We have a keynote set for Smalltalk Solutions 2007, April 30 - May2:
The CTO of InStream Financial, Thomas Gagné will be one of the keynote speakers at the Smalltalk Solutions Conference that will take place in Toronto, Canada on April 30 through May 2.
It's already shaping up to be a great conference - see you there!
Rob Fahrni has bad news about the fruit (especially citrus) crop from California this year:
According to news reports this morning we've had 19 consecutive days below the 32 degree mark. Not good for fruit that matures in the winter months.
Looks like fresh fruit is going to get more expensive.
Wired has a round up on reactions to MS considering having staffers correct inaccuracies in Wikipedia pages:
When a blogger revealed this week that Microsoft Corp. wanted to pay him to fix purported inaccuracies in technical articles on Wikipedia, the software company endured online slams and a rebuke from the Web encyclopedia's founder for behaving unethically.
For all the yelling, I don't know that there's a real problem - an arbitrary MS staffer is no more likely to be biased than an arbitrary Wikipedia author - and it's entirely possible to get wildly negative content from MS haters. Heck - imagine if I were writing the entry on PVP-OPM, given my well documented dislike of Vista DRM.
I can't really fault MS for wanting to "police" the badness. On the other hand, it's easy to see where it could be considered astro-turfing (and MS has been accused of that before). To be honest, I'm not sure what the right answer is here.
Technorati Tags: wikipedia
This is a head scratcher to me - in Florida, there's a proposal to make information on lawyer websites harder to find:
Now, some members of the Board of Governors want to make websites subject to Rule 4-7.2. As reported in The Florida Bar News, “Rule 4-7.2 bars the use of testimonials, language describing the quality of a lawyer or law firm’s services, references to past successes, or language promising results, among others.” However, this information is available to legal consumers when they ask for it.
Chobee Ebbets, chair of the special Bar committee proposing these changes, said lawyers could “still post such information, but it would have to be in a restricted area of their Web sites accessed only by the viewer taking an affirmative action indicating he or she desired further information.” Florida Bar News, January 1, 2007.
Hmm. I dislike the "ambulance chaser" ads on TV as much as the next guy, but websites are not like TV or radio. If I head to a legal site (blog or otherwise), I'm actively seeking information - what benefit does it do me, as an end consumer, to make it harder for me to find what I'm actually looking for? By putting that stuff behind a password wall (or the equivalent), it will disappear from search engines, which will only make my life harder.
I'm not sure what "problem" this is supposed to solve, but it seems awfully silly to me.
Well, this is interesting - Norway is calling Apple's FairPlay DRM (operates only with iTunes and iPods) illegal in Norway - and France and Germany look like they'll join the party:
Apple's digital rights management lock on its iPod device and iTunes software is illegal, the Consumer Ombudsman in Norway has ruled. The blow follows the news that consumer groups in Germany and France are joining Norway's action against Apple.
So... Apple could license FairPlay, or they could just abandon DRM. Apple likes the lockin provided by FairPlay, but they've never been a big fan of licensing their software, either. Makes me wonder - will they decide that the iPod (and iTunes) are a big enough brand now that they can just kiss off DRM, and tell the studios that they had no choice? This will be an interesting one to watch.
More than a fifth (22 per cent) of Windows installations failed tests on their authenticity, according to figures from tests conducted using Microsoft's controversial Windows Genuine Advantage (WGA) tool.
The BSA (Business Software Alliance) has chimed in predictably:
Figures from the WGA audit are lower than those from the Business Software Alliance, which reckons 35 per cent of business software is counterfeit.
Left unasked by the author of the story, or by MS, or by the BSA: How many false positives live in that 1/5th? I'm sure that many of the checks are from systems that have not been installed in accordance with MS licensing - but the real question is, how many have been, and failed anyway?
In a discussion about document formats, Tim Bray says:
Having said that, I still think OOXML is totally bogus; ECMA shouldn’t have gone near it and neither should ISO. The world does not need two ways to say “This paragraph is in 12-point Arial with 1.2em leading and ragged-right justification”. As I argued in 2005 , if you want to capture MS-Office-specific semantics (not a bad thing in principle) the right way to do it is a namespaced layer on top of ODF.
Hmm - should we apply the theory that "the world doesn't need two ways" to products, too? We had Smalltalk; what's the point of Java? We have Java; what's the point of C#? We have MS Office; what's the point of Open Office (and so on). Somehow, I doubt that Bray would agree with any of those assertions (and for good reasons).
And yet there he is, arguing for the "one true format" (which, coincidentally, happens to be the one Sun backs). Like other products, document formats can be good, bad, or indifferent. I haven't looked at either in detail, but I see little harm in letting people who actually care look at them and make an informed decision.
Tim does bring up the conflict of interest inherent in his point later in the post, but still...
Michael isn't sure that servlets are the best way to deal with web requests, so he's exploring pragmas (or tags) as a way of approaching the problem. Here's a snippet of his post; I encourage you to read the whole thing:
So I got thinking about pragmas again - or as Travis wants to call them: "tags". I thought, what if I could just make one class and put all these adhoc response methods in there using tags. So I've come up with a couple of tags you can put in to a subclass of HttpServerTags.TagServer.static_content <get: #('test.txt')> <string: 'text/plain'> ^'Hello World'
In this simple example we say that we respond to a GET request with path of /test.txt and our content type is text/plain. When we request the page, we get back a string of Hello World. Pretty simple and to the point.
The other examples are far more interesting; head on over and check it out
Technorati Tags: web
Beside the new SAP integration the concept of explorative modling using Smalltalk as the modeling language can be a door opener for new Smalltalk projects. While it is only used as a modeling tool customers don't have to be afraid of a new unknown technology. It is strongly seperated. But when he recognizes how fast developing the right thing with Smalltalk is, he'll possibly takes a deeper look. And, when SAP uses it, it can't be so bad. *smile*
If you can see this, you aren't affected (or have the DNS information cached). At present, there's some kind of lookup/router issue,
either within our IT or in the outer ISP that serves us (and I'd guess the latter). affecting Europeans getting to the site (maybe other locations; I just don't know). So this is really more of an explanation after the fact - neither this site, the main CST site, or cincom.com are (or have been) offline - until this gets fixed, they've just been made invisible :/
Nick Carr relates some advice from Dave Sifry, which was directed at aspiring politicians. It fits for anyone though. My stupid college hijinks are invisible unless I decide to tell the stories; the generation passing through school now is leaving a much bigger crumb trail:
According to the FT, Sifry cautioned the attendees that "some Web 2.0 tools could backfire on the next generation of Davos delegates, warning that the Supreme Court justices and presidential candidates of 20-30 years time could be embarrassed by their juvenile MySpace pages and drunken photos on Facebook."
"Googling" job aspirants is fairly standard now - it will be interesting to watch the reactions of interviewing staff as they turn up records of lost weekends (and worse). Will they be forgiving, with an "everyone did that" attitude, or are people actually building up that "permanent record" that our high school teachers warned us about?
Technorati Tags: personal
I've been offered the opportunity to work at Cincom on VisualWorks' VM, and I have accepted it.
I have worked with Cincom's clients, and I know first hand how much trust goes into what I will work with. I am looking forward to take on this unique challenge --- both because of the fun and because of the responsibilities. It is a golden opportunity that I do not want to miss.
We're happy to have him, and we're still on the lookout for more.
Technorati Tags: smalltalk
Engadget notes that there are rumors of un-DRM'd music from major labels out there:
Details are admittedly scant on this one, and we must remember that these disclosures are coming out of France, but nevertheless, word on the street is that "at least one of the four major record companies could move toward the sale of unrestricted digital files in the MP3 format within the coming months." Reportedly, executives of several anonymous technology companies that are meeting at Midem are pondering the move to unrestricted digital music downloads, which just years ago wasn't even a remote possibility.
I wonder - right now, the labels are making most of their digital revenue via iTunes, and Apple is setting the terms there - in ways the labels have not been happy with:
- Support for one device only (the iPod)
- Fixed prices for individual songs
They haven't been able to find a way out of that, and - even though they collect most of the $0.99 themselves (supposedly, they get around $0.70) - they want more. We may actually be at a point where, having tried everything else first, they'll end up "doing the right thing" as a last resort.
