Re: Are there any specific info for the four vague change log entries?
On 22 Feb 2011, at 07:47, Kuge, Tetsuo wrote: Without PR number, it seemed too hard to point out the changed code source code. So I wished that there appear specific PR number from some root Change Log data. Not as such. Check the SVN record for the context of the changes that introduced each of those. Or tell us what problem you're trying to solve and maybe we can tell you something specific. All right. Since I am new to Apache development community, I would like to start from checking SVN and SVN record to ensure if we can identify the relationship with the four Change Log entries. Use the annotated view: http://svn.apache.org/viewvc/httpd/httpd/branches/2.2.x/CHANGES?view=annotate Find the lines you're looking at, and that'll tell you the revision they're in. Now you can see the whole revision and figure out what it's about. You can also look for nearby revisions of the files affected, etc. To take your first example, (1) line 1788, Version 2.1.9, module: mod_proxy Description: *) mod_proxy: Run the request_status hook also if there are no free workers or all workers are in error state. [Ruediger Pluem, Brian Akins brian.akins turner.com] The annotated view tells you this log entry was introduced in revision 306900. So go to http://svn.apache.org/viewvc?view=revisionrevision=306900 and view the change to the mod_proxy.c code, which in this case looks well-commented. If necessary, check nearby revisions that might be related. Note that it's often better to do this looking up in /trunk/ than in 2.2! If you still can't figure it out, that would be a good time to ask here! -- Nick Kew Available for work, contract or permanent http://www.webthing.com/~nick/cv.html
RE: Are there any specific info for the four vague change log entries?
Thank you to instruct a good / specific usage of svn.apache.org. I was about to look into SVN project for some manual page of SVN. Using annotation view to identify the revision number is another good solution than listing up PR numbers. // Using SVN via some scripts like ruby seems attractive for me. -- Regards, Tetsuo. Full name Tetsuo Kuge Hewlett-Packard Japan, Ltd. E-mail tetsuo.k...@hp.com
Re: Time to start planning for httpd 2.3.11-BETA ?
On Feb 14, 2011, at 5:00 PM, William A. Rowe Jr. wrote: On 2/12/2011 10:14 AM, Stefan Fritsch wrote: On Thursday 10 February 2011, Jim Jagielski wrote: Let's commit to pushing for a 2.3.11-BETA... +1 One question: How do we handle truly experimental modules (e.g. mpm_simple, mod_serf). Move them into an experimental subdir before branching or branch first and then move stuff around? Or just mark them as experimental in the docs and maybe in configure? IMHO, all 'experiments' should be provided in all alpha and beta tarballs (disabled by default, obviously). When 2.4 is forked, these can be cast out of the tarball and provided separately until they have the support of the developers. It seems too many legitimate bits have been left in the dumpster of experimental for too long in the past, and having a clear 'promote before GA' policy would help recruit eyeballs to elevate them to code with oversight. +1
Re: Time to start planning for httpd 2.3.11-BETA ?
I think we're about ready... My plan is to TR 2.3.11-beta the start of next week, allowing this week for some final touches... On Feb 10, 2011, at 9:27 AM, Jim Jagielski wrote: Let's commit to pushing for a 2.3.11-BETA...
Re: Time to start planning for httpd 2.3.11-BETA ?
On 22 Feb 2011, at 17:13, Jim Jagielski j...@jagunet.com wrote: I think we're about ready... My plan is to TR 2.3.11-beta the start of next week, allowing this week for some final touches... Remind me, at what point does the API freeze? Regards, Graham --
Re: Time to start planning for httpd 2.3.11-BETA ?
On 2/22/2011 9:43 AM, Graham Leggett wrote: On 22 Feb 2011, at 17:13, Jim Jagielski j...@jagunet.com wrote: I think we're about ready... My plan is to TR 2.3.11-beta the start of next week, allowing this week for some final touches... Remind me, at what point does the API freeze? When 2.3-beta becomes 2.4.0, the API is frozen, and 2.3-beta forks to a 2.5 working branch for forward API actions into 2.6 or 3.0
Re: Time to start planning for httpd 2.3.11-BETA ?
