12:36 a.m.
Hi Folks!
Yes, it's me trying to track down a windoze problem. Can you believe
it? I must be mellowing in my old age. :-)
On Steven's XEmacs installation, the directory,
"~/.xemacs/site-packages" is _not_ part of his load-path, even though
the directory exists. The output from `-debug-paths' didn't show any
paths given at configure time, so he hasn't configured it away.
Is there any reason why a Cygwin build (21.4.13) wouldn't, by default,
have that directory in the load-path?
Steven, it would probably help if you could send the output from
`M-x report-xemacs-bug RET'. And I'd like to see the values of
`user-init-file', `user-init-directory'.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
1:04 a.m.
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
Steven, it would probably help if you could send the output from
`M-x report-xemacs-bug RET'.
Well, this isn't going to help very much:
,----[ report-xemacs-bug failure ]
| Signaling: (void-function nil)
| nil("XEmacs Beta <xemacs-beta(a)xemacs.org>" "[Bug: 21.4.13]
Sample" nil nil nil nil nil)
| compose-mail("XEmacs Beta <xemacs-beta(a)xemacs.org>" "[Bug:
21.4.13] Sample" nil)
| #<compiled-function (topic &optional recent-keys) "...(516)" [map
before-keys user-point mods topic mail-user-agent nil "[Bug: " "] "
compose-mail re-search-forward "^" regexp-quote "$" 1
"================================================================\n" "Dear
Bug Team!\n\n"
"\n\n================================================================\n\nSystem Info
to help track down your bug:\n---------------------------------------\n\n"
Installation-string fboundp append-message (...) #<compiled-function ...
"...(2)" ... 1 "\nCommon Lisp lambda list:\n (lambda &rest
ARGS)\n\n"> clear-message (...) #<compiled-function ... "...(2)" ...
1 "\nCommon Lisp lambda list:\n (lambda &optional LABEL FRAME STDOUT-P
NO-RESTORE)\n\n"> "\n\nLoad-Path Lisp Shadows:\n"
"----------------------\n" format "%s" find-emacs-lisp-shadows
fill-paragraph t "\n" "\n\nInstalled XEmacs Packages:\n"
"-------------------------\n" cl-prettyprint packages-package-list list-modules
"\n\nInstalled Modules:\n" "-----------------\n"
"\n\nFeatures:\n--------\n\n%s" features "\n\n" "Recent
keystrokes:\n-----------------\n\n" key-description move-to-column 50 search-forward
" " ...] 5
("/usr/local/lib/xemacs/xemacs-packages/lisp/net-utils/xemacsbug.elc" . 3939)
(byte-code "ÀÁ ÂÃ!D!" [reverse recent-keys read-string "Bug Subject:
"] 4)>("Sample" [#<keypress-event c> #<keypress-event o>
#<keypress-event m> #<keypress-event p> #<keypress-event o>
#<keypress-event s> #<keypress-event e> #<keypress-event ->
#<keypress-event m> #<keypress-event a> #<keypress-event i>
#<keypress-event l> #<keypress-event return> #<keypress-event control-P>
#<keypress-event control-E> #<keypress-event control-A> #<keypress-event
q> #<misc-user-event (cancel-mode-internal nil)> #<keypress-event (>
#<keypress-event r> #<keypress-event e> #<keypress-event p>
#<keypress-event o> #<keypress-event r> #<keypress-event t>
#<keypress-event -> #<keypress-event x> #<keypress-event e>
#<keypress-event m> #<keypress-event a> #<keypress-event c>
#<keypress-event s> #<keypress-event -> #<keypress-event b>
#<keypress-event u> #<keypress-event f> #<keypress-event space>
#<keypress-event backspace> #<keypress-event backspace> #<keypress-event
g> #<keypress-event space> #<keypress-event '> #<keypress-event
shift-backspace> #<keypress-event "> #<keypress-event F>
#<keypress-event o> #<keypress-event o> #<keypress-event ">
#<keypress-event )> #<keypress-event control-X> ...])
| call-interactively(report-xemacs-bug)
| command-execute(report-xemacs-bug t)
| execute-extended-command(nil)
| call-interactively(execute-extended-command)
`----
Oh, now if I execute (load-library "gnus"), we get this:
And I'd like to see the values of `user-init-file',
`user-init-directory'.
user-init-file: "/home/sharr/.xemacs/init.el"
user-init-directory: "~/.xemacs/"
If you see differences between bug reports of mine with my home path
looking slightly different, it's because it's /home/sharr at work and
/home/seh at home. Otherwise, the setups are the same.