I like this comment from Shaman:
Methodology is killing software development. Pre-occupations with standards and enterprise architecture are taking away the creativity from software. Instead of being the software developer’s best friend, architecture oversight groups are assuming the role of armchair dictators and throwing a spanner in the software development life cycle. When was the last time an enterprise software architect actually coded the recommended reference architecture to see if it was even feasible? And how often is the software architecture updated to keep it relevant with business and technology advances?
Now, I do understand the point behind standards, and why businesses should pay (some) attention to them. As Shaman notes though, too many outfits go overboard, and empower standards groups beyond all reason. The first question you need to ask yourself is this:
Is this helping the customer (internal or external) get their solution sooner?
If the answer is no, then you probably have a problem on your hands.
Technorati Tags: standards
The best explanations of Smalltalk's productivity come from people who have used Smalltalk and mainstream languages in anger:
I just finished writing some blog software using VisualWorks Smalltalk and Seaside. It was unbelievably quick and easy - I shouldn't be too surprised, I *was* a Smalltalker for 5 years, but I last used it on a regular basis over 5 years ago. I'd forgotten how quick things can go. :) I can change some code and immediately hit refresh in the browser to see the change - and if there's a problem, I can click on debug in the web browser and immediately have a debugger up in Smalltalk to figure out the problem.
Seaside is incredible too, I picked it up very quickly reading 3 pages of overview, playing with demos, and then on to coding. There's plenty more to learn, I'm sure, but my web app is up and running with suprisingly little code. Learning/using Java frameworks was never this easy.
I've played a little with Seaside, and it does look amazing.
This story about insurance software brought something to mind - the loss of employee level flexibility due to the type of software being used:
As the recent father of twin babies, Philip B. was relieved to learn that his employer's benefit provider, Sun Life Canada, made the insurance process really simple. Adding the little ones on the plan required no more than a phone call to provide birth dates, names, and that sort of thing. All seemed so easy, until the customer service rep realized what Philip was trying to do: "I'm sorry sir, but we need a different birth date for each of your kids."
"Uhh, er," Philip stuttered, rather puzzled, "they're twins? They were both born on the seventh of May, so they actually do have the same birth date."
"Oh yes, I understand," she said, "but our system cannot handle two people with the same last name born in the same month of the same year on the same plan."
I've seen this kind of thing elsewhere. You walk into a store, and - need to do some kind of one-off thing with an employee - an exchange, perhaps, where you don't have exactly what the software expects. A few years ago, employees could bend the rules a little if things seemed to be otherwise ok. Now - they are ruled by the software. If the software doesn't handle the particular exception that came up, then you as the customer are in for a trip to never-never land. I understand the point of these new systems, but I wonder - has anyone sat down and asked the line employees about the impact on customer satisfaction?
Travis has posted on his SUnitToo tools - he bugged me about it today in the IRC channel when I mentioned that I was adding some tests to my Blog code while I refactored. So, I went ahead and loaded his stuff - it's in the public Store, but (and this stuff is being changed all the time) I had to load the penultimate version:
When prompted, load this version of the pre-req:
So what does that give you? Well, take a look at the browser view with two new tests that haven't been run yet:
The empty beaker indicates that no tests have run, and the tests have "?" next to them. See the changes when I run the tests?
I've chopped the bar at the bottom of the browser, but you get more indications now. So - let me break one of the tests:
That's pretty nice. There are other goodies as well - if you hit "debug", you'll get right to the failing assertion (where the old SUnit lands inside the SUnit code itself). All in all, if you use SUnit in VW or OS8, you should move to SUnitToo. Ohh - one more thing. To do that switch, check the "View" menu and toggle which testing tools to use. You can move back and forth at will, and the new stuff is API compatible with the old.
