The Blog of Frank W. Zammetti - The demented ravings of a software developer

CES was a good show for Palm, no question.  They “brought the thunder”, “made it rain” or whatever the hell it is the cool kids are saying these days.  In fact, I’d go so far as to say it was a great show for them.  They delivered on everything most of us thought they would based on various rumors and hints leading up to the show, and yet still managed to give us a surprise or two.

What did we get?  Let’s run down at least some of it:

• The PDK (Plugin Development Kit).  For code monkeys like me, this is the biggest announcement.  It allows us to write native applications in C/C++ that, from what I think I’ve understood through reading, wind up running essentially as browser plug-ins to the Webkit engine that provides the webOS UI (hence the PDK name).  No question this is cool and opens up a whole new world of applications.  In fact, one of the surprises Palm pulled off was a bunch of titles released into the App Catalog concurrent with the announcement that are pretty awesome.  I’ve downloaded a couple myself already and they are clearly head and shoulders above anything we’ve seen on a webOS device before.  However, all is not right with the world here, but I’m going to come back to this later because it’ll wind up being the main thrust of this post I suspect.

• The Pre and Pixi Plus and Verizon availability.  I used to be a Verizon customer but I’ve been a very happy Sprint customer for a number of years, long before the Pre.  To be frank, I could care less about Verizon and Palm being on their network, except for one key point: it opens up a whole new base of users to sell apps to.  In this regard alone this is a huge bit of news.  We of course knew this was coming for some time, although the devices themselves were kept pretty quiet until the past few days.  I am pretty pissed that I can’t get a Pre Plus as a Sprint customer though and my hope is Palm and Sprint make it available in the not too distant future.  The problem however is if its six months out say, which doesn’t seem like a crazy guess, then I’d bet we’re beginning to talk about a Pre v2 then (or v3, depending on who you ask) and maybe it’s better to just skip the Pre Plus.  The point of course is I want the damned thing NOW, so boo-hiss to Palm on this one.  I’m sure it makes sense from a business perspective… I’d guess Verizon made them being the exclusive provider of the Plus devices part of the deal, and Palm realizes that it’s more important to get webOS-based devices into the hands of more users and that us Sprint customers who get pissed are just a bit of collateral damage.  Makes sense for them.  Sucks for us though.

• The news about webOS 1.4 was where I think there were some surprises.  Certainly I didn’t really anticipate video (although I was hoping).  It sounds like a real nice upgrade all-around and kudos to Palm for continuing their free upgrades at a rapid clip.  Now, where the hell is voice recording guys?  I’m hoping it gets slipped into 1.4, or is there already and they just didn’t mention it.  This is one thing I REALLY miss from my Windows Mobile days and it’s surprising that a modern smartphone doesn’t have it.  Palm, if you’re listening, I can’t imagine for a second this is a difficult or time-consuming thing to implement… slip it into 1.4 and make at least me a very happy person!  Please!  I’ll name my next child after you… Palm Zammetti has a nice ring to it : )

• Flash is coming soon.  This is awesome news and will of course greatly expand the webOS application landscape.  I’ll be very happy to be able to show people Happy Tree Friends, Weebl and Bob and The Lonely Astronaut on-the-go again!  Here’s the one key question though: how is it implemented?  Is it the case that we can visit a web page and have Flash content work?  That’s a given.  However, will I be able to “install” a Flash app and launch it directly from the launcher?  In other words, will I be able to run a Flash application in such a way that it looks no different than any other webOS application?  That’s my hope.  Knowing the architecture of webOS I frankly EXPECT it to work that way.  Now, how “installation” works, I’m not so sure, but at the end of the day since webOS is essentially a browser for all intents and purposes, and Flash is a browser plug-in, I certainly think it SHOULD work this way and if it doesn’t then I’d say Palm has missed an opportunity.  Maybe it’s not up to them, maybe it’s 100% on Adobe, but I know if I was Palm I’d exert as much influence as possible to ensure this was the case.  At this point, if I had to bet, I’d bet it does work this way because it seems too obvious… but we’ll see.

I may have missed one or two things, but I think those are the really major items.

Ok, so now I want to talk about two things that I see as negatives about the whole native app thing that I haven’t seen mention of elsewhere.  First, stability.

Yesterday I was playing Asphalt5, which by the way is an awesome game!  However, it crashed on me twice in a ten minute period.  This was a hard crash too: I had to literally pull the battery to get any sort of reaction.

This is very much the risk you run with native applications.  We can talk all we want about the stability of Linux, how things are compartmentalized and all that jazz, but at the end of the day, its C, its pointers, and that’s a dangerous combination in the hands of lesser developers (hell, even good ones!).  It would be easy to say maybe we’re just looking at one poorly-coded game in my case, and maybe I’m making an unfair generalization from it.  The problem is that it’s a generalization that’s always been true on every platform out there!  The closer you get to bare metal, the more dangerous your code becomes, regardless of what safety the OS provides.  So, we may see some stability issues from this native capability, and unfortunately that’s going to wind up reflecting badly on Palm.  Remember, most users aren’t developers… they don’t know from C, or web versus native development, they just know that their Pre crashed hard playing The Sims, and they’ll blame Palm (and the company who created the game of course, in whatever ratio they deem correct).