--
Steven E. Harris
2:07 a.m.
* Steven E Harris <seh(a)panix.com> writes:
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
> Steven, it would probably help if you could send the output from
> `M-x report-xemacs-bug RET'.
System Info to help track down your bug:
---------------------------------------
uname -a: CYGWIN_NT-5.0 TSUNAMI 1.3.20(0.73/3/2) 2003-02-08 12:10
i686 unknown unknown Cygwin
./configure '--with-dragndrop'
'--with-postgresql=no' '--with-x=no'
'--bindir=/usr/local/bin/i686-pc-cygwin' '--with-site-lisp=yes'
'--with-ipv6-cname=no' '--with-netinstall'
No package-path set... good.
[...]
user-init-file: "/home/sharr/.xemacs/init.el"
user-init-directory: "~/.xemacs/"
So, why on earth is "~/.xemacs/site-packages/" missing from the
load-path. Andy? Adrian? Either of you guys got any ideas?
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
2:29 a.m.
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
* Steven E Harris <seh(a)panix.com> writes:
> Steve Youngs <sryoungs(a)bigpond.net.au> writes:
>> Steven, it would probably help if you could send the output from
>> `M-x report-xemacs-bug RET'.
> System Info to help track down your bug:
> ---------------------------------------
> uname -a: CYGWIN_NT-5.0 TSUNAMI 1.3.20(0.73/3/2) 2003-02-08 12:10 i686 unknown
unknown Cygwin
> ./configure '--with-dragndrop' '--with-postgresql=no'
'--with-x=no' '--bindir=/usr/local/bin/i686-pc-cygwin'
'--with-site-lisp=yes' '--with-ipv6-cname=no' '--with-netinstall'
No package-path set... good.
[...]
> user-init-file: "/home/sharr/.xemacs/init.el"
> user-init-directory: "~/.xemacs/"
So, why on earth is "~/.xemacs/site-packages/" missing from the
load-path. Andy? Adrian? Either of you guys got any ideas?
I haven't been following this thread very closely; do have a $HOME
environment variable set, Steven? What does
(getenv "HOME")
evaluate to? This is just a shot in the dark, of course.
- Vin
4:17 p.m.
Vin Shelton <acs(a)xemacs.org> writes:
do have a $HOME environment variable set, Steven?
Yes, by way of my Windows {2000,XP} user-level environment variable
dialog, set there so that Cygwin picks it up before loading any Cygwin
applications.
What does (getenv "HOME") evaluate to?
"/home/sharr" (at work, Windows 2000)
"/home/seh" (at home, Windows XP)
--
Steven E. Harris
3:30 p.m.
You could try tracing
(packages-find-packages
emacs-data-roots
(packages-compute-package-locations user-init-directory))
to find out where the problem's coming from.
--
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla
5:52 p.m.
Michael Sperber <sperber(a)informatik.uni-tuebingen.de> writes:
You could try tracing
(packages-find-packages
emacs-data-roots
(packages-compute-package-locations user-init-directory))
to find out where the problem's coming from.
I get stuck here:
,----
| Symbol's value as variable is void: emacs-data-roots
`----
The form (packages-compute-package-locations user-init-directory)
looks promising, though:
,----
| (("~/.xemacs/mule-packages" early #<compiled-function nil
"...(4)" [featurep mule] 2>)
| ("~/.xemacs/xemacs-packages" early #<compiled-function nil
"...(2)" [t] 1>)
| ("site-packages" late #<compiled-function nil "...(2)" [t]
1>)
| ("infodock-packages" late #<compiled-function nil "...(4)"
[featurep infodock] 2>)
| ("mule-packages" late #<compiled-function nil "...(4)" [featurep
mule] 2>)
| ("xemacs-packages" late #<compiled-function nil "...(2)" [t]
1>))
`----
--
Steven E. Harris
11:34 p.m.
* Steven Elliot Harris <seh(a)panix.com> writes:
,----
| Symbol's value as variable is void: emacs-data-roots
`----
I haven't looked into this yet (I will later), but...
