I had the honor of being 1 of the lucky 10 invited to attend the Mix n Mash 2007 event held by Microsoft. This is the 2nd one. It’s an event where they invite you to Microsoft’s HQ to see what they are working on, give feedback, and meet Bill Gates.
Here’s a picture of the Mix n Mash invited attendee’s getting their picture taken with Bill Gates (I somehow got in there, booya!)
From left to right:
Jonathan Snook, Julie Lerman, Kelly Goto, Rob Howard, Bill Gates, Molly Holzschlag, Kip Kniskern, Jesse Warden (me!), Keith Peters and Erik Natzke
Last year’s event apparently had a lot of designers who thought the Silverlight content was really over their head. So, this year, they has 2 tracks; 1 designer track and 1 developer track. I told Tim (Microsoft’s Ted Patrick) that I wanted to go to both. In the end I had to choose one. I planned on going to the developer one because I wanted to give the community a play by play of the features Silverlight 2 will have and do a fair comparison with Flash Player 9 & 10.
However, after about 5 virgin fruit punches at Trader Vics, I was really hyped up after talking to various Microsoft employee’s. I believe they are doing everything right with Silverlight, and tons wrong with Blend. After draining every employee I could meet of info, I decided the next morning I’d hit the design track instead. The 4 of us got to meet some of the designers and dev’s working on Web and Blend.
This career histories’ these dev’s have is just unreal. If I’m 80, and can talk even a 100th of how they talked about where they worked, what they worked on, and how those projects affected them, I’ll feel confident I enjoyed my career. I’m not really sure how in the hell I got invited into a small, informal room to talk shop with these pro’s. Yet, the informal nature made it damn fun. Free coffee was pimp, too!
Web seemed like Microsoft’s version of Dreamweaver. Clean, well laid out, and the focus of the conversation was about integrating a good set of CSS features to help raise the bar for the industry. Apparently there is tons of CSS done horribly wrong. They had some cool ideas, and I’m sure we only heard 5% of all of them. At first, I just listened to Snook and Molly. They are both smart, so just absorbing the dialogue was rad. As they started to show PHP code hinting, and SWF embedding, I felt it was my duty to step up. I got shot down pretty quickly about using SWFObject by default for SWF embedding, or even considering using the Embed tag. It’s all good, I have Pimp Cloth Armor +2. I still need to send the Web team an email of the additional parameters one needs to allow a developer to modify to embed Flash content with more control.
Additionally, I asked about what version of PHP code hinting they would support. It sounded like 5. While I appreciated code hinting for PHP, I asked that they also provide a way to improve on those code hints via configuration files in case they don’t get auto-class hinting like Zend for example. If you’ve ever tried to write your own custom code hints for Dreamweaver, attaching a hook to the back of your neck and jumping out of a helicopter with no parachute is more fun. Bottom line, even version 2, if Web is installed on my machine, I’d use it. I primarily use a Mac, but on my home PC, heck yeah, Web looks fun, and their CSS re-factoring tools sound yum. I suck at CSS so any tool that helps me become a better coder by enforcing best practices… woo woo!
Next up was the Blend crew. Some pro’s and cons here. First off, Blend has it rough right out the gate. It’s a phat tool for doing both Vista application design for WPF and Silverlight development ( as subset of WPF ). Creating a tool to support both environments is a challenge. Flex Builder 3 does it pretty well from the Beta 2 that I’ve played with. Choosing the older SDK or the new 3 one is just a radio button, converting older code is a snap if you aren’t using charts, and choosing an application to be an AIR app is just another radio button. The AIR API isn’t monstrous compared to WPF, so Adobe has it a lot easier getting Flex Builder to support the Flex SDK, ActionScript 3, and AIR api’s. Furthermore, people like me don’t really give a flying flip about developing applications for Vista (until customers come calling offering more bling than they do for web appz). The demo’s they showed for Vista were hot, definitely had good designers employed behind their creation, and the syntax just drew me in.