I’ll be honest though: the sooner I see a Commodore 64 emulator ported over the sooner I’ll be 100% in love with the native capability : )

However, there’s a much more subtle negative… or, to put it more correctly, POSSIBLE negative (it very much depends on your point of view, as Obi Wan once said!).

So here we have Palm all along telling us the web application development model is the future.  They’ve built their whole operating system around that concept and in fact have I think effectively bet the future of their company on it because if webOS turns out to be a failure it could ruin them.  From the very start they’ve been telling us that HTML+CSS+JavaScript is all you need to create robust, top-notch applications for webOS, and I know that I for one believed it, in fact did before Palm said anything, and I still very much believe this to be the case.

So, do we now take the native apps as a tacit admission that the web development model doesn’t quite cut it?

I know what you’re going to say: only for a certain class of applications that might be true.  3D games still are not generally possible with web technologies (a few tech demos and even one or two relatively simple games that do exist being exceptions).  And to a large extent I agree with you… what web developers can do increases seemingly every day as the technologies and standards develop, but today, right this minute, there are clear limitations  Why should Palm say that only the web development model is right then for everything?  Why shouldn’t they say that maybe for certain types of applications there needs to be a different approach?

I’m ok with that in general I suppose.  I’m ok with saying “if you’re going to write this sort of application, the standard SDK is fine, and if you’re going to write this other sort of application than maybe you should be using the PDK”.  That makes sense to a large extent.

Where I see it as a negative though is from a philosophical perspective.  Palm is really the first company that has delivered in a mainstream way what many of us have been talking about for a few years: a web technology-based operating system.  To now have to admit it may not be all it’s cracked up to be (if you agree with interpreting it that way) is depressing to those of us that have bought into that model completely.

I was in a sense looking to Palm to push the boundaries on this web development model to new heights because I think they’ve positioned themselves better than any other company to do so.  And I believe they still will do just that, but I think now providing this native capability might serve to muddle the picture somewhat.

To be sure, I think there’s SO much more potential in the web development model, even for games, than what we’re seeing now.  There are a few games in the App Catalog today that actually begin to show what’s possible, and I really wanted to see Palm develop the tools for that model more.  And in fact, I’m 100% sure they are anyway!  If you read the developer site carefully, where they talk about the PDK, they also state that they are working on bringing JavaScript APIs to us developers that give us access to that sort of native power form the web technologies perspective.  Awesome!  That’s what I really want to see!  In fact, I will state for the record that I wish THAT was the primary announcement at CES… as cool as the native apps are, as neat as I think the PDK will be, I would much rather have heard how there were some new JavaScript APIs that gave us more robust hardware access.

I would have preferred that because, philosophically, it’s more inline with what Palm has been preaching all along: the web model.  It’s almost like there’s a dual message now, and from one point of view it’s good, but from another it’s bad.  If the web model was really up to task today we wouldn’t need the PDK.  The fact that it exists in a sense shows that it’s not, and I would have preferred Palm push the boundaries in that area more.

Still, philosophy is debatable, as are these points of view.  Palm did a great job at CES and I for one am VERY excited by what they announced.  I think they made the right moves, all things considered… so long as stability doesn’t prove to be a huge problem with the native apps (only time will tell) and as long as they continue to push the envelope on what the web development model can do.

On the later point, I think they will, but here’s my concern: how easy would it be for Palm to say “ok, we can relax a little now because when people want to code a Twitter client they can use the web model, and when they want to code Halo they can use the native model”.  I think that would be the wrong answer.  The RIGHT answer, to me, is “we’re going to work to make the web model so powerful that you never again need the native model, but FOR NOW, this is the temporary solution we offer”.  That’s cool with me… but over time I should need that PDK less and less, if Palm is doing things right!  If it doesn’t go that way then as far as I can tell they will have admitted the web development model was a mistake.

There’s one other possible third negative that I’m not even sure I agree with, even though I’ve thrown it out there as a possibility in discussions already, and that’s that this native capability might be a bad thing for small-time developers (like me!).  With the EA’s and GameLoft’s of the world in the mix now porting, who can spend so much more on development and, more importantly, ADVERTISING, than I ever could, it’s going to be harder for one and two-man type shops to make a buck.  I think those large development houses might effectively eat our lunch.

Now, some have responded to this by saying that it’s a shame not everyone can be successful but that’s the way the free market works, and I agree.  The difference however is that small-time developers really can’t compete against those bigger companies in most regards.  Like I said, advertising if nothing else.  Oh, to be sure, there WILL be successful small-timers…when it comes to games, a good, fun concept trumps the largest development and advertising budget you can name.  Word of mouth is an extremely powerful thing.

