fSekrit not working on an old XP PC

[MerleOne]

Hi,

A friend of mine says fSekrit.exe v1.40 won't launch on his XP PC.  I'll try to get a capture of the error message and will report here, but is there any known issue on some old XP PC ? I personally never encountered such a behavior.

Thanks.

[c.gingerich]

I have an app that is similar if you want to check it out. It's called Self Note and it supports tabs to allow multi notes. It is free also.

http://starpunch.net/?page=12

[MerleOne]

I have an app that is similar if you want to check it out. It's called Self Note and it supports tabs to allow multi notes. It is free also.

http://starpunch.net/?page=12

-c.gingerich (October 01, 2012, 10:48 AM)

Thanks, will try it.  But I'd still like to know why/if fSekrit has an issue on this old PC.

[Stoic Joker]

FSekrit should run just fine on anything from Win2000 up.

[f0dder]

FSekrit should run just fine on anything from Win2000 up.

-Stoic Joker (October 01, 2012, 02:56 PM)

Actually from Win9x and up - I had to go through a few pains to make that work (and then there were a couple of versions where I accidentally botched the win9x support :$).

So, no, there should be nothing preventing fSekrit working on XP. In fact, I've just tested 1.40 on XP-SP3 (under vmware), and it runs just fine. fSekrit also doesn't use any of the fancy instruction sets, so in theory you should be able to run it on... dunno, probably even a 386

Please do try getting a capture of the error message. Perhaps antivirus? Some of those products don't like fSekrit, either because it's compressed with PECompact, or because of heuristics (copying itself to %TEMP% and launching from there, writing to it's own executable).

c.gingerich, from your site: "256 bit RSA_AES encryption with password". That seems like a weird encryption scheme to choose for an application like this?

[c.gingerich]

F0dder: I'm not good at writing encryption. There was a post on a forum I go to so I decided to use that as it was the only good one I could find that I did not need an external library for.

[MerleOne]

FSekrit should run just fine on anything from Win2000 up.

-Stoic Joker (October 01, 2012, 02:56 PM)

Actually from Win9x and up - I had to go through a few pains to make that work (and then there were a couple of versions where I accidentally botched the win9x support :$).

So, no, there should be nothing preventing fSekrit working on XP. In fact, I've just tested 1.40 on XP-SP3 (under vmware), and it runs just fine. fSekrit also doesn't use any of the fancy instruction sets, so in theory you should be able to run it on... dunno, probably even a 386


Please do try getting a capture of the error message. Perhaps antivirus? Some of those products don't like fSekrit, either because it's compressed with PECompact, or because of heuristics (copying itself to %TEMP% and launching from there, writing to it's own executable).

c.gingerich, from your site: "256 bit RSA_AES encryption with password". That seems like a weird encryption scheme to choose for an application like this?

-f0dder (October 01, 2012, 04:42 PM)

Thanks, I'll look into it .

[f0dder]

F0dder: I'm not good at writing encryption. There was a post on a forum I go to so I decided to use that as it was the only good one I could find that I did not need an external library for.

-c.gingerich (October 01, 2012, 07:30 PM)

- I just have a somewhat hard time seeing how RSA fits in. Your notes are also self-contained and portable, right? Which means you're storing both RSA public+private keys as well as AES symmetric key in the note? Even if the RSA keys are encrypted, I'm not sure it's a good idea publishing the private part (even if that might be more of a theoretical than practical consideration).

I assume you use the user passphrase to generate an AES-256 key to en/decrypt the RSA key, then use the RSA key to en/decrypt a randomly generated AES-256 key, which is then used for the bulk text en/decryption?

PS: Self Note really doesn't like being UPX-decompressed :-)

[c.gingerich]

True. I'll have to look to see if I can find another way.

Nope, it breaks if it is decompressed... trying to hack Self Note, eh? 

[f0dder]

Nope, it breaks if it is decompressed... trying to hack Self Note, eh? 

-c.gingerich (October 02, 2012, 08:37 AM)

Just poking around, as usual

[MerleOne]

Hi all,

I am happy to report that the issue is solved.  Actually, it was a MS Word error.  I had sent the file as a .doc file, to have the webmail accept it.  The receiver had to rename the attached .doc into a .exe to be able to read the message.  But on his PC, extensions of known types were hidden (default XP setting), so when the file (foo.doc) was renamed into a .exe, it was actually renamed into foo.exe.doc. Hence the error message.

The setting has been changed on this PC, and now it works fine.

[f0dder]

Nice to hear

Would probably have been better to .zip up the file? Unless, of course, the webmail either disallows .zip or checks contents for executables as well (*sigh*).

[MerleOne]

Nice to hear

Would probably have been better to .zip up the file? Unless, of course, the webmail either disallows .zip or checks contents for executables as well (*sigh*).

-f0dder (October 03, 2012, 10:10 AM)

Yes, I think it does check .zip for executable files and blocks the mail if one is found.  I could encrypt the zip, in that case I just get a warning, but it's faster to rename as a .doc file.

[x16wda]

Now, everyone silently send up an appropriate word of thanks to Mr. Gates for his decision to hide file extensions by default.

[f0dder]

Now, everyone silently send up an appropriate word of thanks to Mr. Gates for his decision to hide file extensions by default.

-x16wda (October 04, 2012, 06:57 PM)

Dunno if that was Bill's decision, but it's definitely one of the less wise things Microsoft have done. One could argue that we shouldn't use file extensions instead, but rely on file metadata instead, but I'm not convinced about that - extensions survive across different filesystems and transport protocols, are fast to deal with, and are conceptually simple