January 2016 - asdf-devel - mailman3.common-lisp.net

What's the right way to extend ASDF with new symbols?
by Robert Goldman 16 Feb '16

16 Feb '16

Commonly, one wants to extend ASDF with new operation and component classes. But the support in ASDF for referencing such classes all involves automagically interpreting keyword symbols (and some unqualified symbols?) in the ASDF package. If I understand this correctly, this leaves the ASDF extender with the alternatives of either jamming their new symbols into the ASDF package late, and exporting them, or having them live in a new package, which is cleaner, but relegates these new components to second-class status. Consider the following (defpackage my-system-def (:use common-lisp asdf)) (in-package :my-system-def) (defsystem my-system :defsystem-depends-on ("my-asdf-extension") :components ((:my-new-file "file1"))) I don't seem to be able to define MY-NEW-FILE in my-asdf-extension and export it into my-system-def, because my-system-def is already defined by the time my-asdf-extension is loaded. That means I pretty much have to either muck around in the ASDF package (which doesn't scale to multiple different people extending ASDF) or use side effects: (asdf:load-system "my-asdf-extension") (defpackage my-system-def (:use common-lisp asdf my-asdf-extension)) (in-package :my-system-def) (defsystem my-system :components ((:my-new-file "file1"))) Am I right about this? If I am, is there some way to better handle this? Or do we need to rethink how :defsystem-depends-on works?

7 24

Time for 3.1.7 ?
by Faré 12 Feb '16

12 Feb '16

There have been many small fixes since ASDF 3.1.6 in last October, and I suppose it's time for a new minor release. Robert, do you want to handle the release? Are you done with it? What do you think? —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Yesterday I was clever, so I wanted to change the world. Today I am wise, so I am changing myself. — Rumi (1207–1273)

3 6

MAKE function documentation -- example of use?
by Robert Goldman 31 Jan '16

31 Jan '16

The make function description in the documentation is not described well. It seems to imply that there is a default operation for each system that the system developer can specify somehow, without saying how. I see that there's a FIXME suggesting that we need to document the BUILD-OPERATION option for DEFSYSTEM. It would be nice to have an example that shows how MAKE is intended to be used. Not a super-detailed example, just a quickie to show that this is not a feature in search of a use.... Anyone used this? thanks, r

2 2

newbie setup question
by Thomas Lynch 19 Jan '16

19 Jan '16

New at this, sure could use a pointer. I'm not able to get the code to build consistently. It works only after a failure. So I have to do it twice to get it to go. Sure would like to know what is wrong here. These are the directories: /home/LE/sbcl/LE.asd ;holds the definition, as shown below /home/LE/sbcl/src ;holds the lisp source files for my project The system is Debian, udpated today. quicklisp is installed. These are the commands I type: (require :asdf) (pushnew (truename ".") asdf:*central-registry* :test #'equal) (asdf:load-system :LE) (require :LE) (use-package 'LE) in a fresh environment this fails, and dumps me in the debugger, asdf:load-system tells me: "Component :LE not found: Then I exit the debugger, and run it a second time and it works. Not sure, may need quotes the first time, a colon the second, i.e. it is quirky. I have tried a lot of variations of the above, pretty much every example I've been able to find, some have "LE" instead of :LE everywhere. Here is the value of asdf:*central-registry* after the push: * asdf:*central-registry* (#P"/home/LE/sbcl/" #P"/home/mordecai/quicklisp/quicklisp/") Here is the asdf file that is located in /home/LE/sbcl: (in-package :asdf-user) (defsystem #:LE :name "LE" :version "0.1" :author "Thomas W. Lynch <thomas.lynch(a)reasoningtechnology.com>" :description "Lynch's extensions" :depends-on ("local-time") ;; :serial t :components((:module "src" :components ( (:file "package") (:file "fundamental" :depends-on ("package")) (:file "test" :depends-on ("fundamental")) (:file "string" :depends-on ("test")) (:file "functions" :depends-on ("test")) (:file "list-0" :depends-on ("functions")) ;; (:file "list-L" :depends-on ("list-0")) ))))

4 6

Re: Intended behavior of *COMPILE-FILE-WARNINGS-BEHAVIOUR*
by Faré 15 Jan '16

15 Jan '16

[This was my reply that bounced yesterday. I edited it.] On 1/14/16 Jan 14 -1:20 PM, Robert Goldman wrote: > What is the intended behavior when this is bound to :IGNORE? > > I believe that if warnings occur, they should be quashed and > COMPILE-FILE should still return T, although otherwise it would not. > Yes, the intent is that in case of :IGNORE, the compilation should just succeed (assuming an output file was returned). > I'm finding a case where :IGNORE is not enough to make the compile-file > succeed, but I'm grappling with trying to find a minimal test case. So > I was wondering if I was interpreting this variable correctly. > Either no .fasl was created, or you found another bug in UIOP. —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org The uploaded man cut himself from the world. The optimizer saw the simulation would have no more side-effects and optimized it all away.