The form (packages-compute-package-locations user-init-directory)
looks promising, though:
,----
| (("~/.xemacs/mule-packages" early #<compiled-function nil
"...(4)" [featurep mule] 2>)
| ("~/.xemacs/xemacs-packages" early #<compiled-function nil
"...(2)" [t] 1>)
| ("site-packages" late #<compiled-function nil "...(2)" [t]
1>)
| ("infodock-packages" late #<compiled-function nil "...(4)"
[featurep infodock] 2>)
| ("mule-packages" late #<compiled-function nil "...(4)"
[featurep mule] 2>)
| ("xemacs-packages" late #<compiled-function nil "...(2)" [t]
1>))
`----
...*THWACK!* <palm of hand to forehead>. XEmacs 21.4 doesn't support
`site-packages' directory under `user-init-directory'. This is a bug
in XEmacs and needs to be fixed. And it is fixed in a patch that I
submitted months ago that _still_ hasn't been applied.
The reason it didn't dawn on me straight away is that I have been
running with the patch that adds site-packages under
user-init-directory to the default package path for so long that I
forgot that it wasn't in the release version of XEmacs.
Hassle Vin. :-)
Steven, in the meantime you can work around your original problem
with...
(let* ((dir (file-name-as-directory
(expand-file-name "path/to/your/gnus" user-init-directory)))
(autos (expand-file-name "auto-autoloads.elc" dir))
(customs (expand-file-name "custom-load.elc" dir)))
(push dir load-path)
(if (file-exists-p autos)
(load-file autos))
(if (file-exists-p customs)
(load-file customs)))
Please accept my apologies for not realising what the problem was a
lot sooner.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
4:36 p.m.
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
Please accept my apologies for not realising what the problem was a
lot sooner.
No need to apologize. That you even thought to follow up is more than
one can expect. Thank /you/ for figuring this out. I look forward to
21.5, but for now I'm stuck with 21.4.13 as that's the last build
available "out of the box" for Cygwin.
--
Steven E. Harris
6:42 p.m.
Steven Elliot Harris <seh(a)panix.com> writes:
I look forward to 21.5, but for now I'm stuck with 21.4.13 as
that's
the last build available "out of the box" for Cygwin.
If you've got the cygwin tool-chain and don't mind taking a bit of
time, it's pretty easy to build 21.5 from CVS. `M-x build' makes this
especially easy.
Jon
--
But if serving the LORD seems undesirable to you, then choose for
yourselves this day whom you will serve . . . But as for me and my
household, we will serve the LORD.
-- Joshua 24:15 (NIV)
8:33 p.m.
* Steven Elliot Harris <seh(a)panix.com> writes:
Thank /you/ for figuring this out.
You're welcome!
I look forward to 21.5, but for now I'm stuck with 21.4.13 as
that's the last build available "out of the box" for Cygwin.
You may not have to wait as long as you think. My patches that would
have fixed your problem have just gone into 21.4, and so will be in
21.4.15.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
11:35 p.m.
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
You may not have to wait as long as you think. My patches that
would have fixed your problem have just gone into 21.4, and so will
be in 21.4.15.
Do you expect that there will be an 21.4.15 build with an installer
offered for Windows? When 21.4.14 came out, I recall reading that the
Windows maintainer who prepares the installation packages was too busy
to attend to it. Or is it possible to just grab a few files to bring
my 21.4.13 installation up to 21.4.15?
--
Steven E. Harris
1:26 a.m.
New subject: Netinstaller for 21.4.15? [was: Package paths on Cygwin]
>>>> "Steven" == Steven Elliot Harris
<seh(a)panix.com> writes:
Steven> Do you expect that there will be an 21.4.15 build with an
Steven> installer offered for Windows?
Ask Andy ... on second thought, I will. (Cc'ing...) Andy?
Steven> When 21.4.14 came out, I recall reading that the Windows
Steven> maintainer who prepares the installation packages was too
Steven> busy to attend to it.
"Busy" is a relative word in this context. The differences between
.13 and .14 from the point of view of Windows users were rather small.
This is a big difference, and reviewer advocacy will probably go
beyond "strong encouragement" to Andy to the point where maybe
somebody else will step up and do it if necessary.
