From elof at image.dk Mon Aug 2 14:16:48 2004 From: elof at image.dk (=?ISO-8859-1?Q?Kristian_Elof_S=F8rensen?=) Date: Mon, 02 Aug 2004 16:16:48 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e Message-ID: <410E4CD0.4040005@image.dk> Hi I downloaded version 0.2.3 from http://www.common-lisp.net/project/ucw/download.html and installed iterate from the url given in docs/INSTALL, the rest of the stuff I had already. I put symlinks to all four *.asd files in the ucw distro into my systems directory and called (asdf:oos 'asdf:load-op :ucw) However compilation fails. This is on CMUCL 18e on Linun on a PC. re-trying so asdf skips the steps that got successfully compiled the first time gives: CL-USER> (asdf:oos 'asdf:load-op :ucw) ; Python version 1.1, VM version Intel x86 on 02 AUG 04 03:41:05 pm. ; Compiling: /home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.lisp 14 JUN 04 05:29:12 pm ; Byte Compiling Top-Level Form: ; Byte Compiling Top-Level Form: ; Compiling DEFMETHOD SHARED-INITIALIZE :AFTER (STANDARD-COMPONENT T): ; Byte Compiling Top-Level Form: ; Compiling DEFMETHOD/CC GOTO-COMPONENT: ; ; ; File: /home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.lisp ; In: DEFMETHOD/CC CALL-COMPONENT ; (DEFMETHOD/CC CALL-COMPONENT ; (# #) ; "Set the place associated with FROM to TO." ; (LET/CC K # # TO)) ; --> PROGN DEFMETHOD PROGN EVAL-WHEN PCL::LOAD-DEFMETHOD PCL::LOAD-DEFMETHOD ; --> LIST* LAMBDA FUNCTION MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS ; --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET ; --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS PCL::CALL-NEXT-METHOD-BIND LET FLET ; --> PCL::BIND-ARGS LET* BLOCK LET IT.BESE.ARNESI::DRIVE-CPS PROGN LAMBDA ; --> FUNCTION FUNCALL C::%FUNCALL FUNCALL C::%FUNCALL FUNCTION ; ==> ; (FLET (#) ; (PROGN # # #) ; (FUNCALL # TO) ; TO) ; Note: Deleting unused function ; #:G2 ; ; Compiling DEFMETHOD/CC CALL-COMPONENT: ; Compiling DEFMETHOD ANSWER-COMPONENT (STANDARD-COMPONENT T): ; ; ; File: /home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.lisp ; In: DEFMETHOD/CC REPLACE-COMPONENT ; (DEFMETHOD/CC REPLACE-COMPONENT (# #) (LET/CC K # TO)) ; --> PROGN DEFMETHOD PROGN EVAL-WHEN PCL::LOAD-DEFMETHOD PCL::LOAD-DEFMETHOD ; --> LIST* LAMBDA FUNCTION MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS ; --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET ; --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS PCL::CALL-NEXT-METHOD-BIND LET FLET ; --> PCL::BIND-ARGS LET* BLOCK LET IT.BESE.ARNESI::DRIVE-CPS LAMBDA FUNCTION ; --> FUNCALL C::%FUNCALL FUNCALL C::%FUNCALL FUNCTION ; ==> ; (FLET (#) ; (PROGN # # # #) ; TO) ; Note: Deleting unused function ; #:G2 ; ; Compiling DEFMETHOD/CC REPLACE-COMPONENT: ; Byte Compiling Top-Level Form: ; ; Error: (during macroexpansion) ; Execution of a form compiled with errors: ; (LET ((CLASS NIL) ; (IT.BESE.YACLML::ID NIL) ; (IT.BESE.YACLML::STYLE NIL) ..)) ; ; Converted GEN-COMPONENT-SHARED-INITIALIZE. ; Compiling DEFUN GEN-COMPONENT-SHARED-INITIALIZE: ; Converted GEN-COMPONENT-DEFAULT-RENDER-ON. ; Compiling DEFUN GEN-COMPONENT-DEFAULT-RENDER-ON: ; Converted GEN-COMPONENT-NAMED-ACTION. ; Compiling DEFUN GEN-COMPONENT-NAMED-ACTION: ; Converted DEFCOMPONENT. ; Compiling DEFMACRO DEFCOMPONENT: ; Byte Compiling Top-Level Form: ; Compiling DEFCLASS TRANSACTION: ; Compiling DEFCLASS TRANSACTION: ; Compiling DEFMETHOD/CC OPEN-TRANSACTION: ; Byte Compiling Top-Level Form: ; ; ; File: /home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.lisp ; In: DEFMETHOD/CC CLOSE-TRANSACTION ; (DEFMETHOD/CC