- tbnl-devel@common-lisp.net - mailman3.common-lisp.net

[hunchentoot-devel] setuid & setguid
by Nico de Jager 17 Jan '07

17 Jan '07

Hi all Form the Hunchentoot documentation: "On Unix you can use setuid and setgid to change the UID and GID of the process directly after the server has been started. (You might want to do this if you're using a privileged port like 80.) setuid and setgid can be integers (the actual IDs) or strings (for the user and group name respectively)." Does this work on Linux? Will this change the uid and gid of the running lisp image or only of certain threads? In my instance, I start my LispWorks image as root in order to start the server on port 443, and then wish to change the Hunchentoot image to a uid/gid with less privileges. Supplying the :setuid and :setgid parameters have no effect, though, the uid is still shown as root with ps/top. I am running LispWorks 5.0.1 and Debian Sarge backports (kernel 2.6.16) with hunchentoot 0.4.12. Thanks. Nico

2 3

[hunchentoot-devel] New release 0.5.0
by Edi Weitz 17 Jan '07

17 Jan '07

ChangeLog: Version 0.5.0 2007-01-17 More flexible behaviour of RAW-POST-DATA Robustified PARSE-CONTENT-TYPE Download: http://weitz.de/files/hunchentoot.tar.gz Cheers, Edi.

1 0

[hunchentoot-devel] New release 0.4.14
by Edi Weitz 17 Jan '07

17 Jan '07

Changelog: Version 0.4.14 2007-01-17 More meaningful results for RAW-POST-DATA Download: http://weitz.de/files/hunchentoot.tar.gz Cheers, Edi.

