|
C:\> deltree Only good program
|
# ? Oct 23, 2019 20:40 |
|
|
# ? May 16, 2024 17:12 |
|
KILLALLNERDS.EXE
|
# ? Oct 24, 2019 08:39 |
|
debuggers could be made a lot more user-friendly really. tbqh i sort of suspect that beginners would be more likely to use the debugger if you could invoke it with what looks like a function call, like "stopAndInspect(o)" which just breakpoints there, and brings up a nicely explorable watch of the object 'o'. just to remove the modality and often crummy ui of breakpoints from the learning process.
|
# ? Oct 24, 2019 10:04 |
|
all my code works first time op sorry
|
# ? Oct 24, 2019 10:07 |
|
Cybernetic Vermin posted:debuggers could be made a lot more user-friendly really. Yeah. Data breakpoints for instance a super useful but in visual studio (considered by many as having the best c++ debugger ui) you have to input the address (in hex) and size of what you want to break on usually you do that by adding a watch on it, copying the expression, wrapping it into a &() to get the address, and then copying that in the data breakpoint window so they basically already have everything they need to implement a "add data breakpoint" in the context menu when right clicking a variable, and they have had all that for years, but nope, we still have to enter the address by hand the again it took them like 15 years just to make the project properties window resizable
|
# ? Oct 24, 2019 10:23 |
|
Cybernetic Vermin posted:debuggers could be made a lot more user-friendly really. Debugging large data structure flows in clojure is a bitch and a half to the point where you're better off running each component in your pipeline in a REPL and looking for the error instead of the debugging in intellij.
|
# ? Oct 24, 2019 12:25 |
|
Cybernetic Vermin posted:debuggers could be made a lot more user-friendly really. nah, they should just be taught how to use the debugger a little more than not at all. its really not hard to add a break point to a line and then once they figure out that basic stuff they can be taught how to add conditions and watches. telling them to write code to trigger the debugger is going to create bad habits.
|
# ? Oct 24, 2019 15:08 |
|
I had a javascript job a few years ago where all of the developers refused to learn the chrome debugger and instead just console.log'd everywhere. They used CoffeeScript too and were super concerned when I wanted to do the next project without it that learning proper javascript syntax would be some huge undertaking (it wasn't).
|
# ? Oct 24, 2019 15:13 |
|
Shaggar posted:nah, they should just be taught how to use the debugger a little more than not at all. its really not hard to add a break point to a line and then once they figure out that basic stuff they can be taught how to add conditions and watches. i agree in general, but there's been at least one time for me when a "debugger, just loving break here" in compiled code has worked far better than any amount of fancy watches or break conditions. wish I could remember the details though
|
# ? Oct 24, 2019 16:51 |
|
there are definitely situations when you need a super complicated break condition that you just cant fit into a conditional breakpoint and you can just call DebugBreak() inside of a if instead
|
# ? Oct 24, 2019 16:59 |
|
Cybernetic Vermin posted:debuggers could be made a lot more user-friendly really. https://docs.microsoft.com/en-us/dotnet/api/system.diagnostics.debugger.break?view=netframework-4.8
|
# ? Oct 24, 2019 17:23 |
|
Ciaphas posted:technically correct; all programs and programmers are bad coders are bad, programmers are good (and they already checked it worked on paper) yes that means no good software has been produced since the early 80s
|
# ? Oct 24, 2019 17:34 |
|
Zlodo posted:there are definitely situations when you need a super complicated break condition that you just cant fit into a conditional breakpoint and you can just call DebugBreak() inside of a if instead maybe but it should be absolutely your last resort and certainly not taught to new programmers.
|
# ? Oct 24, 2019 18:00 |
|
Zlodo posted:there are definitely situations when you need a super complicated break condition that you just cant fit into a conditional breakpoint and you can just call DebugBreak() inside of a if instead if you're the kind of person that does this, and I think it's fine in moderation, you should also be the kind of person who has a precommit hook to check for the existence of DebugBreak() in the commit
|
# ? Oct 24, 2019 18:25 |
|
rotor posted:if you're the kind of person that does this, and I think it's fine in moderation, you should also be the kind of person who has a precommit hook to check for the existence of DebugBreak() in the commit I don't but that's because perforce is an arse that I cannot understand the flow or functions of for the life of me bloody drat thing
|
# ? Oct 24, 2019 18:30 |
|
perforce: pretty goddamn bad, actually.
