Partition Table for /dev/hdd

               First       Last
 # Type       Sector      Sector   Offset    Length   Filesystem Type (ID) Flag
-- ------- ----------- ----------- ------ ----------- -------------------- ----
 1 Primary           0    81915434     63    81915435 Linux (83)           None
 2 Primary    81915435    86012009      0     4096575 Linux swap / So (82) None
 3 Primary    86012010   167927444      0    81915435 Linux (83)           None
 4 Primary   167927445   398283479      0   230356035 Extended (05)        None
 5 Logical   167927445   208893194     63    40965750*Linux (83)           None
 6 Logical   208893195   249858944     63    40965750*Linux (83)           None
 7 Logical   249858945   290824694     63    40965750*Linux (83)           None
 8 Logical   290824695   331790444     63    40965750*Linux (83)           None
 9 Logical   331790445   398283479     63    66493035 Linux (83)           None

I had posted the parted output previously. This is it:

Minor Start End Type Filesystem Flags
1 0.031 14300.046 primary ext3
3 14300.046 14998.183 primary linux-swap
2 14998.184 76316.594 extended lba
5 14998.214 30004.211 logical fat32
6 30004.242 43355.104 logical ext3 boot
7 43355.136 56705.998 logical ext3
8 56706.029 57278.627 logical linux-swap
9 57278.659 66676.025 logical ext3
10 66676.056 76316.594 logical

Above /dev/hda2 is the extended partition that contains all the logical partitions. In your case, it is /dev/hdd4. My previous posting was tough to read so you may not have noticed it. Thanks.

See my first post.
The partitioning tools *never* do the formatting. It's not their function. You decide what filesystem (type) you want to use, and do an appropriate mkfs.

Has always been the case - the GUIs may obscure this to some extent, but even they require you to specify the fs type.

I raised the question about the cfdisk output you posted, not parted output. I was complaining that the cfdisk output did not give an entry for the extended partition. If you'll look back at your original posting, you'll see that is true. The output I gave above for one of my disks is cfdisk output, which is what I understood you to be asking for. It does have an entry for the extended partition. I didn't have cfdisk on my system, so I had to go find the source and compile it. It turns out that cfdisk suppresses the extended partition for the interactive display, but it does show it when you use the "-P" flag. I had forgotten that detail.

In case it's not clear why it is strange to make a FAT32 logical partition, the only reason anyone would want a Linux FAT32 partition is for the sake of having a common filesystem with MS Windows. But a Linux logical FAT32 partition could never serve that purpose, because MS Windows can't read Linux logical partitions. Besides, these days the ordinary choice for a common system would be NTFS.

Explain yourself please.
Logical partitions are logical partitions. I use FAT32 logicals everyday to exchange data. In fact on a lot of my disks I have nothing but logical partitions.

------------------------------------------------
I raised the question about the cfdisk output you posted, not parted output. I was complaining that the cfdisk output did not give an entry for the extended partition.
-----------------------------------------------
Yes, that is the case. I too didn't know about the -P option. It is only when I used fdisk -l or parted, the extended partition showed up.

The funny thing about cfdisk is there is no way one can create an extended partition with it.

cfdisk allows only primary or logical partition to be selected.

Whenever the first logical partition is opted by the user it will be the 5th position like sda5 or hda5 even if the disk starts from new. The next available primary partition (from either sda1 to sda4) is automatically used up and will not be available in future. The creation of a primary partition thereafter will automatically blocks the growth of logical partitions and so the extended partition boundary becomes fixed.

Thus cfdisk can assist a user's understanding of how a hard disk work.

There's really nothing to explain. If what you've just said is right, then I'm wrong.