Bug report for APR [2004/12/06]
+---+ | Bugzilla Bug ID | | +-+ | | Status: UNC=Unconfirmed NEW=New ASS=Assigned| | | OPN=ReopenedVER=Verified(Skipped Closed/Resolved) | | | +-+ | | | Severity: BLK=Blocker CRI=CriticalMAJ=Major | | | | MIN=Minor NOR=Normal ENH=Enhancement | | | | +-+ | | | | Date Posted | | | | | +--+ | | | | | Description | | | | | | | |13037|New|Nor|2002-09-26|Alloca configure test fails | |14589|New|Nor|2002-11-15|apr_ipsubnet_create causes unaligned access erro| |16056|New|Maj|2003-01-14|Shared memory mutex ownership not correctly esta| |17232|New|Nor|2003-02-20|configure tries to use -lcrypt before testing to s| |19085|Ass|Enh|2003-04-16|configure should detect Netscape LDAP SDK | |20338|New|Blk|2003-05-29|Compile Problem with signal.c on Tru64 4.0G | |20382|New|Nor|2003-05-31|Poor performance on W2000 AS | |20416|New|Nor|2003-06-02|apr_file_eof isn't portable | |22615|New|Min|2003-08-21|MakeEtags script getting a little rusty? | |23051|New|Maj|2003-09-10|bug in apr_file_dup2() under Win32 | |24325|New|Enh|2003-11-01|apr_reslist doesn't work without threads | |24387|New|Nor|2003-11-04|convert_from_ucs()| |24415|New|Nor|2003-11-05|Problem compiling apr-iconv 0.9.4 | |24424|New|Nor|2003-11-05|apr_iconv using native (OS) iconv | |25103|New|Nor|2003-12-01|Purpose of APR_PROGRAM and APR_PROGRAM_ENV is reve| |26240|New|Enh|2004-01-19|APR_SUBDIR_CONFIG does not work with ./configure -| |27654|New|Nor|2004-03-15|Deadlock in apr_thread_cond_wait on WIN32 platform| |27655|New|Enh|2004-03-15|PATCH: Update requested events for a descriptor th| |27846|New|Min|2004-03-22|Return value of apr_uri_parse wrongly documented. | |28205|New|Nor|2004-04-05|2.0.48 (and .49) Compiling Error on SuSE 9.0 AMD 6| |28453|New|Enh|2004-04-18|apr_uri should parse relative to a base URI | |28712|New|Blk|2004-04-30|sdbm under MS-Windows95/98/Me does not work correc| |28714|New|Nor|2004-04-30|apr_file_lock(!APR_FLOCK_NONBLOCK) never block on | |28718|New|Enh|2004-04-30|apr_sdbm_lock does not pass the APR_FLOCK_NONBLOCK| |29858|New|Enh|2004-06-29|Add support for DragonFly BSD | |29985|Opn|Blk|2004-07-08|poll deprecated in OS X 10.3 and incorrectly detec| |30023|New|Nor|2004-07-10|Configure can't find Berkeley DB | |30130|Ass|Enh|2004-07-15|Native support for Sun LDAP (library/includes)| |30182|New|Nor|2004-07-19|apr_file_write_full does not take timeout into acc| |30913|New|Nor|2004-08-28|apr_proc_create() has unexpected side effect | |31878|New|Nor|2004-10-25|Bug in apr_filepath_root (Win32) - path not \0-ter| |32177|New|Enh|2004-11-11|Apache2 missing SendBufferSize implementation on W| |32244|New|Nor|2004-11-15|configure fails to detect FreeBSD 4.10 version cor| |32477|New|Cri|2004-12-01|Compiling source with Visual Studio .NET 2003 | |32498|New|Nor|2004-12-02|R_OK required by exec*() checks | |32520|New|Nor|2004-12-03|apr_time_exp_[gmt|tz|lt] return neg microseconds f| +-+---+---+--+--+ | Total 36 bugs | +---+
Re: svn commit: r109832 -
On Sun, Dec 05, 2004 at 02:08:31PM -0700, Paul Querna wrote: Joe Orton wrote: ...and back to the beginning, did you read the original thread on this topic? No, I did not remember or was not aware of the 'original thread'. I was looking at use iovecs in mod_log_config, and in IRC Justin mentioned that the !HAVE_WRITEV implementation sucks. That started me down the path of trying to fix it. I assume you are talking about this thread in October? http://marc.theaimsgroup.com/?t=10969352322r=1w=2 Yes, indeed. joe
Re: apr_sha1
At 07:53 AM 12/5/2004, Joe Orton wrote: On Sun, Dec 05, 2004 at 12:53:58PM +, Thom May wrote: You don't say which you propose to change, but either way it's an API change not really worth bumping the major number for, I'd reckon. There's an objection to making the md5 functions return void in STATUS, FWIW. Note that on x86 platforms, in general, adding a return type in place of a previous void type is a noop. However, if you code for testing the 'new' return type in 1.1 - the author is free (by our version rules) to assume their code will work back in 1.0 - and the random garbage will cause their code to hiccup if the test the rv. But this is implementation specific, and it's quite possible that on another platform, potentially mainframe compilers for example, adding a return type may break the call/return stack semantics. This simply isn't a good idea until 2.0. Bill
apr_dbd: a generic SQL wrapper
I've spent some time looking at a generic DBD architecture for Apache, and got some feedback from [EMAIL PROTECTED] and folks on IRC. My thoughts as they were a short time ago are at http://www.apache.org/~niq/dbd.html Since writing that I've added builtin support for prepared statements with varargs versions of the query and select functions, and transaction support, to the API. Several people suggested the project should live in APR. At first I dismissed that because it looked like more work than I intended to do. Now I've reduced that task to one of reformulating how to send backend-dependent stuff to connection functions (eg host, port, dbname), which means it can work by just passing in such args as a single string and implementing a simple parser in each driver. Currently I have code for a proposed apr_dbd that compiles and loads cleanly, comprising apr_dbd.[c|h], a MySQL driver apr_dbd_mysql, and a generic module mod_dbd providing a thin httpd layer and dealing with MPM differences. This is a direct alternative to the architecture proposed in my page referenced above. Are folks interested in having this in APR? If yes I'll hack it to a point where it's fit to show and tidy it up. If no I'll keep it all at the module level as originally envisaged. -- Nick Kew
