Hi Sue and Keith.
Nice doc. Looks quite polished, even from the get-go... Here are my
comments:
2.4.4:
I suggest a different name than validate_partition_info, since the
method can change the info, more than just validate it. Maybe
validate_and_align_partition_info()?
2.4.4.4:
For completeness, please add a reference to the doc which defines
om_validate_and_resize_disk_partitions(), since that function is
referenced here. Please do the same for other functions in the existing
installer as well.
Last sentence: rather than refer to the "new library", I would say the
"new library that replaces liborchestrator".
2.4.5:
Similar comment to 2.4.4. The name doesn't reflect that the function
can change its data.
2.5.2.3:
Is there a way to determine the minimum amount of required disk space
dynamically? This would remove the need to update the .image_info file
whenever the system contents change. Furthermore, the size should be
easy to determine since you know how big the running system is, and the
target will be a cpio of it.
2.5.3.3:
Similar comment to 2.5.2.3. Perhaps the recommended size can be
determined by plugging the size of the running system into a formula.
2.5.5.4:
I would at least add an assertion that all data is present, to aid in
debugging.
2.5.6.3:
Nit: I would leave out the last sentence. This spec is not for the
current GUI.
2.5.9.4:
Just curious: how will the percentages passed to update_install_status()
be determined?
2.7.1:
To be clear, you could change:
8 slices on SPARC, 16 slices on x86
to
8 slices on a SPARC disk, 16 slices on an x86 Solaris2 partition
2.8.1:
The menu will have four items, including invocation of the DDU, and a
selection to reboot.
2.8.3:
I thought we'd decided that the text-mode environment would come up with
NWAM enabled by default. That way the DDU can use the network if it has
DHCP enabled. Without any network configuration built into the DDU, at
least NWAM should be enabled.
While on the subject of network screens, can the network screens be
among the first config screens in the installer? That way, one can
start the installer to configure the network, exit back to the menu and
run the DDU to take advantage of the newly-configured network.
I would suggest a very small (1 or 2), but non-zero grub timeout. This
way one can interrupt grub to change kernel parameters for debugging
purposes if necessary. IMO, a small timeout won't be a nuisance and
could be useful.
3.1:
Nit: in order to be consistent with the second paragraph, I would change
"linked against libncurses"
to
"linked against the nCurses library (libncurses)"
Thanks,
Jack
On 09/04/09 16:12, Sue Sohn wrote:
> We have posted the design document for the Text Installer at:
>
> http://www.opensolaris.org/os/project/caiman/TextInstallerProject/design_doc_v1.4.pdf
>
>
>
> Please review and provide comments by COB Tuesday, September 15th.
>
> Thanks,
> Sue
> _______________________________________________
> caiman-discuss mailing list
> caiman-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
