Matthias Becker
Matthias Becker
Pls. also have a look at the failing tests.
> In most cases, computing extra information for a completion item is expensive (and it's actually expensive JDT and Javadoc rendering, and expensive for LSP-based editors), and it often creates...
> Fix CompletionTest to find the completionShell Jörgs Change does not affect this. It's about the element information in the completion popup. Pls. really do test the correct stuff.
> If you like watching text-movies it looks smooth, but any animation any change or alike will catch up the eye so one should just avoid that, it is already...
> and even ABAPs network stuff and did not show a problem. to be honest there is not problem with out "network stuff" because we anyway have a throttling in...
> and rendering it (second delay) can be expensive Pls. provide proof for this. I did test this. Showing the anyway computed information immediately did not cause any issue on...
> Yes for sure everything is always unrelated so why asking ;-) > > It was asked why there are two delays, I explained why in UX design one has...
> and fix code to handle such short delays. This really concerns me a bit. Might there be code outside of platform / JDT / ... that might break due...
> > This really concerns me a bit. Might there be code outside of platform / JDT / ... that might break due to this change? @jukzi what's your feeling...
> but every time is is insisted that it **must be zero** and every one that don't agree is just an non real world idiot This is not true. *I*...