Rollo62

So could you fix it ? I just made a normal XCode update to SDK13.6, did you do this update via command line ? Have you checked if XCode is veryfied ? Like that, or

@Dave Nottage Regarding the iOS13.6 SDK, with XCode 11.6, I'm not sure what is missing there. I just checked the XCode package content, looks like this, seems OK to me at first sight The /System folder is not "empty", but I have not checked out the SDK in my projects yet.

It might have. When I tested .pas extensions, this might occur into strange temporary .$$$ files, when I try to debug and edit. Not sure why that happened, but I moved back to .inc then. Compiling and running should be fine, I am talking about the development process of such file, when opening it from the caller unit.

No problem seen yet, what Platform are you talking about ?

Hi there, related to this post, but I think its a very different issue. As explained, I used XCode 11.5.x and iOS13.5 for quite a while (weeks and months) happily, and then focused a little more on Android, leaving aside iOS, not connecting to XCode during this time. Then after maybe 2-3 weeks later, I've got back to my (unchanged) iOS / XCode / OS setup, and suddenly XCode didn't recognize my phone anymore. I'm using the following setup, and did the following measures to get back on track, but nothing helped so far. Working in host (Macos 10.15.4), current VmWare Fusion (11.5.6) on a current MacBook with enough power and space I didn't change anything, it was running well before when plug in the phone, the USB selection dialog appears, and I can connect to the VM, as usual VmWare shows that iPhone connected to the VM-image, just as it should and as before VM image can show the USB, when also iPhone entry is visible, just as it should The problem is that XCode doesn't recognize the iPhone at all But I can build release packages normally. I've tried rebooting Host, Client VM, iPhone many times I've tried updating all to latest XCode 11.6 and iOS 13.6 I've changes the SIM-Card, to be sure there is no strange hardware error. I've un-trusted the iPhone, but still XCode doesn't react at all This seems not to be an easy one this time, and I'm not sure if this is related to iOS13.6 at all, because this did stop working at iOS13.5 already. Maybe somebody has a clue what could go wrong here ?

I have to give up for the moment, this already costs too much time. The certificates are OK, and I've tried to repair, revoke, renew them several times. The result stays the same: I cannot see my devices in VmWare Fusion - XCode anymore, but on real hardware - XCode it works as usual. For the moment I have to work on hardware Macos again, until I find more time to look into this. This is not an Delphi issue at all, since I'm not even able to do anything in XCode.

Oh, yes. It seems to be another certificate issue: In the Certificates/Devices section I've got this message Then had to confirm: Confirm again: and again ... until you've made yourself really clear that you want to move on development, and not suddenly want to stop it. What is that good for, another level of bu**hit to punish us developers, I never noticed that before ? Maybe this came from the complete change of AppleConnect last year. Anyway, that bad news is that afterwards XCode STILL did not see my devices, although I restartet it already. But I have a reasonable clue now, and try to peek here and there further ... Why dont they give a clear message in XCode then, as Apple knows very well whats going on.

Yes, there are NO devices. Exactly that is the problem. It seems that the phone in the OS is not really distributed to XCode anymore, while I can see that at VM scope and in the Macos guest scope system info. I used an USB2.0 hub to connect it to the USB3.0 ports of the host, that worked well for years, maybe something has changed now. In the VM I have set USB2.0 compatibility. I'm going to check today if it makes a difference if I connect it to the USB3.0 port directy. BTW: When I connect the phone to the host Macos XCode (an older version), then the same phone on the same USB2.0 hub is detected and listed in the Devices & Simulators, as usual. Edit: When I try with an older iPhone 6, also this connects well to the VM, but is not visible for XCode.

I would love to use that even more often, but I'm very careful with this function too. From my experiece this is not very reliable if you use more than one file of same name, which you want to use as default. That depends on the search path which one is taken. The documentation here says that it take the Search library path definitions, I hope that true that its also searches them in the right order. If you change the search path order, then you might see strange errors in that case above. Of course it should be ..\ on Windows, thats what I use there always.

I can agree to that, since I use interposer heavily. Some years ago I have heard that some people have the philosophy to wrap everything into interposers. I'm not going that far yet, but close. I can understand well to have a "safety" layer between the framework and your app. That serves a lot of useful purposes: to easily manage bugfixes at a centralized place for each component to make workarounds around some conceptual limitations of some components to enhance the components with new, custom features they don't interfear with normal designer process, so no need to install custom components That are enough reasons for me to consider interposers everywhere 🙂

Ok, after along way back and forth to the Patch 3, it seems to work more or less. Patch 1 and Patch 2 worked directly, but Patch 3 needed some extra care somehow. I had to re-create .dproj's and re-order units, before it runs without issues. Not sure what the real reasons are, but now all my projects run unter 10.3.3 to 10.4 P1 P2 P3. Whats still open to me is the TChart issue, I hope it gets done soon. https://quality.embarcadero.com/browse/RSP-29414 https://quality.embarcadero.com/browse/RSP-29694 https://quality.embarcadero.com/browse/RSP-28780 https://quality.embarcadero.com/browse/RSP-29414 So I have to disable my chart in the projects. I also faces issues with the Grijjy cloud logger, so I have to turn this off and move to somthing different, but thats acceptable. All in all debugging feels quite sluggish, but compiling and running seems to work then finally. The LSP runs better, so I really hope I'm ready now for Update 1, and I don't need to waste more time in porting stuff.

Try to delete the existing app on the phone, before uploading the new one.

Nope: An update to Macos 10.15.6 doesn't solve it. So maybe I'm foreced to update XCode and/or iOS too, but I try to keep it at the moment, maybe there is another idea howto make XCode detecting my phone again. Updating iOS to 13.6 will also break my XCode connection, and I cannot debug SDK 13.5 there.