CLOSE-TRANSACTION (#) (LET/CC K #)) ; --> PROGN DEFMETHOD PROGN EVAL-WHEN PCL::LOAD-DEFMETHOD PCL::LOAD-DEFMETHOD ; --> LIST* LAMBDA FUNCTION MACROLET PCL::FAST-LEXICAL-METHOD-FUNCTIONS ; --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET ; --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS PCL::CALL-NEXT-METHOD-BIND LET FLET ; --> PCL::BIND-ARGS LET* PCL::PV-BINDING PCL::PV-BINDING1 PCL::PV-ENV LET LET ; --> BLOCK LET IT.BESE.ARNESI::DRIVE-CPS LAMBDA FUNCTION FUNCALL C::%FUNCALL ; --> FUNCALL C::%FUNCALL FUNCTION ; ==> ; (FLET (#) ; (LET # ; # ; # ; #:CPS-LET-G3-4 ..)) ; Note: Deleting unused function ; #:G2 ; ; Compiling DEFMETHOD/CC CLOSE-TRANSACTION: ; Byte Compiling Top-Level Form: ; /home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f written. ; Compilation finished in 0:00:01. Warning: COMPILE-FILE warned while performing # on #. Warning: COMPILE-FILE failed while performing # on #. ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f". In the sldb window: Execution of a form compiled with errors: (DEFMETHOD IT.BESE.UCW:RENDER-ON :AROUND ((IT.BESE.UCW:RESPONSE IT.BESE.UCW:RESPONSE) (IT.BESE.UCW::COMPONENT IT.BESE.UCW:STANDARD-COMPONENT)) (RESTART-CASE (IT.BESE.YACLML:WITH-YACLML-STREAM (IT.BESE.UCW:CONTENT-STREAM IT.BESE.UCW:RESPONSE) (PROG1 (CALL-NEXT-METHOD) (WHEN IT.BESE.UCW::*INSPECT-COMPONENTS* (IT.BESE.YACLML.TAGS:A :HREF (IT.BESE.UCW::ACTION-HREF (LAMBDA () (IT.BESE.UCW::UCW-INSPECT IT.BESE.UCW::COMPONENT IT.BESE.UCW::COMPONENT))) (IT.BESE.YACLML.TAGS:SPAN :STYLE "font-size: 8pt;" "Inspect this " (IT.BESE.YACLML.TAGS:TT (IT.BESE.YACLML.TAGS:AS-HTML (TYPE-OF IT.BESE.UCW::COMPONENT)))))))) (IT.BESE.UCW::RETRY NIL :REPORT "Retry rendering the component." (RETURN-FROM IT.BESE.UCW:RENDER-ON (IT.BESE.UCW:RENDER-ON IT.BESE.UCW:RESPONSE IT.BESE.UCW::COMPONENT))))) [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR] Restarts: 0: [CONTINUE] Return NIL from load of #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f". 1: [RETRY] Retry performing # on #. 2: [ACCEPT] Continue, treating # on # as having been successful. 3: [ABORT] Abort handling SLIME request. 4: [ABORT] Return to Top-Level. Backtrace: 0: (C::DO-CALL # 26 27 4 ...) 1: (COMMON-LISP::FOP-FUNCALL-FOR-EFFECT) 2: (COMMON-LISP::LOAD-GROUP #) 3: (COMMON-LISP::FASLOAD #) 4: (COMMON-LISP::INTERNAL-LOAD #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f" #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f" :ERROR :BINARY) 5: (COMMON-LISP::INTERNAL-LOAD #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f" #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f" :ERROR NIL) 6: (LOAD #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw_0.2.3/src/rerl/standard-component.x86f" :VERBOSE NIL :PRINT ...) 7: ("DEFMETHOD PERFORM (LOAD-OP CL-SOURCE-FILE)" #<#1=unused-arg> #<#1#> # #) 8: ("LAMBDA (PCL::EFFECTIVE-METHOD-GENSYM-0 PCL::EFFECTIVE-METHOD-GENSYM-1 PCL::EFFECTIVE-METHOD-GENSYM-2)" #<#1=unused-arg> #<#1#> # #) 9: (#:G0) 10: (ASDF:OPERATE ASDF:LOAD-OP :UCW) 11: (ASDF:OPERATE 2 ASDF:LOAD-OP :UCW)[:EXTERNAL] --more-- From mb at bese.it Wed Aug 4 13:36:26 2004 From: mb at bese.it (Marco Baringer) Date: Wed, 04 Aug 2004 15:36:26 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e References: <410E4CD0.4040005@image.dk> Message-ID: Kristian Elof S?rensen writes: > However compilation fails. This is on CMUCL 18e on Linun on a PC. there is, in fairly recent versions of cmucl and sbcl, a PCL bug which UCW triggers. try adding the snippet at http://paste.lisp.org/display/1748#3 to your .cmucl-init.lisp file, delete all the x86f files (a full recompilation is the safest thing to do at this point) and retry. if that still fails try with a latest snapshot of ucw (available from ftp://ftp.common-lisp.net/pub/project/ucw/snapshots/), if that still fails shoot me :). for what it's worth two weeks or so ago i had ucw running on linux x86 with very recent cmucl (19a) and sbcl (cvs) with the apache and portableaserve backends. thanks to drew crampsie and jan rychter for the hardware. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From elof at image.dk Thu Aug 5 07:44:56 2004 From: elof at image.dk (=?ISO-8859-1?Q?Kristian_Elof_S=F8rensen?=) Date: Thu, 05 Aug 2004 09:44:56 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e In-Reply-To: References: <410E4CD0.4040005@image.dk> Message-ID: <4111E578.8070800@image.dk> Marco Baringer wrote: >Kristian Elof S?rensen writes: > > > >>However compilation fails. This is on CMUCL 18e on Linun on a PC. >> >> > >there is, in fairly recent versions of cmucl and sbcl, a PCL bug which >UCW triggers. try adding the snippet at >http://paste.lisp.org/display/1748#3 to your .cmucl-init.lisp file, >delete all the x86f files (a full recompilation is the safest thing to >do at this point) and retry. if that still fails try with a latest >snapshot of ucw (available from >ftp://ftp.common-lisp.net/pub/project/ucw/snapshots/), if that still >fails shoot me :). > > > Well I guess it is shooting season then ;-) First off, the code "snippet" that you linked to fails with a "no class named CONDITION-CLASS in both cmucl 18e and in the recently released cmucl 19a. The full dumps are included at the end. Secondly I tried loading the 2004-07-30 snapshowt of ucw, without loading the "code snippet". I did delete all x86f files in between the loads so as to not get version mismatch problems. I would anticipate that the bug the "code snippet" presumeably fixes must nefixed in the released 19a ??? However both 18e and 19a halts the compilation at the same source code form: >for what it's worth two weeks or so ago i had ucw running on linux x86 >with very recent cmucl (19a) and sbcl (cvs) with the apache and >portableaserve backends. thanks to drew crampsie and jan rychter for >the hardware. > > > Which snapshot of ucw was this? compiling ucw-2004-07-30 on cmucl 19a: ; /home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-session.x86f written. ; Compilation finished in 0:00:00. ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-action.x86f". ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-application.x86f". ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-component.x86f". Execution of a form compiled with errors: (DEFMETHOD RENDER-ON :AROUND ((RESPONSE RESPONSE) (COMPONENT STANDARD-COMPONENT)) (RESTART-CASE (WITH-YACLML-STREAM (CONTENT-STREAM RESPONSE) (PROG1 (CALL-NEXT-METHOD) (WHEN *INSPECT-COMPONENTS* (IT.BESE.YACLML.TAGS:A :HREF (ACTION-HREF (LAMBDA () (UCW-INSPECT COMPONENT COMPONENT))) (IT.BESE.YACLML.TAGS:SPAN :STYLE "font-size: 8pt;" "Inspect this " (IT.BESE.YACLML.TAGS:TT (IT.BESE.YACLML.TAGS:AS-HTML (TYPE-OF COMPONENT)))))))) (RETRY NIL :REPORT "Retry rendering the component." (RETURN-FROM RENDER-ON (RENDER-ON RESPONSE COMPONENT))))) [Condition of type KERNEL:SIMPLE-PROGRAM-ERROR] Restarts: 0: [CONTINUE] Return NIL from load of #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-component.x86f". 1: [RETRY ] Retry performing # on #. 2: [ACCEPT ] Continue, treating # on # as having been successful. 3: [ABORT ] Return to Top-Level. Debug (type H for help) (C::DO-CALL # 22 23 53 ...) Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:code/byte-interp.lisp. 0] compiling ucw-2004-07-30 on cmucl 18e: ; /home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-session.x86f written. ; Compilation finished in 0:00:00. ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-action.x86f". ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-application.x86f". ; Loading #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-component.x86f". Execution of a form compiled with errors: (DEFMETHOD RENDER-ON :AROUND ((RESPONSE RESPONSE) (COMPONENT STANDARD-COMPONENT)) (RESTART-CASE (WITH-YACLML-STREAM (CONTENT-STREAM RESPONSE) (PROG1 (CALL-NEXT-METHOD) (WHEN *INSPECT-COMPONENTS* (IT.BESE.YACLML.TAGS:A :HREF (ACTION-HREF (LAMBDA () (UCW-INSPECT COMPONENT COMPONENT))) (IT.BESE.YACLML.TAGS:SPAN :STYLE "font-size: 8pt;" "Inspect this " (IT.BESE.YACLML.TAGS:TT (IT.BESE.YACLML.TAGS:AS-HTML (TYPE-OF COMPONENT)))))))) (RETRY NIL :REPORT "Retry rendering the component." (RETURN-FROM RENDER-ON (RENDER-ON RESPONSE COMPONENT))))) Restarts: 0: [CONTINUE] Return NIL from load of #p"/home/elof/prog/tests/lisp/Foreign_Code/ucw/src/rerl/standard-component.x86f". 1: [RETRY ] Retry performing # on #. 2: [ACCEPT ] Continue, treating # on # as having been successful. 3: [ABORT ] Return to Top-Level. Debug (type H for help) (C::DO-CALL # 56 57 4 ...) Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:code/byte-interp.lisp. 0] Starting lisp with the "code snippet" on 18e: [elof at hugin systems]$ lisp ; Loading #p"/home/elof/.cmucl-init.lisp". ;; Loading #p"/home/elof/prog/tests/lisp/utils.lisp". Error in function PCL::FIND-CLASS-FROM-CELL: No class named: CONDITION-CLASS. Restarts: 0: [CONTINUE] Return NIL from load of "home:.cmucl-init". 1: [ABORT ] Skip remaining initializations. Debug (type H for help) (PCL::FIND-CLASS-FROM-CELL CONDITION-CLASS NIL T) Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:pcl/macros.lisp. 0] Starting lisp with the "code snippet" on 19a: [elof at hugin systems]$ lisp-19a -core /usr/lib/cmucl/lib/lisp-19a.core ; Loading #p"/home/elof/.cmucl-init.lisp". ;; Loading #p"/home/elof/prog/tests/lisp/utils.lisp". ; Compiling LAMBDA (PCL::.PV-CELL. PCL::.NEXT-METHOD-CALL. CLASS): ; Compiling Top-Level Form: ; ; Warning: Undefined type CONDITION-CLASS ; ; In: LAMBDA (PCL::.PV-CELL. PCL::.NEXT-METHOD-CALL. CLASS) ; (PCL::FAST-LEXICAL-METHOD-FUNCTIONS ; (# PCL::.NEXT-METHOD-CALL. # NIL :METHOD-NAME-DECLARATION ...) ; (DECLARE #) ; (PCL::PV-BINDING # #)) ; --> PCL::BIND-FAST-LEXICAL-METHOD-MACROS MACROLET ; --> PCL::BIND-LEXICAL-METHOD-FUNCTIONS PCL::CALL-NEXT-METHOD-BIND LET FLET ; --> PCL::WITH-REBOUND-ORIGINAL-ARGUMENTS LET PCL::BIND-ARGS ; ==> ; (LET* (#) ; (DECLARE #) ; (DECLARE #) ; (PCL::PV-BINDING # #)) ; Warning: Undefined type CONDITION-CLASS ; ; ; [Last message occurs 2 times] ; Warning: 4 more uses of undefined type CONDITION-CLASS. ; ; ; Warning: This type is undefined: ; CONDITION-CLASS ; ; Compilation unit finished. ; 5 warnings Error in function PCL::FIND-CLASS-FROM-CELL: No class named CONDITION-CLASS. [Condition of type SIMPLE-ERROR] Restarts: 0: [CONTINUE] Return NIL from load of "home:.cmucl-init". 