1 0

[hunchentoot-devel] Problem with redirect ...
by Nicolas Lamirault 16 Jan '07

16 Jan '07

hello, i have a problem with my Hunchentoot application. i would like to test redirect, so i create this function : (defun handle-redirect () (hunchentoot:redirect "/clappa/index")) and for debugging : CL-USER> (setf hunchentoot:*catch-errors-p* nil) NIL so, in the browser, http://localhost/mytest/redirect i have this : invalid number of arguments: 1 [Condition of type SB-INT:SIMPLE-PROGRAM-ERROR] Restarts: 0: [TERMINATE-THREAD] Terminate this thread (#<THREAD "hunchentoot-worker-5" {C7572A9}>) Backtrace: 0: ((LAMBDA (SWANK-BACKEND::DEBUGGER-LOOP-FN)) #<FUNCTION (LAMBDA #) {ADF96ED}>) and slime is disconnected. in the buffer *slime-events* i have this : ("invalid number of arguments: 1" " [Condition of type SB-INT:SIMPLE-PROGRAM-ERROR]" nil nil) (("TERMINATE-THREAD" "Terminate this thread (#<THREAD \"hunchentoot-worker-5\" {C7572A9}>)")) ((0 "((LAMBDA (SWANK-BACKEND::DEBUGGER-LOOP-FN))\n #<FUNCTION (LAMBDA #) {ADF96ED}>)") (1 "(SWANK::CALL-WITH-BINDINGS\n ((*PRINT-PRETTY* . T) (*PRINT-LEVEL* . 4) (*PRINT-LENGTH* . 10)\n (*PRINT-CIRCLE* . T) (*PRINT-READABLY*)\n (*PRINT-PPRINT-DISPATCH* . #<SB-PRETTY:PPRINT-DISPATCH-TABLE {B3E7B59}>)\n (*PRINT-GENSYM* . T) (*PRINT-BASE* . 10) (*PRINT-RADIX*)\n (*PRINT-ARRAY* . T) ...)\n #<FUNCTION (LAMBDA #) {ADF96AD}>)") (2 "(SWANK::DEBUG-IN-EMACS #<SB-INT:SIMPLE-PROGRAM-ERROR {CBC8EF1}>)") (3 "((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN))\n #<FUNCTION SWANK:SWANK-DEBUGGER-HOOK>\n #<CLOSURE (LAMBDA #) {CBC8F6D}>)") (4 "(SWANK::CALL-WITH-REDIRECTED-IO\n #<SWANK::CONNECTION {B092C31}>\n #<CLOSURE (LAMBDA #) {CBC8F7D}>)") (5 "(SWANK::CALL-WITH-CONNECTION\n #<SWANK::CONNECTION {B092C31}>\n #<CLOSURE (LAMBDA #) {CBC8F6D}>)") (6 "(INVOKE-DEBUGGER #<SB-INT:SIMPLE-PROGRAM-ERROR {CBC8EF1}>)") (7 "(HUNCHENTOOT::MAYBE-INVOKE-DEBUGGER\n #<SB-INT:SIMPLE-PROGRAM-ERROR {CBC8EF1}>)") (8 "(SIGNAL #<SB-INT:SIMPLE-PROGRAM-ERROR {CBC8EF1}>)") (9 "(ERROR SB-INT:SIMPLE-PROGRAM-ERROR)") (10 "(SB-KERNEL::INVALID-ARG-COUNT-ERROR-HANDLER\n #<unavailable argument>\n #.(SB-SYS:INT-SAP #XB6B130AC)\n #<SB-ALIEN-INTERNALS:ALIEN-VALUE :SAP #XB6B12D9C :TYPE (*\n (STRUCT\n SB-VM::OS-CONTEXT-T-STRUCT))>\n (77))") (11 "(SB-KERNEL:INTERNAL-ERROR\n #.(SB-SYS:INT-SAP #XB6B12D9C)\n #<unavailable argument>)") (12 "(\"foreign function: call_into_lisp\")") (13 "(\"foreign function: funcall2\")") (14 "(\"foreign function: interrupt_internal_error\")") (15 "(\"foreign function: sigtrap_handler\")") (16 "(CLAPPA-WEB::HANDLE-ERROR)") (17 "(HUNCHENTOOT::START-OUTPUT NIL)") (18 "(HUNCHENTOOT::PROCESS-REQUEST\n ((:HOST . \"82.225.177.13:9998\")\n (:USER-AGENT\n . \"Mozilla/5.0 (X11; U; Linux i686; fr; rv:1.8.0.7) Gecko/20060830 Firefox/1.5.0.7 (Debian-1.5.dfsg+1.5.0.7-2)\")\n (:ACCEPT\n . \"text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5\")\n (:ACCEPT-LANGUAGE . \"fr,fr-fr;q=0.8,en-us;q=0.5,en;q=0.3\")\n (:ACCEPT-ENCODING . \"gzip,deflate\")\n (:ACCEPT-CHARSET . \"ISO-8859-1,utf-8;q=0.7,*;q=0.7\")\n (:KEEP-ALIVE . \"300\") (:CONNECTION . \"keep-alive\")) ..)") (19 "(HUNCHENTOOT::PROCESS-CONNECTION\n #<HUNCH someone have any idea how can i debug this ? thanks for any help. -- Nicolas Lamirault

2 4

[hunchentoot-devel] Re: [slime-devel] slime-texi patch
by Jeffrey Cunningham 15 Jan '07

15 Jan '07

Sorry. I'll fix it. --Jeff On Sun Jan 14, 2007 at 08:21:57PM -0800, Bill Clementson wrote: > Hi Jeff, > > You posted to the hunchentoot/tbnl list instead of the slime list. >

1 0

[hunchentoot-devel] Re: [slime-devel] slime-texi patch
by Jeffrey Cunningham 15 Jan '07

