basic misinterpretation when discussing who blogs and who links to who. Dave Winer makes some good points against Chris Nolan, who seems to think that there's a conscious act playing out. Consider this from Chris:

1)This medium was first taken up by techies. Most of them are men. It's not worth going into the statistics on men and women in tech, and the reasons and whyfors. There are more men, that's all you need to know for this conversation.

2)Those men prefer to link and read men like them. As it was in the beginning so shall it ever be. When they wonder where the women bloggers are what they're really saying is "I don’t read any women bloggers."

The point about techies being overwhelmingly male is true enough - you can accept that as a current fact without trying to explain it. Given an overwhelmingly male pool to draw from, is it a huge surprise that so many bloggers (at least the well known ones) are male? Chris' second point simply doesn't work - it's an attempt to create a political explanation for a simple demographics issue. Quick - if I throw 200 people into a room, and 190 of them are male, won't most of the conversations involve men? For that matter, how many of those conversations will be on subjects that the 10 women simply aren't that interested in? And yet, Chris insists on making this into a counting game:

Even though the "blogosphere" has gotten much larger, most of these men are still reading the guys they started out with three years ago., linking to them and talking among themselves. There's talk of broader horizons, but it's pretty much that: Talk. Glenn Reynolds, however, is an exception to this trend. And since he got slapped around last month, Kevin Drum has started to link to more women. Josh Marshall rarely links to women writers. Dave Winer is also stingy.

Chris also falls into the same trap that she accuses the media of falling into - that the "big A-list bloggers" are the only ones worth talking about. She makes fun of this, but her only female example is Wonkette, a woman with all the maturity of a 7th grader who just discovered that talking about sex shocks the parents. Hmm - no mention of people like the Misbehaving.net crew, or Julia Lerman, or Julie Leung - all of whom are far more interesting than Wonkette. That might have something to do with the lack of a potty mouth. For that matter, I suspect that a trip through LiveJournal would turn up a different set of demographics as well. Nolan needs to stop thinking that everything is political, and get out more.

There's some news over on Microsoft Watch about the upcoming IE7 browser. It all sounded pretty good until I got to the last bit:

Partner sources say Microsoft is wavering on the extent to which it plans to support CSS2 with IE 7.0. Developers have been clamoring for Microsoft to update its CSS support to support the latest W3C standards for years. But Microsoft is leaning toward adding some additional CSS2 support to IE 7.0, but not embracing the standard in its entirety, partners say.

So the beatings will continue until web developers just curl up and die, it sounds like.

Tim Bray points to Sun's project coyote, a project to support dynamic languages on the JVM. This is positive - but there's another crucial thing needed - Sun needs to unfreeze the JVM and make it possible to efficiently use a dynamic language on the JVM. Right now, a language like Smalltalk would crawl on the JVM. Tool level support just isn't enough. It's a good first step, and I hope to hear someone make noises about the next one...

Dare Obasanjo attended a syndication talk by Ben Hammersly at ETech - doesn't sound like he came away overwhelmed by Atom. While I don't always agree with Dare, I take the time to read what he writes - it's always well thought out:

At the end of the talk I asked what the story was for versioning both the Atom syndication format and the publishing protocol. Ben floundered somewhat in answering this question but eventually pointed to the version attribute in an Atom feed. I asked how would an application tell from the version attribute if it had encountered a newer but backwards compatible version of the spec or was the intention that clients should only be coded against one version of Atom? His response was that I was 'looking for a technological solution to social problem' and more importantly there was little chance that the Atom specifications would change anyway.

Yeah, right.

During the break, Marc Canter and I talked about the fact that both the Atom syndication format and Atom publishing protocol are simply not rich enough to support existing blogging tools let alone future advances in blogging technologies. For example, in MSN Spaces  we already have data types such as music lists and photo albums which don't fit in the traditional blog entry syndication paradigm that Atom is based upon. More importantly it is unclear how one would even extend to do this in an acceptable way. Similar issues exist with the API. The API already has less functionality existing APIs such as the MetaWeblog API. It is unclear how one would perform the basic act of querying one's blog for a list of categories to populate the drop down list used by a rich client which is a commonly used feature by such tools. Let alone, doing things like managing one's music list or photo album which is what I'd eventually like us to do in MSN Spaces.