1: [ABORT ] Skip remaining initializations. Debug (type H for help) (PCL::FIND-CLASS-FROM-CELL CONDITION-CLASS NIL T) Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:pcl/macros.lisp. From matley at muppetslab.org Sun Aug 8 21:35:27 2004 From: matley at muppetslab.org (Matley) Date: Sun, 8 Aug 2004 23:35:27 +0200 (CEST) Subject: [Bese-devel] Compilation with cmucl 19a Message-ID: <48383.82.50.103.67.1092000927.squirrel@82.50.103.67> After several tries i managed to compile and use ucw under cmucl. The debian version from sid (i dont know why) suffers from the same problem, found by Kristian Sorensen. The 19a(28/7) downloaded from SourceForge with slime(cvs) and portable-aserve works well, but is a bit slow. I used the ucw snapshot of 30/7. The version with mod-lisp(debian package) doesnt work. It got stalled when i get any url. How can i debug it? Thx From mb at bese.it Mon Aug 9 09:43:41 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 09 Aug 2004 11:43:41 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e In-Reply-To: <4111E578.8070800@image.dk> (Kristian Elof =?iso-8859-1?Q?S?= =?iso-8859-1?Q?=F8rensen's?= message of "Thu, 05 Aug 2004 09:44:56 +0200") References: <410E4CD0.4040005@image.dk> <4111E578.8070800@image.dk> Message-ID: Kristian Elof S?rensen writes: > Well I guess it is shooting season then ;-) > > First off, the code "snippet" that you linked to fails with a "no > class named CONDITION-CLASS in both cmucl 18e and in the recently > released cmucl 19a. The full dumps are included at the end. > > Secondly I tried loading the 2004-07-30 snapshowt of ucw, without > loading the "code snippet". I did delete all x86f files in between the > loads so as to not get version mismatch problems. I would anticipate > that the bug the "code snippet" presumeably fixes must nefixed in the > released 19a ??? However both 18e and 19a halts the compilation at the > same source code form: 19a should not require that patch, the issue has already been fixed. i need you to try one more test: CL-USER> (defparameter *j* 0) *J* CL-USER> (define-condition foo () ((a :initform (incf *j*)))) FOO CL-USER> (defmethod bar ((a a)) a) BAR CL-USER> *j* if that last form returns 0, as it should, then the error is elswhere, if it retruns 1 (as it does with the pcl bug) then you need to somehow fix pcl (using cmucl 19a _should_ be enough). -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From mb at bese.it Mon Aug 9 09:45:40 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 09 Aug 2004 11:45:40 +0200 Subject: [Bese-devel] Compilation with cmucl 19a In-Reply-To: <48383.82.50.103.67.1092000927.squirrel@82.50.103.67> (matley@muppetslab.org's message of "Sun, 8 Aug 2004 23:35:27 +0200 (CEST)") References: <48383.82.50.103.67.1092000927.squirrel@82.50.103.67> Message-ID: "Matley" writes: > After several tries i managed to compile and use ucw under cmucl. > > The debian version from sid (i dont know why) suffers from the same > problem, found by Kristian Sorensen. > > The 19a(28/7) downloaded from SourceForge with slime(cvs) and > portable-aserve works well, but is a bit slow. I used the ucw snapshot of > 30/7. you should try evaluating mp:startup-idle-and-top-level-loop (i'm not sure if that's the name, but it's something along those lines). this without this cmucl's event mechanism (upon which the multithreading stuff is built) is rather slow. > The version with mod-lisp(debian package) doesnt work. It got stalled when > i get any url. How can i debug it? you said on irc that mod_lisp does actually work, so i'm not going to addrees this. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From mb at bese.it Mon Aug 9 11:33:51 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 09 Aug 2004 13:33:51 +0200 Subject: [Bese-devel] Linking to resources (css, js) required by components References: Message-ID: Anthony Juckel writes: [sorry for the extremely long delay in responding to this] -------------------------- linking to external files. -------------------------- > First, why link to external files? Two reasons: 1) To make the > rendered page as clean (i.e. free of cruft) as possible and 2) to > allow for browser caching of said pages. 