- asdf-devel - mailman3.common-lisp.net

[asdf-devel] Re: Preview of my ASDF3 article.
by Faré 08 Apr '15

08 Apr '15

I refactored that paper by making separate "short" (8 pages) and "extended" versions (24 pages), thereby fitting the limit while being able to say everything I had on my heart. I submitted the short PDF to ELS 2014. I also moved the papers to a more permanent location. [Source] https://github.com/fare/asdf3-2013/blob/master/asdf3-2014.scrbl [Extended HTML] http://fare.tunes.org/files/asdf3/asdf3-2014.html [Extended PDF] http://fare.tunes.org/files/asdf3/asdf3-2014.pdf [Short HTML] http://fare.tunes.org/files/asdf3/asdf3-els2014.html [Short PDF] http://fare.tunes.org/files/asdf3/asdf3-els2014.pdf I still need to extend the section on pathnames some more, but the short version should be stable, and even the extended version. —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org The worst thing that can happen to a good cause is, not to be skillfully attacked, but to be ineptly defended. — F. Bastiat

2 1

[asdf-devel] errors in ASDF 3.1.0.66
by Anton Vodonosov 08 Apr '15

08 Apr '15

Hello. I've been testing ASDF 3.1.0.66 and unfortunately detected errors: 1. Required method PERFORM not implemented for operation #<STANDARD-CLASS CFFI-GROVEL::PROCESS-OP {599AF195}>, component #<STANDARD-CLASS ASDF/LISP-ACTION:CL-SOURCE-FILE {582A1E35}> 2. New ASDF breaks ECL for Quicklisp. The problem nails down to (require 'sockets). This form works for ECL unless new ASDF is loaded. To reproduce at cl-test-grid.cloud.efficito.com machine: rlwrap ~/lisps/ecl-bin-13.5.1/bin/ecl -norc (require 'sockets) ;;; Loading #P"/home/testgrid/lisps/ecl-bin-13.5.1/lib/ecl-13.5.1/sockets.fas" => ("SB-BSD-SOCKETS" "SOCKETS") ------ compare with: rlwrap ~/lisps/ecl-bin-13.5.1/bin/ecl -norc -load ~/asdf/build/asdf.lisp (require 'sockets) ASDF could not load sockets because In function MAPCAR, the value of the second argument is #P"SYS:SOCKETS.FAS" which is not of the expected type LIST. The report: http://common-lisp.net/project/cl-test-grid/asdf/asdf-diff-33.html Best regards, - Anton

2 5

[asdf-devel] READTABLE isolation in ASDF
by Robert P. Goldman 08 Apr '15

08 Apr '15

Attila Lendvai wrote: > this reminds me of the general idea of backwards compatibility: people > who want to live without this feature can chose not to upgrade, as > always. or if they need the new stuff, then they can just patch asdf > locally to disable the enforcement of this. or asdf could even > accommodate this with a defsystem property that can opt a system out > of this otherwise useful isolation. Yes, that is right; please see my earlier email with next steps: 1. The READTABLE isolation feature WILL NOT BE IN THE NEW RELEASE. 2. When/if it is released, the READTABLE isolation feature will only be an *opt-in* strict mode item. Thanks! R

1 0

[asdf-devel] Pushed 3.1.0.50 -- non-release version
by Robert P. Goldman 08 Apr '15

08 Apr '15

Fixed minor warning issue that caused a test problem on ACL. Added a check for BUNDLE-OP on unsupported platforms (ABCL + Mac OSX; ECL + Mac OSX). Cheers, r

2 1

[asdf-devel] uiop/filesystem:directory-files
by Станислав Кондратьев 08 Apr '15

08 Apr '15

Hello, I tried uiop and discovered two issues. I had installed uiop using quicklisp. The first issue looks like a bug: if I list the files of my home directory in CLISP, every file is included in the list twice as if the directory were listed two times. This takes place no matter if I specify the directory as "~/" or spell it out, and also if I run CLISP in my home directory and use uiop/os:getcwd. However, the function works normally with directories other than the home one. I did not specify wildcards in either case. I have observed this behaviour on Debian and OpenSuse. The second issue looks like a discrepancy between implementations: while in CLISP uiop/filesystem:directory-files returns a list of files (not subdirectories), in SBCL the directories are included into the output. I wonder at least which behaviour is the intended one. Best regards, Stanislav.

