Monday, December 31, 2018

Re: GWT CssResource not propagating width:calc(x) property

After migrating to gss, are you able to calc() without literals ?

On Friday, 19 February 2016 03:48:58 UTC+5:30, Stefan Falk wrote:

Oh!

Thanks Jens and Gilberto! I'm definitely going to check that out!

On Monday, 15 February 2016 14:40:05 UTC+1, Gilberto wrote:

Stefan, convert all your CSS files and <ui:style> declarations to GSS and save yourself from going crazy. With GSS you can use all the CSS (and CSS3!) goodies without pain (including media queries, CSS functions and so on).

Take a look at http://www.gwtproject.org/articles/gss_migration.html and http://www.gwtproject.org/doc/latest/DevGuideGssVsCss.html for hints on how to do the migration.

On Saturday, February 13, 2016 at 9:11:59 PM UTC-2, Jens wrote:

Unfortunately GWT 2.8.0 still does not support that feature. Had to use literal("") as well here.

You are still using Css right? According to https://github.com/google/closure-stylesheets/issues/59 closure stylesheet should support and thus GWT GSS should support it.

-- J.

--
You received this message because you are subscribed to the Google Groups "GWT Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to google-web-toolkit+unsubscribe@googlegroups.com.
To post to this group, send email to google-web-toolkit@googlegroups.com.
Visit this group at https://groups.google.com/group/google-web-toolkit.
For more options, visit https://groups.google.com/d/optout.

Friday, December 21, 2018

Generate user.agent independent code

I prefer to select the one single permutation: <set-property name="user.agent" value="safari"/>

--
You received this message because you are subscribed to the Google Groups "GWT Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to google-web-toolkit+unsubscribe@googlegroups.com.
To post to this group, send email to google-web-toolkit@googlegroups.com.
Visit this group at https://groups.google.com/group/google-web-toolkit.
For more options, visit https://groups.google.com/d/optout.

Wednesday, December 19, 2018

Re: Caused by: java.lang.NoClassDefFoundError: com/google/gson/JsonParseException ?

Ok, thanks for clearing that up.

Re: Caused by: java.lang.NoClassDefFoundError: com/google/gson/JsonParseException ?

For both versions the Maven pom.xml for gwt-servlet artifact declares an optional dependency to GSON. It is optional because it is only needed when you use StackTraceDeobfuscator on the server. So if you use that, you need to add GSON manually to your pom.

Maybe you updated some other libraries too which have included GSON transitively but now don't.

-- J.

Caused by: java.lang.NoClassDefFoundError: com/google/gson/JsonParseException ?

I just upgraded to GWT 2.8.2, that works fine.

I just noticed that in case the deobfuscator throws an exception as it can't find the GSON JsonParseException, why is that?

Caused by: java.lang.NoClassDefFoundError: com/google/gson/JsonParseException

I can't remember that was in GWT 2.8.0, but I am not sure. Neither do I see gwt-server include the gson automatically as dependency when running maven install.

I added it manually to solve it.

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

Yesss, the above solution worked.

Luckily it was the only construction I had, I only had to rewrite this one.

Sorry, there will not be any upgrade to 3.0

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

BTW: what is the correct way to solve this?

Should I make a new class that extends both JavaScriptObject and implements ZtoeslagAppConfig?

Yes, and getPages() would be implemented using JSNI. That also means that AppPage should be a JSO (and might implement in interface). Alternatively you could rewrite your code to use JsInterop annotations and jsinterop-base library. Less boilerplate and you have to do it anyways if you plan to adopt GWT 3.0 in the future.

-- J.

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

BTW: what is the correct way to solve this?

Should I make a new class that extends both JavaScriptObject and implements ZtoeslagAppConfig?

So it would be:

SimpleZtoeslagAppConfig extends JavaScriptObject implements ZtoeslagAppConfig {

}

private SimpleZtoeslagAppConfig getZtoeslagConfig() {

return (SimpleZtoeslagAppConfig) UtilsJsni.getProperty(getJson(), getMiscValues().zorgtoeslag());

}

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

Thanks for the inside Jens, I think that could be it, I will test it.

Tuesday, December 18, 2018

Re: Would a kind soul give me a hint with this error on client side serialization?

I can now answer my own question:

1) compile with "detailed" output script

2) open the page and activate Javascript Developer Tools (in Chrome, F12)

3) put a conditional breakpoint on the like "this$static[signature][2]" with expression

typeof this$static[signature][2] !== "function"

4) execute the remote call.

When debugger stops, check what "instance" variable is holding. This is your problem. Let's say it is the object "MyCustomSerializableClass"

Now, how to solve?

In your service interface, create a method

public MyCustomSerializableClass workaround(MyCustomSerializableClass mc);

In your async service interface, create a method

public void workaround(MyCustomSerializableClass mc, AsyncCallback<MyCustomSerializableClass> asyc);

And add to the my the implementation class

public MyCustomSerializableClass workaround(MyCustomSerializableClass mc) {
  return new MyCustomSerializableClass();
}

Thats it! Solved... And as bonus, GWT 2.6.1 dev-mode came to life again :-)

Hope this saves someone from trouble - took me 5 days 14 hours each to realize all this stuff (may be I'm just slow :-) )...

Regards,

Edson

Em segunda-feira, 17 de dezembro de 2018 15:33:42 UTC-2, Edson Richter escreveu:

I'm receiving the error "this$static[signature][2] is not a function".
I was able to track this error down to the client code:

SerializerBase.java:

native void serialize(SerializationStreamWriter stream, Object instance, String signature) throws SerializationException /*-{ this[signature][2](stream, instance); }-*/; }

What does that means? Would GWT tell me which class is offending the rules? I have literally 250 classes that would be serialized, and could not find the cause.
Also, I'm sure that the client is returning same objects that was send from server (client receives the object, set one BigDecimal value, and return same objects back to server).

Please, give me an advise!

Regards,

Edson

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

Hmm as you are using intersection types this might be related to https://github.com/gwtproject/gwt/commit/1ce1c710c7d3d7bffc246c7f013009b670be44a4 which has been committed between 2.8.0 and 2.8.2.

So it could very well be that now there is a cast check inserted by GWT because you use an intersection type as return type of your getter.

-- J.

Re: ER tool

Can I download these Lienzo based library if it is a open source?

Thanks,

David

On Thursday, December 14, 2017 at 9:10:58 PM UTC+8, manstis wrote:

We use Lienzo extensively over at Drools

Here's a video showing use of our Lienzo based library as a BPMN2 process designer: https://www.youtube.com/watch?v=rDqgTe-DB5E (all GWT and Lienzo).

Our Lienzo based library is also used for some other "diagramming" type of editors:

https://www.youtube.com/watch?v=qBgxVoc2qfw

https://www.youtube.com/watch?v=9wcUU4k0i9I

https://www.youtube.com/watch?v=q0hxyxJzcgI

On Thursday, 14 December 2017 12:10:18 UTC, James wrote:
I like to use GWT or Vaadin to draw ER diagram. I came cross JointJs as well as https://www.lienzo-core.com/lienzo-ks/. There is no GWT wrapper for JointJS. Lienzon seems promising to me. How do you think?

Thanks,

James

Monday, December 17, 2018

Would a kind soul give me a hint with this error on client side serialization?

I'm receiving the error "this$static[signature][2] is not a function".

I was able to track this error down to the client code:

SerializerBase.java:

    native void serialize(SerializationStreamWriter stream, Object instance,
        String signature) throws SerializationException /*-{
      this[signature][2](stream, instance);
    }-*/;
  }

What does that means? Would GWT tell me which class is offending the rules? I have literally 250 classes that would be serialized, and could not find the cause.

Also, I'm sure that the client is returning same objects that was send from server (client receives the object, set one BigDecimal value, and return same objects back to server).

Please, give me an advise!

Regards,

Edson

Re: A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

Anybody any idea please?

Saturday, December 15, 2018

Changes to our Terms of Service and Privacy Policy

Shannon George

We'll soon publish changes to our Terms of Service that may affect the Google services you use. Starting January 22, 2019, services offered under these terms will be offered by Google Ireland Limited instead of Google LLC. We're making similar changes to the Drive, Play, YouTube, and YouTube Paid Service Terms of Service. You can find our new terms here.

At the same time, we're also updating Google's Privacy Policy to specify that Google Ireland Limited will be the data controller responsible for your information and for complying with applicable privacy laws. These updates will take effect if you're based in the European Union, Iceland, Liechtenstein, Norway, or Switzerland (unless otherwise stated in a service-specific privacy notice). Our updated policy is here.

Nothing about your experience in Google services will change. And nothing is changing in terms of your privacy settings, the way your data is processed, nor the purposes of its processing as described in the Google Privacy Policy. However, if you don't want to accept these changes in our terms and Privacy Policy, you can choose to stop using the applicable services.

If you use Family Link to manage a Google Account for users below the age of consent, then please take some time to talk to them about these changes.

You can see and manage your data in the Google Dashboard. And you can always change your privacy settings and data controls anytime in your Google Account.

This email was sent to vaughnjjoan@gmail.com to provide an update about your Google Account.

Friday, December 14, 2018

Snail Mail: ❄️ Crypto Winter 🌬

Slow Ventures Snail Mail

Snail Mail: ❄️ Crypto Winter 🌬

Friday, December 14, 2018

Welcome to another edition of Snail Mail! Today we have a fresh take on the current state of crypto by crypto consultant, Jill Carlson.

