Upgrade to Maps v3

[GoogleCodeExporter]

I would like to see support for Maps v3.

Original issue reported on code.google.com by dusan.ma...@gmail.com on 4 Jun 2010 at 12:23

[GoogleCodeExporter]

In the mean time, see:

http://code.google.com/p/gwt-google-
apis/wiki/MapsFAQ#How_can_I_use_GWT_with_Maps_API_3.0?

Original comment by zundel@google.com on 4 Jun 2010 at 1:03

• Changed state: Accepted
• Added labels: API-Maps

[GoogleCodeExporter]

I'm fully aware of that page, but we use quite a large set of API features. Is 
there any schedule/roadmap, or some 
rough approximation on when it will be available? Is there anybody working on 
it?

Original comment by dusan.ma...@gmail.com on 4 Jun 2010 at 1:13

[GoogleCodeExporter]

There is another open source project out there that has support for a good 
number of features you can use:

http://code.google.com/p/gwt-google-maps-v3/

Original comment by zundel@google.com on 19 Jun 2010 at 3:14

[GoogleCodeExporter]

Now that v2 has been deprecated, I would have thought this would be raised to 
'high' priority? 

It looks here like we are being encouraged to use a third party api over the 
official one.

Would love to get a date on the official api release.

Original comment by tculs...@gmail.com on 25 Aug 2010 at 11:04

[GoogleCodeExporter]

gwt-google-maps-v3 has also been deprecated. This is the message on the project 
page.
"This project will soon be merged into gwt-google-apis and hence is 
depricated".  It would be interesting to get the timeframe.

Original comment by brendanp...@gmail.com on 26 Aug 2010 at 5:49

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Is there any way to track the v3 integration advancement ? 

Since the process for v3 integration is quite long, i would like to be able to 
see at least the shape of the api, i like to work on my code base with future 
migration to v3 in sight, to at least avoid introducing too much compatibility 
problems. 

I'm already pretty worried about the v2->v3 migration, and i am not comfortable 
at all with the silence on this issue.

Original comment by nicolas....@gmail.com on 6 Jan 2011 at 11:11

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

There is no change in status since the FAQ entry (mentioned above as the second 
post in this issue) and several posts on the gwt-google-apis forum.  We are 
updating gwt-google-maps-v3 in a change branch while leaving the 
gwt-google-maps-v3 project stable.

Original comment by zundel@google.com on 6 Jan 2011 at 3:00

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

I really need this!

Thanks,
Brandon Donnelson
http://gwt-examples.googlecode.com

Original comment by branflak...@gmail.com on 6 Feb 2011 at 4:07

[GoogleCodeExporter]

Yes. Uncertainty is a poison pill for developers. Please keep us up to date on 
the path for the version 3.

Original comment by terry.je...@gmail.com on 24 Feb 2011 at 3:59

[GoogleCodeExporter]

I second that, in fact the lack of upgrades here (for me) starts to beg the 
question, if integration is such a big deal, why should developers be 
(considering) using GWT at all. Don't get me wrong, I'm a big GWT fan, BUT if 
we start having issues even with Google (!!) APIs, I think we have a problem. 

Original comment by tculs...@gmail.com on 24 Feb 2011 at 4:43

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

I will comment again on this thread and say that there is no timeline for 
incorporating the Maps v3 bindings into gwt-google-apis.  You can use other 
third party maps v3 bindings or write your own bindings using GWT's built in 
JavaScriptObject and JSNI support.  You are welcome and encouraged to star this 
issue to vote your support.  I maintain the library and I know it is a popular 
request, but adding more comments on the same line obscures my previous updates 
to the status of the problem.

Original comment by zundel@google.com on 24 Feb 2011 at 2:07

[GoogleCodeExporter]

Ok, but you, as a maintainer, are obliged to let your users know what's the 
status. If you say there's no timeline, you should give a reason why there 
isn't one, and when it will be. If there is no interest in this project any 
more from maintainers' side, please let us know, so that we can drop our 
dependencies on gwt-google-apis, or even GWT. That would be a fair behavior, I 
don't see you doing that, all you say is "go honk yourself" (that's what you 
call "use other third party maps v3 bindings or write your own bindings using 
GWT".

