Topre switches are great for debugging code

[keyboardlover]

I noticed this as soon as I started using my Realforce at work. If you're like me and you work on a very large software system, you probably often have to set a lot of breakpoints in a lot of different areas of the code, then step over a lot of them in order to get to a places where you want to step into the code. Because of the capacitive switch, if you hold down F5 (in Visual Studio) it FLIES over those breakpoints - probably at least twice as fast as a rubber dome board. This is because of the capacitive switch triggering the keypress before you bottom out. This has actually made me more productive! More proof that Topre switches kick ass!

[rantenki]

I noticed this as soon as I started using my Realforce at work. If you're like me and you work on a very large software system, you probably often have to set a lot of breakpoints in a lot of different areas of the code, then step over a lot of them in order to get to a places where you want to step into the code. Because of the capacitive switch, if you hold down F5 (in Visual Studio) it FLIES over those breakpoints - probably at least twice as fast as a rubber dome board. This is because of the capacitive switch triggering the keypress before you bottom out. This has actually made me more productive! More proof that Topre switches kick ass!

[msiegel]

shhh, he's justifying expensing one from work

[keyboardlover]

Those settings were the same ones I had when I had a Keytronic though, and it was nowhere near as fast. So how do you explain that?

[itlnstln]

Keyboard firmware.  Some of older ones had a switch or dial to set the internal repeat rate.  It's nothing special about the switches.  As for the activation point, pretty much all mechanical switches share the same advantage.

EDIT: There is no difference in repeat rate between my Cherry G80-3000 and my Topre 103.

[keyboardlover]

Oh ok that makes sense. Thanks for the clarification.

[Fwiffo]

Real men xset r rate 100 255

[woody]

With USB it's all about the keyboard sending first event, which should be in the range of few msecs, then delay and repeat are in the HID driver.

Now with PS/2, you can turbo all the way.

[JBert]

Real programmers use logging. Needs far less breakpoints and can be extracted from running code.
Unless it's code made by your ignorant co-workers who don't know how to code well.

[keyboardlover]

Real programmers use logging. Needs far less breakpoints and can be extracted from running code.
Unless it's code made by your ignorant co-workers who don't know how to code well.

We already log at runtime...how many breakpoints you have depends on how large the software is more than anything else. Unless you're talking about something else? Like logging at compile time (no clue why you'd want to do that).

Edit: since logging occurs at runtime, it's more useful in the maintenance phase than when you're doing new development, so I still don't understand your point =)

[panda-R]

can i buy a topre if i'm not a codemonkey?

[theferenc]

If you do, it breaks the stereotype that exists here. Are you sure you want to go and do something as silly as that?

Though, if I recall correctly, it is most strongly associated with the HHKB, but I've noticed that most people who list a topre board in their sigs tend to be developers of some sort.

Then again, it's gamers and code monkeys on this board, isn't it? Who else cares enough about their keyboards?

[patrickgeekhack]

Show Image

This guy can speak French in Russian.

[keyboardlover]

Personally I would imagine that a writer or journalist would love a Realforce. Pretty much anyone who has to do large amounts of writing - topre switches just makes the whole task much more pleasant =)

[theferenc]

That's true. I tend to bounce between my HHKB and my Customizer 101. Both are in the same layout, so the only real difference for me is the mechanics involved.

There's just something about the sound of the Customizer that I find quite relaxing while I'm writing code. And while it feels so very different from my HHKB, I find both an immense pleasure to use. It really just depends on mood for me.

Edit: If only I could get a 103UB 55g in my preferred layout, I'd have even less money...keyboardlover, do you know if that's even possible?

[panda-R]

i wish i was a journalist, a monkey or a bearded man. Not necessarily in that order.

[keyboardlover]

Edit: If only I could get a 103UB 55g in my preferred layout, I'd have even less money...keyboardlover, do you know if that's even possible?

What exactly is your preferred layout?

[theferenc]

Oh, sorry, it's the UNIX/Sun type 3/HHKB layout, whichever term you prefer.

I'm a bearded UNIX geek, and it just works better for me.

[keyboardlover]

Hmm...how did you configure your Customizer in this layout? Using software? If so, the same software should work with the Realforce.

[theferenc]

I had it custom programmed at the factory. The key remapping itself is fairly easy, but there are physical key changes (splitting the backspace key into 2 distinct switches) that can't be overcome by software, unfortunately. Unicomp also took care of that physical difference for me.



And apparently only the BS keyboards can be (relatively) easily modded that way after the fact.

[keyboardlover]

In that case, I know of no way other than somehow modding it yourself...

[theferenc]

The blank keys on the bottom of my board send the windows key scancodes. And the blank key where ESC usually is? That's caps lock.

Weird, I know, but Jim at Unicomp insisted that caps lock be SOMEWHERE on the board, and since I wasn't using that key, that's where it ended up.

But I definitely prefer my alt keys next to my space bar, so the corner keys went the windows route.

[theferenc]

In that case, I know of no way other than somehow modding it yourself...

Yeah, the idea of modding a $250 keyboard by hand (which would also require getting a complete switch from somewhere else)...not something I really want to be doing.

But it was only $20 extra for Unicomp to do a custom programming job on this one, maybe I can get Topre to do the same thing, for a similar amount...assuming I can find someone there who speaks english, of course.

[theferenc]

Apparently someone there can, since they were able to do it for me. Except they swapped some keycaps for no apparently reason. The '-' and '=' keycaps were swapped when I got the board. The keys sent the correct scan codes for their position, but were labeled incorrectly.

How weird is that?

[JBert]

We already log at runtime...how many breakpoints you have depends on how large the software is more than anything else. Unless you're talking about something else? Like logging at compile time (no clue why you'd want to do that).

Edit: since logging occurs at runtime, it's more useful in the maintenance phase than when you're doing new development, so I still don't understand your point =)

Logging is the poor man's debugger. Consider that every time you need a breakpoint to figure out what your code is doing, some poor schmuck might cuss later on when he's sitting at a customer's office and trying to find out what goes wrong.

Therefore it's far wiser to put some extra logging in your code. When you're debugging your code, you are running it and can use the logging to your advantage, then drop into the debugger to learn the flow of the program.

Of course, all of this assumes you're using a logging framework you can adjust at will - something "decent" like log4j or Logback (Java) or their offspring on other platforms/languages. This technique won't do you any good if you have to go in and comment out code to disable certain output, or if your logging framework causes too much overhead.