Sherlock

I think installation to the physical iDevice would have to be done via Testflight. Not sure if your physical iDevice can be connected to the virtual cloud Mac. @limelect How old exactly are we talking here? iOS 11 should be minimum, better 12.

Well you don't need the Mac physically. You can rent a Mac in a cloud for example here: https://www.macincloud.com/

You need a Mac. the simulator is part of XCode, which only runs on a Mac.

Guys, the german DelphiPraxis has a no-cheat-tool-policy. I'm pretty sure that applies to the english version here as well. So...sniffing is almost OK, modifying must be discussed elsewhere, please.

So I have this application, that runs smoothly on a standard PC, and does quite well on a tablet too. With one exception, the virtual keyboard causes quite the headache. Not only does Windows seem to have two different virtual keyboards, they behave quite differently too: First there is the keyboard that has to be actively opened by the user by clicking on the symbol in the tray area. This works fine, but has the disadvantage of "an additional click". (Image 1) Then there is a keyboard (image 2) that can be opened automatically when FMX.Types.VKAutoShowMode := TVKAutoShowMode.Always; is set. According to the fine manual this will cause the virtual keyboard to show even if there is a physical keyboard present. So as soon as the focus is set on an editable component like TEdit, this keyboard will popup and as soon as said component looses focus the keyboard will disappear. Now here's the tricky part: This keyboard looks different, and on some machines (HP tablets with an i5) takes ages (at least 30 seconds) to "pop up". On my VM that I'm running on a Mac with an i7, this happens instantly. Any user who knows iOS will consider option 1 to be "out of the question" so I use option 2, but for the life of me am not able to speed up the virtual keyboard popup process. This might just become a deal breaker, hence this post. TL;DR: Does anyone here have experience with virtual keyboards on Windows tablets, and knows how to open them up quickly on slower systems? Thanks, Sherlock

Great! Thank you, I'll give that a try.

Hmmm, I don't quite follow. Could you please elaborate?

Guys! Please! I'm trying to work here! Now I have to go get a roll of Kleenex to clean the coffee from my monitors and keyboard... Thank you very much.

Argh! I feel with you.

Why don't you just recover from the backup you made before this experiment?

Just a reminder for the next time anyone installs a Delphi update: TAKE A SNAPSHOT OF THE VM. And if you don't use a VM...reconsider and then just use one from now on. The hassle is reduced to a minimum.

It is strange though that this is having such an impact when switching from 10.3.2 to 10.3.3, don't you think @Dalija Prasnikar?

Just out of curiosity how does an Android Java or an iOS Swift App do this?

Try the 12.4 SDK as David has.

@Dave Nottage For the sake of the use case here, please try changing the name to "Bob the developer" (blanks in the name seem to be an issue).

Which, to be honest, is silly in itself. Backward compatibility is just a means to keep the living dead creeping around longer. I kind of love Apple for their policy of cutting off old stuff when their modern replacement has proven itself (last seen with macOS 10.15 and the 32Bit cut-off). It keeps the system clean, and prevents legacy software from doing stuff in a way that used to be okay but just isn't anymore. Please remember the Gate A20 situation that lasted from the introduction of the 80286 well into the Pentium era.

It is always good to discuss an issue with your peers...if only to see it vanish into thin air.

I knew my job should have been "son". But nooo, I had to go and learn something useful.

Every Tom, Dick and Harry can up a small "Hello World" like project in any new language to GitHub, just to test it. So whenever a groovy new language is out, people with nothing else to do will do just that, making said new language "popular". What bugs me most is this "big company X invented this awesome and totally new language for their project Y because they couldn't have done it without it" meme. Which is just silly. First of all that new language will be just another dialect of some existing language, and secondly in what hell is that a reasonable approach to getting things done kissed and dry?

So...macOS 10.15 has been released, and with it the incapability of running 32Bit applications. Using this tool to https://www.stclairsoft.com/Go64/ check wether or not I have any crucial 32Bit applications before making the dive, I found out, that the PAServer 20.0 (11.2.13.2) is at least in part 32Bit - which gives me a minor rash. What is the verdict on this? Should I stall Catalina installation, or is it OK to go ahead?

Before this gets out of hand, @PeterPanettone, may I suggest looking at the nickname FredS a bit closer and then cross reference it with the person you are looking for. Then have a good laugh and perhaps get some solid help with your issue.

@Joseph MItzen You clearly got my drift!

As long as they are round and colorful I'll take 'em.

There is no official API AFAIK. But you may just as easily call the command line for git itself. Refer to the porcelain and plumbing of git for more information. And of course you would have to distribute a copy of git.

Judging from the provided link alone a webhook is a server running on the client side. The client, when initiating a connection to a "real" server will provide this address to said server. This means, the server will need a client socket itself, and has to be able to connect to that mini server on the client side. It is a nifty concept to reduce open connections on the server side. Also the clients need not ask if a time consuming task is done, they will be notified via web hook. This concept will be rendered useless by firewalls, proxies and other sane security measures. Possibly only something for a LAN application. But in that case, the server might just dump the result of aforementioned lengthy operation into a samba share and the client could poll that every so often. No webhook needed. Unless the result is not dumpable...which I highly doubt.