Original comment by dusan.ma...@gmail.com on 24 Feb 2011 at 2:17

[GoogleCodeExporter]

shame on google ..., but there is 3rd party implementation that works pretty 
fine for me. Check it out. http://code.google.com/p/gwt-maps3/

Original comment by fili...@gmail.com on 24 Feb 2011 at 2:30

[GoogleCodeExporter]

I do not intend to condone honking yourself or others.

Let me try to clarify and summarize from conversations in the group:  This 
issue, number 381 is a request for a new feature as far as the gwt-google-apis 
library is concerned.  Yes, it is something we'd like to add, but there is no 
one working this new feature.  Please do not make future plans expecting it to 
arrive at a certain time because there is no timeline for fixing it or the 
other outstanding new feature requests.

We had a change branch going for a while (see the SVN repository under 
changes/vinays), but now no one is available to work on it.  That change branch 
relies on the JSIO library which is much less efficient than using 
JavaScriptObject overlays.  We were trying to get rid of JSIO and tease out 
other inefficiencies.   There have been some ideas tossed around for 
automatically generating an API from the documentation, but I've only done a 
proof of concept - there are complexities in mapping class hierarchies and 
parameters that are closures in JS to Java constructs.  Its pretty simple to 
look at the design and do it by hand, but difficult to write an automatic rule 
for.  And even if all that were done, to be an official library, we'd need unit 
tests, documentation, examples...

We are maintaining existing features of the API.  It is an open source project 
and if someone wants to make a contribution, let's talk about it on the 
gwt-google-apis group.  

There is more than one third party wrapper checked in to Google Code.  If you 
would like some help wrapping any JavaScript API there are many examples you 
can look to on Google Code and a vibrant GWT community.

Original comment by zundel@google.com on 24 Feb 2011 at 2:50

[GoogleCodeExporter]

Let me add one more comment.  Some developers might be waiting for a 
gwt-google-apis official set of bindings because they hope to be able to 
upgrade to v3 without changing the API.

The Maps v3 API is incompatible with the Maps v2 API and any attempt to make 
new GWT bindings will likely break source and binary compatibility as well.  
The v3 API was carefully crafted to be lightweight enough for use on mobile 
devices, and I'd like to see any official GWT bindings remain true to that 
design goal. 

Original comment by zundel@google.com on 24 Feb 2011 at 3:01

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

I can't speak for others, but the reason I'm looking forward to an upgrade is:

* Maps v2 is deprecated.
* Maps v3 does not require API keys.

I'm perfectly fine with the newer API breaking backwards compatibility.

Original comment by cowwoc...@gmail.com on 24 Feb 2011 at 3:04

[GoogleCodeExporter]

I removed a few comments requesting a release date for this feature (there is 
none) to cut down on the clutter in this issue.  

If you would like to participate in the discussion about contributing to the v3 
bindings that are in changes/vinays, see the groups discussion below.

https://groups.google.com/forum/#!topic/gwt-google-apis/4e9GU638KCc

Original comment by zundel@google.com on 24 Feb 2011 at 3:32

[GoogleCodeExporter]

My organization has made a few web applications
with GWT and the official Maps V2 library.
We also have made mobile/HTML5 apps with GWT,
but not yet with Maps.

I just went to a Google Earth Outreach program where the
presenters emphasized that the Maps V2 API was deprecated,
and that we all (developers in the audience) should switch
to Maps v3. After hearing that, I am hesitant to put much more effort
into GWT + Maps V2, much as I like it.

Over the past two days I have tried using this:
  http://code.google.com/p/gwt-google-maps-v3/
in a new application. I did get it to work, but had
to stop when I came up against the lack of support for
(what were called in V2) TileOverlays. Our map
applications rely on a custom tile layer which appears 
on top of the Google Map Type

Here is a concrete question: I am wondering if it
is worthwhile to keep going with
  http://code.google.com/p/gwt-google-maps-v3/
and try to add the needed tile overlays with
JNDI and ImageMapType. Either that or,
again with JNDI, try to add a V3 KMLLayer
that duplicates the effect of the custom tile layer.
Other than that, at the moment I do not see a
migration path from Maps V2 to Maps V3
for our GWT applications.

