Re Computer'y terms 'formatting' and 'initialisation'

mikeymikec

Lifer
May 19, 2011
20,410
15,108
136
I was watching Ghost in the Shell 2 - Innocence again last night and got to a bit where a character talked about when a commercial android shuts down (permanently), its brain initialises itself to protect the corporation's proprietary software.

It took me back to formatting floppy disks on my Amiga, and on Workbench 1.2 the procedure for formatting was called 'initialisation', which I'm guessing is the superseded term for media formatting these days. In terms of meanings of words, I thought it was interesting because wiping storage could be seen as an ending (regarding say the data getting wiped), yet 'initialise' refers to a beginning. There's probably a word to define a word that means both opposite things at once.
 

lxskllr

No Lifer
Nov 30, 2004
59,427
9,947
126
Kind of pedantic, and you end up getting overly particular about the meaning of words, when the meaning can simply change because people said it did. "Initialization" could refer to the condition hardware rolled off the factory floor in, before the software elves got started. "Formatting" would be the first step the software elves take before loading the software.
 

mikeymikec

Lifer
May 19, 2011
20,410
15,108
136
I wasn't meaning to be pedantic, borderline philosophical if anything :) I'm not demanding that it means one thing or the other.
 
  • Like
Reactions: lxskllr

[DHT]Osiris

Lifer
Dec 15, 2015
17,183
16,333
146
I wasn't meaning to be pedantic, borderline philosophical if anything :) I'm not demanding that it means one thing or the other.
A great well of ire opens up in me when terms are used interchangeably between products when it definitely shouldn't be, such as formatting/initializing/preparing/etc. Doubly so when the vendor decides that the de facto term is inappropriate or simply too plebian and so demands a different term be used (looking at you apple).
 

Red Squirrel

No Lifer
May 24, 2003
70,167
13,573
126
www.anyf.ca
I might remember wrong but I seem to recall in some OSes it was a 2 step process for some media, had to initialize it, then format it. Initialize the disk first, then it would let you format. Under the hood it probably was creating the MBR and initial partition table.
 
  • Like
Reactions: Paperdoc

Charmonium

Lifer
May 15, 2015
10,372
3,436
136
Format means that you're creating a hidden digital structure on media that will then permit normal operations like read, write, delete.

That's a much more complex process than initializing. In that case, all you're doing is filling memory locations such that they don't just contain whatever garbage was in them after manufacture.

Initializing let's you easily determine if specific locations have "possibly" good data that's actually garbage. So if you read a location and get back whatever the initialization character was, you know that location can be overwritten.
 

mikeymikec

Lifer
May 19, 2011
20,410
15,108
136
I might remember wrong but I seem to recall in some OSes it was a 2 step process for some media, had to initialize it, then format it. Initialize the disk first, then it would let you format. Under the hood it probably was creating the MBR and initial partition table.

Interesting thought. I just looked it up in the official Amiga 500 manual and here's what it had to say:

A500 manual said:
1.2.2 Initialise
This option formats a disk for use by the Amiga. It splits the disk into the right number of
tracks/sectors and creates an empty root directory. It can also automatically create a Trashcan
drawer. To use, select the disk to format and select Initialise.
 
  • Haha
Reactions: nakedfrog

manly

Lifer
Jan 25, 2000
13,087
3,852
136
Wiki has a long article on disk formatting, with a short mention of the term "reinitialization."


Seems like Amiga decided to call it one thing in the 1980s; you'd have to dig up an IBM manual from 1973 to see what they called it when they created the 8-inch floppy disk. (Commercially launched in 1971, originally read-only!)


That's a much more complex process than initializing. In that case, all you're doing is filling memory locations such that they don't just contain whatever garbage was in them after manufacture.

Initializing let's you easily determine if specific locations have "possibly" good data that's actually garbage. So if you read a location and get back whatever the initialization character was, you know that location can be overwritten.
This doesn't sound right, although there could be some computer system use case you're thinking about that fits this description.
 