I think it's unlikely that there will be an InstallShield-style
installer, though. Those have never been done by core people, and I
really can't see a good reason for spending effort on them based on
the quality of product and user demand. (My mind could be changed,
that's just my current take.)
Steven> Or is it possible to just grab a few files to bring my
Steven> 21.4.13 installation up to 21.4.15?
Let's put it this way: building XEmacs is much easier and more
reliable than updating core files piecemeal (for one thing, you really
want to redump XEmacs, basically the Lisp equivalent of the link stage
for C programs).
If you really bring bits across a river in a bucket, maybe bandwidth
is an issue. If you have a 14.4kB modem connection or better, I
suggest you bite the bullet and get the Cygwin toolchain and CVS
checkout XEmacs, after which you can _reliably_ update only those
files that need it. (Yes, I have done that, and not just once. It
_was_ worth the wait.)
Although M-x build doesn't carry around as much GUI garbage as VC++,
if you run XEmacs anyway you'll find M-x build to be quite as easy to
use as any installer, at least for this purpose.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
4:22 a.m.
New subject: Netinstaller for 21.4.15? [was: Package paths on Cygwin]
I haven't been tracking this thread, but I plan on doing a build of 21.4.15
for windows. My build machine is currently in for repair, and my day job is
somewhat busy.
andy
-----Original Message-----
From: xemacs-beta-admin(a)xemacs.org
[mailto:xemacs-beta-admin＠xemacs.org]On Behalf Of Stephen J. Turnbull
Sent: Tuesday, January 27, 2004 5:27 PM
To: Steven Elliot Harris
Cc: xemacs-beta(a)xemacs.org; andy(a)xemacs.org
Subject: Netinstaller for 21.4.15? [was: Package paths on Cygwin]
>>>>> "Steven" == Steven Elliot Harris <seh(a)panix.com>
writes:
Steven> Do you expect that there will be an 21.4.15 build with an
Steven> installer offered for Windows?
Ask Andy ... on second thought, I will. (Cc'ing...) Andy?
Steven> When 21.4.14 came out, I recall reading that the Windows
Steven> maintainer who prepares the installation packages was too
Steven> busy to attend to it.
"Busy" is a relative word in this context. The differences between
.13 and .14 from the point of view of Windows users were rather small.
This is a big difference, and reviewer advocacy will probably go
beyond "strong encouragement" to Andy to the point where maybe
somebody else will step up and do it if necessary.
I think it's unlikely that there will be an InstallShield-style
installer, though. Those have never been done by core people, and I
really can't see a good reason for spending effort on them based on
the quality of product and user demand. (My mind could be changed,
that's just my current take.)
Steven> Or is it possible to just grab a few files to bring my
Steven> 21.4.13 installation up to 21.4.15?
Let's put it this way: building XEmacs is much easier and more
reliable than updating core files piecemeal (for one thing, you really
want to redump XEmacs, basically the Lisp equivalent of the link stage
for C programs).
If you really bring bits across a river in a bucket, maybe bandwidth
is an issue. If you have a 14.4kB modem connection or better, I
suggest you bite the bullet and get the Cygwin toolchain and CVS
checkout XEmacs, after which you can _reliably_ update only those
files that need it. (Yes, I have done that, and not just once. It
_was_ worth the wait.)
Although M-x build doesn't carry around as much GUI garbage as VC++,
if you run XEmacs anyway you'll find M-x build to be quite as easy to
use as any installer, at least for this purpose.
--
Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573
JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
5:37 p.m.
New subject: Netinstaller for 21.4.15?
"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
Although M-x build doesn't carry around as much GUI garbage as
VC++,
if you run XEmacs anyway you'll find M-x build to be quite as easy
to use as any installer, at least for this purpose.
I installed the `build' package and worked through M-x build as an
alternative to the build I tried by hand (mentioned in my parallel
message to Steve Youngs). M-x build is very impressive.
Unfortunately, the build failed in the same way as my first try: It
looks like I need an XPM package that I don't have installed right
now. I do have the Cygwin package "xpm-nox" installed, described as
"X11 pixmap library -- no X required." Cygwin's Package Search reveals
that the xpm.h is provided in the xpm-nox package. In fact, I do have
xpm.h, but it's sitting in /usr/include/noX/xpm.h rather than the
expected .../X11/xpm.h. I'll try finding where I can tell XEmacs to
look elsewhere for this file.