Any advice appreciated.
(Another thing they said at the Google Outreach program
was this: use Google Groups to say what features
you want.)

Original comment by jhmalp...@gmail.com on 27 Feb 2011 at 8:26

[GoogleCodeExporter]

What is the status of this issue? 

Original comment by racam...@gmail.com on 23 Sep 2011 at 7:26

[GoogleCodeExporter]

With the new limits & pricing that will be introduced on introduced on October 
1st, 2011 this issue should definately get more priority. See pricing: 
http://code.google.com/apis/maps/faq.html#tos_pricing

Google will charge 2.5 fold for using the Maps v2 yet at the same time they do 
not offer an official GWT Maps API for v3. IMO there should either be a v3 api 
for GWT or using v2 should not be punished by making v2 users pay 250% of the 
v3 price.

Original comment by simon.w....@gmail.com on 23 Sep 2011 at 7:54

[GoogleCodeExporter]

Now that Google Maps v3 supports poly editing 
(http://code.google.com/p/gmaps-api-issues/issues/detail?id=2432&can=1&q=poly%20
editing&colspec=ID%20Type%20Status%20Introduced%20Fixed%20Summary%20Stars%20ApiT
ype%20Internal), can we expect some forward movement with associated GWT 
bindings? I would think that poly editing was the big, missing GMap feature 
that was making the development of GWT bindings seem premature.

Original comment by cdwo...@gmail.com on 17 Nov 2011 at 8:41

[GoogleCodeExporter]

I have the feeling Google has little interest in further developing GWT 
bindings for their APIs. We've been begging for this to happen for a year and 
we've gotten nothing. In fact, it makes me very nervous using GWT in future 
projects. I feel like it could be killed off at any second like Wave was...

Original comment by ad...@dualsportmaps.com on 17 Nov 2011 at 8:53

[GoogleCodeExporter]

I second that.

Original comment by jones.no...@gmail.com on 17 Nov 2011 at 9:05

[GoogleCodeExporter]

We dropped GWT completely and I'd advise others to as well before you get too 
far in, as this does not bode well for anything GWT.

Original comment by e...@centigonsolutions.com on 17 Nov 2011 at 9:51

[GoogleCodeExporter]

I agree. I am seriously looking into replacing GWT with play / coffeescript.

Original comment by jacques....@gmail.com on 18 Nov 2011 at 3:08

[GoogleCodeExporter]

Guys, be reasonable. Would you blame GCC team for lack of development in libpng 
? There simply is not big enough market for GWT Google APIs. GWT itself, afaik 
is keeping up quite well. It has HTML5 features, and the core of it - 
cross-compilation works very well.

Original comment by dusan.ma...@gmail.com on 18 Nov 2011 at 3:13

[GoogleCodeExporter]

I agree with Dusan. Guys, just use gwt-openlayers and you can use any Map API 
you like. Replaying GWT with something that has no IDE support is ... eh .. not 
very smart.

Original comment by LinuxC...@gmail.com on 18 Nov 2011 at 3:15

[GoogleCodeExporter]

Google Maps v3 was released May 2009; 2.5 years later, GWT bindings still do 
not exist. That is not a subtle hint, that's an emphatic "NO" from Google. 

Original comment by r...@simplesmartads.com on 18 Nov 2011 at 3:18

[GoogleCodeExporter]

I don't think this has anything to do with the market size of GWT Google APIs. 
I think it says more about the GWT team not being sufficiently funded to do 
some of the basic things, probably because of the market size of GWT. They been 
given enough funding to continue developing the core product but they are not 
being given enough funding to do any more than that. At some point in the past 
they did have the funding to make GWT bindings for their APIs but I guess the 
product didn't take off as they hoped and so it's slowly being drained of life. 
It's a serious shame because I really cannot stand working with weakly typed 
scripting languages for products that will have a long shelf life. They are 
"easy" to get into but become a nightmare to maintain for the same reasons they 
are easy. For those of us who are investing for the long term, GWT is (or was) 
a great product. The absence of API support from Google speaks loud and clear 
about their commitment to its future though. I'm worried because I've got a 
huge investment in GWT and I think it's about to circle the drain.