3) because you can't generate jpegs via yaclml code. 4) because doing it well is not as easy it looks and so we should solve this issue within the framework. this whole thing would be really easy if html allowed us to put javascript and css links anywhere in the page, unfortunetly we have to put css stuff in the body of the HEAD tag, and since we're supposed to put js stuff there as well (we could get away with putting SCRPIT tags more or lessall over the place though) we'll do that. what this requires is that we be able, when rendering the HEAD tag, to know _all_ of the components in the page. i had, up until now, avoided making component nesting exlicit because i didn't really know what kind of API i wanted, now we're going to have to do it. so, here's my current idea: (defgeneric MAP-SUB-COMPONENTS (component function)) (defmacro DO-SUB-COMPONENTS ((symbol component &optinal return-value) &body body) ...) this will allow us to build something whereby each component can decide to include some javascript or it's own special css file and have the proper html put in the head section. this requires that components define methods on MAP-SUB-COMPONENTS instead of just calling RENDER-ON or just saw chris double's nice intro to partial contiuations: http://www.double.co.nz/scheme/partial-continuations/partial-continuations.html for those of you who are intersted in this kind of thing ucw's "contiuations" are actually partial contiuations (though we lack a first order "mark"s) which are lexically limited to the with-call/cc macro. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From elof at image.dk Mon Aug 9 16:04:40 2004 From: elof at image.dk (=?ISO-8859-1?Q?Kristian_Elof_S=F8rensen?=) Date: Mon, 09 Aug 2004 18:04:40 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e In-Reply-To: References: <410E4CD0.4040005@image.dk> <4111E578.8070800@image.dk> Message-ID: <4117A098.5010601@image.dk> Marco Baringer wrote: >Kristian Elof S?rensen writes: > > > >>Well I guess it is shooting season then ;-) >> >>First off, the code "snippet" that you linked to fails with a "no >>class named CONDITION-CLASS in both cmucl 18e and in the recently >>released cmucl 19a. The full dumps are included at the end. >> >>Secondly I tried loading the 2004-07-30 snapshowt of ucw, without >>loading the "code snippet". I did delete all x86f files in between the >>loads so as to not get version mismatch problems. I would anticipate >>that the bug the "code snippet" presumeably fixes must nefixed in the >>released 19a ??? However both 18e and 19a halts the compilation at the >>same source code form: >> >> > >19a should not require that patch, the issue has already been fixed. > >i need you to try one more test: > >CL-USER> (defparameter *j* 0) >*J* >CL-USER> (define-condition foo () > ((a :initform (incf *j*)))) >FOO >CL-USER> (defmethod bar ((a a)) a) >BAR >CL-USER> *j* > >if that last form returns 0, as it should, then the error is elswhere, >if it retruns 1 (as it does with the pcl bug) then you need to somehow >fix pcl (using cmucl 19a _should_ be enough). > > On both 18e and 19a the defmethod call fails with a "no class named A" error. Perhaps you intended the code to be slightly different than what you wrote? Kristian From mb at bese.it Mon Aug 9 16:05:35 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 09 Aug 2004 18:05:35 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e In-Reply-To: <4117A098.5010601@image.dk> (Kristian Elof =?iso-8859-1?Q?S?