Mark Sibly

Forum Replies Created

Viewing 15 posts - 136 through 150 (of 1,431 total)
  • Author
    Posts
  • in reply to: TestPace X – Under development #13830

    Mark Sibly
    Keymaster

    Yay, nice looking effort there!

    in reply to: Window drag pausing #13829

    Mark Sibly
    Keymaster

    You’ll need to stick your update code into OnRender to do this as this is the only callback that continues to be processed while windows goes ‘modal’.

    in reply to: New iOS device for Mark #13827

    Mark Sibly
    Keymaster

    Yay, just scored an  Apple iPhone 6s 16GB Gold Mint Condition Refurbished with 1 Year warranty for 360NZD! I will be spending the leftover funds on some semi-urgent plumbing if that’s OK with everyone…

    iphone6s is a LOT less than 8, and this article kinda sealed the deal: https://www.stuff.co.nz/technology/gadgets/97911184/10-reasons-to-buy-an-iphone-6s-instead-of-an-iphone-8

    Thanks for the generosity everyone, I hope to have the ‘display zoom’ issue fixed very soon.

    in reply to: New iOS device for Mark #13824

    Mark Sibly
    Keymaster

    Yikes, iphone8 is expensive!

    iphone6s might be the way to go…

    in reply to: New iOS device for Mark #13805

    Mark Sibly
    Keymaster

    Do all iphone6’s have the ‘display zoom’ feature?

    Do all versions of ios have the display zoom feature?

    Just don’t wanna accidentally buy one without it!

    in reply to: New iOS device for Mark #13783

    Mark Sibly
    Keymaster

    Hey everyone, thanks for the offer of help!

    It’s true, my ipad2 and nvidia shield are likely to be ‘retired’ pretty soon, ie: no more OS updates, so if I want to keep testing on real devices (‘if’…!) I’ll need to buy one. Frankly, I would rather  just stick with simulators and leave it up to the community to do the real world testing…I tried the ‘buy lots of HW’ route with monkey-x and it really just made me go broker faster so I wont be going down that road again.

    Also, there appears to be a bit of a glitch with iphone6 and ‘display zoom’ mode. Unfortunately, inexplicably, iphone simulator does not offer this feature so to fix it I’ll need a ‘real’ iphone6, so my choice of which real device to get is sort of made for me on the ios front. I would actually rather it was a newer model as iphone6 will itself likely be canned in the near future. Are we up to 8 now? Or is it 10?!?

    There are however some iphone6’s available on trademe.co.nz (NZ’s answer to ebay) some with very low reserves, some a bit beaten up etc, so I’ll have a go at buying one online ‘ultra cheap’ there, ie: <100NZD.

    That’s if Diffrenzy hasn’t already fixed this issue? He had already found one related bug and now the reported display size is only off by 3 pixels. Could be a rounding error or some sort? The actual issue is here:

    https://github.com/blitz-research/monkey2/issues/331

    in reply to: ANDROID – bbvariant.h:63: error: undefined reference to #13730

    Mark Sibly
    Keymaster

    Whew!

    in reply to: WIP Jsonifier! #13728

    Mark Sibly
    Keymaster

    > Ooo! Is it possible to add blobs?

    Json doesn’t support ‘blobs’, but you can roll your own.

    For example, the gltf format allows you to use ‘data:’ URLs, and has an ‘accessor’ system that allows for data to be stored in external files. These aren’t actually part of json, but implemented ‘on top of’ json. I don’t see why we can’t implement/steal similar systems though.

    Here’s a new WIP version that features an attempt at serializing constructors. Really rough and there’s lots to sort out here but it’s looking promising. Note that the LoadC and CreateC functions are really just ‘encapsulating’ the function call ‘C.Load( path:String )’ and the simple ctor ‘New C( Vec2i )’. I have  no idea how this still will be implemented in practice yet though!

    in reply to: Simple mojo3d view culler #13727

    Mark Sibly
    Keymaster

    Develop branch should be updated now.

    Note: if git starts giving you weird errors http messages on windows when you ‘git push’, install this:

    https://github.com/Microsoft/Git-Credential-Manager-for-Windows/releases/tag/v1.14.0

    Modern software is rubbish…


    Mark Sibly
    Keymaster

    Try adding Super.New( "Window" ) at the top of the ctor, eg::

    Pyro needs a minor fix here due to a recent monkey2 bugfix.  ScreenManager currently has 2 ‘default’ constructors, a bit like this:

    So if you just go ‘New ScreenManager’ without any args, there’s no way to tell which one to use. Monkey2 should really check for this and cause an error when compiling ScreenManager, not just when compiling a class that extends ScreenManager. I’m not sure if this is 100% possible when compiling ALL overloads with default parameters though, will look into it. If nothing else, the error should be better!

    Also, pyro will need to be tweaked here.

    in reply to: Mojo3D CameraClsMode Equivalent #13704

    Mark Sibly
    Keymaster

    This is not currently possible but I’ll look into it.


    Mark Sibly
    Keymaster

    What do I need to #Import to run that? Would a full runnable example be possible?

    in reply to: Curry! #13702

    Mark Sibly
    Keymaster

    I have been on a diet since Jan 2 and that looks really, really good!

    in reply to: Emscripten: Save Pixmap #13670

    Mark Sibly
    Keymaster

    You can’t do this yet sorry – I need to learn how! Any advice most welcome…

    The emscripten stuff is currently very much isolated from the javascript side, it’s really just 100% pure c++ right now! In addition, I presume sending ‘file downloads’ requires being able to send http ‘headers’ too, which means you need to be running a server etc.

    in reply to: Compare json values. #13667

    Mark Sibly
    Keymaster

    Yes, adding a JsonNull class is probably a better way to do the null value thing, and would prevent at least that particular stack explosion. Are there any others?

    I’m still a bit weird about comparing references in fancy ways, just been bitten by it so many times in c++ perhaps. I probably just need to go ahead and do it?

    Also in your code you check y=NullValue and return 1 if true, but x can be null itself and then we should get zero in result.

    Where in the code? There is a check in the ‘root’ CompareJon for when x and y are both null and it returns 0 there. This is the only place where x can be null, and in all other cases, x is always > null.

    The little RndJson test seems to work OK too.

Viewing 15 posts - 136 through 150 (of 1,431 total)