1 0

Re: Intended behavior of *COMPILE-FILE-WARNINGS-BEHAVIOUR*
by Robert Goldman 15 Jan '16

15 Jan '16

[... this message also bounced yesterday...] On 1/14/16 Jan 14 -1:20 PM, Robert Goldman wrote: > What is the intended behavior when this is bound to :IGNORE? > > I believe that if warnings occur, they should be quashed and > COMPILE-FILE should still return T, although otherwise it would not. > > I'm finding a case where :IGNORE is not enough to make the compile-file > succeed, but I'm grappling with trying to find a minimal test case. So > I was wondering if I was interpreting this variable correctly. > > thank you > The relevant bit of the manual describes acceptable values for this variable and COMPILE-FILE-FAILURE-BEHAVIOUR, but does not explain their semantics: ASDF checks for warnings and errors when a file is compiled. The variables *COMPILE-FILE-WARNINGS-BEHAVIOUR* and *COMPILE-FILE-FAILURE-BEHAVIOUR* control the handling of any such events. The valid values for these variables are ':error', ':warn', and ':ignore'.

1 0

Intended behavior of *COMPILE-FILE-WARNINGS-BEHAVIOUR*
by Robert Goldman 15 Jan '16

15 Jan '16

[This message bounced yesterday for some reason] What is the intended behavior when this is bound to :IGNORE? I believe that if warnings occur, they should be quashed and COMPILE-FILE should still return T, although otherwise it would not. I'm finding a case where :IGNORE is not enough to make the compile-file succeed, but I'm grappling with trying to find a minimal test case. So I was wondering if I was interpreting this variable correctly. thank you

1 0

Test
by Erik Huelsmann 15 Jan '16

15 Jan '16

Test

1 0

Where is cl-launch.asd?
by Robert Goldman 12 Jan '16

12 Jan '16

I tried to load the asdf tools interactively, under SLIME, and they fail to launch, although everything seems to run from make. I get this error: Component "cl-launch/dispatch" not found, required by #<PACKAGE-INFERRED-SYSTEM "cl-scripting/commands"> and, indeed: ~/lisp/asdf/ext/cl-launch $ find . -name '*.asd' ~/lisp/asdf/ext/cl-launch $ I see this: ~/lisp/asdf/ext/cl-launch $ git status HEAD detached at 220b8ca nothing to commit, working directory clean ~/lisp/asdf/ext/cl-launch $ Is this another oddity of git submodules? Is the version of cl-launch in our git repo not sufficiently up-to-date? Actually, when I look at xcvb/cl-launch, there is no cl-launch.asd there, either. This is crippling my ability to fix 1532464

4 6

Moving towards resignation
by Robert Goldman 12 Jan '16

12 Jan '16

I am now very unhappy with my role as maintainer. Over my repeated expressions of concern, the minimakefile -- which aims to replace normal shell tools with CL-based scripting facilities -- has been merged with core ASDF. Since then, almost all of my ASDF-related time has gone into wrestling with said scripting facilities. I don't mind if people think it's cool to try to make a CL scripting facility. But it's not my thing. I just want my tools to work. The CL scripting facility is, IMO, at an "alpha" level in which it works acceptably in the developer's environment, but can be expected to hiccup seriously anywhere else. I don't want to fight it any more. So I'm offering the community two choices: 1. I "un-merge" the minimakefile facilities and go back to vanilla CL and the standard Unix toolkit for maintenance. 2. Someone emerges to take over the maintenance of ASDF in its current shape, and I go back to being a simple contributor. I barely have enough time to keep up my responsibilities as maintainer; I simply don't have the time to maintain cl-launch, cl-scripting, yadda yadda yadda, as well as ASDF. Even if I did, I wouldn't want to, because this is not a direction that interests me personally. I don't say it's a wrong direction, just that it's not my thing. Fare likes this structure, and he is still the master of ASDF. Me handing off ASDF is more survivable than driving Fare away. So I believe that finding a maintainer who is intrigued by the new direction, and enjoys moving it forward, is a better choice than going back and ripping out the new framework (in retrospect, we should have *added* the new scripting facilities, rather than *replacing* the old). So consider this a call for my replacement. It's been an interesting ride, and I have learned a lot in the process, but it's probably time for someone else, who's more motivated by the new infrastructure. Best, R

3 2