poizan42

41 ms·

481.
by poizan42 10y ago
As long as it's correctly prefixed with X- I think it should be fine.
482.
by poizan42 10y ago
What is TRR?
483.
by poizan42 10y ago
That uClinux got merged into the mainline kernel and is basically the nommu support today, at least as far as I understand it.
484.
by poizan42 10y ago
- also ATmega1284P and ATmega644A mentioned in the article doesn't support external code memory nor changing the code memory after boot. However they do support a boot loader that can load code initially. So you would have to implement
485.
by poizan42 10y ago
Linux can already run on MMU-less systems (CONFIG_NOMMU). GCC already supports AVRs, though you may have to tweak the backend a bit so ints are 32-bit wide etc. These are not the big issues in porting the Linux kernel to an AVR. The major p
486.
by poizan42 10y ago
> Disassembly by a team that results in new code written by the same team is not legal in any country that enforces copyrights. It's not illegal in itself. But you are easily opening yourself up to claims that the code you have writ
487.
by poizan42 10y ago
You mean, like, USB OTG that basically every smartphone has?
488.
by poizan42 10y ago
"A while ago"? That was at least back in NT 3.5
489.
by poizan42 10y ago
> FWIW, old Office documents were actually CFBF (Compound File Binary Format) files That's actually the "new" binary formats. The usage of CFBF seems to have been introduced in Office 4.2 (at least Excel 5.0 is the first E
490.
by poizan42 10y ago
I can tell from experience that the Linux implementation of ASLR however is completely worthless. Why? Because the executable you launch itself isn't randomized. The executable must be completely trivial for there to not be enough usab
491.
by poizan42 10y ago
It's for ETW. It's not doing any telemetry unless you use it for telemetry. Also I think they removed the initialization code from being linked in by default in the latest update.
492.
by poizan42 10y ago
Every picture contains some amount of noise. It should at least in theory be equivalent to the Halting problem to decide whether this noise was generated by an algorithm or truly random. But there's a lot of practical issues here. If y
493.
by poizan42 10y ago
> It would be awesome if you could easily write code that intermixed statements dealing with things at a lower level (malloc, pointer arithmetic, etc.) and those that access high level abstractions (GC'ed objects, virtual dispatch,
494.
by poizan42 10y ago
Not if you use the /clr:pure flag
495.
by poizan42 10y ago
Completely offtopic, but am I the only one who don't like the usage of ! in the strcmp checks? To me conceptually !cond is checking for falsity, but the return value from strcmp is not logically a boolean
496.
by poizan42 10y ago
So if you have a million vectors then you have wasted 8GB. It only makes sense if you have a few large vectors.
497.
by poizan42 10y ago
What do you mean by that? If you scramble the keys as they enter the input system then you can't write anything. If you want the user to be able to use use the keyboard then you can also read what's getting sent to userspace.
498.
by poizan42 10y ago
The amount of name reuse in software in recent years really frustrates. It makes it so much harder to find the information you are looking for. Besides that I really don't get why perforce are renaming their version control system yo H
499.
by poizan42 10y ago
That should be exposed as an option so the user can choose whether they need performance or reliability.
500.
by poizan42 10y ago
The version they consider "stable" is rarely updated. That just means they have a very high standard for when it's considered stable. If you look through the changelog you'll see that the largest gaps between any release
501.
by poizan42 10y ago
> 7zip is not very well maintained, whereas PeaZip has regular releases. 16.00 2016-05-10 15.14 2015-12-31 15.12 2015-11-19 15.11 beta 2015-11-14 15.10 beta 2015-11-01 15.09 beta 2015-10-16 15.08 be
502.
by poizan42 10y ago
Android has Log.WTF which of course stands for What a Terrible Failure (and totally not that other word, the similarity is a complete coincidence /s )
503.
by poizan42 10y ago
> thats not true if you inspect the details of the specific ABI and cheat a little or use some inline assembler... Some x86 C compilers supports calling conventions that lets you put things in registers, e.g. Microsofts "fastcall&qu
504.
by poizan42 10y ago
The claim was: The same goes for C++ or Swift. Getting the function pointer for a method can be involved, but all functions adhere to the C ABI. This is clearly false since at least one C++ compiler generates functions that does not adhere
505.
by poizan42 10y ago
At least MSVC on x86 uses "thiscall" by default which puts the this pointer in ecx. You cannot call those functions from C without specific compiler support.
506.
by poizan42 10y ago
I would think that mechanical harddrives will become a niche product for workloads requiring frequent writes in the not so far future. Also as large flash is getting cheaper it becomes more feasible to use a lot of it for over-provisioning
507.
by poizan42 10y ago
So? Use CreateProcess on NT then Edit: Actually it already does: https://github.com/ImageMagick/ImageMagick/blob/master/Magic...
508.
by poizan42 10y ago
There are already 2TB SSDs on the market, although quite expensive, but give it a few years and mechanical hard drives are a thing of the past.
509.
by poizan42 11y ago
Those processes are weird compared to normal processes. The main place the user mode information about a process is tracked in a normal windows process is through the Process Environment Block (PEB). You get the list of modules and commandl
510.
by poizan42 11y ago
I found this much better at explaining what went down before the 3.6 release: https://tung.github.io/DynaHack/2016/02/18/stepping-down-fro...

More ›