Is it easy for the less experienced programmers? Yep.
Does it cut down on cross-browser problems? To some extent, but there are a couple of gaping holes, e.g. standardising text-size and the broken box, but you'd have that anyway.
Is it a good solution for any size of project? Some debate on this, but MHO is that it is great for small web sites but currently not able to cope with large sites on IE. Some people say they have yet to have this proved to them, but my site is getting slow to load on IE and it's not that big. It's a fairly simple concept to allow modularisation (and thereby allow secure and non-secure areas validated serverside) but I imagine it would mean a fairly significant change to the GWT architecture, so whether we get that is anyone's guess, and the Google team are traditionally just a tad tight-lipped about future plans. Like Tony Blair said "I don't make predictions. I never have, and I never will." :-)
But it's still in beta... So what, so is Gmail. Tha beta tag means nothing. Would saying it's gone gold make it work better? Of course not. Just evaluate it and make a decision.
But the compiler isn't open-source... Most things aren't but they still get used.
And will GWT be free for ever? Well this version will :-) But look at what Google give away for free right now. They won't charge for that or everyone will switch away, and the cost of GWT over the cost of the rest of what they do pales into insignificance, and think of all the programmers and companies who would never trust Google again if they did? Think of the loss of goodwill. But Google aren't going to promise fee jam today, free jam tomorrow and free jam for ever, are they?
Would I recommend it? For smaller sites and sites that can be broken into discreet areas of that size or smaller, yes. If they fix the modularisation for larger projects, then yes for them too.
But whatever they do, there will always be times where another solution is more appropriate.
Hi, I'm using GWT for an intranet application. I am currently using 1.1.10. I have not upgraded to the latest because this release is stable for my needs. At some point, I'm sure I will upgrade, but that isn't a major concern for me. I'm not really worried about the compiler being closed because I use other closed technologies as well. I wouldn't have a problem using GWT for a public application (if I had a public application to do). I don't know what to tell you as far as how to pitch this to your team. I found GWT and started using it and my results with it was enough of a pitch for my team.
HTH, Chad
On 12/5/06, Adrian Marti <marti.adr...@gmail.com> wrote:
> thanks for the reply, anyone else?
-- There are two types of people in the world: * Those who need closure
Adrian Marti wrote: > Basically i need to assure the group that GWT is going to be around for > a little awhile and that we can start using this technology now.
This is a wise question, and one that has been asked a lot before, because it seems too good to be true. But it isn't...it's just simply true. GWT is free, and it will remain so if not get even freer. See Bret Taylor's post on a similar thread: http://groups-beta.google.com/group/Google-Web-Toolkit/browse_thread/...
Not to mention that the GWT team is growing quickly with some really fantastic engineers.
> What exactly does RC mean with respect to GWT being Beta (perpetual > beta?)?
It's definitely weird to have an RC release within a Beta product, but I assure you that GWT won't be perpetually "Beta". With the big adoption we've seen, it's hard to keep in mind that GWT has only been available for less than 7 months :-) Coming out of Beta is one of the top things on our mind.
> Is this tech production ready? Howabout just for an internal app where > users may be a little more lenient? Anyone using it for real business > apps?
Here's how I would think about. Even when there are bugs in GWT, choosing an approach other than GWT invites your team to write a whole host of other kinds of bugs in your own code that GWT would make impossible (or at least very unlikely), including all sorts of subtle JavaScript errors and browser-specific DOM errors. Not to mention that it's much harder to debug JS than it is to debug Java source with GWT. In other words, deciding against GWT in favor of writing tons of JavaScript by hand isn't really a way to mitigate risk overall, it's just trading one sort of risk for another. However, by choosing GWT, you can benefit from the fact that there's a strong team of extremely devoted full-time Google engineers working on your behalf to fix bugs and create big new features that will benefit your project in both the short and long term.
> Is the closed-source compiler a sticking point for anyone else? How can > i convince others to just trust Google to fix bugs etc.?
First, there are really very few compiler bugs, and even fewer that are likely to affect you in normal work. Second, we take compiler bugs very seriously since right now you cannot easily fix them yourself. So, fixing compiler bugs is one of our highest prioritiest, arguably *the* highest. Beyond that, all the real surface area of GWT is in the user libraries, which are already open source. The source is included in gwt-user.jar, and it's very easy to understand, debug, and patch that source yourself -- which can be useful if you run into a bug that is in your critical path that we won't be able to fix in a timeframe that works for you.
We never want you to be stuck waiting on us, and we explicitly prioritize our own work to keep the product in a state where teams like yours can, in the worst case, help themselves.
> I am really enjoying using the toolkit and appreciate all the work > being done.
Thanks! We are really happy about the architecture, and we think we can extend it to add some very exciting enhancements, many of which have sprung from some of the constructive discussions that occur on this group.
> Adrian Marti wrote: > > Basically i need to assure the group that GWT is going to be around for > > a little awhile and that we can start using this technology now.
> This is a wise question, and one that has been asked a lot before, > because it seems too good to be true. But it isn't...it's just simply > true. GWT is free, and it will remain so if not get even freer. See > Bret Taylor's post on a similar thread: > http://groups-beta.google.com/group/Google-Web-Toolkit/browse_thread/...
> Not to mention that the GWT team is growing quickly with some really > fantastic engineers.
hi bruce,
there is, and there will continue to be a huge cloud of fear, uncertainty and doubt cast over gwt. imho your improved features, team size, team talent, regression test suite and imminent open source of compiler/hosted mode env will not move these clouds away any time soon.
of course there is one event that can eliminate all FUD:
the next release of GMail client powered _entirely_ on gwt. once this happens the gwt team will be forced to be agile, be more flexible with their architecture, forced to listen and respond to the customer.
if and when this event ever happens, the industry will instinctively know that the future of gwt is in good hands.
pls eat your own dog food on one of your largest investments and FUD will be no longer.
> there is, and there will continue to be a huge cloud of fear, > uncertainty and doubt cast over gwt. imho your improved features, team > size, team talent, regression test suite and imminent open source of > compiler/hosted mode env will not move these clouds away any time soon.
What a downer comment. I'm having a hard time understanding where this is coming from. Aren't you being a bit dramatic here? What makes you say there's a huge cloud of FUD? I mean, seriously, we're working very hard to give you a great (imho) tool for free. What's the problem?
> the next release of GMail client powered _entirely_ on gwt. once this > happens the gwt team will be forced to be agile, be more flexible with > their architecture, forced to listen and respond to the customer.
Why is this your litmus test? And why isn't your own experience with GWT (and everyone else's) enough to convince you that GWT is worthy of your trust? To draw an analogy, did you wait for Sun to rewrite Solaris in Java before starting to develop in Java?
> if and when this event ever happens, the industry will instinctively > know that the future of gwt is in good hands.
> pls eat your own dog food on one of your largest investments and FUD > will be no longer.
Who is perpetuating the FUD you're talking about? We're just trying to provide a very useful tool to web developers. That's it.
> To draw an analogy, did you wait for Sun to rewrite Solaris
in Java before starting to develop in Java?
That would be more like writing the google search engine, clustered FS and adsense in GWT. Ash is just talking about a parade product everybody knows, like GMail. When I go to a customer and praise GWT as an argument for fast and stable development, he'll ask me 'and who else is using it?'. I can't answer much to that. And when he asks me 'and why isn't google using it?' I've even less to counter with.
I trust GWT and have switched most of my web projects to it, why don't you Bruce?
Surely a better way to look at this is that applications like GMail and Docs & Spreadsheets are actually driving GWT forward? The knowledge and technology that Google has invested in the development of those applications is the foundation of GWT.
Making GWT a general purpose framework is necessarily limiting in some respects whereas the very thing that makes an application like GMail great is that Google has innovative engineers trying out new things. If those applications are constrained by using GWT then you'll probably see less innovative stuff propogating into the framework, or at least it will slow it down.
I do agree that it would be nice to be able to point to a reference application but I don't agree that Google switching their products to GWT would be beneficial to the near term development of the framework, nor would it bring any compelling arguments for uptake of the framework on any specific project - citing Google as using a Google product is not a convincing argument for the long term support and development of the product as free and open source.
> What a downer comment. I'm having a hard time understanding where this > is coming from. Aren't you being a bit dramatic here? What makes you
this is coming from a gwt evangalist that advocates gwt at a number conferences/events. perhaps it is a overly dramatic, but there is a deep seeded msg in here that has fortunately hit a nerve.
> say there's a huge cloud of FUD? I mean, seriously, we're working very > hard to give you a great (imho) tool for free. What's the problem?
im not debating whether ur team is working hard or not. nor am i debating whether the tool is free. the problem is the question "which non-trivial application does google use gwt for?"
> > the next release of GMail client powered _entirely_ on gwt. ...
> Why is this your litmus test? And why isn't your own experience with
welcome to my reality. like george, i get challenged by a number of people on "the litmus test". i dont understand why u arent repeatedly challenged with the same issue.
> GWT (and everyone else's) enough to convince you that GWT is worthy of > your trust? To draw an analogy, did you wait for Sun to rewrite Solaris
frankly, gwt earned my trust when Scott Blum[1] provided me with some personal support when i was looking down the barrel some time ago.
> in Java before starting to develop in Java?
no, but i did wait for the collabnet boys to maintain the subversion codebase under subversion before i and a mountain of other people managed their investments/assets under that configuration management tool.
two things earn my trust: 1. someone helps me with an issue that i was unable to solve by myself 2. when someone eats their own dog food on their own critical asset
> > pls eat your own dog food on one of your largest investments and FUD > > will be no longer.
> Who is perpetuating the FUD you're talking about? We're just trying to
im telling you about a consistent feed that i have been getting. go ahead, validate this with other gwt evangelists in other countries. im sure this is not a unique case.
imho the FUD is an effect, caused by a lack of trust internel to google.
> provide a very useful tool to web developers. That's it.
i never questioned the usefulness of the tool. i have been smoking gwt since release. i have delivered 4 gwt applications. i have injected byte-code into the gwt-compiler to get around initial compiler defects. i know the good, bad and the ugly about gwt. we could go on and on, however, these are not the points for this discussion.
what i am trying to do is give u a viewport into my reality, one that is not living in denial.
I'm a bit more upbeat about GWT, I jumped on to try it out almost at day 1, and still think that it's a great product. It has it's downsides, but for me (a JAVA programmer, familiar with Swing but not that familiar with web programming) it's been almost perfect, I get to re-use a lot of my Swing experience and it all fits together nicely.
Getting back on topic though, I have to (at least partially) agree with some of these guys. GWT's reputation suffers (rightly or wrongly) from there being no highly visible Google apps written in it. I hear your argument about Solaris and Java, but I think it's not valid. Java was not designed as an operating system language, so why would you write an operating system in it? GWT on the other hand was designed for AJAX development, like Gmail. I can understand why you wouldn't take the time to re-write Gmail in GWT (that would be crazy), but it really would do you guys the world of good if you started using GWT for future Google projects. It would lend it credibility, and probably add to it's stability, ease of use, and widget libraries and API.
Don't get me wrong, I love GWT for so many reasons, but it's reputation in the community (IMHO) is not that great. I would be interested in your thoughts about these things...
My Solaris-in-Java example was dumb. I take full responsibility for the failed analogy :-)
The actual point I was trying to make was that you don't rewrite existing products like GMail in the blink of an eye, and since GMail already does work well, there's not enough incentive to rewrite it in GWT in the near future.
GWT has great support within Google, and there are many products in development at Google with GWT, but just don't be disappointed if a GWT-based GMail doesn't appear very soon. As this thread has expanded, I now understand that you didn't literally mean GMail, but that you're looking for more examples of successful products built with GWT. That makes perfect sense, and these are starting to show up and will grow over time. Other than the example you probably already know about (http://images.google.com/imagelabeler/), not that much has shipped from Google yet.
Thanks for taking the time to address my concern directly. I can definitely use direct support from the tech lead of the product i am trying to push :) . I think we are all on the same side here, some of us just having a little tougher time pushing this brand new technology, no matter how excellent/exciting/promising it is.
Bruce: I'm glad we understand each other a bit better now. Don't get us wrong, we are generally huge fans of what you guys have done with GWT (my company has already built 3 fairly large products using it).
I guess we're all looking forward to the day we can point to some well known sites on the internet and say "Look, this is built in GWT. It's fast, user friendly and cross-browser compatible". I think that's the main thing the community is lacking at the moment. Our hope was that perhaps there would be a new product from Google some day that could show it off. Something on the same scale as Gmail, Google Maps or Google Spreadsheets.