Sounds like events moved past the Atom API while it was being put together. Podcasting walked right past it. Their idea seems to be that you should just slap Base64 encoded content into the feed - uh huh - I sure want to download a 10MB screencast (more like 15 after being encoded, but never mind) each time I look at the feed. Looks like the Atom API is worth ignoring as something to actually use...

CNET reports that TiVO and Comcast have reached a deal - looks like the Comcast branded DVR's may soon be TiVO devices:

The companies announced the agreement Tuesday, saying they are working to make TiVo's DVR service and interactive advertising capability available over Comcast's cable network. The first of their co-developed products will be available in mid- to late-2006 and will use the TiVo brand. Financial terms of the deal were not disclosed, but the alliance is long-term and nonexclusive.

I could end up with a hybrid TiVO/ReplayTV thing going. Hmm

I guess the Red Sox didn't go to DisneyWorld after winning the Series last year; now we know what they did instead :)

Now this is classic stuff. With most of these you have to "wait for it" at the end. Attempt 8 is especially good :) Hat tip Troy

Dave Winer points to an anti-AutoLink article by Danny Sullivan, where the lack of awareness continues full throttle:

All-in-all, Butler is just the latest example of the "mess" AutoLink created when it was released, as I wrote earlier. It came out, then we got an AutoLink killing script, a supposed way to kill that script, now a tool some will use to fight back at Google plus heaps of bad PR for Google continuing.

Two years ago, the company pulled the related searches feature that its own AdSense publishers hated within 48 hours. We don't need months more of testing AutoLink for Google to realize it needs to make some significant changes to please publishers and not just the usual noises of always considering feedback. Let's get on with an actual solution, starting with an immediate opt-out.

Wow, you mean Butler would have been impossible without Google AutoLink? Huh. Danny, better send your highlighters to me. There's no telling what misleading damage you could do to books and magazines with them. While you're at it, turn off any pop up blockers you have. You're only causing an escalation of the tools war with it.

Danny wants opt out. Umm, bear in mind that the only pages modified are those on the local screen. To fully disable client controlled markup changes, how far are you willing to go? Browser DRM-like mods that disable useful functionality? Like, say, copying text so that it can be referenced in a blog post? Heck, by copying a small portion of Danny's post and reproducing it (without his permission, I might add) - I've quite possibly run afoul of the scheme he wants.

Here's the thing. Load AutoLink, then press the button and look at the page. Now look at the page on your server. It hasn't changed, has it? Good. Then leave my view of the content alone

If you have been patching the 3.8 release of BottomFeeder with the update stream, you may have had difficulties getting some feeds to process recently. I accidentally let some code that had VW 7.3 dependencies slip into the update stream, and that was causing problems. So - if you've seen such problems, check for updates - and grab the latest revs of these two packages:

• Http-Overrides
• NetResourcesHTTP

No need to restart; you can check the load and continue box. That should solve the problem. Sorry about that!

Our protocol team asked me to put up a survey for them - it's online now. Please let us know what you think is important by taking the survey.

Tim Marman responded to my rants on AutoLink this way:

Copyright exists to protect the owner of the content, not the consumer. It reserves in the holder the exclusive rights to reproduce, distribute, perform and display the work, and to license these exclusive rights to others as the holder sees fit. Just because the end user knows the changes are being made doesn't mean they are authorized to do so.

To which I say this - if that's the case, we better ban scissors, tape, glue, and highlighters. Armed with those, I can modify the way the holder's content is reproduced, displayed, and distributed. God forbid I should markup a book or magazine with a highlighter and then give it to someone else to read - I've just gone against the wishes of the copyright holder! Time to lock up high school and college students, I guess. Oh, wait - the RIAA is already trying that.

