Re-compressing F96-CE_2 sfs's using zstd made the system faster on my 11 yo laptop

[Feek]

F96-CE (xz compression) runs a bit slower than vdpup64 (zstd compression) on my 11 year old Thinkpad (DDR3).

The mksquashfs command in F96-CE_2 does not seem to support the zstd compression type.
I therefore re-compressed with the zstd type from vdpup64.
After booting, the zstd decompression in ram seems to be fully functional and the F96-CE_2 then behaves much faster on my machine (starting and running programs, searching the file system).

1 small drawback: sfs files compressed in zstd take a bit more space.
However, there is plenty of disk space, so why not have a faster system when the option exists.

Thanks for developing the renovated fossa.

[backi]

@Feek wrote:

The mksquashfs command in F96-CE_2 does not seem to support the zstd compression type.
I therefore re-compressed with the zstd type from vdpup64.

Maybe this Tool could be helpful.
PackIt-1.22 - a flexible Archiving/Compressing GUI
https://oldforum.puppylinux.com/viewtopic.php?t=89211

packit-1.22.pet:
https://oldforum.puppylinux.com/downloa ... p?id=71048

[Clarity]

Forgive me for asking, but

• how does one measure performance of this report?

• which boot options would be used to show the performance difference?

• Is this referencing boot operations or normal desktop operations or some particular subsystems?

Curious as how this was measured.

[Feek]

Forgive me for asking, but

• how does one measure performance of this report?

• which boot options would be used to show the performance difference?

• Is this referencing boot operations or normal desktop operations or some particular subsystems?

Curious as how this was measured.

Clarity,
thanks for your questions.
I meant normal desktop operations (opening and running applications, searching the file system).
I don't know if it can be measured in any way.... I can just see and feel the speed of the system when using it .

To be honest, FP-96 with xz compression works normally on my hardware. If I didn't have the comparison with vdpup64 and if I didn't have the info about zstd from dimkr, I wouldn't have thought of it at all.
Besides, it probably also depends on the hardware. On new and very powerful machines, the difference between xz and zstd can be imperceptible.

[dimkr]

A main SFS compressed with zstd level 19 tends to be around 8-12% bigger compared to xz with the same block size. On my Thinkpad X220, which doesn't feel slow and runs modern browsers, IDEs and virtual machines just fine, applications open instantly with zstd, but ~1s with xz. On my Eee PC 1001PX applications take ages to start with xz, but usually 1-2s with zstd. In addition, xz decompression keeps the single CPU core busy, so applications freeze every time I start an application or read big files from the main SFS: this doesn't happen with zstd.

Personally, I think the modest increase in size is worth it. Yes, this increase in size also translates into increased RAM usage with pfix=copy or slower read from disk if you use pfix=nocopy and boot from and old and slow mechanical drive, but the difference in responsiveness is hard not to notice.