--
Steven E. Harris
6:01 p.m.
New subject: Netinstaller for 21.4.15?
Steven Elliot Harris <seh(a)panix.com> writes:
I'll try finding where I can tell XEmacs to look elsewhere for
this
file.
I added /usr/include/noX to the site include path and got a lot
farther -- all the way to this crash:
Loading site-load...
Finding pointers to doc strings...
Finding pointers to doc strings...done
Dumping under the name xemacs
.data and .bss reversed
found bss - keeping 30a5d0 of 30b800 bytes
found data - keeping 85d0 of 2bba00 bytes
writing file header
writing optional header
new data size is 5c7200, >= 5c7200
writing text header (unchanged)
writing .data header
writing following data header
copying executable up to data section ... 2788f8 bytes
dumping .data section... 85d0 bytes
padding .data ... 800 bytes
dumping .bss into executable... 30a5d0 bytes
padding .bss ... a30 bytes
dumping following data section... 2000 bytes
writing rest of file
make[2]: Leaving directory `/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/src'
./xemacs -no-packages -batch -no-autoloads -l update-elc-2.el -f batch-update-elc-2
/home/sharr/temp/xemacs-cvs/xemacs-HEAD/src/../lisp
Fatal error (11).
Your files have been auto-saved.
Use `M-x recover-session' to recover them.
[Bug report admonishment...]
Lisp backtrace follows:
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
string-match("\\(^\\(RCS\\|CVS\\|SCCS\\)$\\)\\|\\(^\\(etc\\|info\\|man\\|lisp\\|lib-src\\|bin\\|pkginfo\\)$\\)"
"etc")
# bind (raw-entries reverse-dirs directory path exclude-regexp max-depth directories)
paths-find-recursive-path(("/home/sharr/.xemacs/site-packages/") 1
"\\(^\\(RCS\\|CVS\\|SCCS\\)$\\)\\|\\(^\\(etc\\|info\\|man\\|lisp\\|lib-src\\|bin\\|pkginfo\\)$\\)")
# bind (directories)
packages-find-packages-in-directories(("/home/sharr/.xemacs/site-packages/"))
# bind (name roots)
packages-find-packages-by-name(("/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/" "/home/sharr/usr/local/")
"~/.xemacs/site-packages")
# bind (thunk a-time name)
#<compiled-function (name a-time thunk) "...(18)" [name roots thunk a-time
time packages packages-find-packages-by-name] 4>("~/.xemacs/site-packages"
early #<compiled-function nil "...(2)" [t] 1>)
apply(#<compiled-function (name a-time thunk) "...(18)" [name roots thunk
a-time time packages packages-find-packages-by-name] 4>
("~/.xemacs/site-packages" early #<compiled-function nil "...(2)"
[t] 1>))
# bind (consumer list)
packages-deconstruct(("~/.xemacs/site-packages" early #<compiled-function
nil "...(2)" [t] 1>) #<compiled-function (name a-time thunk)
"...(18)" [name roots thunk a-time time packages packages-find-packages-by-name]
4>)
# bind (packages default time package-locations roots)
packages-find-packages-at-time(("/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/" "/home/sharr/usr/local/")
(("~/.xemacs/site-packages" early #<compiled-function nil "...(2)"
[t] 1>) ("~/.xemacs/infodock-packages" early #<compiled-function nil
"...(4)" [featurep infodock] 2>) ("~/.xemacs/mule-packages" early
#<compiled-function nil "...(4)" [featurep mule] 2>)
("~/.xemacs/xemacs-packages" early #<compiled-function nil "...(2)"
[t] 1>) ("site-packages" late #<compiled-function nil "...(2)"
[t] 1>) ("infodock-packages" late #<compiled-function nil
"...(4)" [featurep infodock] 2>) ("mule-packages" late
#<compiled-function nil "...(4)" [featurep mule] 2>)
("xemacs-packages" late #<compiled-function nil "...(2)" [t]
1>)) early nil)
# bind (configure-last-packages configure-late-packages configure-early-packages)
#<compiled-function (configure-early-packages configure-late-packages
configure-last-packages) "...(20)" [configure-last-packages
configure-late-packages configure-early-packages package-locations roots
packages-find-packages-at-time early late last] 7>(nil nil nil)
apply(#<compiled-function (configure-early-packages configure-late-packages
configure-last-packages) "...(20)" [configure-last-packages