For example, their binding syntax was sick! Instead of just being a variable like Flex does it, it actually had parameters IN the binding. The 2nd one was one way or two way binding. OMG, I write so much bloody code to do 2-way bindings in Flex. For the simple stuff, you can use 2 binding tags, or just another bind… but for the complex stuff, you end up writing your own getter/setters. WPF’s way is just… awesome. So awesome I forgot what the 3rd parameter does. Adobe, copy that shiz.
…alas, I didn’t really get a feel for what the app had for Silverlight 2 creation. You can only show so much in an hour when someone keeps running their mouth about how they hate the Solutions Panel. We used the Blend September preview on a project, and I felt it was good. The integration between Design & Blend needed some work, but you COULD get your designs there, just not 100% intact. What really beefed me about Blend, though, was their timeline. With XAML (the XML markup language for WPF & Silverlight apps) having all kinds of wonderful options with regards to animations being separate elements and applicable to objects either as integrated or re-usable resources, I was really irritated that the current timeline in Blend doesn’t leverage that framework to it’s fullest. Furthermore, the community has asked Adobe countless times over the years for Flash to have an After Effects like timeline option. I told Samuel Wan almost like 2 years ago that they really need to use that as a selling point for Blend as an advantage. I guess they think that LiveMotion had that, and didn’t out-sell Flash, so why should they? I still disagree with the current incarnation; to me it needs a lot of work. The newest Blend preview I saw had a bezier tweening panel for some of the keyframes. Dope; it was big, and for a specific keyframe. Regardless, I still think it’d be cooler if it were integrated into the timeline itself. Once you go After Effects, you never go back. I liked what I saw. It was hilarious to see the project manager take it personal that they didn’t have a specific month (October?) preview.
I debated for an hour about that damn Solutions Panel. Maybe the Visual Studio guys like it, but after 1 project using that SOB, I hate it. I argued that we need to have a regular explorer window in the program. I don’t want to remove it, rather just for those of us web developers who are fully capable of managing our files in Windows Explorer / Mac Finder, we don’t need some glorified panel to hand-hold us. I’ve never seen an app do well that built their own explorer like metaphor whereas the ones that do good use the OS based explorer as the base, and built atop it. A lot of the text editors like EditPlus, TextMate, etc. do it very well.
Raise your hand if you think, in retrospect, the Project Panel in Flash MX 2004 was a complete failure? Of those who raised your hand, do you dig GProject? I know I do. Grant Skinner built his own, and it did well. I’d argue, though, that it was merely Grant’s skill and long history of Flash that made it successful. Most of us would of preferred just a simple explorer window we could point to our project directory, give the IDE some class info, and call it a day. Kinda like FlexBuilder….
Their counter-argument was that they need the Solution’s Panel for 2 reasons. First, it allows the same project to be opened in Visual Studio. Fair enough. The second reason was that it allowed them to determine what assets needed to be compiled in the Silverlight app. Whatever. Give me an explorer window, and parse the XAML or C# to know what assets I used. Either that, or make it not suck; like work like Windows normally works. Drag and drop, a REAL folder structure, not some made up one, and auto-refresh.
Another funny part was when I asked why their XAML didn’t use proper namespaces. The only namespaces it used was for custom stuff or some attributes. They replied that it’s perfectly valid XML to not show the namespaces unless you are writing elements outside of the default. Then Molly chimed in her agreement with that “tone” that implies, “What the f$*2 are you talking about?”. It’s pretty awesome to have one of the project leads of Blend and the web standards queen shoot you down at the same time. Owned. Naturally, I’m left wondering why in the heck MXML in Flex has the namespaces like that… and what’s the default namespace for MXML? Grr!
I could rant all day about Microsoft Surface. All I’m going to say is, first, YouTube doesn’t do it justice; you have to see it in person. Second, it changes everything. For those who can’t take the epic sentence, here’s some sarcasm to compensate.
Lastly, we got to introduce ourselves to Bill Gates and ask him a question or two. I thought long and hard about what to ask, and every question I came up with could probably be found to be already asked and answered by crawling the web. Although I felt pressured to ask something relevant technically to the Flash & Flex community, I just couldn’t find something that was appropriate. So, I asked about his decision to focus on the Gates Foundation full time. Keith covered all the questions in his series of blog posts. I’m re-posting my question here with a few corrections since the transcriptionist couldn’t hear me too well.
