> No. My opinion is: "Don't mess with date/timezone, at all". Not because it
> is hard to do or easy to screw up. Actually, carefully written it is quite
> easy and safe.
WARNING: Skip to the last two sentences for the technical stuff.
90% of this post is a philosophical answer to your innuendo & questions! :)
I agree. I can already set the time zone (as I showed you with examples).
I'm only trying to execute the tzutil command from within Dialog to do so.
I love Dialog.
I love the power of Dialog.
But I am not a programmer.
If I was a programmer, I'd be telling you how to set the time zone.
Not asking for advice on how to run the tzutil command from within Dialog.
> IMHO, most people who "require" such a function are just trolls who try
> to avoid certain kill filters. In my first answer I mentioned a possible
> reason outside this scope. (If I wouldn't still give you the benefit of
> doubt, we wouldn't discuss this matter, at all.)
How do I defend myself from such an accusation?
If I wanted to troll, I would have done so long ago. I realize that anyone
who wishes his privacy does similar stuff as do trolls, but wasting time on
this troll issue is like asking me if I rob little old ladies, or if I'm
gay, or if I'm a communist.
The evidence is as flimsy on all three.
If I wanted to troll, just as if I had wanted to rob little old ladies, I
wouldn't need to nor care to set the time zone more elegantly from within
Dialog to do so.
> But in any case, such obfuscation techniques usually don't protect in the
> long run. Word choice, spelling, grammar, presence or absence of certain
> typos help to classify postings, even if they seem to come from different
> authors. These posts can then be analyzed by timing patterns: Core sleeping
> time, work time, lunch, time for shopping, holidays, feast days and so on.
> This helps to guess the timezone.
Same as above.
The evidence you have that I wish to troll is the same as the evidence you
have that I wish to rob little old ladies or that I'm a communist or that
I'm gay. There's no difference in the accusations with respect to evidence.
If I was trolling, or robbing little old ladies, or a communist, or gay, or
whatever you want to accuse me of on zero evidence, would my denial make
any different do you anyway.
You can either believe I care about privacy, or you don't believe it.
If you don't believe it, then stop responding to my queries.
And I'll stop responding to yours.
It's really that simple.
You either trust me at my word or you don't.
But let's not waste time on this issue if you don't trust my word.
I value your knowledge and you know that because I specifically called you
out as one of the top three Dialog-knowledgeable people in this ng.
But if we spend the entire conversation accusing each other of robbing
little old ladies on zero evidence, then let's stop wasting everyone's time
who has to wade through this dialog.
> When simulating postings from a certain timezone, you'd also need to
> simulate local language patterns. Changing to flexible timezones would
> single out your postings, if you (maybe automatically) choose timezones
> where usually nobody posts to the Internet (on a specific subject). So,
> neither automatic nor manual changes prevent from creating identifiable
> patterns. Most people would neglect their own timezone (at all or more
> often than others). Again, a pattern by itself.
Ah. I think I understand your objection to header privacy.
It's not only philosophical. It's also logical.
I'm not sure you're aware that there are different kinds of privacy.
a. There is header privacy
b. There is body privacy
c. (there are plenty of others, e.g., newsgroup privacy)
Privacy is not always simple just like, I'm sure you're aware, browser
fingerprinting privacy is not always simple.
In terms of privacy, the HEADER is completely different than the BODY.
Personally, I'm not seeking BODY privacy.
I'm only seeking HEADER privacy.
Why, you would naturally ask?
My answer is the threat model.
Who is my antagonist?
Whom am I hiding from?
My answer is simple, and logical, and (I believe) perfectly reasonable.
But if you want to make this a conversation of the philosophy of the
various threat models, this thread is going to go off topic fast.
A. The threat model from robotic harvesters (like Google, Facebook, Amazon,
etc.) is completely different than the threat model from people (like you,
Michael Bauerle or Jernej Simoncic), which itself is completley different
than the thread model from a well-funded adversary such as a government
entity.
You know, of course, for example, that China has been harvesting Linked-In
information en masse, right?
How are they harvesting all that private information?
They're doing it with robots, of course,
They don't have individuals reading everyones' BODY of their messages and
then determining who those people are from the BODy of the message.
People are sufficiently sloppy that all a robot adversary needs is the
HEADER of the Usenet message.
For privacy against the robotic threat model, the BODY is probably
meaningless. Sure, a really well funded adversary will find that every
message has a "unibomber manifesto" fingerprint in slang, misspellings,
idiom, word choice, etc.
My threat model for this question of privacy is the robotic harvesters.
Those robotic harvesters are most likely HEADER harvesters.
We can wax philosophically whether that is a logical assessment of the
threat model, and you can rightly ask if I know of any current robotic
header harvesters, where, to go more deeply into this type of adversary, we
should change the subject line to something like:
OT: Philosophy 101. What is the robotic harvester adversary threat model?
