Re: /dev/tcp

"Carlos E.R." <[email protected]>
Newsgroups comp.os.linux.misc
Organization Tebibyte_Retro_Gaming
Message-ID <[email protected]>
On 2026-08-24 06:31, c186282 wrote:
> On 8/23/26 07:50, Carlos E.R. wrote:
>> On 2026-08-23 09:37, c186282 wrote:
>>> On 8/23/26 00:01, Rich wrote:
>>>> vallor <[email protected]> wrote:
>>>>> At Thu, 20 Aug 2026 23:30:40 -0000 (UTC), Lawrence D’Oliveiro 
>>>>> <[email protected]> wrote:


>>>    ANYway, /dev/tcp does have uses. It's worth exploring.
>>>    Maybe YOU can find a use to suit YOUR needs/desires.
>>>    I used it for a custom disk-scan/wipe app (had to
>>>    use a 'C' module because huge track/cluster/byte numbers
>>>    were needed for modern mag drives - the main app was
>>>    in Lazarus for the pretty display and buttons).
>>>
>>>    Note Lazarus/FPC *says* it supports huge seek numbers
>>>    but DOESN'T ... limited to maybe 4gb.
>>
>> I wrote a program that writes bytes on big partitions.
>>
>>
>> excerpted:
>>
>> const
>>       sourcefilename = './BigRandom';
>>       rawdest: string = '';     // example: '/dev/sda11'
>>
>>       shuflecount = 1024;
>>       arraysize = 1023;
>>       ChunkSz=1048576;
>> type
>>       tCacho= array [1..ChunkSz] of byte;
>>
>> var
>>     Fin, Fout: file of tCacho;
>>     gotresult: Word;
>>
>>     BigData : array [1..shuflecount] of tCacho;
>>     OutputIndex : Int64;    // counts MiB written.
>>
>>
>> begin
>>
>> // writes one MiB
>>        {$I-}write(Fout,BigData[RandomIndex]);{$I+}
>>        gotresult:= ioresult;
>>
>>
>> The purpose of the program is to fast fill a partition or disk with 
>> nearly true random data. It is as fast as dd could be.
> 
>    Um ... how many GB could that cope with ?

I have done entire disks sized terabytes. But I don't actually use "seek".

> 
>    'Seek' counts can sometimes be supplimented
>    by an 'offset' value - so basically you're
>    getting the mult of two doubles. I didn't
>    do it that way, and thus needed very large
>    sector/track/cluster integers. An external 'C'
>    pgm using really large integer types was needed.

Can you seek not a byte, but a record or array?

Seek(var F; N: Int64)

File can be a file of some variable that is not a byte. And

type Int64 = - 9223372036854775808..9223372036854775807;

(why signed?)

Why not QWord?

type QWord = 0..18446744073709551615;


>    Anyway, it worked.
> 
>    My app could ALSO just LOOK at data, kinda put
>    it into a "ghex"-looking sidebar (but with
>    ASCII translation if possible). So you could
>    just look, OR zap. Anyway, nothing TOO complex
>    aside from the sheer size of modern mag disks.
> 
>    As useful utility. I encourage people to write
>    their own.
> 
>    DID add one odd, maybe questionable, feature ...
>    you could spec on the CL x-number of gigs to
>    TOTALLY wipe at the beginning ... and then
>    a couple numbers for "shotgun" blasting of
>    later tracks. Most of the file-table/system
>    stuff tends to be early on, and data with
>    holes in it is "less useful". This made the
>    "wipes" much faster.
> 


-- 
Cheers, Carlos.
ES🇪🇸, EU🇪🇺;
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.