Original comment by ad...@dualsportmaps.com on 18 Nov 2011 at 3:42

[GoogleCodeExporter]

You might have noticed that http://www.google.com/flights/ is built with GWT 
and uses Maps v3. So it might rather be that they prefer doing it in JSNI than 
with the gwt-google-apis (I haven't looked at the compiled JS code closely, so 
I really can't tell). Either that, or they use the 
changes/vinays/gwt-google-maps-v3 branch or the gwt-google-maps-v3 project (or 
they completed the port and/or wrote a new one internally that they haven't yet 
open sourced).

One thing I can say for sure though (it's been said publicly on the GWT forum, 
and reinforced privately to me too): Google uses GWT heavily internally, on big 
projects; and the GWT community is vibrant. So please stop throwing FUD on the 
future of GWT.

Original comment by t.broyer on 18 Nov 2011 at 4:40

[GoogleCodeExporter]

Rather than telling us that Google really cares about GWT, perhaps you could 
push our concerns up the chain and try to secure some funding for creating GWT 
bindings for your APIs. It should be obvious at this point that your user base 
is getting worried about your commitment to the project because of this 
oversight. 

You cannot make people stop worrying by telling us everything is fine. We see 
smoke, so its natural for us to believe there is a fire. You could just put out 
the metaphorical fire, by creating the bindings, rather than just insulting us 
by saying that smoke is a normal part of life. 

We are the vibrant GWT community. We are scared. You need to know about it or 
there won't be a vibrant community. This isn't FUD, its an intervention. We are 
worried about you.

Original comment by ad...@dualsportmaps.com on 18 Nov 2011 at 4:51

[GoogleCodeExporter]

Ridiculous, you're bragging about lack of bindings for Maps API, which is JS. 
There is well documented and heavily used way of using JS libraries in GWT. I 
use Maps v3 in my GWT project thanks to gwt-google-maps-v3, and yes, it wasn't 
updated for more than a year. Guess why? No one missed anything in it. If you 
miss anything, then please stick with the OSS principles, and as a developer 
implement it yourself, and submit a patch. Apparently you're a developer, so 
please stop sniffing the smoke, and join the project. Please read again a 
comment #20. Good luck

Original comment by dusan.ma...@gmail.com on 18 Nov 2011 at 5:02

[GoogleCodeExporter]

Comment #11 is why people don't want to rely on gwt-google-maps-v3. It 
explicitly says that the final version will be different than 
gwt-google-maps-v3. I try to be careful when I choose my dependencies, and use 
official channels when I can, especially if it is explicitly told me that 
everything will be rolled out as normal.

The fact is, this issue is still opened instead of being simply marked "won't 
fix", while no informations nor advancement is given. It if would have been 
closed, some would have just whine a little instead of raging, and everybody 
would have jumped happily using gwt-google-maps-v3.

Original comment by nico...@justmagic.com on 18 Nov 2011 at 5:22

[GoogleCodeExporter]

A complete rewrite of all Maps related stuff seems like a pretty large burden 
to place on users. I know when I started using the Maps V2 API I was expecting 
that the V3 API bindings would come out and be a simple swapping of jars in the 
project (minus the API key). I've not used this library you speak of because as 
its not a Google project. I like to be conservative when choosing my 
dependencies. 

It's probably naive of me to expect backward compatibility, even from Google. 
Just switching from GWT 2.3 to 2.4 crashes all the RequestFactory clients I 
have in the field. Yet another significant burden placed on users. I recall all 
the inculcation in university on the important of backward compatibility. I 
guess that message didn't make it to the RequestFactory folks. Considering 
comment #20's suggestion that the likely way the API would be implemented would 
be through some automated process, that makes it unlikely that it will be 
backward compatible either. As this threads status is still "accepted" I'm 
hoping V3 API bindings are create officially and that backward compatibility is 
on their mind. 

It would be nice if Google would remove the uncertainty from this situation and 
post up a timeline and expectation for completion or tell us "we are not in the 
business of GWT binding creation anymore". Either way, our pathes forward would 
be easier to navigate. 