|
# ? Oct 24, 2019 19:39 |
|
debugger UI is very important and vastly underrated. having a window where you watch memory changing, one that shows you a backtrace, one that shows mixed code and assembly, one that shows local variables, and one that shows the cpu registers is a world away from a command line gdb where you have to reinspect everything you care about after every step.
|
# ? Oct 24, 2019 22:14 |
|
I wrote this thread because my current project doesn’t have a debugger because dumb reasons and I’m using printf() and I want to die
|
# ? Oct 24, 2019 22:15 |
|
debugger in Intellij idea is, as with most of the jetbrains stuff, super good.
|
# ? Oct 24, 2019 22:32 |
|
command line debuggers are impossibly stupid
|
# ? Oct 24, 2019 23:40 |
|
don’t sign
|
# ? Oct 24, 2019 23:42 |
|
the only time ive ever used a debugger is in a web dev on 3rd party code do people just not let their poo poo break and dump out gigantic traces? talking about BUSINESS code here, nothing of consequence
|
# ? Oct 24, 2019 23:44 |
|
tbh printf() is easier a lot of times debuggers are good and they have their place also they're a pain in the rear end compared to printfs, plus a lot of time you need to reduce your optimization levels to get good-quality debug tracing
|
# ? Oct 24, 2019 23:52 |
|
I'm writing a kernel module right now, and printk is way easier than trying to gdb that poo poo
|
# ? Oct 24, 2019 23:54 |
|
I will die on this hill
|
# ? Oct 24, 2019 23:55 |
|
Poopernickel posted:I'm writing a kernel module right now, and printk is way easier than trying to gdb that poo poo oh yeah I forgot about that, for kernel modules printk is a godsend e: though I'm pretty sure it's not the proper way to debug a kernel module, it would be using a remote debugger I think Private Speech fucked around with this message at 00:14 on Oct 25, 2019 |
# ? Oct 25, 2019 00:05 |
|
Poopernickel posted:I'm writing a kernel module right now, and printk is way easier than trying to gdb that poo poo Linux kernel debugging is a POS, so printk is your only option.
|
# ? Oct 25, 2019 00:50 |
|
trying to imagine debugging an os kernel and laughing at the thought
|
# ? Oct 25, 2019 01:01 |
|
i prefer pintf() where you just think really hard about it down at your local
|
# ? Oct 25, 2019 01:06 |
|
akadajet posted:trying to imagine debugging an os kernel and laughing at the thought just stop all cores with jtag and debug, what’s the problem?
|
# ? Oct 25, 2019 02:47 |
|
The Management posted:just stop all cores with jtag and debug, what’s the problem? if it's drivers for an x86/amd64 linux then that's kinda not always an option, but you can get a remote serial debug interface set up, it's just a giant pain e: windows driver debugging is ime actually a bit easier, but the inability to read through the kernel source sucks Private Speech fucked around with this message at 03:06 on Oct 25, 2019 |
# ? Oct 25, 2019 02:53 |
|
sometimes code interacts with external stuff not controlled by the debugger and that is sensitive to timing (hardware, servers, other clients in peer to peer applications, etc.)
|
# ? Oct 25, 2019 03:17 |
|
actually I only write pure functions with no side effects at all, philistine
|
# ? Oct 25, 2019 04:28 |
|
good luck debugging the lambda calculus
|
# ? Oct 25, 2019 05:38 |
|
is the xcode debugger good?
|
# ? Oct 25, 2019 05:57 |
|
BONGHITZ posted:is the xcode debugger good? it's a gui on top of lldb it works but it's not great
|
# ? Oct 25, 2019 06:05 |
|
osx has lldb and dtrace which is p good
|
# ? Oct 25, 2019 06:07 |
|
pram posted:osx has lldb and dtrace which is p good lldb is terrible.
|
# ? Oct 25, 2019 15:53 |
|
i meant dtrace is good
|
# ? Oct 25, 2019 15:58 |
|
|
# ? May 16, 2024 17:12 |
|
The Management posted:lldb is terrible. nope
|
# ? Oct 25, 2019 18:37 |