> > > From what I can tell, most of the really good teams are using Android > Studio + Gradle or IntelliJ + Maven already. >
I think that is a bit early to say. I would say we are a "good" team, but we haven't successfully been able to migrate over due to our builds simply not working in gradle. In fact, I would guess that there are many developers who have tried to switch to gradle but failed due to some issue. For example, our team's app collection is what I would consider a "large" project (+1,000,000 lines of code), and we have tried switching to gradle on 3 ocassions. We are actually in the middle of a conversion attempt right now but we may not be able to complete the conversion because gradle won't build a few of our projects unless we run assemble on it twice. Why this happens, I have no idea. I am confident the bugs will all be solved with the new build system, but it still isn't ready for prime time. I think the Google developers are just being responsible and waiting for gradle to be ready before wasting peoples time if things don't work. Trust me when I say, I am eager to move away from the bear that is Eclipse. We deal with crashes on an almost daily basis. On Wednesday, July 2, 2014 11:18:31 AM UTC-4, Kevin Schultz wrote: > > I think the future of Eclipse+ADT is pretty clear to anyone following the > development of Android Studio. It is going to be a second class citizen at > best. I honestly hope that it gets killed off completely so as not to pull > resources away from Studio development. The lack of clarity on this is a > big problem though. > > I was a bit surprised to find out how many developers have not yet > switched to Gradle / Android Studio. I took an informal poll at a NYC > Android Developers meetup this spring and from the show of hands it was > about 50/50 Eclipse+ADT vs Android Studio. That night we gave 2 talks on > why you should switch to Studio & on how to use Gradle, and I know that > gave some people the nudge to switch over, but I was really expecting more > people to have switched. > > I think the people that haven't yet switched fall into two camps. > > #1) Eclipse users. A couple years ago I was at a company that wrote > Eclipse plugins, so you can imagine that I'm a pretty big fan of Eclipse. I > had never tried IntelliJ at all prior to Android Studio. Switching seemed > of dubious value, even though I had heard from a lot of Java developers > that IntelliJ was superior. Having now made the switch, I vastly prefer > IntelliJ. The refactoring tools are better, and the stability of the IDE is > much better. But I think there will be a lot of developers that no matter > what the feature set of the Android tools just don't want to learn IntelliJ > after years of using Eclipse. I have never met anyone that has used > IntelliJ extensively but still prefers Eclipse, but I have met many that > simply have never used IntelliJ and don't want to invest the time to learn > something new. I don't think there is much the Android tools team can do > about getting this group of people to switch other than officially killing > off the Eclipse plugin. Some people won't be happy, but you can't please > everyone. > > #2) Those that think Android Studio + Gradle are not yet stable enough. I > think this is the bulk of those that haven't yet switched, and I think this > is largely a result of the messaging from the tools team. > > My personal experience has been extremely positive. My team was trying out > Studio after each release since I/O 2013, and we started getting everything > in our app working sometime in August 2013. We switched our master branch > over to Gradle and all team members switched to Android Studio in about > October 2013. Since then we have been using Android Studio all day every > day for production work and haven't looked back. We have lost maybe 2 full > days spread over the last 8 months to fixing things after changes to the > Gradle plugin causing us to have to rework our build files. That is not a > big deal in light of the fact that all of the features in Gradle have saved > us a massive amount of time: switching our dependencies to Maven instead of > git submodules, package naming, build config, build variants, etc. Perhaps > most importantly, Android Studio itself is significantly more stable today > than Eclipse + ADT ever was. > > But up until I/O 2014 the developer.android.com documentation was > effectively a link to the download page with a big red warning that this > wasn't for production use. A lot of people are taking that at face value > and not even trying Studio. The "beta" tag and better documentation that > was posted will help, but I still don't think it is enough. The impression > of Studio from those that haven't tried it is that it crashes all the time, > or that Gradle changes all the time (and yes, it changes, but generally > it's not hard to keep up with). > > Android Studio is the better option for production development today. From > what I can tell, most of the really good teams are using Android Studio + > Gradle or IntelliJ + Maven already. If it's good enough for the use of > teams making apps for millions of users, then we shouldn't be discouraging > developers from using Studio. I think larger teams have the resources to > try out multiple build systems and IDEs and do their own cost/benefit > analysis. Generally the conclusion is that Studio is better than > Eclipse/ADT/Ant, and a polished IntelliJ + Maven setup is comparable or > better than Studio at this time. But individual developers and new entrants > to the platform don't have that luxury. They go by what the documentation > recommends, and right now, even with the changes after I/O 2014, I think it > pushes people towards Eclipse + ADT. > > That is unfortunate because individual developers and hobbyists will > benefit the most from Studio + Gradle. The ease of pulling in dependencies > & leveraging libraries is highly valuable to newcomers to the community. > All of the additional Android refactoring tools & the improved visual > design tools makes individual developers more productive and eliminate > bugs. And learning 2 tools over the next 2 years is a waste of time for > people that are already trying to learn the entire Android platform. *I > think Google is being overly conservative with pushing the new tools and > dramatically underestimating the costs of the current status quo. * > > I think Google should make it clear that they endorse Studio over > Eclipse+ADT for use today. I imagine there was a lot of thought put into > the exact wording on the developers site, but "[Studio] will be the > official Android IDE once it's ready" is not a strong vote of confidence. > In my opinion, in addition to the wording about the fact that in the future > it will be the official IDE, it should be clear that it is in use by many > developers already and that the Studio is plenty stable (in fact, more > stable than Eclipse). The only downside to using Gradle is that the API is > evolving and you sometimes have to update your build.gradle files. That's > not a big ask from developers in the context of everything else that is > constantly evolving in the ecosystem. > > And if Google intends to sunset Eclipse+ADT when Studio is "ready", that > should be in the current messaging as well. Right now the cost/benefit > question is 'are the new features worth the cost of learning a new tool'. > If you know that you will have to switch eventually, then it changes your > thought process to 'well I have to learn this someday anyway, and I'll get > the new features today'. > > > On Tuesday, July 1, 2014 7:49:38 PM UTC-4, Csaba Kozák wrote: >> >> Thanks for both of your answers! >> >> I should not have bring the problems of SDK 23 here. It affects the >> Gradle build system too, also i was sure that you guys will fix it. Thanks >> for the first patch, appreciated! >> >> Xavier is very, very diplomatic, but i guess this means ADT will indeed >> go away. I think it may be too much to expect maintaining both of the IDEs, >> i just a little but confused about the lack of information about ADT here. >> > -- You received this message because you are subscribed to the Google Groups "adt-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to adt-dev+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
