Opened 23 years ago
Closed 21 years ago
#237 closed defect (wontfix)
provide self diagnostic script(s)
Reported by: | Owned by: | ||
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Book | Version: | CVS |
Severity: | normal | Keywords: | |
Cc: |
Description
after finishing part of lfs system (say, chapter 5), run a script that test your LFS system so-far to make sure it's mounted properly, all programs are in place, all are statically linked, etc.
Change History (12)
comment:1 by , 23 years ago
comment:2 by , 23 years ago
comment:3 by , 23 years ago
Owner: | changed from | to
---|---|
Status: | new → assigned |
It was a P5 but there've a lot of problems lately with regards to Glibc installations which may have to do with bad chapter 5 installations. Let's just start working on this stuff and get at least log files out there so people can compare it against their systems. Adding a script to test for non-static programs would be helpful too.
comment:5 by , 22 years ago
Priority: | lowest → high |
---|
comment:6 by , 22 years ago
Owner: | changed from | to
---|---|
Priority: | high → lowest |
Status: | assigned → new |
comment:7 by , 22 years ago
Cc: | added |
---|
comment:9 by , 21 years ago
Priority: | lowest → normal |
---|
Raising to P3 - would be nice to get done soon. The checklfs mentioned in this bug should not be used - it was written with the $LFS/static procedure in mind.
Greg had posted something (URL eludes me at the moment) a while back which may be a better basis to start from again.
comment:10 by , 21 years ago
Cc: | removed |
---|
comment:11 by , 21 years ago
Greg had posted something (URL eludes me at the moment) a while back which may be a better basis to start from again.
No I didn't, I ended up implementing a few "sanity checks" and placed them into the book at strategic places. This would seem more logical and educational to me than providing self-diagnostic scripts. The current Ch 5 build method is a lot saner than the old one, so the need for self-diagnostic scripts is not really there anymore IMHO.
I recommend closing this bug.
comment:12 by , 21 years ago
Resolution: | → wontfix |
---|---|
Status: | new → closed |
It so seems there's nothing left to really do then now that the sanity checks are in place. Anything else that could be done (such as verifying an installation by checking against a known log file) falls under an ALFS implementation if you ask me. Closing this bug.
What we can also provide here are the "configure", "make" and "make install" outputs from every package we install. This will give people a reference to check their own output against. The more compile-savvy people may get a clue from the configure output difference to fix some of their own problems.