configure-late-packages configure-early-packages package-locations roots
packages-find-packages-at-time early late last] 7> (nil nil nil))
# bind (consumer list)
packages-deconstruct((nil nil nil) #<compiled-function (configure-early-packages
configure-late-packages configure-last-packages) "...(20)"
[configure-last-packages configure-late-packages configure-early-packages
package-locations roots packages-find-packages-at-time early late last] 7>)
# bind (envvar-value package-locations roots)
packages-find-packages(("/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/" "/home/sharr/usr/local/")
(("~/.xemacs/site-packages" early #<compiled-function nil "...(2)"
[t] 1>) ("~/.xemacs/infodock-packages" early #<compiled-function nil
"...(4)" [featurep infodock] 2>) ("~/.xemacs/mule-packages" early
#<compiled-function nil "...(4)" [featurep mule] 2>)
("~/.xemacs/xemacs-packages" early #<compiled-function nil "...(2)"
[t] 1>) ("site-packages" late #<compiled-function nil "...(2)"
[t] 1>) ("infodock-packages" late #<compiled-function nil
"...(4)" [featurep infodock] 2>) ("mule-packages" late
#<compiled-function nil "...(4)" [featurep mule] 2>)
("xemacs-packages" late #<compiled-function nil "...(2)" [t]
1>)))
# bind (earlyp latep lastp earlyp-lp latep-lp lastp-lp set-global-package-paths
inhibit-packages)
startup-find-load-path((early late last) t)
# bind (called-early inhibit-packages)
startup-setup-paths(t nil)
normal-top-level()
# (condition-case ... . error)
# (catch top-level ...)
Signal 11
make[1]: *** [update-elc-2] Error 139
make[1]: Leaving directory `/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/src'
make: *** [src] Error 2
I'll prepare the requested bug report.
--
Steven E. Harris
7:12 p.m.
New subject: Netinstaller for 21.4.15?
--- Steven Elliot Harris <seh(a)panix.com> wrote:
Steven Elliot Harris <seh(a)panix.com> writes:
> I'll try finding where I can tell XEmacs to look elsewhere for this
> file.
I added /usr/include/noX to the site include path and got a lot
farther -- all the way to this crash:
Loading site-load...
Finding pointers to doc strings...
Finding pointers to doc strings...done
Dumping under the name xemacs
.data and .bss reversed
found bss - keeping 30a5d0 of 30b800 bytes
found data - keeping 85d0 of 2bba00 bytes
writing file header
writing optional header
new data size is 5c7200, >= 5c7200
writing text header (unchanged)
writing .data header
writing following data header
copying executable up to data section ... 2788f8 bytes
dumping .data section... 85d0 bytes
padding .data ... 800 bytes
dumping .bss into executable... 30a5d0 bytes
padding .bss ... a30 bytes
dumping following data section... 2000 bytes
writing rest of file
make[2]: Leaving directory `/mnt/c/Documents and Settings/sharr/My
Documents/temp/xemacs-cvs/xemacs-HEAD/src'
./xemacs -no-packages -batch -no-autoloads -l update-elc-2.el -f
batch-update-elc-2 /home/sharr/temp/xemacs-cvs/xemacs-HEAD/src/../lisp
Fatal error (11).
Did you use the --pdump argument to configure? This is where it crashes when I
don't use it.
--Rick
8:07 p.m.
New subject: Netinstaller for 21.4.15?
Rick Rankin <rick_rankin(a)yahoo.com> writes:
Did you use the --pdump argument to configure? This is where it
crashes when I don't use it.
I added --pdump, fixed up some missing libraries, and got a build that
starts up. There are some problems with the expected package paths (as
I'm just running XEmacs from the build directory, not yet committing
to a "make install"), and I noticed errors related to finding several
auto-autoloads.el* files for my various Gnus builds.
Other than that, I've been wasting time trying to get the default font
back to the Courier New family.
This whole `M-x build' system /looks/ like it's set up to enable a
repeatable build process. What files do I need to save so that I can
start up a fresh XEmacs, run `M-x build', and get all the same options
back?
--
Steven E. Harris
4:19 a.m.