Welcome to the world that the anti-AutoLink forces are unwittingly trying to build...

Phil Ringnalda spots something ironic from Google:

It's maybe not quite as bad as when they randomly banned themselves from crawling, but moving your blog every five or six posts, and changing the URL for your feed in between, all without ever saying anything about it, and hiding from search engines, is what you do if you're being stalked, or hiding your journal from the 'rents, not if you are documenting trying to build a platform.

I can verify that trying to find documentation on Google's Atom support is a job for the intrepid. It's as if they had decided that going with Atom was a bad idea, but admitting that anything Dave Winer ever said was right was just too painful for them.

Patrick Logan asks an interesting question about the .NET and JVM competition:

I'm curious whether the interesting race is the emerging-yet-not-fully-realized race to "fully" support dynamic languages on these platforms.

First, one of the two of them would have to recognize that dynamic language support is more than an afterthought.

Who knew that "24" was going to turn into a "first I torture you, and then it's a buddy flick" thing. And for extra fun, CTU is permanently staffed with people with personality problems and conflicts. If it wasn't for the pacing, no one would watch this show at all. It doesn't just jump the shark - it does the hokey pokey over it.

Martin Fowler has some interesting things to say about static and dynamic typing. It's not arant, and it's not an evangelical piece, either. Here's an interesting segment, but I'd suggest that you read all of it:

Another area where static typing is useful is that it allows programming environments to be much more helpful. The revelation here (as in so many things) was IntelliJ. With an IDE like this I really felt the type system was helping me. Even simple things like auto-completion are greatly helped by static types, and leading IDEs can do much more than that.

Despite this, there's still something particularly satisfying about programming in languages like Smalltalk and Ruby - and I think it has a great deal to do with the dynamic typing. Chatting at Camp 4 Coffee with Bruce Eckel we both agreed that one of the most frustrating things about the static/dynamic typing debate is that it's very hard to put into words the advantages of working in a dynamically typed language. Somehow things just seem to flow better when you're programming in that environment, even when I'm doing my Ruby in emacs instead of IntelliJ. (Smalltalk, of course, has both the language and a lovely programming environment.)

Mind you, there are loadable components for VisualWorks that give you auto-completion.

Scoble reminded me of the Newsweek article on blogging, and it seems that I failed to abuse the dumbest part of it:

The top-down mainstream media have to some degree found the will and the means to administer such care. But is there a way to promote diversity online, given the built-in decentralization of the blog world? Jenkins, whose comment started the discussion, says that any approach is fine—except inaction. "You can't wait for it to just happen," he says. Appropriately enough, the best ideas rely on individual choices. MacKinnon is involved in a project called Global Voices, to highlight bloggers from around the world. And at the Harvard conference, Suitt challenged people to each find 10 bloggers who weren't male, white or English-speaking—and link to them. "Don't you think," she says, "that out of 8 million blogs, there could be 50 new voices worth hearing?" Definitely. Now let's see if the blogosphere can self-organize itself to find them.

Huh? I'm supposed to run off an find 10 bloggers who aren't male, white, or English speaking - and link to them? Forget the first two criteria - I really don't care about the ethnicity or gender of an author - it's the content that I care about. Which points out the utter stupidity of Levy and Jenkin's suggestion - if I can't read it, I'm sure not going to link to it. Why? Unless I can understand the content, how do I know what I'm linking to? Do Levy and Jenkins recommend restaurants without having tried the food? Do they recommend movies without having see them? How about TV shows? Unless content is in a form I can read, I can't evaluate it.

For me, that limits things to English, and a small set of Spanish language content (i.e., ones that don't go beyond my now meager (through lack of use) vocabulary. Sure, I could find a blogger speaking Farsi or Japanese with really cool graphics. But unless I have some idea of what they are saying, I'm not going to recommend them. Likewise, if those hypothetical bloggers don't speak English, I have no expectation that they would ever link to me.

