Re: [alfs-log] [ALFS Trac] #1700: Remove && in BLFS scripts

2019-05-16 Thread ALFS Trac
#1700: Remove && in BLFS scripts
-+---
 Reporter:  pierre   |   Owner:  pierre
 Type:  enhancement  |  Status:  assigned
 Priority:  normal   |   Milestone:  3.0
Component:  ablfs| Version:  SVN
 Severity:  normal   |  Resolution:
 Keywords:   |
-+---
Changes (by pierre):

 * owner:  alfs-log@… => pierre
 * status:  new => assigned


Comment:

 I think I'll remove the &&
 except when preceded by [...] or
 test ...

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


[alfs-log] [ALFS Trac] #1721: Some installation instructions are not generated correctly when using porg

2019-05-16 Thread ALFS Trac
#1721: Some installation instructions are not generated correctly when using 
porg
+
 Reporter:  pierre  |  Owner:  alfs-log@…
 Type:  defect  | Status:  new
 Priority:  normal  |  Milestone:  3.0
Component:  ablfs   |Version:  SVN
 Severity:  normal  |   Keywords:
+
 Generating installation instructions for BLFS, taking into account whether
 running as user or root, whether using porg, destdir package management or
 nothing is a complicated task because markup in the book is not very
 accurate:
 - The only indication telling a  tag is an installation
 instructions is that  has a role="root" attribute, and that the
 tag occurs in a sect2 with role="installation" attribute. But some of
 those tags are not installation instructions: for example, they can be
 user or group adding, or some database updating (mime, desktop, icons,
 ...).
 - There may be several  tags, which should be grouped
 in order to generate the porg logging correctly. But if they are
 interspersed with config instructions, this is hard...

 This may lead to buggy instructions, for example beginning with porg
 opening stanzas and finishing without closing stanzas, or converse...
 Also, some single quotes may be escaped (as `'\''`), when not needed,
 leading to errors when running the script...

 This can be improved in two directions:
 - improve the stylesheet generating those instructions.
 - improve the book itself so that different types of instructions are more
 clearly identified (and of course adapt the stylesheet to use that).

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1719: Some files may be not logged when using cmake+porg

2019-05-16 Thread ALFS Trac
#1719: Some files may be not logged when using cmake+porg
+-
 Reporter:  pierre  |   Owner:  alfs-log@…
 Type:  defect  |  Status:  new
 Priority:  normal  |   Milestone:  3.0
Component:  jhalfs  | Version:  SVN
 Severity:  normal  |  Resolution:
 Keywords:  |
+-

Comment (by pierre):

 There are two possibilities for avoiding that (maybe others, of course):
 - One brutal way is to remove files when they are there (use '''porg -rb
 $JH_PKG_DIR''' before install, or even before cmake).
 - Another possibility is to use a DESTDIR install (hopefully, the DESTDIR
 is empty).

 There are several drawbacks with the first approach: some packages are
 needed for compiling/installing, so cannot be removed from the system,
 even just before installation (occurs only in LFS, I think. One such
 package is coreutils. Of course, also glibc, and a couple of others).
 Also, some packages change library versions at each update, so that
 removing the old library breaks all packages compiled against that library
 (example poppler).

 I'm sure the issue does not exist with autotools: existing files are
 overwritten. I do not know for meson.

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1720: unset MAKELEVEL in scriptlets

2019-05-16 Thread ALFS Trac
#1720: unset MAKELEVEL in scriptlets
+---
 Reporter:  pierre  |   Owner:  pierre
 Type:  defect  |  Status:  assigned
 Priority:  normal  |   Milestone:  3.0
Component:  jhalfs  | Version:  SVN
 Severity:  normal  |  Resolution:
 Keywords:  |
+---
Changes (by pierre):

 * owner:  alfs-log@… => pierre
 * status:  new => assigned


--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1714: Add some pages from "After LFS Configuration Issues" to ablfs

2019-05-16 Thread ALFS Trac
#1714: Add some pages from "After LFS Configuration Issues" to ablfs
-+-
 Reporter:  pierre   |   Owner:  alfs-log@…
 Type:  enhancement  |  Status:  closed
 Priority:  normal   |   Milestone:  3.0
Component:  ablfs| Version:  SVN
 Severity:  normal   |  Resolution:  fixed
 Keywords:   |
-+-
Changes (by pierre):

 * status:  new => closed
 * resolution:   => fixed


Comment:

 Done at r4064

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1700: Remove && in BLFS scripts

2019-05-16 Thread ALFS Trac
#1700: Remove && in BLFS scripts
-+-
 Reporter:  pierre   |   Owner:  alfs-log@…
 Type:  enhancement  |  Status:  new
 Priority:  normal   |   Milestone:  3.0
Component:  ablfs| Version:  SVN
 Severity:  normal   |  Resolution:
 Keywords:   |
-+-

Comment (by pierre):

 This should be easy enough. of course, only && at the end of lines should
 be removed. But it is possible the a dev intends to have:
 {{{
 long instruction A &&
   instruction B
 }}}
 to be equivalent to:
 {{{
 if long instruction A; then
   instruction B
 }}}
 How to know? Ask dev to not use the && construct? When ''long instruction
 A'' is of the form ''[ -letter something ]'', that may be detected, but
 otherwise?

