Skip to contentSkip to conversation

1.2.6.1 (Was Re: 1.2.6 released. No more queue littering ...)

DISCUSSION Mads Martin Joergensen 2 participants 5 messages May 9, 2005

lore
Share & tools
lore
#0 lore raw
Quoted 8 lines · Mads Martin Joergensen
Quoted 1 line

o If prefix but no Subject: header is present, add one (Jakob Hirsch)

There's a bug here. In case we meet a subject already we forget to clear the flag. This patch should fix it, so that no 2 subjects are present:

do_all_the_voodo_here.c 9 May 2005 07:13:51 -0000 line 148

Quoted 7 lines
do_all_the_voodo_here.c 9 May 2005 07:13:51 -0000 renamed from do_all_the_voodo_here.c 8 May 2005 16:18:02 -0000 1.24 +1 -0
@@ -142,6 +142,7 @@
                /* Add Subject: prefix if wanted */
                if(prefix) {
                        if(strncasecmp(hdrline, "Subject:", 8) = 0) {
+                               subject_present = 1;
                                unqp = cleanquotedp(hdrline + 8);
                                if(strstr(hdrline + 8, prefix) = NULL &&
                                   strstr(unqp, prefix) = NULL) {

So there was another condition causing potential wrong MAIL FROM: on the sub-subscriber mails, and since the Subject: bug is grave enough I've released 1.2.6.1. Sorry for the mess.

Signature · 4 lines
Mads Martin Joergensen, http://mmj.dk
"Why make things difficult, when it is possible to make them cryptic
 and totally illogical, with just a little bit more effort?"
                                -- A. P. J.
#1 lore raw
Quoted 2 lines · Mads Martin Joergensen

So there was another condition causing potential wrong MAIL FROM: on the sub-subscriber mails, and since the Subject: bug is grave enough I've

You mean it's possible that listaddr is not set (or not correct) when generate_subscribed in mlmmj-sub.c is called? I don't see that right now. What am I missing?

#2 lore raw
In reply to Mads Martin Joergensen
Quoted 6 lines · Jakob Hirsch
Quoted 2 lines

So there was another condition causing potential wrong MAIL FROM: on the sub-subscriber mails, and since the Subject: bug is grave enough I've

You mean it's possible that listaddr is not set (or not correct) when generate_subscribed in mlmmj-sub.c is called? I don't see that right now. What am I missing?

It was unsub-notsubscribed, sorry. Anyway; Here's an excerpt from mlmmj-unsub.c (1.2.6 one) line 432:

 if(is_subbed_in(subddirname, address)) {
         /* Address is not subscribed */
         myfree(subddirname);
         myfree(listaddr);

         printf("%s is not subscribed to %s.\n", address, listaddr);
         generate_notsubscribed(listdir, listaddr, address, mlmmjsend);

         exit(EXIT_SUCCESS);

}

Signature · 4 lines
Mads Martin Joergensen, http://mmj.dk
"Why make things difficult, when it is possible to make them cryptic
 and totally illogical, with just a little bit more effort?"
                                -- A. P. J.
#3 lore raw
In reply to Mads Martin Joergensen
Quoted 5 lines · Mads Martin Joergensen
Quoted 3 lines

You mean it's possible that listaddr is not set (or not correct) when generate_subscribed in mlmmj-sub.c is called? I don't see that right now. What am I missing?

It was unsub-notsubscribed, sorry. Anyway; Here's an excerpt from mlmmj-unsub.c (1.2.6 one) line 432:

The change is in both mlmmj-unsub.c and mlmmj-unsub.c.

Quoted 7 lines
 if(is_subbed_in(subddirname, address)) {
         /* Address is not subscribed */
         myfree(subddirname);
         myfree(listaddr);

         printf("%s is not subscribed to %s.\n", address, listaddr);
         generate_notsubscribed(listdir, listaddr, address, mlmmjsend);

Because of myfree(listaddr) and using listaddr after that, right? Then using listaddr in printf is also not ok.
Wouldn't it be easier to move to myfree(listaddr) below the printf and generate_notsubscribed?

And in mlmmj-sub.c listaddr is not freed in the block where generate_subscribed is called.

#4 lore raw
In reply to Mads Martin Joergensen
Quoted 10 lines · Jakob Hirsch
Quoted 7 lines
if(is_subbed_in(subddirname, address)) {
        /* Address is not subscribed */
        myfree(subddirname);
        myfree(listaddr);

        printf("%s is not subscribed to %s.\n", address, listaddr);
        generate_notsubscribed(listdir, listaddr, address, mlmmjsend);

Because of myfree(listaddr) and using listaddr after that, right? Then using listaddr in printf is also not ok.

Yes, the printf is wrong too. *sigh* I need to move this to mempools ASAP.

Quoted 2 lines

Wouldn't it be easier to move to myfree(listaddr) below the printf and generate_notsubscribed?

I would rather only always rely on listdir, since we're sure it's not gone underneath us.

Quoted 2 lines

And in mlmmj-sub.c listaddr is not freed in the block where generate_subscribed is called.

No, the problem only existed in mlmmj-unsub.c

Signature · 4 lines
Mads Martin Joergensen, http://mmj.dk
"Why make things difficult, when it is possible to make them cryptic
 and totally illogical, with just a little bit more effort?"
                                -- A. P. J.