I'm reminded again why I'm not renewing my Newsweek subscription - with great thinkers like Levy talking to equally challenged bozos like Jenkins, I'm sure not missing much.

The submission deadlines for OOPSLA 2005 are fast approaching - you can submit tutorials here. Anyone giving a tutorial at StS 2005 should try and give it at OOPSLA.

I love these guys. They think they have a clever argument against Google's Auto-Link; it still hasn't occurred to them that the features is:

• Optional (I have to install it)
• Limited to IE (The toolbar doesn't work elsewhere)
• The Links have to be manually requested
• The Links are clearly marked

Or heck, maybe they do. When you are reduced to ad-homeneim attacks in place of facts, it's a sign that your argument really stinks. So witness their latest attempt - they've decided that all of us with a differing view live in the moron family tree. Hmm, that's clever. I haven't seen an argument that sophisticated since, oh, maybe 7th grade.

Then there are the obvious problems. They put up a QuickTime link without labelling it as such, and couldn't be bothered to have the link actually look like a link. That sure made it easy to start your video - especially when the yellow click to play text (which would be a link in any page that wanted to be obvious) isn't a link.

Then add in the fact that they can't quite figure out how to setup a QuickTime download. If I dare to use Firefox instead of the great and powerful IE, it leaves turds all over my screen. If you're going to make a technology point, it would help if you showed some basic awareness of the technology you use.

And hey - they actually noticed my specific complaint, but didn't understand it. They just want this feature to go away, because the poor, unwashed masses will be baffled by the (dare I say again, clearly marked??) added links. Apparently, that's not good enough.

So... the idea is, any content from the producer is sacrosanct. The end consumer cannot use tools to customize the presentation - even if said tools do not change the content for anyone else. Well then - better outlaw highlighters, because marking up a book changes the emphasis from the implicit desires of the content producer - and any future owner of the book will be misled by it. Better outlaw anything that can create end user customizations - those cut and paste exercises kids do with magazines? It's just wrong - those horrid little children are mucking with the editorial content in ways the producer didn't sanction.

Ooh, their big argument is that 90% of the users don't know how to install or uninstall toolbars? Well heck, then how the heck did you get the toolbar in the first place, you cluestick lacking puzzlewits? Last time I looked, IE didn't come with the Google Toolbar installed. Heck, IE comes from MS, and they have their own competing search service. Are you laboring under some misconception that MS is going to start bundling the Google toolbar? You have to go out of your way to install the toolbar, so - if your argument is correct - 90% of the end users can't install the blasted thing. Which also means that the small percentage who do install it know full well what they are getting. Not to mention that it also means that the small percentage installing it can figure out how to uninstall it. Are you telling me that the small percentage of the cognoscenti who understand installation/deinstallation of optional tools need to be protected from themselves? Quick, take away my fireplace and matches too - I might hurt myself.

Ahh, we find the proposed solution from these guys - an industry standard API whereby content producers can opt in or out of various end user customizations. Hoo boy. We've seen where that kind of crap goes - it's happening with music and video now, with DRM. Want to move the music bits from device A to device B? Heck no, you have to jump through the hoops so kindly provided by the RIAA. Now these guys want to do that to all the written work on the net as well? I can see it now - I want to copy text - no, wait, that producer banned that action, so to do that I have to find my archived copy of a pre-DRM'd browser to copy the text - assuming that BBN hasn't fully gotten its way and changed the base protocols such that they don't work at all.

If content producers can opt out of end user controlled customizations, it really means that content consumers get their fair use rights tossed out the door. That'll make a lot of existing media giants happy as pie, and it'll baffle the bright folks at BBN, who will have no idea how that ever happened. "But we thought we were saving links", they'll cry.