Original comment by ad...@dualsportmaps.com on 18 Nov 2011 at 5:41

[GoogleCodeExporter]

Up until a couple days ago, Maps V3 was nowhere near the functionality of
Maps V2. And the Maps team can attest that getting V3 to this point took
way longer than they had planned. The guys writing the GWT Google API
bindings are beholden to the milestones of the individual API teams (didn't
I just see preliminary bindings released for Google+?). Why would any of us
prioritize bindings to an API that's very unfinished, especially when the
existing V2 bindings suffice and the deprecation period is pretty far from
over, at least in Google time?

Backward compatibility aside, I'd like to hear more about zundel's idea of
dynamically generating the bindings from documentation. Without
understanding or looking too deeply, this process sounds very reminiscent
of what the Discovery API does. Is there any more discussion of this
concept, vis-a-vis generating Maps bindings? Sounds like the last piece
that's needed is the Maps guys to finish a JSON schema, since I don't see
it in the API Explorer. If that is indeed the case, we should be starting
(and starring) a thread over at Maps to request expedited movement on the
Maps API JSON schema.

Original comment by cdwo...@gmail.com on 18 Nov 2011 at 6:08

[GoogleCodeExporter]

I've begun writing my own maps v3 wrapper. For instance some of my notes today 
on the api loading routine. http://c.gwt-examples.com/javascriptinjection - You 
can keep an eye on this site for updates on progress. 

Brandon Donnelson
http://gwt-examples.googlecode.com

Original comment by branflak...@gmail.com on 4 Dec 2011 at 5:45

[GoogleCodeExporter]

http://c.gwt-examples.com/home/wrapping - evidence of my v3 wrapping (LatLng) 
class... 

Original comment by branflak...@gmail.com on 4 Dec 2011 at 7:28

[GoogleCodeExporter]

Way cool. Thanks!

Original comment by treethin...@gmail.com on 4 Dec 2011 at 9:10

[GoogleCodeExporter]

I've heard rumor there is a new maps api in testing internally?

This is my current progress on writing a third party maps v3 api. I'm walking 
the classes slowly. 
https://docs.google.com/spreadsheet/ccc?key=0AnyjOBpyOTUWdHpQZTl1UWJUUGxEcjBzYlB
EcUVGZUE Note this is not written in stone and may change.)

Brandon Donnelson
http://c.gwt-examples.com
http://gwt-examples.googlecode.com

Original comment by branflak...@gmail.com on 15 Dec 2011 at 5:17

[GoogleCodeExporter]

I also heard that rumor, first hand from a Google employee (I don't know if 
that makes it more or less reliable though). He said that it was an 
"automatically" generated library and currently about a month old. 

Original comment by ad...@dualsportmaps.com on 15 Dec 2011 at 1:12

[GoogleCodeExporter]

FYI: Update on progress >

- I'm producing some quick demos to flush out some of the rendering basics. 
http://gonevertical-apis.appspot.com

- http://c.gwt-examples.com/api/google-maps-v3 - Consolidated links to all my 
map stuff

- 
https://docs.google.com/spreadsheet/ccc?key=0AnyjOBpyOTUWdHpQZTl1UWJUUGxEcjBzYlB
EcUVGZUE - Build progress

- I can't promise a repository commit date yet. It's possible <= 2[+] weeks. 
(no promise yet, but watch the build progress on spreadsheet if your interested)

Original comment by branflak...@gmail.com on 4 Jan 2012 at 5:53

[GoogleCodeExporter]

Will that be integrated in the official GWT Google APIs ?
Have you already some code to play with, before more stable classes ?
Thanks.

Original comment by alexis.h...@gmail.com on 5 Jan 2012 at 7:16

[GoogleCodeExporter]

My first goal at the moment is to write the code. I'll submit it at some point 
if it can meet or beat expectations.

I haven't committed my source code yet, but soon. I'm going to finish the 
library and StreetView classes and then evaluate if I can release before 
wrapping the service classes. 

Brandon

Original comment by branflak...@gmail.com on 5 Jan 2012 at 3:03