Menu
▾
▴
naviserver-tracker — SourceForge Tracker Notifications
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
(61) |
Mar
(153) |
Apr
(39) |
May
(10) |
Jun
(15) |
Jul
(15) |
Aug
(2) |
Sep
|
Oct
(17) |
Nov
(2) |
Dec
(13) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
(18) |
Feb
(9) |
Mar
|
Apr
|
May
|
Jun
|
Jul
(7) |
Aug
(1) |
Sep
(2) |
Oct
|
Nov
(1) |
Dec
|
| 2007 |
Jan
(8) |
Feb
(3) |
Mar
|
Apr
|
May
(2) |
Jun
|
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2008 |
Jan
|
Feb
|
Mar
|
Apr
(6) |
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: SourceForge.net <no...@so...> - 2005-10-28 16:02:57
|
Bugs item #1340838, was opened at 2005-10-28 15:57 Message generated for change (Comment added) made by eide You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340838&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current >Status: Closed Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_rename deletes file Initial Comment: ns_rename, set up in "compat.tcl", deletes a file if it is given as source and target. ns_rename <filename> <filename> The underlying "file rename" command would notice that but the file gets deleted before the call to it. ---------------------------------------------------------------------- >Comment By: Bernd Eidenschink (eide) Date: 2005-10-28 16:02 Message: Logged In: YES user_id=197910 Added a check if "from" and "to" match and throw an error if so. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340838&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-28 15:57:50
|
Bugs item #1340838, was opened at 2005-10-28 15:57 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340838&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_rename deletes file Initial Comment: ns_rename, set up in "compat.tcl", deletes a file if it is given as source and target. ns_rename <filename> <filename> The underlying "file rename" command would notice that but the file gets deleted before the call to it. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340838&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-28 14:33:17
|
Bugs item #1340708, was opened at 2005-10-28 13:46 Message generated for change (Comment added) made by eide You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: C-API Group: Current Status: Closed Resolution: Invalid Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_ftruncate broken? Initial Comment: Example: <file> = File with 100 chars Test 1: set f [open <file> w] ns_ftruncate $f 200 close $f File size = 200 bytes completely with "0" bytes. Test 2: set f [open <file> w] ns_ftruncate $f 50 close $f File size = 50 bytes completely with "0" bytes. ns_truncate works as expected when appending or truncating. Or do I just expect the wrong thing from ns_ftruncate? ---------------------------------------------------------------------- >Comment By: Bernd Eidenschink (eide) Date: 2005-10-28 14:33 Message: Logged In: YES user_id=197910 Oh heck! Sorry for stealing your time! Completely overlooked the a/w args... ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-10-28 13:51 Message: Logged In: YES user_id=95086 You are opening with "w" which truncates the file initially. If you wanted to open existing file for writing, without zeroing the file you need to: open file a+ open file w+ or similar. See "man n open" for more info. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-28 14:29:57
|
Bugs item #1340755, was opened at 2005-10-28 14:29 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340755&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_httptime not locale safe Initial Comment: ns_httptime is supposed to return a date in a predictable format, an english locale one; but it does not, see comment in httptime.c: /* * This will most likely break if the locale is not an english one. * The format is RFC 1123: "Sun, 06 Nov 1997 09:12:45 GMT" */ The TCL api uses it in "util.tcl" (ns_setexpires) and "sendmail.tcl" (DATE header). This is sometimes causing mails to be sorted wrong in MUAs or misleading http proxies, depending on locale set. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340755&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-28 13:51:59
|
Bugs item #1340708, was opened at 2005-10-28 15:46 Message generated for change (Settings changed) made by vasiljevic You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: C-API Group: Current >Status: Closed >Resolution: Invalid Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_ftruncate broken? Initial Comment: Example: <file> = File with 100 chars Test 1: set f [open <file> w] ns_ftruncate $f 200 close $f File size = 200 bytes completely with "0" bytes. Test 2: set f [open <file> w] ns_ftruncate $f 50 close $f File size = 50 bytes completely with "0" bytes. ns_truncate works as expected when appending or truncating. Or do I just expect the wrong thing from ns_ftruncate? ---------------------------------------------------------------------- >Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-10-28 15:51 Message: Logged In: YES user_id=95086 You are opening with "w" which truncates the file initially. If you wanted to open existing file for writing, without zeroing the file you need to: open file a+ open file w+ or similar. See "man n open" for more info. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-28 13:46:41