Tell you what, let's set up an out of the way house for Better Bad News and their ilk. We'll send them print outs of content, and we'll make sure they don't have access to any highlighters. That way, they can be happy in their ignorance, and the rest of us can move along.

I implemented support for the mt API this afternoon - that was simple. I then made the mistake of looking at the Atom API. Gah. First up, I went to Blogger, since they supposedly support the Atom API now. Following the crumbs from their site (through a developer blog last updated in 2004, fills me with confidence), I ran across this draft of the spec. That's labelled 0.9, dated December 2003. Hmm, that seems old. So I wandered over (with trepidation) to the Atom Wiki, and found this IETF draft. Well. That's labelled 0.2 (now part of the IETF system), and is dated September 2004. Hmm again

So did Blogger implement the old rev? Who knows? They sure aren't saying - or if they are, they hid the instructions in a sub-basement somewhere. From there, I wandered back to the page explaining the latest draft from the IETF, wherein we learn to be afraid:

The current revision of the AtomAPI is contains quite a few changes from previous revisions. All of those changes have been talked about seperately, but when put together represent a fairly large change to the API. Here is Quick Reference for the AtomAPI which also highlights those changes.

So... do I bother dealing with whatever Blogger (Google) did, knowing full well that the more recent drafts have changed a lot? Do I not worry about that, since a working implementation will trump the fevered dreams of the Atom crowd? Do I just punt for now, figuring that it's mostly irrelevant?

Gads, at this point I hope that Microsoft comes out with an API. I must be desperate...

It needs more testing, but I've added client and server support for the mt (and xml-rpc based API) blog posting api. The Silt server now supports the api points, and the posting tool does as well. I'm sure there will be issues (there always are!), but it wasn't too hard. I suppose I have to seriously look at the Atom stuff soon...

Chris Petrilli explains why web UI's are not the end-all, be-all that so many would like us to believe they are:

I’ve been working on an application, for a while, off and on, and have at various points contemplated what the UI would look like. Now, please understand that I’m not Tog or Donald Norman. Instead, I’m simply a user who on occasion notices good and bad design. Most web-based applications are only “good” if you put them in the context of artificially low expectations and the general crap that permeates the market.

Instead, I’ve decided that rich user interaction is critical. Not the kind of stuff that uses XmlHttpRequest, but the kind of thing that uses drag and drop, contextual menus, inspectors, and other generally accepted practices in the UI world. For example, if a user is looking at a set of data, and has a chart open, why should they not be allowed to select and drag that data onto the chart and have the chart adapt to their new data? Perhaps, even it should even ask the user whether to integrate the data, or build a comparable graph with the new data.

Web interfaces have their uses, but there are also areas where they are a huge, huge leap backwards.

The screencast I did last week was a success - I've had a lot of hits on that link so far. Since people seem to be interested, I think I'm going to start doing this on a regular basis - I'll have it again on Wednesday.

The media hasn't been happy with the ongoing fact checking that the Blogsphere provides, and they've been casting about for something negative to hang on it. Here's the latest bunch of crap hurled against the wall by Newsweek - they wonder why blogging is dominated by white males:

At a recent Harvard conference on bloggers and the media, the most pungent statement came from cyberspace. Rebecca MacKinnon, writing about the conference as it happened, got a response on the "comments" space of her blog from someone concerned that if the voices of bloggers overwhelm those of traditional media, "we will throw out some of the best... journalism of the 21st century." The comment was from Keith Jenkins, an African-American blogger who is also an editor at The Washington Post Magazine [a sister publication of NEWSWEEK]. "It has taken 'mainstream media' a very long time to get to [the] point of inclusion," Jenkins wrote. "My fear is that the overwhelmingly white and male American blogosphere... will return us to a day where the dialogue about issues was a predominantly white-only one."

