skypher/paktahn

unhandled error condition has been signalled: Makepkg failed (status 1)

Closed this issue · 7 comments

gour commented

Hi,

for the sake of experiment I wanted to install 'tupac', but it fails when it wants to re-install yaourt not recognizing it's already installed.

See below:

[gour@gaura-nitai paktahn] paktahn tupac
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
1 aur/qupac 0.1.3-1
Qt4 frontend for tupac. Features: search as you type, install/remove queue, pre-download in order to view install scripts content before installing. MOCKUP RELEASE: ONLY INSTALL FUNCTIONAL
2 aur/tupac 0.5.6.1-1
A cached pacman implementation that boosts some pacman operations: faster searches, AND searches, aur support, colored output, system sanity check, frontend friendly and more...
=> -----------------------------------------------------------
=> Enter numbers (e.g. '1,2-5,6') of packages to be installed.
=> -----------------------------------------------------------
[1-2] => 2
Installing package "tupac" from AUR.
--2009-08-13 07:21:51-- http://aur.archlinux.org/packages/tupac/tupac.tar.gz
Resolving aur.archlinux.org... 208.92.232.29
Connecting to aur.archlinux.org|208.92.232.29|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 739 [application/x-tgz]
Saving to: `tupac.tar.gz'

100%[======================================================================================================>] 739 --.-K/s in 0s

2009-08-13 07:21:52 (32,0 MB/s) - `tupac.tar.gz' saved [739/739]

tupac
tupac/PKGBUILD
Edit PKGBUILD (Y/n) n

Continue building "tupac" (Y/n) y
Installing binary package "gcc" from repository "core".
warning: gcc-4.4.1-1 is up to date -- skipping
local database is up to date
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
Installing binary package "pacman" from repository "core".
warning: pacman-3.3.0-3 is up to date -- skipping
local database is up to date
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
Installing package "yaourt" from AUR.
--2009-08-13 07:22:01-- http://aur.archlinux.org/packages/yaourt/yaourt.tar.gz
Resolving aur.archlinux.org... 208.92.232.29
Connecting to aur.archlinux.org|208.92.232.29|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 852 [application/x-tgz]
Saving to: `yaourt.tar.gz'

100%[======================================================================================================>] 852 --.-K/s in 0s

2009-08-13 07:22:02 (53,8 MB/s) - `yaourt.tar.gz' saved [852/852]

yaourt
yaourt/PKGBUILD
yaourt/yaourt.install

Edit PKGBUILD (Y/n) n

Continue building "yaourt" (Y/n) y
Installing binary package "wget" from repository "core".
warning: wget-1.11.4-2 is up to date -- skipping
local database is up to date
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
INFO: AUR: No results found
Installing binary package "diffutils" from repository "core".
warning: diffutils-2.8.1-6 is up to date -- skipping
local database is up to date
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
Installing binary package "pacman" from repository "core".
warning: pacman-3.3.0-3 is up to date -- skipping
local database is up to date
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
Installing binary package "gettext" from repository "core".
warning: gettext-0.17-3 is up to date -- skipping
local database is up to date
==> ERROR: A package has already been built. (use -f to overwrite)

Date/time: 2009-08-13-07:22An unhandled error condition has been signalled: Makepkg failed (status 1)

0: (SB-DEBUG::MAP-BACKTRACE #<CLOSURE (LAMBDA (SB-DEBUG::FRAME)) {100511F959}>)[:EXTERNAL]
1: (BACKTRACE 1152921504606846975 #<SYNONYM-STREAM :SYMBOL TERMINAL-IO {10002007A1}>)
2: (TRIVIAL-BACKTRACE:PRINT-BACKTRACE-TO-STREAM #<SYNONYM-STREAM :SYMBOL TERMINAL-IO {10002007A1}>)
3: (TRIVIAL-BACKTRACE:PRINT-BACKTRACE #<SIMPLE-ERROR {1005118741}>)[:EXTERNAL]
4: (PAKTAHN::DEFAULT-ERROR-HANDLER #<SIMPLE-ERROR {1005118741}>)
5: (SIGNAL #<SIMPLE-ERROR {1005118741}>)[:EXTERNAL]
6: (ERROR "Makepkg failed (status ~D)")[:EXTERNAL]
7: (PAKTAHN::RUN-MAKEPKG)
8: (PAKTAHN::INSTALL-AUR-PACKAGE "yaourt")
9: (PAKTAHN::INSTALL-PACKAGE "yaourt" :DB-NAME NIL)
10: (PAKTAHN::INSTALL-DEPENDENCIES ("gcc" "pacman" "yaourt" "php" "libxml2"))
11: (PAKTAHN::INSTALL-AUR-PACKAGE "tupac")
12: (PAKTAHN::INSTALL-PACKAGE "tupac" :DB-NAME "aur")
13: (PAKTAHN::SEARCH-AND-INSTALL-PACKAGES "tupac")
14: (PAKTAHN::MAIN ("tupac"))
15: (PAKTAHN::CORE-MAIN)
16: ((LABELS SB-IMPL::RESTART-LISP))

debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread" RUNNING {100424C021}>: Makepkg failed (status 1)

Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL.