New subject: Netinstaller for 21.4.15?
>>>> "Steven" == Steven E Harris
<seharris(a)raytheon.com> writes:
Steven> This whole `M-x build' system /looks/ like it's set up to
Steven> enable a repeatable build process. What files do I need to
Steven> save so that I can start up a fresh XEmacs, run `M-x
Steven> build', and get all the same options back?
Adrian Aichner is the author and maintainer, but as he hasn't
commented yet, you might try
M-x customize-group RET build RET.
If you can short-circuit the ./configure stage, then you can always
repeat a build by using ./config.status --recheck.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
1:32 a.m.
* Steven Elliot Harris <seh(a)panix.com> writes:
Do you expect that there will be an 21.4.15 build with an installer
offered for Windows?
I would imagine so. I'm actually a little shocked that XEmacs.org
hasn't released one for 21.4.14. Sounds like it got forgotten.
Or is it possible to just grab a few files to bring my 21.4.13
installation up to 21.4.15?
No.
But why don't you just build it from source?
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
4:54 p.m.
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
But why don't you just build it from source?
I tried and got stuck. I'm not sure I'll have time to figure out how
to build my own XEmacs that works as well as the packaged 21.4.13
does.
Here's what I did:
o Check out 21.5 from CVS.
o Run configure, providing my local installation prefix and no other
arguments, just to see what that would get built.
o Run make.
o Start thinking about all the Cygwin packages I'll need to install to
get this done.
Keep in mind that my Cygwin installation mostly contains library or
runtime packages but not necessarily the corresponding "devel"
packages, for I don't build too many programs from scratch. That may
explain some of the warnings and the failure evident in the build
transcript attached below.
Is there some way to figure out what arguments were provided to
configure in an existing XEmacs build? I'd like to see how the
packaged 21.4.13 was built.
--
Steven E. Harris
6 p.m.
--- Steven Elliot Harris <seh(a)panix.com> wrote:
Steve Youngs <sryoungs(a)bigpond.net.au> writes:
> But why don't you just build it from source?
I tried and got stuck. I'm not sure I'll have time to figure out how
to build my own XEmacs that works as well as the packaged 21.4.13
does.
Here's what I did:
o Check out 21.5 from CVS.
o Run configure, providing my local installation prefix and no other
arguments, just to see what that would get built.
o Run make.
o Start thinking about all the Cygwin packages I'll need to install to
get this done.
Keep in mind that my Cygwin installation mostly contains library or
runtime packages but not necessarily the corresponding "devel"
packages, for I don't build too many programs from scratch. That may
explain some of the warnings and the failure evident in the build
transcript attached below.
Is there some way to figure out what arguments were provided to
configure in an existing XEmacs build? I'd like to see how the
packaged 21.4.13 was built.
From build-msw-release.sh in the top-level directory:
./configure --with-dragndrop --with-postgresql=no --with-x=no \
--bindir=/usr/local/bin/i686-pc-cygwin --with-site-lisp=yes \
--with-ipv6-cname=no --with-netinstall
However, I've never been able to get it to work without adding the --pdump
option. FWIW, here's the configure options I use:
prefix=/tools/local
./configure --cflags='-g -O' \
--without-x11 \
--with-msw \
--with-dragndrop \
--with-xpm \
--pdump \
--prefix=$prefix \
--site-prefixes=$prefix \
--package-path=$prefix/lib/xemacs/packages
Rick
10:10 p.m.
* Steven Elliot Harris <seh(a)panix.com> writes:
Is there some way to figure out what arguments were provided to
configure in an existing XEmacs build?
`M-x describe-installation RET'
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
1:33 a.m.
* Steven Elliot Harris <seh(a)panix.com> writes:
| Symbol's value as variable is void: emacs-data-roots
This is because that variable _only_ exists in XEmacs 21.5.
--
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
| XEmacs - The only _______ you'll ever need. |
| Fill in the blank, yes, it's THAT good! |
|------------------------------<sryoungs(a)bigpond.net.au>---|
7603
days inactive
7619
days old
23 comments
9 participants
participants (9)
-
Andy Piper -
Jon Ericson -
Michael Sperber -
Rick Rankin -
Stephen J. Turnbull -
Steve Youngs -
Steven E. Harris -
Steven E. Harris -
Vin Shelton