> Intersecting (even small) unusual answering time lags with network outage
> in certain regions of the Internet can narrow down your location further.
> Some of these lags are probably induced on purpose. - Not because of you,
> specifically, but to identify a whole bunch of users.
As I said, there are MANY ways to "fingerprint" a user.
The HEADER is one of the easiest ways though.
We could discuss the philosophy of header privacy in other terms but again,
we'd be engaging in a conversation of no ending if we ask what value the
specific header lines provide, in reality.
For my question, for example, does it matter what nntp server I use?
For my question, which is a technical one, does it matter what my name is?
Does my email address make any different to the answer to the question?
Why would my nntp posting host matter to a human for a technical question?
What technical significance to the question is my Time Zone?
Why would my Usenet agent matter when I said it clearly in the OP body?
I can easily argue, logically so, that the header provides almost no value
to the respondent outside of the subject line.
I can even more easily argue that the header provides a robotic adversary
with lots of value, interestingly, also outside of the subject line.
In fact, what's interesting, philosophically, is that the subject line
provides almost zero value to robotic adversaries, while the subject line
is ALL the value to the human respondent.
Hence
A. The SUBJECT line is the only meaningful header for technical questions.
B. The rest of the lines provide almost zero value to human respondents.
C. Yet, the rest of the lines are what robotic adversaries would harvest.
> The quickest answer times with long typed texts done by you can be an age
> indicator (typing speed). The list goes on and on.
Again, I make zero attempt to obfuscate the body of the message.
Obviously I'm erudite, for example. Obviously I'm well educated.
Obviously I'm experienced at Usenet. Obviously I know you are well known
for Dialog expertise. (I can list the obvious for a long time.)
Do you think I can't purposefully misspell words if I wanted to?
But what value would that provide me when I ask a technical question?
Do you think this is some kind of Nigerian Bank Scam?
I have said clearly that I wish for privacy in my headers.
I have explained above (albeit only recently) that the adversary is a
robot.
Do I know of any of these robotic adversaries?
Nope.
Do I think they exist?
Yep.
Do you?
I don't know. Do you?
>> I noted you said that this timezone (and date) is provided to
>> OnBeforeSendingMessage so it might be too late to run the tzutil command in
>> OnBeforeSendingMessage.
>
> The whole raw ready-for-transmission message (headers and body) is handed
> to OnBeforeSendingMessage. If you replace the whole date line with sth.
> else (no matter, how the new date has been deduced/created) the new date
> and time will be the ones seen by any server and client, afterwards.
This is really the FIRST on-topic sentence in your entire post! :)
I need to repeat that I didn't want to mess with the entire date line.
I only wanted to change the time zone.
Since you've already explained that there is no way to change just the time
zone in the date line (and I'm not an idiot so I understood that), then I
give up on that approach.
The only other viable approach that I can think of is simply to set Dialog
to run a Windows command before I compose the article.
There is one other approach, but since I am not a programmer, it's out of
my league, which is to run a sequence of 'date', 'time', and 'tzutil'
commands and use the results from those commands to re-define the entire
date header - which I've said many times is not what I wish to do.
>> I guess I could set a custom script in Dialog to simply call tzutil.exe.
>> Do you know, offhand, the syntax to call a Windows command from a script?
>
> Reference the ShellExecute function from shell32.dll.
Thank you for that shell-execute advice to run the tzutil.exe command from
Dialog, as this is only the SECOND on-topic sentence in your post. :)
Searching, the n.s.r archives first
https://groups.google.com/forum/#!forum/news.software.readers
I find only one instance of those two keywords, from the year 2002
https://groups.google.com/d/msg/news.software.readers/JxcXUgy_jQw/0mjETtpHuxYJ
which even the venerable Marcus Monnig participated on in those days!
-------------------
http://dialog.datalist.org/scripts/ScriptShowHelpMaximized.html
http://dialog.datalist.org/scripts/ReadArticleInGoogleGroups2.html
-------------------
Program ShowHelp;
uses
Forms;
const
SW_NORMAL = 1;
SW_MAXIMIZE = 3;
SW_MINIMIZE = 6;
function ShellExecute(hwnd: LongWord; lpOperation: pchar;
lpFile: pchar; lpParams: pchar;
lpDir: pchar; nShow: integer): LongWord;
external 'ShellE...@shell32.dll stdcall';
Begin
ShellExecute(Application.MainForm.Handle, 'open',
pchar(ExtractFilePath(Application.ExeName) +
'dialog.chm'), '', '', SW_MAXIMIZE);
End.
-------------------
Is that the kind of use model for "tzutil.exe" you were kindly suggesting?