Got forwarded this and want to subscribe? Go to https://slow.claims.

🍿Q+A: Jill Carlson on The State of Crypto

Jill, Seth Rosenberg and Slow's Alex Marcus recently wrote a book, The Existential Millennial. They were hanging last weekend and naturally Jill and Alex started talking crypto. The conversation turned pretty interesting so they figured they would share it with our readers! Enjoy!

Alex: The crypto bubble has popped. What is your take on the good, the bad and the ugly?

Jill: So the good thing is that the noise is gone. This time last year there was just so much confusion out there. Everyone wanted a piece of the action often for the wrong reasons. With all the hype came a lot of distraction. Now that it's quieter it is a lot easier to buckle down, focus, and build.

The bad, of course, is that the money is also gone. It's a lot harder for projects to fundraise -- even the good ones.

And this leads us to the ugly. Many projects did not adequately prepare for the boom times to end. They managed their burn rates poorly. They did not anticipate regulatory headwinds. They partied like it was 1929. This is now leading to layoffs at startups like Consensys and other mainstream projects folding.

AM: Most blockchain companies who accepted BTC and ETH when fundraising have lost serious amounts of capital in recent months. How are companies going to continue to drive growth with less runway and resources?

JC: A lot of founders got it in their heads last year that they would sit on the crypto they used to raise funds. They did this for any number of reasons... Some thought they could be crypto hedge fund managers while also running their project. Some thought it did not jive with their ideology to hold fiat currency (yes seriously). Of course really what they did was double down on their exposure to the space. It was the equivalent of starting a new social networking app and deciding to invest all the funds you raised into Facebook stock. Probably not a great idea to have that much correlated risk.

There are a lot of other problems that come along with this model as well. Making long term strategic decisions and even hiring become problematic as your burn rate suddenly becomes a moving target. Your runway might be swinging around 30% in a week.

The biggest issue, though, is that many of these projects don't have revenue strategies. The idea was that their ETH and the value of their own token would continue to rise over time and they wouldn't need revenue. That's turning out to have been shortsighted. It will be an interesting year ahead as we watch these free open source software projects experiment with finding a suitable business model.

AM: Despite the bear market some projects are making more technical progress than ever. How can those companies expect to successfully launch at the current market cap?

JC: The teams that are innovating will continue to do what they have always done, chipping away making technical progress. It doesn't take a $100mm ICO to be able to innovate and successfully ship a cryptocurrency or related product. Ethereum and Zcash both taught us this back in 2014.

That said, even the companies that are making real progress are going to have a real challenge when it comes to capturing the attention, hearts, and minds of their potential users. When cryptocurrency is going through a hype cycle, it's easy -- everyone can't wait to get their hands on what you have to offer. We see cryptocurrencies competing with each other in bull markets for market cap share. When the market slows down, however, the crypto world wakes up and remembers that coins are not competing with each other -- they are competing against the status quo.

AM: The bear market has hurt both builders and investors. But investors are seeing deals at better prices with more power over deal terms. Is now a good time to be an investor in projects (not crypto)?

JC: Yes. There was an odd dynamic throughout the middle of this year in which the prices of publicly traded cryptocurrencies had retraced dramatically from the highs (e.g. Bitcoin went from $20,000 to $6,000) but private deals and SAFTs had not yet repriced and mega rounds were still getting done. I'm glad those days of dislocation finally seem to be behind us.

There's nothing like investing in a down market! And yet few seem to want to. The very fact that investors are in a more competitive position now speaks to how many have left the crypto market (at least until the next hype cycle returns).

AM: There are still no blockchain applications with real traction. Will 2019 bring our first popular dapp or will it be a year during which we continue to build out protocols and infrastructure?

JC: Crypto has a problem. The problem is that many of the great minds in the space have a tendency to be optimizers. They are absolutely brilliant when it comes to looking at something and saying "oh if I tweak this then I can make that one thing better." We see this play out even in the nicknames projects are given -- Web 3, Ethereum 2.0, "a better Tether", etc.

There are not nearly enough people, however, looking at protocols and possible applications from first principles. There are few who start with a problem statement and work to create the best product to solve that problem. Until more innovators start to do this, we won't see a dapp or product really gain traction.

I think we are stuck in the protocol and infrastructure phase a little while longer. And when we finally do see a dapp or product take off, I suspect it won't look like anything we are anticipating today. I think it might look pretty mundane, actually, and we might not even realize that it's using something called a blockchain in it's backend.