Now this may be the perfect food - the caffeinated donut. With this, you can give up juggling the mocha and the chocolate creme in your rental car :)
It's time for the weekly look at the logs - BottomFeeder downloads proceeded at the respectable clip of 244/day. The details:
On to the HTML page accesses, by tool:
|Tool||Percentage of Accesses|
Opera continues to creep up there. Finally, the Syndication accesses by tool:
|Tool||Percentage of Accesses|
|Net News Wire||7.2%|
|Google Feed Fetcher||5.1%|
|RSS 2 Email||1%|
iTunes is new to that list. And the Planet Smalltalk numbers still look awfully poll heavy to me...
|Smalltalk Solutions 2007 is coming up sooner than you think - April 30-May 2 in Toronto, Canada at it360. Head on over here to register (more properly, pre-register - registration isn't quite live yet) for the conference.|
We'll have a lot of people at the conference, and I'll be live-blogging - and I think I'll try to do some ad-hoc podcasts from the show with my digital recorder - we'll see how that works out. See you there!
The patent claims the invention of “a facility” (the object test bench) that “receives an instantiated object, displays the instantiated object visually, receives a command from a developer relating to the instantiated object, and provides a result corresponding to the received command. As an example, the facility invokes a method provided by the instantiated object or retrieves a value of a property of the instantiated object.”
Michael at least mentions that this feature of his tool (BlueJ) was "inspired" by Smalltalk - but hey - I've got a news flash for him - the above is as good a "pocket definition" of a Smalltalk inspector as anything I've read (sure, most ST inspectors are not graphical - however, I've seen plenty of extensions going back to the early 90's that are). So let me take a moment to yell at Dan Fernandez, one of the MS people involved in this patent: Have you so much as looked at Smalltalk, ever? Maybe if you had, your team wouldn't be applying for dumb patents for ideas they didn't invent. Kind of like the MS morons I mentioned here, who apparently filed their patents relating to RSS without ever looking at various online readers, like BlogLines.
I guess MS simply believes that tons of money and scary lawyers makes them right. Sadly, they might be right - but not terribly ethical.
Scoble complains that he's not getting the audience he thinks he should since he headed to PodTech:
Another point? That I’ve become less interesting since leaving Microsoft. Well, I think that’s because Engadget isn’t watching ScobleShow.com . Including an interview with the Zune team that wasn’t linked to by Engadget or Gizmodo, either.
Over there, in just four months, I’ve posted more than 120 videos , gotten more than 70 interviews with Silicon Valley CEOs, and had some real interesting ones with Retrevo and gang at CES, among others. But, nah, that’s not as interesting as seeing inside Microsoft, is it?
The reality is, it's not as interesting at a business level. Why? Well, having Microsoft open up that way draws an audience, and - simply by dint of their size - they are going to get attention. PodTech is a small entity, so it's going to take more work just to get noticed. That's just the way it is.
Second though, I'll say something that I've brought up to Scoble before - video is not the best format for an awful lot of the stuff he's producing. Take most of these videos - to consume those, I have to be at my PC, with a screen. Yes, I could have a portable device like a Zune or a video iPod - but the point is, the video demands attention. When do an awful lot of people consume this kind of thing?
- While Driving
- While Exercising
In other words, at times when they can listen, but not watch. I get the distinct impression that an awful lot of video blogs use a camera simply because they can - not because of any value that the image adds. It's not that Scoble should eliminate the video; but he should add an audio-only component. I bet his audience would grow as a result.
- implementing = without implementing hash
- poor hash functions causing many collisions in Sets and Dictionaries
- multiple initialization with super new initialize
- failing to call super initialize
- improper copying (shallow copy vs. deep copy vs. partial copy)
- condition ifTrue: [^true] ifFalse: [^false]
- at:put: returns the element, not the collection
- failing to use 'yourself'
- not using streams and using collections instead which memory bounds you and creates work for the garbage collector
- collections created too small
- Not understanding the Process Model (specific to Cincom Smalltalk)
Today, I'll call out the three *cough* inventors *cough* whose names are on it: Guatam Goenka, Partho Das, and Umesh Unnikrishnan. Here's a question - do any of you have any pride, or do you enjoy taking credit for work that was done well over a decade ago by other people?