3 3

[asdf-devel] component-load-dependencies
by Didier Verna 08 Apr '15

08 Apr '15

Hello, currently Declt advertises component dependencies by using COMPONENT-LOAD-DEPENDENCIES. This function however only returns a list of symbols naming the components. What would be the best way to retrieve a list of component objects instead ? Thanks! -- Resistance is futile. You will be jazzimilated. Lisp, Jazz, Aïkido: http://www.didierverna.info

2 6

[asdf-devel] ASDF 3.0.3 ported to MKCL 1.1.7
by Jean-Claude Beaudoin 08 Apr '15

08 Apr '15

Hello ASDF maintainers, I just completed a port of ASDF 3.0.3 to what will soon be released as MKCL 1.1.7. It currently passes all the 48 tests of yours on Linux (Ubuntu 12.04) and I will very soon test it on Win32 and Win64 (I expect a routine ride there). I would like to submit back to you the few lines of ASDF code I had to conditionalize to get this port working OK. Can we do this through a "Pull request" on github? I see both a fare/asdf and a rpgoldman/asdf repositories that look like the head of ASDF evolution which one should I fork, merge with and then send the pull? Cheers, JCB

3 5

Re: [asdf-devel] Advanced Configuration
by FAU 08 Apr '15

08 Apr '15

I corrected it to your advice accordingly. (defmethod output-files ((operation compile-op) (component foo-file)) (multiple-value-bind (pathnames translatep) (call-next-method) (values (mapcar #'(lambda (p) (make-pathname* :directory (append (pathname-directory p) (list (cond ((featurep :foo1.0) "1.0") ((featurep :foo1.1) "1.1") (t (error "No foo feature found."))))) :defaults p)) pathnames) translatep))) On Sat, 2013-11-16 at 17:34 -0500, Faré wrote: > On Sat, Nov 16, 2013 at 5:21 PM, FAU <fau(a)riseup.net> wrote: > > (defmethod output-files ((operation compile-op) (component foo-file)) > > (let ((l (multiple-value-list (call-next-method)))) > > (apply #'values (mapcar #'(lambda (p) > > (make-pathname* :directory (append (pathname-directory p) (list # > > +foo1.0 "1.0" #+foo1.1 "1.1")) :defaults p)) > > (car l)) > > (cdr l)))) > > > Inasmuch as I think that runtime conditionals should be used rather than > compile-time conditionals wherever performance permits (see also featurep), > your general scheme is valid. However, you can should be using > (multiple-value-bind (pathnames translatedp) ... > (values ...)) > There are only two values returned in this protocol. > > —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org > When my contradictors' ideas aren't an incoherent mess I can easily disprove, > I call them for the ideologists they are; but they are so dogmatic, > they don't change their mind after I insult them. >

1 0

[asdf-devel] Ready to release?
by Faré 08 Apr '15

08 Apr '15

I tweaked the upgrade to punt more on abcl, and it looks like it's working. cmucl, on the other hand, is desperately borking on upgrade, with or without punt. Apart from that, asdf 3.0.2.28 passes all tests for me on linux, including upgrade tests (except on cmucl), for ccl clisp sbcl cmucl allegro allegromodern abcl I also ran tests (but not upgrade tests yet) successfully for: ecl ecl_bytecodes lispworks xcl A few minor patches ago, I also tested that it can be compiled and run correctly on genera. I admit I'm growing impatient for a release 3.0.3 of ASDF. —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org He who refuses to do arithmetic is doomed to talk nonsense. — John McCarthy, in his webpage on Progress and Sustainability

2 10

Re: [asdf-devel] Re: Bug#723977: missing symlink in /usr/share/common-lisp/systems
by Faré 08 Apr '15

08 Apr '15

>>: fahree >: rpgoldman >> Which UIOP patches were you thinking of? I fixed a simple one >> regarding ensure-function. I see other portability issues with >> directory* and symlinks, but I am not competent to address them. > > I'm thinking of 1205555 https://bugs.launchpad.net/asdf/+bug/1205555 and > 1205653 https://bugs.launchpad.net/asdf/+bug/1205653 > > I was hoping to get some work done on these, but that target is slipping. > I was thinking about them as well. Notice how the goals are opposite: one wants symlinks to be followed, the other not to. I admit to trying to introducing the avoidance of recursing through symlinks, because in some cases it could cause infinite recursion, etc., especially when some software likes to have an uplink to a directory full of dependencies, and the software is in the same directory. Although I would like there to be a semi-standard way to avoid symlinks, I understand it might not be possible, and the anti-symlink code might have to be removed, and such software will have to be amended. You're the maintainer, you get to decide. > I just tested on an old and crufty linux box. Worked for ccl sbcl cmucl > allegro allegromodern. ECL fails the test on my box. Goes into an > infinite loop compiling test/file3.lisp -- gets an error in compilation, > retries, gets an error .... > > This may be the fact that I'm trying this on an ancient OpenSUSE box, > with a distro version that's been EOLed, so that may be me, not ASDF. > > If you could test with ECL, and it works for you, I am going to chalk > this up to my ancient distro. > Which test script causes that? It's working here on ecl-13.5.1-914ce253-linux-x64 > I don't have a good way to tell if ABCL is busted on Mac or our bundle > op is busted on Mac. > Maybe we can convince someone on the ABCL team to run our tests? —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Procrastination is great. It gives me a lot more time to do things that I'm never going to do.

1 0