The cCLan list has apparently become the asdf-devel list.
Is this still where the upstream for split-sequence is? If so, can this patch be applied? If not, who's in charge?
[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ] "Politics" is made up of two words, "Poli", which is Greek for "many", and "tics", which are blood sucking insects. -- Gore Vidal
Faré fahree@gmail.com writes:
The cCLan list has apparently become the asdf-devel list.
Is this still where the upstream for split-sequence is? If so, can this patch be applied? If not, who's in charge?
I guess I am, unless someone else wants to be.
diff -rN -u old-split-sequence/build.xcvb new-split-sequence/build.xcvb --- old-split-sequence/build.xcvb 1969-12-31 19:00:00.000000000 -0500 +++ new-split-sequence/build.xcvb 2009-10-22 20:10:35.114171499 -0400 @@ -0,0 +1,4 @@ +#+xcvb +(module (:fullname "split-sequence"
:depends-on ("split-sequence"):supersedes-asdf ("split-sequence")))
I don't have a problem with this bit. (I don't see why you need a reader conditional, but OK, whatever)
--- old-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.110170150 -0400 +++ new-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.114171499 -0400 @@ -50,6 +50,8 @@ ;;; * (split-sequence #; ";oo;bar;ba;" :start 1 :end 9) ;;; -> ("oo" "bar" "b"), 9
+#+xcvb (module ())
(defpackage "SPLIT-SEQUENCE" (:use "CL") (:nicknames "PARTITION")
This is... no. Please find some way of not infesting code with build-related metainformation.
Best,
Christophe
2009/10/23 Christophe Rhodes csr21@cantab.net:
Faré fahree@gmail.com writes:
The cCLan list has apparently become the asdf-devel list.
Is this still where the upstream for split-sequence is? If so, can this patch be applied? If not, who's in charge?
I guess I am, unless someone else wants to be.
OK.
--- old-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.110170150 -0400 +++ new-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.114171499 -0400 @@ -50,6 +50,8 @@ ;;; * (split-sequence #; ";oo;bar;ba;" :start 1 :end 9) ;;; -> ("oo" "bar" "b"), 9
+#+xcvb (module ())
This is... no. Please find some way of not infesting code with build-related metainformation.
OK, I'll put this much-requested feature near the top of my TODO list.
[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ] Only presidents, editors, and people with tapeworms have the right to use the editorial "we." -- Mark Twain
Faré writes:
--- old-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.110170150 -0400 +++ new-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.114171499 -0400 @@ -50,6 +50,8 @@ ;;; * (split-sequence #; ";oo;bar;ba;" :start 1 :end 9) ;;; -> ("oo" "bar" "b"), 9
+#+xcvb (module ())
This is... no. Please find some way of not infesting code with build-related metainformation.
OK, I'll put this much-requested feature near the top of my TODO list.
If every .lisp file has to begin with such an expression, can't you simply introduce a file-as-module defaulting scope?
-T.
From: "Tobias C. Rittweiler" tcr@freebits.de Subject: Re: [asdf-devel] XCVB patch for split-sequence Date: Fri, 23 Oct 2009 10:00:27 +0200
Faré writes:
--- old-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.110170150 -0400 +++ new-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.114171499 -0400 @@ -50,6 +50,8 @@ ;;; * (split-sequence #; ";oo;bar;ba;" :start 1 :end 9) ;;; -> ("oo" "bar" "b"), 9
+#+xcvb (module ())
This is... no. Please find some way of not infesting code with build-related metainformation.
OK, I'll put this much-requested feature near the top of my TODO list.
If every .lisp file has to begin with such an expression, can't you simply introduce a file-as-module defaulting scope?
Where would the inter-file dependencies go, then?
Let's suppose you implied that the module form would be specified if only there actually are any dependencies to speak of.
This implies that you have to use heuristics while interpreting the first form -- is it in the (module (&key &allow-other-keys)) form or not.
Is this acceptable to Faré?
regards, Samium Gromoff -- _deepfire-at-feelingofgreen.ru O< ascii ribbon campaign - stop html mail - www.asciiribbon.org
From: Samium Gromoff _deepfire@feelingofgreen.ru
From: "Tobias C. Rittweiler" tcr@freebits.de
Faré writes:
--- old-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.110170150 -0400 +++ new-split-sequence/split-sequence.lisp 2009-10-22 20:10:35.114171499 -0400 @@ -50,6 +50,8 @@ ;;; * (split-sequence #; ";oo;bar;ba;" :start 1 :end 9) ;;; -> ("oo" "bar" "b"), 9
+#+xcvb (module ())
This is... no. Please find some way of not infesting code with build-related metainformation.
OK, I'll put this much-requested feature near the top of my TODO list.
If every .lisp file has to begin with such an expression, can't you simply introduce a file-as-module defaulting scope?
Where would the inter-file dependencies go, then?
Let's suppose you implied that the module form would be specified if only there actually are any dependencies to speak of.
This implies that you have to use heuristics while interpreting the first form -- is it in the (module (&key &allow-other-keys)) form or not.
Is this acceptable to Faré?
Moreover, the share of leaf modules is only meaningful in simple systems, I believe -- you won't save much in the general case.
So, why bother?
regards, Samium Gromoff -- _deepfire-at-feelingofgreen.ru O< ascii ribbon campaign - stop html mail - www.asciiribbon.org
2009/10/23 Tobias C. Rittweiler tcr@freebits.de:
+#+xcvb (module ())
This is... no. Please find some way of not infesting code with build-related metainformation.
OK, I'll put this much-requested feature near the top of my TODO list.
If every .lisp file has to begin with such an expression, can't you simply introduce a file-as-module defaulting scope?
Not sure what you mean. Unlike say PLT module forms, the XCVB module form does not enclose the whole file, it's just the first form in the file, protected by a #+xcvb so you can keep compiling normally without xcvb.
I need a place to specify the dependencies in a SEXP language, if possible in the file itself (so I can trivially use a Make backend), so that's the obvious solution. A non-trivial Make backend or Make-less backend is of course possible and makes it possible to have dependencies in a central file if you want. I'll be working on the standalone backend next, which will use a cache of tthsum instead of file timestamps to detect change, at which point it will be easier to have file dependency information be centralized, for those maintainers who consider dependency information as "pollution".
PS: I first sent my mail to the cclan-list on sourceforge, and was surprised to receive a note redirecting me to the asdf-devel mailing-list as its successor. What about all the cclan packages that are not asdf? (are there any of them beside split-sequence?)
PPS: Unless it's on-topic for cclan or asdf, please follow-up to xcvb-devel@common-lisp.net.
[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ] Classical Liberalism: the only truly subversive ideology.
-
Christophe Rhodes -
Faré -
Samium Gromoff -
Tobias C. Rittweiler