Cincom's EMEA team (and Cincom's partner, the Heeg company) were at OOP 2007 last week - here are some photos that Georg Heeg sent me. below, that's Monika Laurent of Cincom talking to Ralk Ehret of SAP (who spoke at the users conference) in the center:
Same shot, slightly different perspective. That's Yvonne Schickel and Helge Nowak at the far left corner (both of Cincom):
Finally, the same shot again, from further back - off to the right in all three shots in Andreas Toenne of the Heeg company.
Vassili Bykov has posted a nice modification to the Refactoring BRowser - I think I'll keep this one loaded:
VB-Sketch-DontModeMeIn is an attempt to fix that within the constraints of the existing RB and the UI framework.
Load it. Close any open browsers. Open a new one. Look at a method source. Start changing it. Without accepting, click on another method. Or another class. Or search for a class. The button with a red circle and white dots at the bottom of the browser is the link back to the postponed edit left behind. The same works for class comments or any other tools. For example, while writing a class comment you can have a peek at class methods to remember what that instance creation method was called.
There's a press release out on the upcoming Smalltalk Solutions 2007 conference:
January 26, 2007…ITWorld Expo Canada and The Smalltalk Industry Council (STIC) are pleased to announce that Smalltalk Solutions 2007 will be hosted for the second consecutive year by LinuxWorld Canada and IT360 (www.it360.ca) in Toronto, April 30 May 2, 2007 at the Metro Toronto Convention Centre.
See you there!
Technorati Tags: sts2007
Remember the WKP lawsuit filed against a Maine blogger who was being persistent about digging into the way money was being spent by the Maine tourism office? Well, he's kept digging over the last few months, and found out that the lawsuit wasn't solely from WKP, the PR agency - turns out that the director of the government office he was investigating initiated the whole thing:
From: Lewis, Dann [mailto:Dann.Lewis@maine.gov]
Sent: Tuesday, February 28, 2006 10:52 AM
To: firstname.lastname@example.org; email@example.com;
Cc: Lyons, Steve
Subject: Far enough?
Tom- et al: Has this gone far enough? This apparently is spreading
into the blogosphere, which may cause permanent damage to all concerned.
1. How can we shut this guy up?
2.How can we mount a counter offensive to get the facts out?
From: Lewis, Dann [mailto:Dann.Lewis@maine.gov]
Sent: Tuesday, February 28, 2006 11:12 AM
To: ‘Nancy Marshall’; firstname.lastname@example.org; email@example.com; firstname.lastname@example.org
Cc: Lyons, Steve; Charlene Williams
Subject: RE: Far enough?
I don’t think anything short of a restraining order & lawsuit will
stop this train wreck.
Will be talking to WKP later today and then try to get
Well, well, well. How has this Lewis guy been paid back for this fine service to the people of Maine? The Boston Globe reports that he's been canned:
Maine tourism director given the boot after nearly 12 years
Serves him right. Agree or disagree with Lance Dutson, he's providing an independent viewpoint into an area of government that doesn't get a lot of lights shone on it. We need more guys like Lance, not fewer.
Oh, and a snide aside - the WKP website still sucks.
On Friday, an alert reader emailed me about a new article by Michael Kölling, the creator of BlueJ, about a patent issued by Microsoft for features in Object Test Bench that are comparable to BlueJ's Object Bench. I'll post the full "anatomy of a firedrill" some time later, but for now we can officially say that the patent application was a mistake and one that should not have happened. To fix this, Microsoft will be removing the patent application in question. Our sincere apologies to Michael Kölling and the BlueJ community.
Kudos to them for responding quickly - this blogosphere thing seems to have some impact on things :)
Ed Felton explains how the music labels are going to have to eat a lot of crow when they start selling un-DRM'd mp3 files:
Why did the Times (and many commentators) mistake MP3 for “unrestricted”? Because the industry has created a conventional wisdom that (1) MP3 = lawless copying, (2) copyright is a dead letter unless backed by DRM, and (3) DRM successfully reduces copying. If you believe these things, then the fact that copyright still applies to MP3s is not even worth mentioning.