But, if that’s ALL you’ve got going for you, it’s a HUGE gamble to even develop a game.

I’m not sure how this is going to play out… some claim that it hasn’t had a negative impact on the Apple app store,   However, I’ve looked at a number of reports about the top-selling games of 2009 for the iPhone, and guess what?  Most of the top-selling games come from two sources: EA and GameLoft.  Hmm, I just might be on to something here, unfortunately.  Believe me, as someone who develops games and has a few projects gearing up that I hope to make some money off of on webOS, I certainly HOPE I’m wrong and I even WANT to be wrong!

One other thing I want to touch on, and that’s Ares.  I was going to write a rather lengthy blog post about that a week or so ago, but I decided not to because I wanted to think about it a bit more.

Here’s my basic opinion… technologically, it’s cool as hell.  The developers working on it deserve all the credit in the world, they are doing fantastic work.  And, it’s going to be useful for some people.  In fact, just being able to do some of the layout and have the basic skeleton code generated, that would be worth it to ME, someone who otherwise wouldn’t need it.

If I was running the show at Palm, would I have built Ares?  Probably not and here’s why: I don’t really think it’s necessary.  The barrier to entry with webOS is already so low that I don’t think Ares buys you much there.  As far as source control and emulation and all that stuff?  Any developer worth their salt has those bases covered anyway.
Who is Ares really for?  I don’t think it’s really for someone like me who has a strong web development background already.  The SDK and Eclipse plug-in gives me all I really need.  No, I think it’s for people that aren’t as accomplished a developer.  If I was running the show, I think I would have had a different set of priorities, a different resource utilization map (I also got some inside information on this aspect of it, which I won’t divulge because it was a private conversation… suffice it to say some reasonable arguments were made in rebuttal to my thoughts here… one thing I will say is that Palm certainly seems to have a vision and isn’t just doing things for the sake of doing things… whether a given decision turns out to be right or wrong, good or bad, having a vision is a key to success).

And here’s where that could be a problem: Palm has said for a while that they want quality over quantity in the App Catalog.  We’ve already seen hints that they’ve strayed from that a bit… might Ares push that goal further away?  If less talented developers can easily develop for webOS, will that result in a ton more “craplets” out there?  Maybe.

My suspicion about Ares is that at some point down the road we’re going to look at it and say “wow, Ares was a cool product, a neat idea, but there wasn’t much point to the exercise”.  Now, I want to make sure I say here that I could very well be wrong!  I spoke to someone at Palm about this and he mentioned that Ares is a forward-looking product, they see it as something that will be appreciated more down the road.  I think that’s a fair thought, and it might even prove to be correct… it may at some point down the road be recognized as a stroke of genius.  As we sit here today looking at it though, I personally doubt it.

Still, like I said, Ares is very impressive from a technology standpoint and I very much want to commend the team working on it.  Even if it winds up being a bust in terms of real-world usage I do think it’s pushing boundaries and showing us what a new model of developer tooling might look like.  In and of itself, that’s very much worth something.

One last point before I go… I’m going to deduct a few merit points from Palm for one thing: STILL NO **USABLE** FREAKING AUDIO API FOR JAVASCRIPT APPLICATIONS?!? I’m getting sick of silent games, or games using the same set of system sounds that everyone else uses.  Please, for the love of all that is good in the world, audio API yesterday please!

1. Chris says:

   January 9, 2010 at 5:49 pm

   Wow! Really interesting post.

   I was feeling a little slighted by Palm, because I, too, bought into the CSS/HTML/JavaScript model that they presented as the future. It felt like Ares, and then the subsequent PDK announcement, erased (or diluted, at the very least) that vision. And I’m certainly no superstar developer. I say this as a person who was heavily involved in web design and development until around 2001 (the bad economy led me to a new career). I had been itching to get back into the game, and WebOS felt like the best opportunity when it came out. Now, I’m left feeling confused and concerned. And possibly stuck in my terrible accidental career.

   But enough about me. Your post confirmed my fears, yet somehow also made me feel better about it all. Let’s hope the PDK is indeed a “stopgap” until the JS APIs are there. And let’s hope Ares is taking us someplace better.

   Until then, I’m sticking with your book.
   

   Reply to this Comment
2. Palm at CES 2010 – 3D, Video, Pre and Pixi Plus and More | Pre 101 says:

   January 10, 2010 at 1:40 am

   [...] sure that the capabilities of the Canvas element have been fully realized by game developers.  As Frank Zammetti pointed out, there is a tremendous potential for applications developed with existing Web [...]

   Reply to this Comment
3. Ben Galbraith says:

   January 12, 2010 at 2:55 pm

   Chris,

   The PDK is an important piece to the puzzle, but it in no way lessens our commitment to the Web as the native platform for the device.

   Best,

   Ben
   Director, Developer Relations, Palm

   Reply to this Comment