Jesse Warden: So, your a (almost called him a dude) man who has really influenced a lot of software industries and then as soon as you announced the Gates Foundation, the media jumped on it and talked about how you were influencing philanthropic endeavors. Was there a turning point that made you feel like you wanted to do that full-time or was it a culmination thing? How did you get to the point where you now want to focus on it full-time?
BILL GATES: Well, I said even in my 20s that I didn’t think somebody who was 60-years old should be deciding the technology strategy for a state of the art software company. So, you know, I’m pretty old now, and it’s good for somebody else to come along. I’m still the guy who thinks, hey, saving a few bytes of memory is a worthwhile thing, let’s spend a few hours and save a few bytes here. Nowadays just saving a megabyte, it’s not worth it, don’t waste your time. So, you get a certain mindset about what makes sense. To me, computing power is still a little scary. Now, compared to the people before me, I was a revolutionary in how I thought about those things.
So, there comes a point where somebody should pick things up. Since the age of 17 I was CEO and sort of chief software architect until seven years ago, and then I got — that was a big change, maybe not as big but it was a big change, because then Steve took over and has run the company since the year 2000, and that was sort of getting ready for the idea of running Microsoft in a different way.
The foundation stuff has turned out to be very exciting, and I’d say there’s one thing in common with the foundation and Microsoft. When we announced a software-centric vision of the world, 1975, there weren’t a lot of people involved in that, and it’s been exciting to see it sort of developed as we dreamed that it would, and many people coming in and contributing to that.
When the foundation got going in global health, there just wasn’t that much — there was nothing going on in terms of Malaria, TB. We have 20 diseases that we do. And we think we’re going to conquer the majority of those in some reasonable period of time.
So, my ability to spend full time on that, go get the pharmaceutical companies more involved, get the best scientists more involved, back some risky approaches there, I may be more unique in terms of facilitating some of that than I am in software.
And sometime in my fifties I was going to make a change, and so about a year and a half ago, we picked the date and told the world, and we’re on track for that to happen.
So, there’s nothing magic about the date, but it seemed like Microsoft was in good shape, Ray Ozzie was the person stepping up to take on a big part of what I’d one uniquely, and so we made the choice.
Afterwards, a few of us went out to finish the night out. It was really cool to meet a lot of new people in my industry; that’s one of the reasons why I got into consulting in the first place. They were all smart and clearly dug their professions, being passionate about what they do. I got to learn a little more about some of the Microsoft people, and how they got into their roles.
By the end of the day, I had pretty much lost my voice. I had never talked about software for so long, and so vigorously. Insanely fun!
All in all, the trip was a blast. I met new people, learned a lot, and hopefully helped Microsoft out. Most importantly, it really changed my perception of Microsoft. I never grew up in the world a lot of other people did with regards to Microsoft being a bully against the openness of the web. So, I never had a negative perception of them to begin with. Instead, I grew up with Corel, Adobe, and Macromedia software so just never really cared about their products because I never used them much. Yeah, Word, Excel, and XP, but I didn’t make my career on that software. Anyway, you hang out in the web world for a length of time, and there is enough Microsoft vitriol that some is bound to get splashed on you. It was funny to hear some of Microsoft employee’s, especially the new ones, talk about their shock, and ways of dealing with the hate and anger thrown their way the minute they say they work at Microsoft. This is apparent visually in the body language of those that are in parts of their designer division. Even some of the contractors who merely consult on some projects for Microsoft recount on how they borrow notorious death threats from colleagues with similiar Microsoft ties, and use those as de-sensitizing measures by posting print-outs of them on their wall to constantly remind them that “someone else has it worse”. Really wild. Regardless, they are a lot like Adobe. They are cool, having people there who clearly love what they do, and are insanely smart.