The industry will find these views particularly inconvenient when it is ready to sell MP3s. Having long argued that customers can’t be trusted with MP3s, the industry will have to ask the same customers to use MP3s responsibly. Having argued that DRM is necessary to its business -- to the point of asking Congress for DRM mandates — it will now have to ask artists and investors to accept DRM-free sales.
Here's what I wonder: Will David Geffen's head explode when this happens?
Technorati Tags: music
I've posted on the crap-ware that is Vista's DRM before; now, the day of the launch (yawn), comes news that PVP-OPM has been hacked. Meaning: it's still a pain in the neck for law abiding consumers using the OS, and no problem at all for pirates wanting to bypass it.
I accidentally walked myself into an interesting corner recently. I noticed that every version of VisualWorks I use on my laptop (running Windows XP) thought its home directory was C:\vw7.4.1. I spent time searching the registry, but that didn't turn up anything helpful. Finally, Bob Westergaard, one of our engineers, reminded me that I might be picking up an environment variable. Sure enough: when I opened a DOS prompt and typed "Set", I saw the setting.
I had completely forgotten how or when I had set that, but Bob pointed me to the System properties tool in Windows. That's where things got confusing. Here's what I saw when I opened that up and checked the Environment:
Notice the lack of any setting for VISUALWORKS. This gave me pause, but hey - this is Windows, weird crap happens. So I set the variable, hit "Ok". came back in, cleared the variable, hit "Ok". Problem solved, with some Windows oddness tossed in.
On today's Smalltalk Daily, we take a look at Travis' SUnitToo(ls) package, which you can load from the public store repository. It adds some nice things to testing in Cincom Smalltalk, giving you more and better visual cues. I should note that the package has some additional features beyond what I covered in the screencast; check the package comment for details.
So today's the day. After 5+ years, Windows Vista launches and it seems that there barely is a whimper. In really looking the last two days, there seems to be no real online and blogger buzz, and no real excitement. Where was that vaunted Billion dollar ad campain? Why were there no TV ads? Anything? In other words, we're missing the "wow." It wasn't supposed to like this. What we saw in 2003 was exciting. Robert Scoble was recruiting his Team 99, there was a lot of excitement and many of us in the Microsoft community were truly excited. We were told it was going to be the biggest launch events in history, they were going to spend billions, it would be everywhere, people would be lining up just like Windows 95. What we got in Vista was so much less than in Longhorn 2003.
Later on he explains that he likes Vista, and sees the upgrade as a "no brainer" (Given the obnoxious DRM in Vista, I beg to differ...). Anyway - this was certainly no marketing coup. Vista was years late, and is generating zero buzz. I still want a Macbook Pro, and have less than no interest in Vista - which is not the kind of reaction MS was hoping for.
Technorati Tags: Vista
Dave Winer must be the only guy on the planet with a newsreader that can't handle Atom:
I was going to post a note saying that it would be better if you redirected to the new feed, but then I saw that your new feed isn't RSS, to which I ask -- why??
Do you want to lose subscribers?
Because that's what happened. I can't read your feed anymore Ted. I'll survive, but I will miss your posts.
I've never been a huge fan of Atom (although, as I posted awhile back, I fully understand why it exists) - but it's not as if it's hard to find a tool that handles RSS and Atom.
I'm posting some open positions that Precision has available - contact them (see below) for details:
Precision Systems currently has Smalltalk positions open all across North America, from Los Angeles to New York and everywhere in between! Please contact Vicki Ross if you’re interested in any of the following positions:
Northern New Jersey – multiple positions
Senior Smalltalk Developer, permanent, 6 month contract-to-hire and 12+ month contract
New York, NY – multiple positions
Smalltalk Developer, Smalltalk Team Lead, Smalltalk/Java Developer, contract and permanent
Smalltalk Developer, Permanent
Smalltalk Developer, permanent or contract-to-hire
Northern California – 2 positions
Smalltalk ObjectStudio Developer, contract: 6 months; 25% telecommute possible
Smalltalk ControlWORKS Developer, contract: 2 to 6 months
Software Engineering Manager, permanent
Texas – 2 positions
Smalltalk Developer, contract or 6 month contract-to-hire
.Net Developer, Smalltalk a plus, permanent
Milwaukee, WI – 2 positions
Senior Smalltalk Developer, Junior Programmer/Analyst, permanent
Don’t forget to inform your co-workers and friends; for any new and successful referral to Precision we will pay you $1,000!
