The importer is supposed to enable accounts that were enabled in Pidgin. Were these accounts enabled in Pidgin when you imported them?

No, we don't have code for fetching that setting from Pidgin, confirming the bug.

the accounts are enabled in pidgin yes

Created attachment 171829 [details] [review]
proposed fix

Review of attachment 171829 [details] [review]:

Thanks a lot for your patch. I inlined some comments.
You should really learn to use git, it would make your life much easier. :)

::: empathy-2.32.0.1/src/empathy-import-pidgin.c
@@ +95,3 @@
 #define PIDGIN_ACCOUNT_TAG_SETTINGS "settings"
+#define PIDGIN_SETTING_PROP_UI "ui"
+#define PIDGIN_SETTING_PROP_NAME "name"

What are these settings exactly?

@@ +195,3 @@
+  tag_ui = (gchar *) xmlGetProp (settings, (xmlChar *) PIDGIN_SETTING_PROP_UI);
+
+  if (tag_ui && !tp_strdiff (tag_ui, "gtk-gaim"))

why gtk-gaim?

@@ +203,3 @@
+          type = (gchar *) xmlGetProp (setting,
+              (xmlChar *) PIDGIN_SETTING_PROP_TYPE);
+          if (!tp_strdiff (name, "auto-login") && !tp_strdiff (type, "bool"))

Please add some comments explaining the logic.

Created attachment 172096 [details] [review]
updated patch

>What are these settings exactly?

Pidgin stores the required parameter separately from the general settings:
        <settings>
            <setting name='check-mail' type='bool'>0</setting>
            <setting name='buddy_icon_timestamp' type='int'>0</setting>
            etc ...
        </settings>
        <settings ui='gtk-gaim'>
            <setting name='auto-login' type='bool'>1</setting>
        </settings>
PIDGIN_SETTING_PROP_UI is the name of the property indicating that this tag is UI settings.
PIDGIN_SETTING_PROP_NAME is the name of the 'name' property of the 'setting' tag. BTW, there was a minor issue in the original code where this value was mixed up with the name of the 'name' tag, which is not the same thing.

>why gtk-gaim?
No idea why, but this value is still valid for the latest version (2.7.3).

>Please add some comments explaining the logic.
I've updated the patch.

Looks great, I merged your patch to master. Thanks a lot!

This problem has been fixed in the development version. The fix will be available in the next major software release. Thank you for your bug report.

*** Bug 627659 has been marked as a duplicate of this bug. ***