15 Jan '07

Hmmm. This is strange. I was given commit privileges specifically for doing extensive work on doc/slime.texi. CVS lets me log in alright and seems to accept the changes - prompts me for comments and everything. But I just checked out a fresh copy and none of the changes I made this month (several sessions) are there. And yet, the changes I made in December are there (for example, the image on the front page). I don't know what to make of it. I'm going to post a patch to the list which incorporates the changes you made plus those I've made. (attached) Here's a changelog entry for it: 2007-01-14 Jeff Cunningham <jeffrey(a)cunningham.net> * doc/slime.texi: Update of function/key index with links. --Jeff On Sun Jan 14, 2007 at 07:54:38PM -0800, Bill Clementson wrote: > Hi Jeff, > > On 1/14/07, Jeffrey Cunningham <jeffrey(a)cunningham.net> wrote: > >On Sun Jan 14, 2007 at 05:10:32PM -0800, Bill Clementson wrote: > >> Attached is a patch for slime-texi. Here's the changelog entry: > >> > >> 2007-01-14 Bill Clementson <billclem(a)gmail.com> > >> > >> * doc/slime.texi: Cleanup of documentation typos/errors. > > > >Hi Bill; > > > >I noticed your patch is applied against doc/slime.texi from Jan 1 > >whereas I checked a number of changes into CVS for that file on Jan > >11. Should I do something beyond email the list that I have made > >changes? I am unfamiliar with what is customary in this list. Should I > >also send out a patch as you have done? > > > >I applied your patch to my current version and it worked fine. > > I did a cvs update before I made the changes to slime.texi, so my > patch should be against the latest version of slime.texi. Looking at > the changelog, it appears that the last patch to slime.texi was made > on 2006-12-24, so I don't know what happened to your changes. I just > did a "cvs co slime" as well to see what I got on a fresh install and > the date on slime.texi is also 2006-12-24. Perhaps your changes were > never applied? Since there are a limited number of people with commit > privileges, it's generally a good idea to double-check that any patch > that you email in is actually applied as sometimes they "fall through > the cracks". > > -- > Bill Clementson >

1 0

[hunchentoot-devel] New release 0.4.13
by Edi Weitz 14 Jan '07

14 Jan '07

ChangeLog: Version 0.4.13 2007-01-14 Added favicon.ico to example website (thanks to Yoni Rabkin Katzenell, Toby, and Uwe von Loh) Download: http://weitz.de/files/hunchentoot.tar.gz Cheers, Edi.

1 0

[hunchentoot-devel] favicon.ico patch for test.lisp
by Yoni Rabkin Katzenell 14 Jan '07

14 Jan '07

Some time ago I added two lines to test.lisp which add a favicon to the "menu" page. Maybe someone here will find it useful in a Google search. You don't actually need the patch, but it is the shortest method of explaining what to add where. -- "Cut your own wood and it will warm you twice"

6 12

[hunchentoot-devel] How to pass define-easy-handler symbol?
by Jeffrey Cunningham 14 Jan '07

14 Jan '07

This is probably a dumb question, but I haven't been able to figure it out yet. Suppose I want to call 'DEFINE-EASY-HANDLER inside another function and pass the symbol as an argument. I tried stuff like this: (defun some-function (name) (define-easy-handler ((intern name) (concatenate 'string name ".html") :default-request-type :post) ((formarg :parameter-type :keyword) (html-generator name formarg))) But it wants a real symbol. Then I tried passing names as symbols and turning them into strings where I needed. That works in the sense that it creates the page, but each one is bound to the same symbol no matter what symbols I give it, so only the last page created works. Is there a way to do this? Thanks for any help. --Jeff

2 6

[hunchentoot-devel] Re: hunchentoot specials
by Edi Weitz 11 Jan '07

11 Jan '07

[Cc to mailing list.] Hi! On Wed, 10 Jan 2007 16:38:32 +0100, Gábor Melis <mega(a)retes.hu> wrote: > I've been migrating some of my stuff over to hunchentoot from > paserve and I'm generally happy with it. Good... :) > One thing that strikes me as odd is the number of specials that > govern server behaviour. I'm not talking about *request* and *reply* > since they are dynamic extent by nature, but *log-...*, > *default-...*, and well ... almost everything else. It seems to me > that they make it very hard to run more than one hunchentoot server > with different settings. I'm thinking virtual hosts. > > To keep it short: > > - do you agree? Yes. The reason is Hunchentoot's heritage from TBNL which was a single-server library. And the fact that I myself almost never use more than one server per image. > - if yes, do you think it will be changed? I'm pretty busy right now, so don't hold your breath. But, yes, it will be changed one day. > - would you take patches in this direction if I ever happen to get > around to do it? Sure. However, I get a lot of patches where I finally end up re-writing most of what I received which really doesn't save time. What I'd like to see is: - Follow the coding and indentation conventions I use elsewhere in the code. - Don't use tab characters. - Every function/variable/class/slot/etc. should have a reasonable documentation string. - The patch should include updates to the HTML documentation as well (if necessary). Yes, I'm anal about this, sorry. Anyway - apart from that I'm always happy to incorporate patches. And if I like the new code enough, I might also end up doing the work listed above myself. But it'll be faster if someone else has already done so. Thanks, Edi.

1 0