Jon Udell tries to hook up Outlook calendars and Google calendars - and demonstrates the issues that non-propeller heads are going to run into. The interesting thing is, the issues aren't really due to MS or Google actively playing walled garden games; both support iCal, for instance. It's just that neither one really concentrates on interop, so the steps just aren't that easy. The post may end up being highly valuable for me - I don't use Outlook, but "everyone" else at Cincom does.
On today's Smalltalk Daily, I walk through a simple server I created in Cincom Smalltalk quite awhile ago as an example. Today, I just walked through the basics of how it's set up, and how it handles requests. Tomorrow, we'll take a look at process handling in the context of this little server.
Technorati Tags: smalltalk
I just watched last week's "24" on the cable box DVR, and I really do have to classify it as a guilty pleasure. The plotlines are fairly absurd, the plot twists make no sense - but it moves relentlessly. Even with the silly stuff, it's entertaining.
Runar Jordahl read the paper I pointed to this morning, and drew the same conclusion as the author of that paper did:
As discussed here, native threads are based on shared state. In the discussion, threads are (correctly) defined as “shared-state concurrency”, while processes are “threads without shared state”. The problem with threads is that the shared memory is modified non-deterministic. Since threads run in parallel and each thread might be scheduled differently per execution, every execution of the program tends to produce different intermediate results. This means the danger of deadlock or other side effects are not easily detected through testing. Reproducing a problem can be impossible. Contrast this with the green thread model used in most Smalltalk images: With green threads the program runs deterministic.
Running multiple Smalltalk images basically means running multiple processes. The processes will not share data directly in memory. (VisualWorks images can share static data, but that is another story. Of course, this cannot lead to the discussed problems.) An implementation using a divide and conquer algorithm with multiple images can also easily be designed to run deterministic. Testing, debugging, and understanding such a setup will be a lot easier than a (native) thread implementation.
I've been explaining this to people who want multi-threaded Cincom Smalltalk VM's for a long, long time now - the end result of that wish would be a less stable VM that had subtle bugs. A better answer: use multiple images and interprocess communication (such as Opentalk) to get the job done.
I finally finished "The End of the Old Order" - it takes a long time to read a 600+ page book when you do it in 10-20 page snippets at bedtime. The most interesting part was the last chapter, which covered the peace negotiations of late 1805. The allied coalition had Napolean in a much worse position than they knew; but Francis (Austria) was utterly demoralized by Austerlitz, as was Alexander (Russia). Frederick William (Prussia) got taken out by his faithless envoy, Haugwitz. It's a fascinating look at the difference between reality and perception, and I look forward to Kagan's coverage of the rest of Napoleonic era.
That still leaves me with a huge stack of books though :) Presently, I'm reading this one: "Cathedral, Forge, and Waterwheel", which covers medieval technology. A lot more happened during that era than a lot of people realize, and this book makes that point well. In the meantime, I've still got "Ghengis Khan" and "Flags of our Fathers" queued up.
But wait: there's more! I keep getting books as holiday and birthday gifts (never mind the gift cards I have in my wallet), so there's a bunch of good stuff in the pile: "Pirate Coast: Thomas Jefferson, the First Marines, and the Secret Mission of 1805". That's a topic I'm very interested in reading about; there are echos of that one today.
I've also got the end of the Stuart era on tap, with "Ungrateful Daughters". That era of English history is fascinating - I look forward to this one. Over to the American West with "Empire Express", which tells the tale of the trans-continental railroad. Then there's Sean Wilentz' "The Rise of American Democracy: Jefferson to Lincoln". Still in US history, there's "Gentleman Revolutionary", by Richard Brookhiser. That's about Gouverneur Morris, one of the lesser known founding fathers.
Whew - that's a lot of reading - I'd best get started :)
Technorati Tags: history