soegaard2
2020-5-9 09:11:18

The mysterious “double title” phenomenon that @spdegabrielle experienced was due to macOS which now supports tabbed windows. Could these be used to implement “proper” tabs in DrRacket?

https://developer.apple.com/design/human-interface-guidelines/macos/windows-and-views/tab-views/


laurent.orseau
2020-5-9 09:16:09

Is there a particular aspect of the current tabs you don’t find ‘proper’?


soegaard2
2020-5-9 09:21:02

Proper might be the wrong word. In most cases it is beneficial to use the system provided gui elements. It will make the program look better.

One thing I am missing from the current tabs in DrRacket: • I can’t grab a tab and rearrange the order.


laurent.orseau
2020-5-9 09:22:10

Right. You can use Ctrl-Shift-PageUp/Down though (Or Cmd-… on mac?)


soegaard2
2020-5-9 09:24:19

What is it bound to? (I haven’t got PageUp and PageDown on my laptop)


spdegabrielle
2020-5-9 09:25:49

@laurent.orseau

> Is there a particular aspect of the current tabs you don’t find ‘proper’?

The can be dragged, re-arranged, docked and undocked from the tabs bar.

I was also able to dock a frame%


spdegabrielle
2020-5-9 09:26:35

It was a fun bug. I miss it.


laurent.orseau
2020-5-9 09:26:35

@soegaard2 Move tab left or right


soegaard2
2020-5-9 09:28:32

That can helps. I am used to (in Chrome) to grab a tab with the mouse and move it.


laurent.orseau
2020-5-9 09:29:06

If you Click on the “Tab” menu, it should give you the correct keybinding


laurent.orseau
2020-5-9 09:29:25

(maybe Ctrl-Shift-PageUp is one on my custom bindings though)


soegaard2
2020-5-9 09:30:39

It’s alt-cmd-[ and alt-cmd-] here.


spdegabrielle
2020-5-9 10:17:13

I’m tempted to log ‘proper tabs’ as a feature request but to my shame my c skills amount to ch1 of K&r and a couple of tutorials, and I don’t think I should be making requests to a OSS project if I can’t contribute to their resolution.


laurent.orseau
2020-5-9 10:21:49

I think you should. It shows interest. Maybe more people will join to show their interest, and maybe someone knows just what to do to do it fast.



spdegabrielle
2020-5-9 10:32:44

p.kushwaha97
2020-5-9 12:03:16

I’m using rackunit, and after a racket version upgrade, my tests started to fail. I’m getting expected (chaperone-of? <val>) received: <val> I referred to impersonators and chaperones doc, and it seems that test-equal? should say both of these are equal. What might be going wrong?


p.kushwaha97
2020-5-9 12:03:34

I’m not explicitly using chaperones in my code, so some library I’m using must’ve started to use it.


spasoevi
2020-5-9 12:45:26

@spasoevi has joined the channel


p.kushwaha97
2020-5-9 13:32:18

I also have another issue with rackunit. Despite passing the ’verbose option, I get same output as the ’quiet option.


p.kushwaha97
2020-5-9 13:52:57

I currently use spacemacs with racket. It’s a pretty neat combo (when it’s not slowing down). It might be worth comparing the feature set!


greg
2020-5-9 18:19:22

racket-xp-mode now also shows annotations given to the online-check-syntax logger. So e.g. for typed/racket you can see the type annotations for sub-expressions. When check-syntax also produces an annotation for the same span, they are now combined and both shown.


greg
2020-5-9 18:21:53

Still to-do: Submit a PR to better index and inter-link the documentation about the online-check-syntax logger and the mouse-over-tooltips syntax properties.



soegaard2
2020-5-9 18:57:54

Any ideas?


sorawee
2020-5-9 19:18:46

In my experience with big-bang, the performance problem almost always comes from big rational number that keeps getting bigger and bigger as the program runs


sorawee
2020-5-9 19:19:13

But I also didn’t use big-bang that much


sorawee
2020-5-9 19:19:59

File an issue? Can you also provide a program that exhibits this problem?


p.kushwaha97
2020-5-9 19:24:05

I’ll try to create a reproducible subprogram and do that. I thought it might not be a bug and some problem with my understanding so asked here first.


gfb
2020-5-9 22:35:17

I should make a reddit account, but until then feel free to communicate any of the following that you feel like : big-bang and 2htdp/image are optimized for teaching, not performance, Having taught with it, I don’t see any obvious red flags in that code. As far as running it outside of DrRacket, you get a lot of the benefit of that by going to the language customization and select “No debugging or profiling”. Even if you switch to gracket, you might have to have done that first to get benefits, or remove the compiled directory or run make to remove already-compiled code.


samth
2020-5-10 00:53:41

Unfortunately most of the advice there is not going to be helpful. Slow big-bang programs are caused by 1. too-complex image structures, which need freeze 2. Rationals, as @sorawee said 3. Algorithmic problems 4. Big-bang being slow


gfb
2020-5-10 01:53:37

Right, and none of 1–3 apply here unfortunately, especially since (for #1) the state is not an accumulating image nor (for #2) an exact number, and the algorithm is very simple. Is any of this folk-knowledge written up somewhere?


gfb
2020-5-10 01:56:02

I’ve given them https://github.com/daniig/baseline_racket/pull/1 now, which addresses the rest of their question and some of the non-helpful advice.


samth
2020-5-10 02:04:19

That is all good advice


samth
2020-5-10 02:04:36

The real problem is just that big-bang is too slow, probably