Maybe somewhat related to this, I just fell (again) into the other Apple trap. By leaving aside the working XCode VM, only for some days/weeks, to test and prepare Android. Today I try moving back to the current system (same VM, same XCode, sam OS 10.15.5, same iPhone (13.5), etc. The iPhone is suddenly no more detected by XCode, even that iPhone is connected to the VM and listed in the USB. (usualy restart, reboot OS and Phone already done). I had similar cases before, where current, working setups ( of the former versions ), suddenly disappear to function with XCode and RadStudio. And yes, IT IS NOT A DELPHI ISSUE. The general recommendation to this is, to update to the latest versions (at least thats what Apple says ). If I do that, then I'm back to entry 1. of this thread, and the new versions cause other issues. So at the moment I try to install the new OS 10.15.6 only, with the hope that this might be sufficient to detect iPhone again. This is typical Apple, to me it looks like a modern way of planned obsolescense, which is absolutely annoying. I only can say, if I wouldn't have to provide iOS solutions, I would recommend to everybody to stay away from Apple purchases. Usually I'm quite neutral, but Apple already costs me days of my lifetime, for their daily nonsenses. Go and better buy Android products

Then you should care, that "visual form editing capabilities" is what I meant by RAD.

I can think of a world with RAD designers, as the best and fastes way to create views, prototypes, and maybe even events and bindigs. Plus a code-centric, MVVM enabled, way to produce the business logic behind in a modern way. Why should that be so hard to bring both concepts (RAD & MVxx ) together in a nice marriage, I never understood that. ? Just open all the options in the framework, and let people create the way they want (and let them also grow with their skills in the same tool). Even if thats a complete rewrite of most the components, that should be worth it. From all other visual designers I have checked so far, the RadStudio I like the most. Thats why I'm on RadStudio. Maybe thats only because of old habits and laziness, but I think the current concept is still best. Maybe the question is not RAD or MVVM, but RAD plus MVVM.

VSC is a great tool, seems to be well thought from ground up, with a zillion users out there. Also consisting of a large ecosystem, with plugins and many supporters. Most important maybe, as said before, the young guys are there: If you want catch attraction to Delphi code in the young generation, then this is a good place to go. VsCode is Open-Platform, mostly used on Macos, because of its coolness factor. By the way: Worth to mention, its free and open source IMHO this is the new way to go, all other IDE look old fashioned against this ( Ok, I know Atom, etc. too, but the winner takes it all).

Maybe Delphi should be designed as a plugin for VsCode, and FMX could be an extension for PlatformIO ( Oh my, TMS is halfway on the direction to that )

Maybe that helps too. I basically moved to TListView, since I find the more reliable supported than TListBox. Although TListBox would be handy in most places.

Yes, they are in VsCode anyway. Installing another Plugin is no big deal. Not sure if in VsCode marketplace is some kind of "Ad" publishing for new plugins. If so, would be maybe great to try to reach all those JS, Node.js developers out there.

@Dave Nottage Yes, no direct solution. But maybe the intent filter could lead to some insights, howto achieve that goal in Java and Delphi.

Good guess IMHO. I only hope that the 75% Delphi-only developer have enough vision to predict the future of desktop-only apps. Many of them might be forced by THEIR customers, to offer a mobile or web-centric solution. See some posts before, the TMS WebCore, what is possible with a little vision, braveness and motivation even for a small company. You don't need the crystal clear glass bowl, to predict the upcoming future, the frozen glass bowl will be sufficient

Have you checked out the Kastri project ?

OnItemClick Your second post shows that too, is that the solution or does it have an issue ?

Not sure what the question really is. procedure TForm1.FormCreate(Sender: TObject); var AppEventService: IFMXApplicationEventService; begin if TPlatformServices.Current.SupportsPlatformService(IFMXApplicationEventService, AppEventService) then AppEventService.SetApplicationEventHandler(HandleAppEvent); // Register the type of intent action that we want to be able to receive. // Note: A corresponding <action> tag must also exist in the <intent-filter> section of AndroidManifest.template.xml. MainActivity.registerIntentAction(TJIntent.JavaClass.ACTION_VIEW); TMessageManager.DefaultManager.SubscribeToMessage(TMessageReceivedNotification, HandleActivityMessage); end; procedure TForm1.HandleActivityMessage(const Sender: TObject; const M: TMessage); begin if M is TMessageReceivedNotification then HandleIntentAction(TMessageReceivedNotification(M).Value); end; function TForm1.HandleAppEvent(AAppEvent: TApplicationEvent; AContext: TObject): Boolean; var StartupIntent: JIntent; begin Result := False; if AAppEvent = TApplicationEvent.BecameActive then begin StartupIntent := MainActivity.getIntent; if StartupIntent <> nil then HandleIntentAction(StartupIntent); end; end; function TForm1.HandleIntentAction(const Data: JIntent): Boolean; var Extras: JBundle; begin Result := False; if Data <> nil then begin Memo1.ClearContent; Extras := Data.getExtras; if Extras <> nil then Memo1.Text := JStringToString(Extras.getString(TJIntent.JavaClass.EXTRA_TEXT)); Invalidate; end; end; HandleIntentAction(const Data: JIntent): Boolean; Is the Action function that is doing the work when the intent is proceeded. While TMessageManager.DefaultManager.SubscribeToMessage(TMessageReceivedNotification, HandleActivityMessage); registers the Handler. and this can be called from two sides, 1. from the handler: HandleIntentAction(TMessageReceivedNotification(M).Value); By the registered callback . 2. StartupIntent := MainActivity.getIntent; if StartupIntent <> nil then HandleIntentAction(StartupIntent); During startup of the app: when the OS puts the app into foreground by selecting the intent. Just looked fastly over this, what exactly is missing ?