Need a last second holiday gift for your _______ millennial? Get a copy of The Existential Millennial. It is a 135-page, hardcover Mad Lib Book for Millennials with illustrations from the talented Shui Pei Luu.

All proceeds are going to suicide prevention and brain health non-profits.

😋 Short Thoughts and Novelties...

Speaking of Millennials...they are leading the shift away from meat to other plant-based alternatives. Almond milk, cauliflower rice, gluten-free bagels, cashew cheese...as Mutagu from Zoolander might say - "Plants are so hot right now."
A year and a half after Trump reversed the Obama administration's decision to block the TransCanada's Keystone XL pipeline, a federal district judge in Montana has halted the project again. Meanwhile Shenzhen is officially the world's first city with a fully electric bus fleet and soon all 22,000 of their taxis will be too!
We are huge fans of human + machine systems. We even threw a conference focused on the topic last week (content from that to come!). Apple News might be the AAI app that touches the most people in the world. It is read regularly by roughly 90 million people!
Slow portfolio company Fragments has changed its name to Ampleforth after the the "1894" character whose job is to translate poetry into Newspeak. Ampleforth aims to preserve the unit of account and money properties of its tokens. Instead of pegging to a fixed supply of fiat capital, Ampleforth's Amples have an elastic supply based on demand.

Want moar? Check us out on:
Facebook, Twitter, Medium or our Website

Signup for Snail Mail Newsletters at http://slow.claims

Our mailing address is:
1006 Kearny St, San Francisco, CA 94133

Don't Want These Emails?
unsubscribe from this list

Donate BTC: 3BJW4B6GGpoQrjeom6RpVtkza3XPw2qjoK
Donate ETH: 0xD7599b3D15805aDF3144676914964e8fff53C925

Thursday, December 13, 2018

A class cast exception when upgrading from GWT 2.8.0 to 2.8.2 ?

I just upgraded from GWT 2.8.0 to 2.8.2 and get a class cast exception which I don't understand, please some help.

It goes wrong in the following line of code:

ZtoeslagAppConfig this.bla = new JsoZtoeslagAppConfig(getZtoeslagConfig());

All relevant code:

private <T extends JavaScriptObject & ZtoeslagAppConfig> T getZtoeslagConfig() {

return (T) UtilsJsni.getProperty(getJson(), getMiscValues().zorgtoeslag());

}

private JavaScriptObject getJson() {

return this.json;

}

public interface ZtoeslagAppConfig {

AppPage[] getPages();

}

public final class JsoZtoeslagAppConfig implements ZtoeslagAppConfig {

private final JavaScriptObject jsoConfig;

public <T extends JavaScriptObject & ZtoeslagAppConfig> JsoZtoeslagAppConfig(final T jsoConfig) {

super(jsoConfig);

}

public AppPage[] getPages() {

return UtilsJsArrays.toArray(getJsoPages());

}

private JsArray<? extends JsoAppPage> getJsoPages() {

return UtilsJsni.getPropertyAsArray(getJsoConfig(), getMiscValues().pages());

}

private JavaScriptObject getJsoConfig() {

return this.jsoConfig;

}

During super dev mode, in Chrome the values of this.json above:

json_1_g$:

newsletter: "825eceb44f"

zorgtoeslag: {payments: Array(3), pages: Array(3)}

Just before calling the constructor in:

new JsoZtoeslagAppConfig(getZtoeslagConfig());

It is calling a gwt method for casting that seems to fail:

static Object castTo(Object src, JavaScriptObject dstId) {

checkType(src == null || canCast(src, dstId));

In this method, I see that dstId has the value 1891, and src is some json value.

When performing the above body it calls the following method, that I don't directly understand as I don't see the method call in the above function:

static native boolean jsEquals(Object a, Object b) /*-{

return a == b;

}-*/;

I think this method fails as a is some json and b has the value null.

What goes wrong here?

Bazel rule for gwt dev-mode

Hi All,

In our project, we want to run gwt dev-mode via bazel. We use this rule for building gwt for production, but it is not very useful for our dev-mode because of our project structure. In our project, we have more than 100 source roots, and each source root could contain several bazel packages. So, we have to add more than 100 java_roots in rule definition, which is not very well.

Also in our project, we have several different gwt application. Even if we put all our sources in one source root we still will want dev-mode to see only sources which are in deps for this gwt application.

So putting the entire source root on classpath does not sound like a good idea.

Right now we wrote our own dev-mode rule, in this rule, we collect all sources on which gwt application depends via bazel aspect and then put symlinks to all these source files into one folder and add this folder to the classpath.

It works very well, we have source updates extremely fast, but we still have one issue - we do not see new files (e.g. files which had been created after dev mode was started).

Does anyone have any ideas, how we could solve this issue? Or maybe someone knows a better way to run dev-mode via bazel?