On Feb 22, 2011, at 10:45 AM, William A. Rowe Jr. wrote: On 2/22/2011 9:43 AM, Graham Leggett wrote: On 22 Feb 2011, at 17:13, Jim Jagielski j...@jagunet.com wrote: I think we're about ready... My plan is to TR 2.3.11-beta the start of next week, allowing this week for some final touches... Remind me, at what point does the API freeze? When 2.3-beta becomes 2.4.0, the API is frozen, Yeppers... Right now, it's just slushy ;)
Configurable suexec bin
Hello there. Some days ago I found that I'm tired of original suexec which is shipped with apache. I have two issues: 1.I'd like to configure it with config file, not with rebuilding, because I use modern OS with package system and don't want to depend on self-compiled components. 2.I'd like to use apache2+fcgid+suexec+php5. But with original suexec I had to put dumb script to every users docroot, which only execs /usr/bin/php-cgi. So I just want to allow suexec execute some commands out of docroot tree and owned by the users other that one we setuid to. That's why I created a little bit patched version of suexec, called suexec-conf: https://github.com/gerasiov/suexec-conf Please take a look at this small project and comment. I tried not to lessen security a lot, but may be I've made some errors. I plan to upload it to Debian, so I could use apache+fcgid+suexec+php5 out of the box on any Debian server. PS Yes, I know about mpm-itk, but for some reason, I prefere threaded mpm + suexec -- Best regards, Alexander GQ Gerasiov Contacts: e-mail:g...@cs.msu.su Jabber: g...@jabber.ru Homepage: http://gq.net.ru ICQ: 7272757 PGP fingerprint: 04B5 9D90 DF7C C2AB CD49 BAEA CA87 E9E8 2AAC 33F1
Re: Configurable suexec bin
On Tue, 2011-02-22 at 20:04 +0300, Alexander GQ Gerasiov wrote: because I use modern OS with package system and don't want to depend on Now there's a contradiction. 2.I'd like to use apache2+fcgid+suexec+php5. But with original suexec I apache2 ?? whats that sarcasm? it is called apache ... be it v1.x or 2.x or future 3.x ... I plan to upload it to Debian, so I could use apache+fcgid+suexec+php5 out of the box on any Debian server. ahh, if thats your version of modern, id HATE to see what you called old signature.asc Description: This is a digitally signed message part
Re: Configurable suexec bin
Hello, Noel. On Wed, 23 Feb 2011 09:27:42 +1000 Noel Butler noel.but...@ausics.net wrote: On Tue, 2011-02-22 at 20:04 +0300, Alexander GQ Gerasiov wrote: because I use modern OS with package system and don't want to depend on Now there's a contradiction. No it is not. You just need to separate me as software maintainer and me as OS user. Anyway this is not the subject for debates, I suppose. 2.I'd like to use apache2+fcgid+suexec+php5. But with original suexec I apache2 ?? whats that sarcasm? it is called apache ... be it v1.x or 2.x or future 3.x ... Nope, I dont understand your aggression, but in some places there are exist (or there were in the past, doesn't matter) apache1 for the reason of compatibility and stability, and apache2 as the current version which is modern, but could lead to some incompatibilities. And I don't see any problem here. I plan to upload it to Debian, so I could use apache+fcgid+suexec+php5 out of the box on any Debian server. ahh, if thats your version of modern, id HATE to see what you called old As I can see, you said nothing concrete, your post is full of unexpected rage. That's why I think you're a troll or just a developer with some mental problems. Next time I'll just ignore you. PS If anyone sees an aggression or provocation in this mail or in previous one, you're wrong, there isnt any. -- Cheers, Alexander GQ Gerasiov Contacts: e-mail:g...@cs.msu.su Jabber: g...@jabber.ru Homepage: http://gq.net.ru ICQ: 7272757 PGP fingerprint: 04B5 9D90 DF7C C2AB CD49 BAEA CA87 E9E8 2AAC 33F1
Re: Configurable suexec bin
On Wed, 2011-02-23 at 02:46 +0300, Alexander GQ Gerasiov wrote: Hello, Noel. If you have seen apache2 it is only because debian (and its clones) call it that, I've seen an apache tree call it that yet. But, thats a pretty typical response from a debian troll, debian is right and everyone else is wrong *sigh* signature.asc Description: This is a digitally signed message part
Re: Configurable suexec bin
On Wed, 2011-02-23 at 11:55 +1000, Noel Butler wrote: If you have seen apache2 it is only because debian (and its clones) call it that, I've seen an apache tree call it that yet. Correction, I've *never* seen... signature.asc Description: This is a digitally signed message part
Re: svn commit: r1073520 - /httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c
On 02/22/2011 10:43 PM, j...@apache.org wrote:
Author: jim
Date: Tue Feb 22 21:43:44 2011
New Revision: 1073520
URL: http://svn.apache.org/viewvc?rev=1073520view=rev
Log:
Be at least somewhat more RESTful... Use POST for changing stuff.
Modified:
httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c
Modified: httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c
URL:
http://svn.apache.org/viewvc/httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c?rev=1073520r1=1073519r2=1073520view=diff
==
--- httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c (original)
+++ httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c Tue Feb 22 21:43:44
2011
@@ -882,28 +930,27 @@ static int balancer_handler(request_rec
}
}
-if (r-args) {
-char *args = apr_pstrdup(r-pool, r-args);
-char *tok, *val;
-while (args *args) {
-if ((val = ap_strchr(args, '='))) {
-*val++ = '\0';
-if ((tok = ap_strchr(val, '')))
-*tok++ = '\0';
-/*
- * Special case: workers are allowed path information
- */
-if ((access_status = ap_unescape_url(val)) != OK)
-if ((strcmp(args, w) strcmp(args, b_nwrkr)) ||
(access_status != HTTP_NOT_FOUND))
-return access_status;
-apr_table_setn(params, args, val);
-args = tok;
-}
-else
-return HTTP_BAD_REQUEST;
-}
-}
+if (r-args (r-method_number == M_GET)) {
+const char *allowed[] = { w, b, nonce, NULL };
+ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, r-server, parsing
r-args);
+push2table(r-args, params, allowed, r-pool);
+}
+if (r-method_number == M_POST) {
+apr_bucket_brigade *ib;
+apr_size_t len;
+char *buf = apr_pcalloc(r-pool, 1024);;
+
+ib = apr_brigade_create(r-connection-pool,
r-connection-bucket_alloc);
+rv = ap_get_brigade(r-input_filters, ib, AP_MODE_READBYTES,
+APR_BLOCK_READ, 1024);
Couldn't this cause a one by off error later on as buf is only 1024 bytes long
and we need
a terminating '\0'?
+if (rv != APR_SUCCESS) {
+return HTTP_INTERNAL_SERVER_ERROR;
+}
+apr_brigade_flatten(ib, buf, len);
+buf[len] = '\0';
+push2table(buf, params, NULL, r-pool);
+}
if ((name = apr_table_get(params, b)))
bsel = ap_proxy_get_balancer(r-pool, conf,
apr_pstrcat(r-pool, BALANCER_PREFIX, name, NULL));
