zdrv in "humongous" initramfs

'steve p' has rebuilt Wary with the main .sfs and the zdrv .sfs inside the initramfs, which is what we used to call a "humongous" Puppy. Steve posted a question:

http://bkhome.org/archive/blog2/201010/woof-uploaded-october-29-2010.html

Puppy is supposed to be able to handle a 'zdrv.sfs' placed inside the initramfs, however I looked at the 'init' script and yes, there is a logic fault. The main 'puppy.sfs' gets loaded ok, but not the 'zdrv.sfs'.

If you used Woof to build a Puppy with only a main .sfs, no zdrv.sfs, then it would be ok.

Anyway, I have fixed the 'init' script, it should now load the zdrv.sfs.

Regarding id-strings appended to 'vmlinuz' and 'puppy.sfs' (or, if use a traditional name such as wary_094.sfs then id-string is not required), this is documented in earlier blog posts:

http://bkhome.org/archive/blog2/201009/rationalised-puppy-filenames.html
http://bkhome.org/archive/blog2/201009/simplified-puppy-filenames.html

Steve also asked a question about how the init script determines the boot partition. There are (?) some unusual boot situations where you might not have 'vmlinuz' in the boot partition. Or what you want Puppy to think is the boot partition. I can't really think of such a situation off hand, but yes, any file named 'vmlinuz' with the id-string appended to the end (no carriage-return) would fool the init script into thinking that is the boot partition (and sub-directory).

---

Comments

Newest WebM decoder gains massive speedup - up to 40% faster (!)
Username: happypuppy
WebM decoding gets much faster with the newest build from git: http://blog.webmproject.org/2010/10/vp8-codec-sdk-aylesbury-release.html

Tags: woof