Smalltalk Tidbits, Industry Rants

Gilad Bracha on extending system code:

One way of achieving this would be to actually go and add the necessary methods to the class Symbol, so that symbols could behave like parsers. I know otherwise intelligent people who are prepared to argue for this approach. As I said, Smalltalkers would call these additions extension methods, but I find the more informal term monkey patching conveys a better intuition.

Typically, one wants to deliver a set of such methods as a unit, to be installed when a certain class or library gets loaded. So these changes are often provided as a patch that is applied dynamically. Not a problem in Smalltalk or Ruby or Python (though I gathered from the Pythoners in Krakow that they, to their credit, frown on the practice).

Apparently, there is a need to explain why monkey patching is a really bad idea. For starters, the methods in one monkey’s patch might conflict with those in some other monkey’s patch. In our example, the sequencing operator for parsers conflicts with that for symbols.

Hmm. First, version control systems exist partly for that reason - to let developers look at exactly those kinds of conflicts. Second, would Bracha rather see developers in Smalltalk (et. al.) use the same kind of "support class" nonsense that Java developers have been locked into for years?

Sure, you can get in trouble with conflicting extensions - in BottomFeeder, I ran into exactly that problem when I started using Antony Blakey's improved Look and Feel policy. Here's the thing though: I spoke to Antony, and, after we talked a bit, we dealt with the problem. So let me see: my solution involved more and better communication between developers. Bracha's involves "just say no".

Call me crazy, but I think my approach seems far more practical in the real world. At the end of the day, his approach comes from the school of thought I like to call "the library developer is god".

Hat tip Patrick Logan

Technorati Tags: monkey patching, smalltalk

Now that the latest Cincom Smalltalk is out, I've released the latest BottomFeeder - version 4.5.

What's new? The upgrade to VW 7.6, and this - you can share subscriptions for podcasts with iTunes more easily now (Windows and Mac). I'll be adding more functionality in that direction over time.

Technorati Tags: smalltalk, RSS, Atom

Now this is a neat story - the light put out by this plasma bulb is truly impressive. I don't like compact flourescents, but I could definitely deal with this technology

Technorati Tags: light, light bulbs

Andres Valloud has started posting video from Smalltalks 2007 to YouTube.

As promised, here's the video for the "Is Software Practice Advancing?" panel from SPA 2008. I've got the video available three ways:

• Higher quality mp4 (740 MB)
• Medium quality mp4 (150 MB)
• Medium quality wmv (170 MB)

This is the same content I posted in audio only form earlier, so you should see that post for the notes.

Technorati Tags: software practice, smalltalk, software development, spa2008

Two things struck me reading this ancient (1995) piece from Newsweek:

• How many things he got so, so wrong (newspapers, online shopping)
• How common the same kind of complaint is today

Hat tip Michael for the link

This podcast is taken from the panel discussion at SPA2008: "Is Software Practice Advancing?" The panel was moderated by John Daniels, and started with a small group. After about 20 minutes, they asked volunteers from the audience to join by submitting names to a random pick - Peter Deutsch got picked first, and I followed a few minutes later. I'll also have video from this panel later on; I have to get it saved down to a reasonable size first.

It was a fun conversation, with the debate topic centered on whether things have improved in terms of software practices over the last 15 years. There was a fair amount of skepticism on the panel, and it was a lively discussion.

As always, please send feedback to smalltalkpodcasts@cincom.com - or visit us on Facebook or Ning. You can subscribe in iTunes, and please try and cast a vote for the podcast over at Podcast Alley. One caveat about this edition; it's long, about 74 minutes. I usually keep the episodes shorter than that, but I didn't want to break this panel discussion into parts.

We have been interviewing candidates for a bit now, but we still have 3 open slots for Smalltalkers:

• 3043, Smalltalk Developer
• 3045, Smalltalk Developer
• 3051, Smalltalk Developer

If you're interested, go ahead and send an email to employme@cincom.com/ We're doing some big stuff with Smalltalk here, and - as Randal Schwartz is saying - this is a fun time to be involved.

Technorati Tags: smalltalk

This week, I have a panel discussion from SPA 2008: "Is Software Development Advancing?" It was a fun discussion, and a little over 30 minutes in, I got invited onto the panel. That was an interesting aspect of the panel; as topics came up, they asked for volunteers for the panel, we put our names on cards, and they picked one. It was fun, and I'll also be releasing the video. Stay tuned!

Michael was inspired by Randy's latest work on the NewPrerequisiteEngine, and integrated the work into the browser's pre-req properties tab. See here and here for screen captures.

Now that we have this facility, I was disappointed with how behind the entire prerequisites UI was... so I've updated it and published to public store as NewPrerequisiteEngine 34. I'll post up some screenshots of it. It's nothing fancy, in fact, it's a major simplification over the old UI. But it lets you force prerequisites, mark prerequisites that are there as optional - and of course, compute.

Technorati Tags: cincom smalltalk

Patrick Logan makes a good point: dynamic languages are productive, but they are even more so if you have good tools. Adding class extensions stop being "monkeypatching" if you have tools that support the idea:

People developing large systems in dynamic languages, and people providing dynamic languages being used to build large systems have to also realize this:

You are no longer working with a "scripting" language. You need to demand and provide really good tools. Examples can be found on the internets, read about the Smalltalk and Lisp environments from way back when. Someday you can become as smug as we are, or maybe as brilliant as the people who made them for us.

Technorati Tags: dynamic languages, smalltalk

I've seen a lot of bad ideas, but this is way up there on the ick-o-meter:

Stuff like this:

For only $50, Sony will uninstall some of the trial software it loads onto new Vaio laptops.

Fire up those browsers and FTP clients: ObjectStudio 7.1.3NC and VisualWorks 7.6NC are ready for download. We'll have ObjectStudio 8.1 NC ready soon; we're running through the last bits of Vista certification now.

For VW 7.6 and the forthcoming OS 8.1, Seaside is now supported - and Vista is supported for both products (Microsoft certification is pending for ObjectStudio).

Technorati Tags: visualworks, objectstudio

David Meerman Scott makes an interesting point: if the audience you are going after is networked, then traditional media mentions are way, way less useful than you think they are.

Technorati Tags: marketing, publicity

Randal Schwartz on Smalltalk:

If there is any year for Smalltalk to regain a commercial visibility, this will be it.  I mean, look at all the things coming together:

• the OLPC XO is putting Smalltalk into the hands of thousands of young kids
• Cincom and Gemstone are stepping up to support Seaside in a big way
• Gemstone is offering the single-instance free commercial license and GLASS quickstart appliance
• Squeak's license is finally getting cleaned up
• Seaside is reaching a nice level of maturity
• Seaside running on GNU Smalltalk for those that want a command-line environment
• Croquet is maturing, even being adopted as a commercial "virtual meeting" space
• Ruby on Rails has reestablished dynamic languages as useful for the web

He's right - from command line and files to full bore environments, Seaside has it all, and it runs portably across all major Smalltalk implementations. On the relational database side, that connection is handled by GLORP - which is also OSS and portable.

Come on in, the water is fine :)

Technorati Tags: seaside

I find this post endlessly amusing. Consider Firefox - a large C/C++ application that's been plagued over the years by memory bloat and leaks. So now we see one of the lead developers touting the progress they've made - by building a poor man's garbage collector:

Some leaks are harder to fix than others. One of the most difficult ones is where two objects have references to each other, holding each other alive. This is called a cycle, and cycles are bad. In previous versions, we've used very complex and annoying code to manually break cycles at the right times, but getting the code right and maintaining it always proved to be difficult. For Gecko 1.9, we've implemented an automated cycle collector that can recognize cycles in the in-memory object graph and break them automatically. This is great for our code as we can get rid of lots of complexity. It is especially significant for extensions, which can often inadvertently introduce cycles without knowing it because they have access to all of Firefox's internals. It isn't reasonable to expect all those authors to write code to manually break the cycles themselves.

There are collectors out there for C and C++ (not to mention languages that come with that already implemented) - so I guess the older question I could ask is, will they end up with most of a Lisp system built by the time they're done :)

Technorati Tags: garbage collection

As usual, SPA 2008 was a great event - and I'd like to thank everyone involved who helped make it that way - but especially Eoin Woods, who has made it possible for me to bring a Smalltalk tutorial to SPA for two years now. I'm already looking forward to next year. Thanks everyone, and I'll see you all again in St. Neots.

Technorati Tags: smalltalk

Michael shows us that it's time to push your assumptions about what Smalltalk is suited for aside:

so without further rambling, here is a short game play video of FreqWars as it stands now

Technorati Tags: graphics, fun, games

John Nolan is running the last session of the day (for me, anyway) - when was "yes" the wrong answer? What happens when you say no, and how can you say it effectively? Looks like an interactive session, no there likely won't be many notes here :)

First exercise - we paired off and shared anecdotes about projects where "no" would have been the right answer at some point (guess which one I used?) So why can yes be wrong?

• negative outcomes in mid to long terms
• compromised responsibilities
• project planning treated as negotiation
• unchalleneged assent
• inappropriate assumption of competence
• honesty gap
• teaching the customer to say no if they don't know how
• punished by rewards (bad incentives)
• yes as a habit
• "we can't say no now" - too much committed to project
• status/position prevents assessment

So why is no difficult? Going back to pairs to discuss why "no" didn't happen.

• the need to be the "go to guy"
• pressure to "get on with it"
• positional (rather than principles) based negotiation
• fear of saying no (fear of consequences, real or imagined)
• reputation for negativity (boy who cries wolf)
• social stigma (not a team player)
• embarassment/social interaction
• avoiding confrontation/controversy
• not empowered
• "once is not enough" - must repeat instead of once and feeling you have done your duty
• cultural expectations
• no is an absolute, while yes is (can be) variable
• charismatic people can be difficult to say no to
• as a project manager, fear that a "no" will be seen as a personal failure (especially around cancellation)
• fear that some/all of team may take the blame

The pressure to say yes: personal examples and group examples (another exercise).

The evils of "win-win" - the book "Getting to Yes" being an example:

• emotional compromise
• avoidance of conflict at all points
• ubiquitous, unchallenged, accepted attitude
• "yes" seen as the goal - seeking agreement rather than solution

John recommends (not with complete agreement) a book: "Start with No".

So getting back from a break, what happens when you say no - Immediate:

• delays, long discussions
• increased complexity
• pressure to go back to "yes"
• emotional reactions/black mail
• relief
• anxiety
• increased/improved authority/influence
• decreased authority/influence

Longer term:

• lowered or raised risk
• improved or reduced communications
• improved or worsened relationships
• curiousity as to final result (mostly for external consultants who leave)
• can cause redundency
• improved influence/respect
• loss of respect
• "the world did not end"
• increased openess/better future confidence going forward
• business survival prospects improved (saying no to bad business)
• lingering bitterness

One group conclusion: most of the short term consequences/reaction are negative, but the long term ones work out positively. A photo of the discussion:

So what's wrong with saying no: I think it can be summed up with conflict avoidance, and the risk that you might not be right.

"It's not what you do, it's the way you do it" To sum up: How do we do it?

• get them to say no for you - walk them through the argument, draw it out until they reach the same conclusion (or, possibly they convince you that they are right after all)
• explain consequences
• influence with "because" techniques

Make sure you have enough information to make a decision. If not, defer. However, make sure you start getting the information you need. Avoid passive/aggressive behavior - there's a difference between options and procrastination :)

When you do have to say no:

• make it personal: take responsibility
• keep it short - and make sure it comes out as no
• prepare for the likely/possible outcomes
• making it personal does not mean making it confrontational
• remember that you likely don't have common goals
• make sure you understand what the question you are answering actually is
• when in doubt, say no. You can always change to yes later (the reverse is hard)
• have empathy, and be honest

I did a brief post on the panel two days ago, and I have video that I'll be posting later - in the meantime, Radio Free Agile has a summary.

Technorati Tags: software productivity

After the keynote, this is an interesting slowdown - a "roundtable" discussion of various software boks. First up: Beck's "Smalltalk Best Practice Patterns". The book dates back to 1997, so the discussion is almost nostalgic in nature - a lot of the people here started in Smalltalk, and they have fond memories of their work in it.

The panel liked Beck's book - the take away seems to be that yes, it's Smalltalk specific - but a lot of the lessons in the book are generally applicable and have to do with communicating intent between developers through well written code.

The group discussed other books, and it was a nice break from the more highly interactive "think" sessions after Tuesday evening :)

The second item was a seminal paper from 1972 - "On the criteria to be used in decomposing systems into modules" by David Parnas" - the group agreed that while the ideas expressed are now "old hat", it was ground breaking for its time.

Great line about one of the other books ('"Paradoxes of Group Life") - "I must say, I didn't understand this review any better than I understood the book". This book got a very general thumbs down as being incomprehensible for the general reader.

Next: "Software Requirements and Specifications" (Jackson) - Interesting take on this one: don't read it straight through, follow the "root map" through it. Basically, it's a website in printed form. Interesting take on this one, too: many of the people in the group thought the book had some good ideas, but that Jackson himself is irritating.

The last book taken up: "The Social Atom". This one isn't technical, it's about group behavior, but on a more approachable basis than something "technical" on the subject (like "Paradoxes of Group Life"). Michael Feathers brought this book up, and it was mentioned in his talk. People liked this book, but there seemed to be a sense of "what do I take away from it?". At the same time, people liked it. Heh - "I hated this book until I came in here".

Technorati Tags: books, reviews

The Wednesday morning sessions at SPA are always interesting: there's the late night whiskey session on Tuesday, followed by the mandatory checkout by ten on Wednesday. As a result, the morning keynote is always a bit sedate - at least for the audience.

This is a cool talk - Michael is bringing together the economics of legacy code and the book "Freakonomics", which I enjoyed immensely. It looks like I should also look at "The Social Atom", another book he's mentioning. Another reference for the talk - Brian Foote's "Big Ball of Mud" talk.

As with the other keynotes, you'll be able to se the video (or hear the audio only for it) sometime in the next few weeks as I process and post the content.

Technorati Tags: michael feathers, legacy code

Gordon Weakliem doesn't much care for this post from Joel Spolsky, and points out Mark Pilgrim's take-down of it. The problem is, if Pilgrim read Spolsky's post, he didn't understand any of the words. He has a lot of fun with Spolsky's premise, which is that standards often mean nothing if practice has made it irrelevant - and for good or ill, that is the way it is in HTML and CSS. Where does Pilgrm go? Places like this: (his "translation" of Spolsky's points)

I demand documented standards with open reference implementations. That's why I only develop with Microsoft technologies.

That doesn't address the actual issue though. The reality is, if you want your web page to work with the browsers people actually use, which is more relevant: the supposed standards, or the working reality? Whether you like it or not, real people don't care about the W3C specs, they care about the website they just visited and the content there. Are you interested abstract purity, or making a sale?

Pilgrim (and Weakliem) have picked the former. Most people don't give it a first thought, much less a second.

Update: Gordon thinks I'm missing the point. Here's the thing: Pilgrim seems to think that Spolsky is wrong because he develops for Windows. Gordon thinks he's wrong because he seems to believe that adhering to the spec will magically cause old web pages to get fixed. The old web pages won't get fixed, people will blame IE8 for not rendering them, and Pilgrim will still be a buffoon who is way too enamored by how clever he thinks he is.

But other than that, sure :)

Technorati Tags: standards, browsers

I just attended a fascinating workshop here at SPA 2008 - a set of pairing exercises. It was run by John Daniels, Laura Hill, and John Cleal, and it was one of the best run sessions I've been at here (which is saying a lot - they've all been pretty good).

We split into two groups (one of 4, the other of 5), and then ran through four exercises. In some we worked alone, and in others we paired - and after each one, we compared notes and filled in a questionnaire. The exercises were:

• Descriptive writing - describe the painting (below) in 200 words, as if you were explaining it to a blind person. I did this one solo
• Scheduling - given a home redesign job, schedule the tasks optimally to take the least time, assuming that personnel costs are no issue. I did that one solo.
• Personnel - interview a problem employee, and see if you can come to a mutually agreeable solution - I did that in a pair
• Manual - we had a lego model (fairly complex) to build. I did that in a pair.

The interesting conclusion we came to (in my group, at least) was that the personnel issue was best dealt with by more than one person. Having a second person in the room kept the discussion away from emotion, and I think that's something I'll take away from this. The Lego exercise merely documented my well known spatial issues :)

It was a fun workshop - if they run it again, I'd recommend it.

Technorati Tags: pairing

Now, is this great PR for Apple, or what?

Viewers of the Charlie Rose show tonight were stunned to see the normally composed Rose looking like he’d just been in a bar fight. He has a very bad black eye and a bandage over part of his forehead. I contacted the show’s producers to hear what happened. Earlier today, they said, Rose tripped in a pothole while walking on 59th Street in Manhattan. He was carrying a newly purchased MacBook Air and made a quick (but ultimately flawed) decision while falling: sacrifice the face, protect the computer. “In doing so, he pretty much hit the pavement face first, unfortunately,” they said.

Tha sad part is, I understand :)

Technorati Tags: macbook, macbook air

So what is this field all about? It's about capturing/discovering/mapping the social graph, and then providing (hopefully useful) services through that. One irritation: you have to enter this data over and over again (and transmit that pain to all your friends) for each new service. There are solutions to that problem, but - as Scoble has demonstrated - these are controversial).

This all depends on the API provided by the platform, and at the moment there's Facebook and Open Social (Google and partners). Bebo has adopted the Facebook API, so it seems to be the more relevant right now.

Key things to remember about your Facebook (et. al.) application:

• Runs on your server
• communicates with the service via a set of HTTP APIs

Facebook has a fairly clean API and set of "Facebook tags" (FBML) that you get access to. The presentation just went to a demo on that.

Now I'm getting into the whole "back channel at the conference" thing :) - Jane Chandler (one of the attendees) sent along a set of useful links that cover some of this ground:

• Periodic table of the web
• An internet mind map
• A map of the network
• The Web is Agreement

It's fairly easy to publish an app - promoting it is a whole different ballgame, and will (or won't) get adopted on its own merits. There's also a separate platform for mobile devices, a data store API, and a new API for Facebook pages. Revenue model: advertising with shared revenue (Facebook's terms). Will it broaden to more traditional business applications? Not clear yet, but the thinking seems to be yes-ish. Possibilities? Combination of location data from mobiles and social graph data from the service.

Technorati Tags: facebook, social media

Tuesday morning at SPA 2008, and it's time for another keynote - this time with Peter Deutsch. While I've interacted with Peter on and off over the years on Smalltalk mailing lists, I hadn't met him in person before this conference, so it's been nice to be able to put a face with the name and personality.

As with yesterday's panel, I'm filming the discussion, so I won't be following every bit here in this post. However, here's the url Peter references for the presentation:

http://www.major2nd.com/users/ghost/papers/spa2008.html

Heh - with such a rich background to draw on, Peter has only hit on two of his 8 points by the halfway mark in his talk. Doesn't matter, really - it's an interesting talk. You'll be able to see it (or listen to the audio only) when I post it.

Update: Looks like the link is bad; Peter has informed us that it will work after the conference, when he gets back to his office.

Technorati Tags: language, software+development

Travis has posted an implementation of some of the nonsense we talked about last weekend. Fun Stuff!

Anyway, in flippancy, I suggested that one way of encoding a large range of at: someIndex methods tersely was to simply use roman numeral selectors. The nice thing about roman numerals is that they start at one, so they match Smalltalk code well. I was then challenged to make it do tail based access as well. I had said I would use case for this distinction, but in the end decided to use a trailing underscore to indicate that the selector should go from the back of the collection, rather than the front. Mixed case is supported.

What a pile of manure, huh? But such is the normal kind of fun hack for me. And it turns out that NormalManure is a nice anagram for "Roman Numeral", so I published it in the Open Repository under the name of NormalManure. The other part of the blog title composed of roman numeral characters (my second candidate for a goofy package name).

SPA now has a Facebook group - you can join the conversation there.

Technorati Tags: facebook

It's after lunch, and there's a panel session: Has the practice of software development improved? Looking back over the last couple of decades, the initial answer seems to be mostly no - the languages and tools haven't really moved forward all that much, outside of the automation of a few pieces of grunt work. If you look across the field, on productivity, predictablity (et. al.), it's all been small beans.

Rather than take notes on each part of the talk, I'm filming it - I'll post the video later on, after I have time to do some editing and preparation on it.

And no, the room was not empty :) The entire crowd avoided the front two rows of seats.

Technorati Tags: software, software development, productivity, tools, languages

If you watch Stargate SG-1, you know about the replicators. Well, in the movie that just came out (which wraps up the ORI thread), we learn why the replicators are so devious - the little buggers are Javascript powered. Clearly, to stop them the team should hav e deployed a SOA solution :)

First session at SPA - three hours on the concept of code debt. I was a little nervous about the time - but I need not have. The time flew by, as the presenters (Peter Marks and David Harvey) did a great job. The exercise they led off with set the tone perfectly: We split into two groups, with each group off into pairts. Each group got a set of Javascript (with issues and tests) to fix. One group had well factored, easy to read code, the one I was in had poorly factored, very hard to read code.

If that didn't express the idea of code debt, nothing would :) From there we covered a lot of ground, and had a few exercises to consider things that cause code debt. As is often the case for a SPA session, the outputs ended up as sticky notes on a tack board:

It was a great way to start the week - I'm looking forward to the rest of it. Here are two shots of the presenters:

Technorati Tags: code+debt, software+development

I really like this conference - it's small, and has a lot of the same feel that Smalltalk Solutions has (without being focused on any one topic). Things opened up this morning with the introductory plenary - I snapped a few photos:

Technorati Tags: conference

One of the interesting aspects of the SPA conference is that sessions are expected to come to a conclusion, or generate some kind of output. For the tutorial yesterday, I have the code I was using to walk through Seaside. We didn't get as far as the posted code did; we didn't really get to dealing with Ajax at all.

Anyway, the full example system can be downloaded here, as a parcel (in a zip file). Any questions, send them to me, James Robertson.

Technorati Tags: smalltalk, seaside, outputs

This week we had an ad-hoc talk with Travis Griggs (after a number of technical difficulties involving bandwidth issues here at the conference hotel). Michael ended up hosting the skype chat, and the recording proceeded fine after we did that.

We ranged fairly widely, but the talk centered around API design in Smalltalk - with a few wild proposals tossed out in the form of argument by exaggeration :) It was a fun talk, and I hope you enjoy it.

As usual, if you have questions, please send them to smalltalkpodcasts@cincom.com - or visit us on Facebook or Ning. You can subscribe at iTunes, and please remember to go cast a vote for the podcast at Podcast Alley.

Technorati Tags: smalltalk, code smell, api

We actually did get a podcast recorded this weekend - Michael had to host it, as my bandwidth at the conference hotel pretty much sucks. However, it was a lot of fun - Travis joined the call and we had a great time bouncing thoughts around. I should be able to get it posted in the next day or two.

The tutorial went well here at SPA - and just in time, I saw this announcement on the Seaside mailing list:

Dear all, We would like to announce that we are about to publish an extended and improved version of our Seaside tutorial as a printed volume via Lulu. We plan the book "An Introduction to Seaside" to be finished by mid-April. Thanks to everyone who provided feedback and suggestions for improvement! Michael Perscheid on behalf of the HPI Software Architecture Group

This is a reference to this fine Seaside 2.8 tutorial. Great news!

Technorati Tags: smalltalk, spa2008

Speaking of whole new worlds of web development (see my last post), I'm giving a Seaside tutorial here at SPA 2008 in about 2 hours. It should be a lot of fun!

Technorati Tags: seaside, smalltalk