Or maybe the garbage collector… even if it has finished to use an object, there’s no guarantee that it will deallocate the memory space reserved for this object (frequently it is not done before a spécific call (bad method) or the end of the program). Also there is a difference between the memory in use you can see in your task manager and the reserved memory by your application. Did you catch the portion of code that launch the exception ? Is there a profiler in the visual studio xx you use ?
On RSA try to have a key as long as you can. When I implemented it the key length was of 7. If a hacker sees a message of 49 numbers, it has its first information… then an analysis with frequency of letters could break the beautiful work of the RSA formula… in C++, I used NTL to manipulate big numbers. And try to apply Euclide calculation/algorithm once by yourself it is funny xD
With a length of 8 number by character, it should be safe (with message and/or passwords minimum length greater than 8 too)
Thanks again for the update Flavien. I can’t express how much better it is now to hear things are progressing and actually being able to see how development is unfolding. But Please don’t burn yourself out on daily reports. 2 or 3 reports a week is more than enough! Or, perhaps have one DEV write one update a week and that would cover not only the info/PR department but free up your time. Anyway, I will not say anything more on the matter but simply say thank you for the hard work and letting us hang with the cool kids.
10ms for the TLS handshake really sounds alot. If you were using an apache tomcat server I would suggest to configure a non-blocking connector. In case of a blocking (default) connector at tomcat, the maximum number of simultaneous connections is limited by the maximum number of threads.
Regarding your idea with the CA callback, you could run Wireshark to record and analyze all TCP connections. Also after installing the cert on the clientside (at windows?) you should be able to skip any round-trip.
For me, UV unwrapping is the number one time consuming and boring task in 3D modelling. This looks like an amazing tool. If I had one criticism it would be nice to see larger islands, that might cost of a bit of wasted texture space but make it easier for the artist. That also depends on how you’re painting the textures, do you paint on the 2D UV map or do you use a 3D painting tool to paint on to the model directly?
Yeah, the initial unwrap isn’t very good. I didn’t want to spend too much time in it when we’re going to replace the model down the road. Just need something good enough.
I use Substance Designer and Painter. I still have to give the new Quixel suite a go since they added 3d painting as well.
I can do most stuff with tri-planar projections in SD. Thats a good workaround for the sub par unwrap. There aren’t any obvious visible seams. (yet >_> )
Thanks for the update Jan. Really like the overall shape and I can already tell it will look great after texturing. Stuff like this is more exciting than the game itself…well, for me anyway.
Cough…cough, anyway, always willing to shamelessly promote myself, here is some stuff I made; some semi-good and some just for a good laugh:
I just want to say I’m impressed by the response to community feedback, and the way you folks whipped project updates right around so quickly. Well done. Loving it!
I can’t help but feel you, @INovaeFlavien, for the WPF work. It is much easier to do cool looking stuff on WPF than with WinForm… but man it does contain some nice hurdles sometimes
Have you considered using a framework for your UI?
Not to make any ad, but I know “Infragistics” makes some nice work there and has some decent support reactivity / web documentation. There are others, of course, and maybe the license cost isn’t worth it for “only” a launcher.