= =?iso-8859-1?Q?=F8rensen's?= message of "Mon, 09 Aug 2004 18:04:40 +0200") References: <410E4CD0.4040005@image.dk> <4111E578.8070800@image.dk> <4117A098.5010601@image.dk> Message-ID: Kristian Elof S?rensen writes: > On both 18e and 19a the defmethod call fails with a "no class named A" > error. Perhaps you intended the code to be slightly different than > what you wrote? yes, typo. (that should teach to always test what i write) the defmethod form should have been: (defmethod a ((f foo)) f) i checked and this issue has definetly been fixed in cmucl 19a. which is weird since the backtrace you posted makes it look like the error is this... -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From elof at image.dk Mon Aug 9 19:05:37 2004 From: elof at image.dk (=?ISO-8859-1?Q?Kristian_Elof_S=F8rensen?=) Date: Mon, 09 Aug 2004 21:05:37 +0200 Subject: [Bese-devel] compilation error for ucw 0.2.3 with cmucl18e In-Reply-To: References: <410E4CD0.4040005@image.dk> <4111E578.8070800@image.dk> <4117A098.5010601@image.dk> Message-ID: <4117CB01.7020903@image.dk> Marco Baringer wrote: >Kristian Elof S?rensen writes: > > > >>On both 18e and 19a the defmethod call fails with a "no class named A" >>error. Perhaps you intended the code to be slightly different than >>what you wrote? >> >> > >yes, typo. (that should teach to always test what i write) > >the defmethod form should have been: > >(defmethod a ((f foo)) f) > >i checked and this issue has definetly been fixed in cmucl 19a. which >is weird since the backtrace you posted makes it look like the error >is this... > > > Using this version *j* is 0 at the end using both 18e and 19a [elof at hugin elof]$ lisp-18e -core /usr/lib/cmucl/lib/lisp-18e.core -noinit CMU Common Lisp 18e, running on hugin With core: /usr/lib/cmucl/lib/lisp-18e.core Dumped on: Thu, 2003-04-03 15:47:12+02:00 on orion See for support information. Loaded subsystems: Python 1.1, target Intel x86 CLOS 18e (based on PCL September 16 92 PCL (f)) * #+cmu18 t T * (defparameter *j* 0) *J* * (define-condition foo () ((a :initform (incf *j*)))) FOO * (defmethod a ((f foo)) f) # * *j* 0 [elof at hugin elof]$ lisp-19a -core /usr/lib/cmucl/lib/lisp-19a.core -noinit CMU Common Lisp 19a, running on hugin With core: /usr/lib/cmucl/lib/lisp-19a.core Dumped on: Wed, 2004-07-28 18:51:48+02:00 on lorien See for support information. Loaded subsystems: Python 1.1, target Intel x86 CLOS based on Gerd's PCL 2004/04/14 03:32:47 * #+cmu19 t T * (defparameter *j* 0) *J* * (define-condition foo () ((a :initform (incf *j*)))) FOO * (defmethod a ((f foo)) f) ; Compiling LAMBDA (PCL::.PV-CELL. PCL::.NEXT-METHOD-CALL. F): ; Compiling Top-Level Form: # * *j* 0 From mb at bese.it Thu Aug 12 19:55:46 2004 From: mb at bese.it (Marco Baringer) Date: Thu, 12 Aug 2004 21:55:46 +0200 Subject: [Bese-devel] happy birthday ucw! Message-ID: i noticed that the first commit the ucw's cvs repository was on july 26 2003. happy birthday ucw! [holy shit. have i really been messing with this thing for a year already?!?] -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From mb at bese.it Fri Aug 13 15:27:39 2004 From: mb at bese.it (Marco Baringer) Date: Fri, 13 Aug 2004 17:27:39 +0200 Subject: [Bese-devel] the mod_lisp backend Message-ID: start up ucw with the mod lisp backend and run these two commands: $ ab -n 20 -c 20 'http://localhost/ucw/examples/index.ucw' $ ab -n 20 -c 1 'http://localhost/ucw/examples/index.ucw' on my machine the first command requires, on average, 2709ms per request while the second requires 130ms per request, about an order of magnitude less. the key is that the mod_lisp backend is currently single threaded, which means that if we get hit with 20 requests at once that last request will take a _long_ time to run. i'm not great at multithreaded programming (i ignore it as much as possible), suggestions on what the best may to go about multi threading the mod_lisp backend would be very helpfull. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From mb at bese.it Fri Aug 13 15:34:46 2004 From: mb at bese.it (Marco Baringer) Date: Fri, 13 Aug 2004 17:34:46 +0200 Subject: [Bese-devel] ucw is slow. Message-ID: maybe it's not geologically slow, but ucw is still pretty dawn slow. it's acceptable for some types of apps, those where processing and db access dwarf any kind of io and rerl overhead, however for "simple" things ucw is seriously lacking, and it's starting to show. anybody with a good profiler (and the know-how to make good use of it) want to see where ucw's hot spots are? -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From mb at bese.it Mon Aug 16 12:35:34 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 16 Aug 2004 14:35:34 +0200 Subject: [Bese-devel] a wiki developed with ucw Message-ID: matley (aka Luigi Panzeri) is developing a wiki with ucw: http://www.ghislieri.it/ucw/devwiki/index.ucw another ucw app to play around with :) p.s. - (i don't know if i've ever mentioned this or not) i registered the channel #ucw on irc.freenode.net. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen From matley at muppetslab.org Mon Aug 16 14:51:23 2004 From: matley at muppetslab.org (Matley) Date: Mon, 16 Aug 2004 16:51:23 +0200 (CEST) Subject: [Bese-devel] with-slots and action calls Message-ID: <51911.82.50.98.225.1092667883.squirrel@82.50.98.225> defactions containing a with-slots and a call to another action, doesn't compile (on the last ucw snapshot): e.g. (defaction foo ((cp standard-component)) ()) (defaction bar ((cp standard-component)) (with-slots (place) cp (foo cp))) This is the error: * (defaction foo ((cp standard-component)) ()) (defaction bar ((cp standard-com ponent)) (with-slots (place) cp (foo cp))) ; Compiling LAMBDA (PCL::.PV-CELL. PCL::.NEXT-METHOD-CALL. CP): ; Compiling Top-Level Form: FOO * Type-error in KERNEL::OBJECT-NOT-SYMBOL-ERROR-HANDLER: (CP) is not of type SYMBOL [Condition of type TYPE-ERROR] Restarts: 0: [ABORT] Return to Top-Level. Debug (type H for help) (C::PROCESS-TYPE-DECLARATION (PCL::VARIABLE-REBINDING (CP)) #S(C::LEXENV :FUNCTIONS (# # # # # ...) :VARIABLES (# # # # # ...) :BLOCKS (#) :TAGS NIL :TYPE-RESTRICTIONS (# #) ...) (#>)) Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:compiler/ir1tran.lisp. 0] From mb at bese.it Mon Aug 16 15:13:06 2004 From: mb at bese.it (Marco Baringer) Date: Mon, 16 Aug 2004 17:13:06 +0200 Subject: [Bese-devel] On compling ucw on cmucl Message-ID: ucw works fine under cmucl, however it does generate a few warnings. you should get 1 or two warnings about the function ACTION-HREF being undefined (code which calls it is compiled before the function itself gets defined). you should also get a few notes about cmucl deleting unused local (FLET) functions, this is an artifact left by the cps transformer. so if these are the only notes and warnings you see everything is fine, anything else is cause for concern. -- -Marco Ring the bells that still can ring. Forget your perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen