Re: Question about delay constraint to exit an ISR at DIRQL.
From: Brad Miller (nospam_at_nospam.net)
Date: 05/27/04
- Next message: JaffaLink: "Re: Developing NDIS driver for XP, starting help"
- Previous message: Tobias Erichsen: "User-mode midi-driver"
- In reply to: Ray Trent: "Re: Question about delay constraint to exit an ISR at DIRQL."
- Next in thread: Stephan Wolf: "Re: Question about delay constraint to exit an ISR at DIRQL."
- Messages sorted by: [ date ] [ thread ]
Date: Thu, 27 May 2004 19:23:53 +0100
On Thu, 27 May 2004 09:57:03 -0700, Ray Trent
<rat@synaptics.com.spamblock> wrote:
>The point of the previous poster was that the .exe files that MS ships
>*are* valid .zip files, just with a different extension. The .exe part
>(typically, in these cases) is marked as a part of the zip header in a
>clever way so that zip ignores it.
Firstly 'So what?' isn't a 'point' its a retort. And a rather rude
one at that. For me, It conveys an absolute lack of interest and a
certain amount of contempt.
However, I'm very impressed in your ability to interpret it as an
informative lesson in the innovative PE structure of executable zips!
All I could see was a curt and valueless response.
However, Since you have so gracefully furnished me with a semantic
analysis of the words 'So what' - let me point out a few things:
1. It isn't a valid .ZIP file, its a executable. A decompression
engine wrapped around an embedded zip file. This is actually a quite
a large difference.
2. My unzipping application is never called in the default action, and
so in this case it hardly makes sense to say that 'zip ignores the
executable header' rather zip IS the executable header. Its also an
uneccessary contagion vector.
Sure, its common for decompression apps to skip the header engine and
extract the embedded zip. Only if you open the file from within the
app though. If you're going to suggest that we should do this then we
obviously don't need the executable header. Getting rid of it
enforces good practice. It should AT LEAST be a download choice.
3. Regardless of all this - it has never been considered best practice
to distribute standard archives in an executable form. The more
executables you pass the more likely you will invite infection. If a
standard .zip is added to the user may be suspicious of the .exe and
may excercise caution - not so if an executable zip is infected.
If I get any replies saying any decent user has AV then you've missed
my point completely since AV is a reactive and easily beaten
technology and hardly a defence for sloppy sharing practices.
Similarly, claiming that M$ is a trusted source doesn't quite cut it -
THEY shouldn't be condoning the sharing of executable formats where
not entirely neccessary either. Lead by example.
I'm sure many of you will disagree, some quite violently.
Brad.
>> On Tue, 25 May 2004 17:54:42 +0200, Thorsten Jens <usenet@thodi.de>
>> wrote:
>>
>>
>>>On 2004-05-25, Alexander Grigoriev <alegr@earthlink.net> wrote:
>>>
>>>>I could never understand, why MS distributes documents as EXE files, rather
>>>>than ZIP files. Everybody has one or another Unzip utility.
>>>
>>>So what?
>>
>>
>> ummm, portability for starters. No need to be grumpy Thor -
>> Alexander has a valid point. I find it an uneccessary nuisance too
>> sometimes.
>>
>> Brad.
- Next message: JaffaLink: "Re: Developing NDIS driver for XP, starting help"
- Previous message: Tobias Erichsen: "User-mode midi-driver"
- In reply to: Ray Trent: "Re: Question about delay constraint to exit an ISR at DIRQL."
- Next in thread: Stephan Wolf: "Re: Question about delay constraint to exit an ISR at DIRQL."
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