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1713: Use sect1info tags to get tarball names and versions in LFS

2019-05-16 Thread ALFS Trac
#1713: Use sect1info tags to get tarball names and versions in LFS
-+---
 Reporter:  pierre   |   Owner:  pierre
 Type:  enhancement  |  Status:  assigned
 Priority:  normal   |   Milestone:  3.0
Component:  jhalfs   | Version:  SVN
 Severity:  normal   |  Resolution:
 Keywords:   |
-+---
Changes (by pierre):

 * owner:  alfs-log@… => pierre
 * status:  new => assigned


--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


[alfs-log] [ALFS Trac] #1720: unset MAKELEVEL in scriptlets

2019-05-16 Thread ALFS Trac
#1720: unset MAKELEVEL in scriptlets
+
 Reporter:  pierre  |  Owner:  alfs-log@…
 Type:  defect  | Status:  new
 Priority:  normal  |  Milestone:  3.0
Component:  jhalfs  |Version:  SVN
 Severity:  normal  |   Keywords:
+
 jhalfs (and blfs extension) run scriptlets from a Makefile, so that
 MAKELEVEL=1 when starting "make" in a build directory. This can lead to
 weird errors, as explained in postgresql INSTALL file:
 {{{
 If you want to invoke the build from another makefile rather than
manually, you must unset MAKELEVEL or set it to zero, for instance
like this:
 build-postgresql:
 $(MAKE) -C postgresql MAKELEVEL=0 all

Failure to do that can lead to strange error messages, typically
about missing header files.
 }}}

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1721: Some installation instructions are not generated correctly when using porg

2019-05-16 Thread ALFS Trac
#1721: Some installation instructions are not generated correctly when using 
porg
+-
 Reporter:  pierre  |   Owner:  alfs-log@…
 Type:  defect  |  Status:  closed
 Priority:  normal  |   Milestone:  3.0
Component:  ablfs   | Version:  SVN
 Severity:  normal  |  Resolution:  fixed
 Keywords:  |
+-
Changes (by pierre):

 * status:  new => closed
 * resolution:   => fixed


Comment:

 Fixed at r4093

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


[alfs-log] [ALFS Trac] #1719: Some files may be not logged when using cmake+porg

2019-05-16 Thread ALFS Trac
#1719: Some files may be not logged when using cmake+porg
+
 Reporter:  pierre  |  Owner:  alfs-log@…
 Type:  defect  | Status:  new
 Priority:  normal  |  Milestone:  3.0
Component:  jhalfs  |Version:  SVN
 Severity:  normal  |   Keywords:
+
 When a file is already present on the system, cmake does not reinstall it.
 SO porg does not log it.

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1720: unset MAKELEVEL in scriptlets

2019-05-16 Thread ALFS Trac
#1720: unset MAKELEVEL in scriptlets
+-
 Reporter:  pierre  |   Owner:  alfs-log@…
 Type:  defect  |  Status:  new
 Priority:  normal  |   Milestone:  3.0
Component:  jhalfs  | Version:  SVN
 Severity:  normal  |  Resolution:
 Keywords:  |
+-

Comment (by pierre):

 I guess MAKELEVEL=1 is harmless in most cases, but for the few cases it is
 not (postgresql is one such case), we should unset MAKELEVEL (or set it to
 0) in the generated scriptlets.

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page


Re: [alfs-log] [ALFS Trac] #1721: Some installation instructions are not generated correctly when using porg

2019-05-16 Thread ALFS Trac
#1721: Some installation instructions are not generated correctly when using 
porg
+-
 Reporter:  pierre  |   Owner:  alfs-log@…
 Type:  defect  |  Status:  new
 Priority:  normal  |   Milestone:  3.0
Component:  ablfs   | Version:  SVN
 Severity:  normal  |  Resolution:
 Keywords:  |
+-

Comment (by pierre):

 One first thing to do is to have more than two modes for outputting
 instructions (right now, with only have mode="root", and no mode). Also,
 checking that there is no role"root" before for outputting porg stanzas
 may be not enough, same for no role="root" after for closing...

--
Ticket URL: 
ALFS Trac 
Automated Linux From Scratch
-- 
http://lists.linuxfromscratch.org/listinfo/alfs-log
Unsubscribe: See the above information page