(no restarts: If you didn't do this on purpose, please report it as a bug.)

(PAKTAHN::RUN-MAKEPKG)
source:
error finding source: The source file no longer exists:
/var/abs/local/system/paktahn/src/aur.lisp

Sincerely,
Gour

The current implementation takes care of removing AUR directories so you seem to have some residual ones from earlier. Deleting them should work.

Note: since we're not heeding any dep information now I decided that it would be best to bluntly try reinstalling packages.

Do you think we should change that and assume that already installed packages are alright?

gour commented

I do not have any residual folder, omly built package in my local repo...

Otoh, having something like:

[gour@gaura-nitai paktahn] paktahn tupac
INFO: Cache is out of date, refreshing.
INFO: Building cache for db "local"... done.
INFO: Building cache for db "gaur"... done.
INFO: Building cache for db "core"... done.
INFO: Building cache for db "extra"... done.
INFO: Building cache for db "community"... done.
INFO: Writing cache meta data... done.
INFO: Writing cache contents... done.
1 aur/qupac 0.1.3-1
Qt4 frontend for tupac. Features: search as you type, install/remove queue, pre-download in order to view install scripts content before installing. MOCKUP RELEASE: ONLY INSTALL FUNCTIONAL
2 aur/tupac 0.5.6.1-1
A cached pacman implementation that boosts some pacman operations: faster searches, AND searches, aur support, colored output, system sanity check, frontend friendly and more...
=> -----------------------------------------------------------
=> Enter numbers (e.g. '1,2-5,6') of packages to be installed.
=> -----------------------------------------------------------
[1-2] => 2
Installing package "tupac" from AUR.
--2009-08-13 11:31:32-- http://aur.archlinux.org/packages/tupac/tupac.tar.gz
Resolving aur.archlinux.org... 208.92.232.29
Connecting to aur.archlinux.org|208.92.232.29|:80... connected.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable

The file is already fully retrieved; nothing to do.

tupac
tupac/PKGBUILD
Edit PKGBUILD (Y/n) n

Continue building "tupac" (Y/n) y
Installing binary package "gcc" from repository "core".
..

i.e. reinstalling already installed pkgs is not OK.

My standpoint is that paktahn should get proper dep-resolution asap. ;)

That's the main point of using it over scripting wrappers.

What do you think?

btw, how does paktahn handle local user-repo?

The file is already fully retrieved; nothing to do.

This looks like you do have the AUR folder already. Check /var/tmp/tupac.

My standpoint is that paktahn should get proper dep-resolution asap. ;)

I'm not even sure yaourt does that. Anyway I agree, but it's going to take some time. Until then I have decided to make Paktahn skip already installed packages.

btw, how does paktahn handle local user-repo?

Not any different from sync repos. Should it?

gour commented

This looks like you do have the AUR folder already. Check /var/tmp/tupac.

Ahh, forgot about /var/tmp...just looking in my local repo at /avr/abs/local :-(

I'm not even sure yaourt does that. Anyway I agree, but it's going to take some time.

Heh, that's why we want something better than yaourt. How much work is to bind the whole libalpm?

Until then I have decided to make Paktahn skip already installed packages.

That's ok.

Not any different from sync repos. Should it?

No, this is fine.

Ahh, forgot about /var/tmp...just looking in my local repo at /avr/abs/local :-(

Yeah well, it's not really obvious unless you read the source... :/

Heh, that's why we want something better than yaourt. How much work is to bind the whole libalpm?

Not that much, but I'd prefer to put my time into other things right now. Do you think always working directly with libalpm would have an advantage over calling out to Pacman, though?

gour commented

Do you think always working directly with libalpm would have an advantage over
calling out to Pacman, though?

I'd say so, according to my understanding, i.e. there should be the reason why the pacman's functionality is abstracted into libalpm, and it sounds good from the (newbie)Haskeller's perspective. :-)