From rjs at fdy2.demon.co.uk Sun Nov 8 21:59:03 2009 From: rjs at fdy2.demon.co.uk (rjs at fdy2.demon.co.uk) Date: Sun, 08 Nov 2009 21:59:03 +0000 Subject: [movitz-devel] Build problem Message-ID: I am having problems building a Movitz image, with SBCL (and CMUCL) I get the following error when compiling memref.lisp. I can't see anything wrong with the code itself but it looks as if an 'ECASE form within another one isn't being expanded properly. There was a change to basic-macros.lisp in July, are there some other diffs waiting to be committed ? Robert Swindells ;; While Movitz compiling MEMREF in losp/muerte/memref.lisp: debugger invoked on a COMMON-LISP:SIMPLE-ERROR: Don't know how to compile compound form ((HOST LITTLE) (MEMREF OBJECT OFFSET INDEX INDEX TYPE UNSIGNED-BYTE16 ENDIAN LITTLE PHYSICALP T)) Type HELP for debugger help, or (SB-EXT:QUIT) to exit from SBCL. restarts (invokable by number or by possibly-abbreviated name): 0: [RETRY-FUNOBJ ] Retry compilation of MEMREF. 1: [SKIP-TOPLEVEL-FORM] Skip the compilation of top-level form DEFUN MEMREF. 2: [RETRY ] Restart Movitz compilation of "losp/muerte/memref.lisp". 3: [CONTINUE ] Skip Movitz compilation of "losp/muerte/memref.lisp". 4: Skip the compilation of top-level form REQUIRE MUERTE/MEMREF. 5: Restart Movitz compilation of "losp/muerte/common-lisp.lisp". 6: Skip Movitz compilation of "losp/muerte/common-lisp.lisp". 7: Skip the compilation of top-level form REQUIRE MUERTE/COMMON-LISP. 8: Restart Movitz compilation of "losp/common-lisp.lisp". 9: Skip Movitz compilation of "losp/common-lisp.lisp". 10: Skip the compilation of top-level form REQUIRE COMMON-LISP. 11: Restart Movitz compilation of #P"losp/los0.lisp". 12: Skip Movitz compilation of #P"losp/los0.lisp". 13: [ABORT ] Exit debugger, returning to top level. (MOVITZ::COMPILE-CONS ((:HOST :LITTLE) (MEMREF OBJECT OFFSET :INDEX INDEX :TYPE :UNSIGNED-BYTE16 :ENDIAN :LITTLE :PHYSICALP T)) # # NIL :EBX NIL) 0] From rvedam at gmail.com Sat Nov 21 19:11:51 2009 From: rvedam at gmail.com (Ram Vedam) Date: Sat, 21 Nov 2009 13:11:51 -0600 Subject: [movitz-devel] compiling LiCE for Movitz Message-ID: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> I don't know if there is a lice-devel mailing list so I thought I post this here and get pointed in the right direction. I downloaded and tried compiling LiCE for movitz, but it would seem that the new version requires a couple of macros to be implemented inside Movitz before it can successfully hook into it. My question is: has anyone tried compiling the new version of LiCE on the movitz platform and if so what sort of issues are you having? Currently, I'm compiling a movitz image using SBCL and it is complaining that "WITH-OPEN-FILE is not implemented yet". Is anyone else having this problem? Ram -------------- next part -------------- An HTML attachment was scrubbed... URL: From sabetts at gmail.com Sat Nov 21 19:31:55 2009 From: sabetts at gmail.com (Shawn Betts) Date: Sat, 21 Nov 2009 11:31:55 -0800 Subject: [movitz-devel] compiling LiCE for Movitz In-Reply-To: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> References: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> Message-ID: <62e3e0b30911211131y48800215g3af7f0506db5cf05@mail.gmail.com> 2009/11/21 Ram Vedam : > I don't know if there is a lice-devel mailing list so I thought I post this > here and get pointed in the right direction. I downloaded and tried > compiling LiCE for movitz, but it would seem that the new version requires a > couple of macros to be implemented inside Movitz before it can successfully > hook into it. My question is: has anyone tried compiling the new version of > LiCE on the movitz platform and if so what sort of issues are you having? > Currently, I'm compiling a movitz image using SBCL and it is complaining > that "WITH-OPEN-FILE is not implemented yet". Is anyone else having this > problem? Hi Ram, Only one version worked on Movitz. I compiled a demo binary and then moved on. Since movitz was/is still very young and I wanted a Real Editor I chose to drop support for movitz and focus on getting it doing useful things on unix OSes. I believe it will still compile with sbcl and the appropriate libraries but I haven't touched it in years. -Shawn From rvedam at gmail.com Sat Nov 21 20:21:10 2009 From: rvedam at gmail.com (Ram Vedam) Date: Sat, 21 Nov 2009 14:21:10 -0600 Subject: [movitz-devel] compiling LiCE for Movitz In-Reply-To: References: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> Message-ID: <2ead50800911211221r6426d05cuf33a9ba781b592e1@mail.gmail.com> Have there been any attempts at writing a filesystem for Movitz? I think that LiCE would definitely compile with no adjustment to Shawn's source if there was a filesystem on Movitz, since it would mean that those macros / functions associated with file I/O would have been implemented. In retrospect, I guess I should have started with that first. I haven't seen any references to an attempted filesystem inside the Movitz source, but I may be looking in the wrong place. Ram On Sat, Nov 21, 2009 at 1:34 PM, Andreas Davour wrote: > On Sat, 21 Nov 2009, Ram Vedam wrote: > > I don't know if there is a lice-devel mailing list so I thought I post >> this >> here and get pointed in the right direction. I downloaded and tried >> compiling LiCE for movitz, but it would seem that the new version requires >> a >> couple of macros to be implemented inside Movitz before it can >> successfully >> hook into it. My question is: has anyone tried compiling the new version >> of >> LiCE on the movitz platform and if so what sort of issues are you having? >> Currently, I'm compiling a movitz image using SBCL and it is complaining >> that "WITH-OPEN-FILE is not implemented yet". Is anyone else having this >> problem? >> > > This list have been really quiet for a long time now. Hopefully somebody > come out and give some feedback on your problem. Personally I don't have > much experience to share. Shawn Betts is, I believe the originator or LiCE > so maybe he knows something. > > I actually started to delve deeper into the heart of movitz to develop some > of the more interesting parts, but I got lost in the source. I don't have > much talent and experience navigating a source maze. > > /andreas > > -- > A: Because it fouls the order in which people normally read text. > Q: Why is top-posting such a bad thing? > A: Top-posting. > Q: What is the most annoying thing on usenet and in e-mail? > -------------- next part -------------- An HTML attachment was scrubbed... URL: From rvedam at gmail.com Sat Nov 21 23:27:20 2009 From: rvedam at gmail.com (Ram Vedam) Date: Sat, 21 Nov 2009 17:27:20 -0600 Subject: [movitz-devel] compiling LiCE for Movitz In-Reply-To: References: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> <2ead50800911211221r6426d05cuf33a9ba781b592e1@mail.gmail.com> Message-ID: <2ead50800911211527q6d3d7dc9p4f28e2f10004ce88@mail.gmail.com> I found the tmp directory with an ext2fs implementation. I'm just starting to look at it. Ram On Sat, Nov 21, 2009 at 4:02 PM, Andreas Davour wrote: > On Sat, 21 Nov 2009, Ram Vedam wrote: > > Have there been any attempts at writing a filesystem for Movitz? I think >> that LiCE would definitely compile with no adjustment to Shawn's source if >> there was a filesystem on Movitz, since it would mean that those macros / >> functions associated with file I/O would have been implemented. In >> retrospect, I guess I should have started with that first. I haven't seen >> any references to an attempted filesystem inside the Movitz source, but I >> may be looking in the wrong place. >> > > There used to be a tmp directory with some stuff about ext3. I haven't > checked out the source in a long time, since I dropped out of lisp for a > while. My source is two computers (generations!) away... > > Filesystems was exactly what I was about to do for movitz, and other have > also tried. It sure would be nice to have. I have so much to do these days, > but if I could get the lisp bug again I think some basis could be laid quite > easy. > > > /andreas > > -- > A: Because it fouls the order in which people normally read text. > Q: Why is top-posting such a bad thing? > A: Top-posting. > Q: What is the most annoying thing on usenet and in e-mail? > -------------- next part -------------- An HTML attachment was scrubbed... URL: From anontrol at gmail.com Sun Nov 22 04:56:53 2009 From: anontrol at gmail.com (anon trol) Date: Sun, 22 Nov 2009 13:56:53 +0900 Subject: [movitz-devel] compiling LiCE for Movitz In-Reply-To: <62e3e0b30911211131y48800215g3af7f0506db5cf05@mail.gmail.com> References: <2ead50800911211111g641f50bbrde3419aabfff3385@mail.gmail.com> <62e3e0b30911211131y48800215g3af7f0506db5cf05@mail.gmail.com> Message-ID: On Sun, Nov 22, 2009 at 4:31 AM, Shawn Betts wrote: > > but I haven't touched it in years. > > How very unfortunate. Seems like it would be an ideal candidate (with a lot of work) for the virtualization and cloud computing memes that have popped up in the subsequent years (you might even be able to market it to AWS users who wanted to trim down their monthly bills). All the overhead running other lisp systems on various OS platforms seems like such an extreme waste... and we all know that x86 is headed to microcontroller land shortly. I'd *really* like to see a small but complete lisp operating system make it to this venue. I might actually get back into embedded programming if that happened. -------------- next part -------------- An HTML attachment was scrubbed... URL: From frode at netfonds.no Mon Nov 23 12:11:58 2009 From: frode at netfonds.no (Frode V. Fjeld) Date: Mon, 23 Nov 2009 13:11:58 +0100 Subject: [movitz-devel] Build problem References: Message-ID: <82zl6d784x.fsf@netfonds.no> Andreas Davour writes: > On Sun, 8 Nov 2009, rjs at fdy2.demon.co.uk wrote: > >> I am having problems building a Movitz image, with SBCL (and CMUCL) I >> get the following error when compiling memref.lisp. >> >> I can't see anything wrong with the code itself but it looks as if >> an 'ECASE form within another one isn't being expanded properly. >> >> There was a change to basic-macros.lisp in July, are there some other >> diffs waiting to be committed ? > > Did anyone every say anything about this? I checked in something that hopefully fixes it, but I had some trouble posting here about it and then forgot.. let's see what happens with this message :) -- Frode V. Fjeld Netfonds Bank ASA From rvedam at gmail.com Fri Nov 27 04:16:27 2009 From: rvedam at gmail.com (Ram Vedam) Date: Thu, 26 Nov 2009 22:16:27 -0600 Subject: [movitz-devel] Filesystem question Message-ID: <2ead50800911262016h758361cif900136c4d995cc6@mail.gmail.com> Hi, I tried creating an image with the ext2fs.lisp file that is inside the tmp directory to see to what degree has the filesystem being implemented, and when I boot into the image using QEMU, the image just goes into an endless cycle of rebooting. Has anyone had any luck trying to create an image using this port of ext2? If so, besides the endless rebooting (which I'm currently trying to figure out how to deal with) are there any other known issues or is it just that it isn't completed? Ram -------------- next part -------------- An HTML attachment was scrubbed... URL: From frode at netfonds.no Fri Nov 27 08:04:53 2009 From: frode at netfonds.no (Frode V. Fjeld) Date: Fri, 27 Nov 2009 09:04:53 +0100 Subject: [movitz-devel] Filesystem question References: <2ead50800911262016h758361cif900136c4d995cc6@mail.gmail.com> Message-ID: <82zl68qtp6.fsf@netfonds.no> Ram Vedam writes: > I tried creating an image with the ext2fs.lisp file that is inside the > tmp directory to see to what degree has the filesystem being > implemented, and when I boot into the image using QEMU, the image just > goes into an endless cycle of rebooting. Has anyone had any luck > trying to create an image using this port of ext2? If so, besides the > endless rebooting (which I'm currently trying to figure out how to > deal with) are there any other known issues or is it just that it > isn't completed? I would suppose that it just isn't completed. When you generate the image, take note of how early the ext2 file is "loaded" at boot-time. A typical cause of boot-time crash/reboot is that certain things (like memory allocation) is attempted before crucial parts of the run-time have been set up. -- Frode V. Fjeld From metawilm at gmail.com Sat Nov 28 09:18:23 2009 From: metawilm at gmail.com (Willem Broekema) Date: Sat, 28 Nov 2009 10:18:23 +0100 Subject: [movitz-devel] unfold-circular-list Message-ID: There's a bug in unfold-circular-list, in that it calls subseq on a circular list: subseq is defined as taking a "proper sequence". Below is a patch. - Willem diff --git a/parse.lisp b/parse.lisp index 4ee75b9..ab711b9 100644 --- a/parse.lisp +++ b/parse.lisp @@ -46,7 +46,7 @@ (loop for x on list as i upfrom 0 as cdr-index = (find-cdr list (cdr x) i) until cdr-index - finally (return (values (subseq list 0 (1+ i)) + finally (return (values (loop repeat (1+ i) collect (pop list)) cdr-index))))) (defun symbol-package-fix-cl (symbol)