What really drove it home for me, though, was the internal strife; the healthy kind. When I was about to go work at IBM, I had peers exclaim how horrible it was in their web technology division. This couldn’t be further from the truth; meaning, if you actually do software development vs. extremely simple and tedious HTML updates. Clearly, IBM is so big, that you can’t judge a company by one employee in one part. Having an opinion of a company so large is personal, it makes it really hard to stereotype. My experience there was really positive. Naturally, a company that large has a lot of internal companies vying for their own self-interest. So, it was really cool to hear about how Microsoft has these internal mailing lists, with lots of traffic, and various other politics that play out over time across these areas of the company’s product lines. The company’s best interests are fiercely debated. No one employee I talked to could (fairly) give me a sense of how much one concept was agreed on percentage wise.
For example, a lot(?) in Microsoft believe, like I do, there is room in the Rich Internet Application market to do things right that others have done wrong. To provide good tools and technology for people to build cool experiences atop of. Like, say… downloading fonts and using them on the client in Sliverlight 1.0. The counter to that is those who believe they should instead be nurturing their existing .NET developer market towards these tools, and allow them to leverage their existing skillset in the RIA arena. The example here is how Blend and Visual Studio’s integration model ENFORCES the code-behind methodology. You design, layout, and transition your app in Blend, and write the code for it in Visual Studio. Each view or component has associated code that makes it work in an associated .cs file; aka, code behind the design, that is visually represented in an associated .xaml file.
There are 3 ways to do this in the Flex world, and I hate them all. Although inheritance still is bleh compared to composition, when you grow up extending GUI base classes to create your View’s in Model View Controller, it’s really hard to relate to these mature Java & C++ programmers who come in and go, “OMG… that is so gross! You need to separate your View and it’s implementation.” Say what? The 3 ways are: MXML is all GUI, and ActionScript is externally referenced in your mx:Script tag; MXML is all GUI, and you extend that MXML class with ActionScript to handle the implementation; or the reverse; you write ActionScript implementation code, and extend that with MXML for layout.
The down-side is the constant “throw it over the wall” mentality that I see in the perception of work-flows. The best teams (villages?) are those that have designers and developers working together, regardless of tools. Having Blend be the experience creation liaison between Design and Visual Studio isn’t a good substitute for a qualified Interaction Designer. If you read Grant Skinner’s initial entry about Thermo, Adobe’s tool in development targeted at Designers that will allow them to leverage the Flex SDK from a design perspective, and actually be empowered to design good Flex apps, you can see how again, you run into dangers when you try to supplement a tool for a qualified person. Unfortunately, it’s hard to sell a client on needing another $20,000 so you can afford the proper Information Architecture contractor on a project. “What? We’re already paying for a DBA, back-end developer, front end coder, designer, and project/client manager… what the hell is an IA really going to add?” The same goes for hiring Interaction Designers, or any really good designer / developer hybrid who can specialize mitigating work and designs between the designers and developers on the team. No tool can replace the importance of those people on projects. No one can get around production art (aka break up).
Regardless, I still like the tools. To me, Flash MX 2004 proved that the tool is mightier than the run-time. Yes, AS1 bytecode was slow, but when you give someone a tool to leverage a class based (ActionScript 2) language to create really compelling and powerful web applications… that just says a lot about how effective good tools can be at empowering people to leverage runtimes. Allowing those tools to integrate, and integrate well, is what really helps facilitate good team interaction in a work-flow that works. I hereby request Microsoft refrain from using the term “throw it over the wall” and instead focus on Expression’s mantra being “tool integration”. There’s something special about how Blend and Visual Studio can share the same project. You don’t see Flex, Flash, Photoshop, or Fireworks doing that shiz. You need to leverage that, build upon that, and empower teams, especially those who have never done hardcore RIA work before, to be successful on their own merits and not have the marketing suggest they retreat into their own camps. Instead, it’s a collaborative effort.
My vote for representatives from the Flash/Flex community for next year (2008) are:
Developers = Grant Skinner, Colin Moock, Darron Schall, Chafic Kazoun, Claus Wahlers, Joe Berkovitz, Doug McCune, Ben Stucki
Designers = Joshua Davis, Bradley Grosh, Hillman Curtis, Your Mom
Thanks again Microsoft, it was dope! I hope I helped out.