Oh please. One of the joys of the net is the "on the internet, no one know if you're a dog" phenomenon. This all seems to stem from the Technorati top 100 blogs list. You know what I notice about that list? It seems to be a pretty eclectic mix of opinion, across politics, technology, media - and a bunch of other topics. Compare that to the MSM, where I can get homogeneous group think without regard to the outlet I pick. I neither know nor care whether the top 100 listed there are black, white, asian (what have you). What I'm happy about is the intellectual diversity shown.

That's sadly lacking in, say, Newsweek - where I can find a nicely homogenized set of ideas every week. Here's a tip for Jenkins - when you find some people who don't all subscribe to the same worldview, maybe you'll stop bleeding readers. In the meantime, you can watch me not renew my subscription to Newsweek - I no longer need to read it in order to know what the content will be each week. That's kind of a problem.

Ever wonder how Scoble seems to see all the negative MS press so fast? He leverages RSS via various search engines. I do the same thing, and I'll be talking about how and why this is useful at StS 2005. In product management and marketing terms, what you don't know can kill you.

Apparently, Microsoft said the following at VSLive (no link - can't get SDTimes to resolve this morning):

Developers should build server-side applications that can operate through a browser using ASP.NET 2.0 - but if the server detects that the client is a Windows box with the .NET framework installed, developers should push down a Windows Forms client application instead

Sure we should. Let's see how many ways I can multiply the number of inbound support calls, hmm? Never mind those of us using Apache and Linux - and even people using a Windows server - the MS theory would double your web development costs. Apparently, resources are so cheap in Redmond that they forget how the rest of us live.

This ComputerWorld piece doesn't fill me with warm fuzzies - Wells Fargo (and other banks) have deployed web-enabled ATM's:

Wells Fargo & Co. last week said it has completed a five-year project to Web-enable its 6,200 ATMs in 23 states. The Windows-based infrastructure is designed to allow Wells Fargo to update its entire network remotely when it needs to do things like add new languages and enable customers to make envelope-free deposits.

Sure, Windows can be secured (although in practice it rarely is). The web enablement is what worries me. I'm sure that the banks will figure out how to secure them - it's just a matter of how bad the damage gets to be while they go through their learning curve.

I saw this gem on the AIM service this morning in eWeek:

The revamped terms of service, which apply only to users who downloaded the free AIM software on or after Feb. 5, 2004, gives AOL the right to "reproduce, display, perform, distribute, adapt and promote" all content distributed across the chat network by users.

"You waive any right to privacy. You waive any right to inspect or approve uses of the content or to be compensated for any such uses," according to the AIM terms-of-service.

Although the user will retain ownership of the content passed through the AIM network, the terms give AOL ownership of "all right, title and interest in any compilation, collective work or other derivative work created by AOL using or incorporating this [user] content.

I'm no expert, but doesn't that fly in the face of copyright? If I write it, I own the rights, so far as I understand it.

If you browse the blogs here with IE, you'll have noticed that the category is no longer showing for each post. This is some kind of oddball css issue with IE - it's not a problem with any other browser I've tested (in other words, with standards compliant ones). We will try to address this, since a fix on this end will certainly come before the IE team can spell css, much less implement it properly...

The writers of Battlestar Galactica understand dramatic tension better than a lot of the shows I watch. I just watched Friday's episode on the Replay, and I spent the entire show on the edge of my seat. And then the end conversation between Gaius and the Cylon in his head was just... freaky. I'm getting very, very curious as to where this religious angle is going. The only sure thing is that Gaius is getting further and further "out there".

I'm not even sure how to classify this:

Apparently, the Surf Junky popup ads can be blocked using Mozilla Firefox and because of this Surf Junky now block this browser. However, with my User Agent Switcher extension you can fool them into thinking that you are browsing with Internet Explorer. Combine this with an extension that reloads the page on a regular basis and you have a pretty efficient way of abusing the Surf Junky system.

So, there's a "service" that people with too much free time sign up for (at $0.75 per hour, no less). The service works by throwing pop-ups at you. Using Firefox is verboten, since it blocks pop-ups - but if you use the useful tool, you can scam the service.

It's all way too bizarre...