Paperdoc

Platinum Member
Aug 17, 2006
2,459
353
126
My understanding is this.
1. In the very early days of floppy disks and not much alternatives, a floppy disk off the manufacturing line had NO info of any kind. At first use a FORMAT operation actually wrote ALL of the data Tracks, each subdivided into Sectors by particular signals (and possibly filled with all zeros). As its second step that operation would then use specified tracks/sectors at the start of the disk to write real data that defined the very first Sector, wrote info to a specified set of Sectors to create the space for data to be the Directory of files, and then another set of Sectors to create the File Allocation Table which records which Sectors are in use by each file. Each computer system maker (e.g., Apple, Commodore, IBM, etc.) had their own proprietary system of how such space was defined and used (even down to the spacing of Tracks on a disk), and many had several versions of that as floppy storage evolved. So all floppies started completely blank until the Tracks and Sectors were first written. The process always used the entire capacity of the floppy disk for ONE "disk" or Volume.
2. Early hard disks were handled in virtually the same manner since each computer maker had their proprietary system for those.
3. As the IBM PC design and its several iterations began to dominate the field, a version of floppy disks became popular that already had all of its Tracks and Sectors written on the disk in one of several standard IBM layouts, but still void of data. Similarly, hard drives came that way from the factory. As each was installed, one had to run a Format operation to write to specified locations the data structures that created the Root Directory and the File Allocation Table. One item in the Root Directory was the spec for the numbers of Tracks and Sectors on the disk. But still, that Format process set the disk up to use ALL of its space for a single Volume. For hard disks, though, that Head and Sector count actually were encoded in the hardware BIOS chip of the drive along with a few other essential items. Together these were called the "Disk Type", of which there were about 47 distinct varieties.
4. As the sizes of HDD's and some floppy disks increased, the details of the File System used became outmoded and newer versions were introduced. So the Format process started to include in the Root Directory a specification of which File System it used.
5. As even larger Hard Drives were developed the system of identifying hardware parameters (Tracks, Sectors, Landing Zone, etc) changed to avoid the limited list of Disk Types. Instead, each HDD had all that data in its BIOS chip along with a process to spew that out when requested. In parallel, computer drive control systems had a routine to request that info and set itself up to deal with each HDD correctly. About the same time a new capability was introduced so that the large space available could be sub-divided into separate Volumes on the same HDD unit. To accommodate this the preparation process was altered into two steps. FIRST one had to INITIALIZE the HDD. The unit still came with all Tracks and Sectors already written by the manufacturer but with no data. Initialization wrote to a specific location at the start of the disk a brief table containing the name and size of the first Partition (the Master Partition) and its starting and ending points on the disk. It had extra space for the addition of up to three more Secondary Partitions' names and locations. Thereafter one had to run the FORMAT operation on EACH Partition individually. That would do the traditional Format tasks for that one Partition only. SOME OS's offered a tool called Preparation or even Initialization that combined the two processes, ensuring that this system would check the existing structures defined on the HDD and avoid encroaching on previously-defined Partitions. These tools also included, of course, a tool to selectively remove an existing Partition, but removing the Primary Partition would essentially remove them all.
6. Since the time when disks arrived from the manufacturers with all Tracks and Sectors created but empty, (stage 3 above), the FORMAT operation did NOT write anything to those empty tracks. So those tracks were untouched when the disk was re-Formatterd, and old data might be recoverable, although trying to organize a sequence of Sectors into a whole File is very difficult. More recently, though, FORMAT has been changed to write all zeroes into all Sectors and then go back and re-read that "info" to verify that the Sector CAN record and read back data accurately. This is a way for the drive unit to identify and mark for non-use any faulty Sectors as a part of the Format process. That way you are less likely to lose data to a "Bad Sector" nobody knew about. But this also means that recovering old data from a disk that has been re-Formatted is extremely unlikely.
 
Last edited:
  • Like
Reactions: mikeymikec

mikeymikec

Lifer
May 19, 2011
20,410
15,108
136
An Amiga double-density floppy disk can store 880KB. The PC equivalent is 720KB IIRC. Some PC floppy drives can have their hardware altered to work as Amiga drives.
 

Charmonium

Lifer
May 15, 2015
10,372
3,436
136
Wiki has a long article on disk formatting, with a short mention of the term "reinitialization."


Seems like Amiga decided to call it one thing in the 1980s; you'd have to dig up an IBM manual from 1973 to see what they called it when they created the 8-inch floppy disk. (Commercially launched in 1971, originally read-only!)



This doesn't sound right, although there could be some computer system use case you're thinking about that fits this description.
That's some leftover bs from C and probably Assembler. Originally, when you turned a computer on, you couldn't really be sure what was in the storage locations you planned to use. So it was normal practice to initialize. I'm pretty sure this was true for mainframe programs but the COBOL compiler probably did this for you. On early PCs, I'm not completely sure.