|
Bugs item #1340708, was opened at 2005-10-28 13:46 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: C-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_ftruncate broken? Initial Comment: Example: <file> = File with 100 chars Test 1: set f [open <file> w] ns_ftruncate $f 200 close $f File size = 200 bytes completely with "0" bytes. Test 2: set f [open <file> w] ns_ftruncate $f 50 close $f File size = 50 bytes completely with "0" bytes. ns_truncate works as expected when appending or truncating. Or do I just expect the wrong thing from ns_ftruncate? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1340708&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-22 12:54:54
|
Bugs item #1230150, was opened at 2005-06-29 23:42 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1230150&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current >Status: Closed >Resolution: Invalid Priority: 5 Submitted By: Bernd Eidenschink (eide) >Assigned to: Stephen Deasey (sdeasey) Summary: ns_log Fatal does not shutdown server Initial Comment: In nsd/log.c one noted side effect of the severity flag "fatal" given to ns_log is "WILL CAUSE AOLSERVER TO EXIT!". AFAIK this does not work, e.g. in command mode NaviServer happily logs but does not exit. ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-08-01 23:09 Message: Logged In: YES user_id=87254 Is anyone buying this, can I close this bug? I recently added some code to make Tcl log Tcl_Panic messages to the server log, then abort(). Which is unlike Ns_Fatal, which logs the message and then calls _exit()... ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-08 04:10 Message: Logged In: YES user_id=87254 The Fatal log level is used only twice in the source code: once in the Ns_Fatal command which then calls _exit(), and once in the fatal signal handler which then calls abort(). I don't think you can have Ns_Log exit the process. Tcl code should not, in general, be logging fatal errors because, in general, it's hard to get Tcl fataly wrong. I was saying it would be a bad design to ns_log fatal, not that there aren't other ways to shutdown the server and we need to hide them... :-) The only thing I'd change is the documentation: don't mention 'fatal' as a valid level in the Tcl docs. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-07-07 01:03 Message: Logged In: YES user_id=95086 The "fatal" severity would/should trigger server to go down. Even from Tcl scripts. Otherwise the "fatal" has no real meaning. ---------------------------------------------------------------------- Comment By: Bernd Eidenschink (eide) Date: 2005-07-06 04:31 Message: Logged In: YES user_id=197910 >I'm not sure whether it's a good idea for Tcl procs to >shutdown the server like that. It should be prety difficult >to get into a 'Fatal' situation with Tcl... But they can already do 'ns_shutdown (-restart)'. And many, many more nasty things ;-) IMHO, without a general concept to disarm the developers and their scripts, it should shutdown the server. At least it was always documented that way (in the ancient docs) and people believed it :-) Or we make it a config option: ns_param ShutdownAtFatal true ? ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-06 03:26 Message: Logged In: YES user_id=87254 This comment goes with the function Ns_Fatal(), which does indeed call _exit(1) which should shutdown the server :-) I'm not sure whether it's a good idea for Tcl procs to shutdown the server like that. It should be prety difficult to get into a 'Fatal' situation with Tcl... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1230150&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-22 12:51:40
|
Bugs item #1333276, was opened at 2005-10-20 08:17 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333276&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Bernd Eidenschink (eide) >Assigned to: Stephen Deasey (sdeasey) Summary: 'ns_env unset' removes value, not key Initial Comment: As seen in command line mode: % % ns_env set LALALA 123 % ns_env exists LALALA 1 % ns_env get LALALA 123 % % ns_env unset LALALA % ns_env get LALALA % ns_env exists LALALA 1 % This is contrary to the usual use of variables (and arrays) where an 'unset' command will remove the variable (not only value) and 'exists' will return 0. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333276&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-22 12:51:07
|
Bugs item #1333280, was opened at 2005-10-20 08:21 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333280&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Bernd Eidenschink (eide) >Assigned to: Stephen Deasey (sdeasey) Summary: ns_env get -nocomplain complains; ns_env get will not Initial Comment: Using 'ns_env get' behaves like 'ns_env get -nocomplain' while the latter complains: % % ns_env get AAA % ns_env get -nocomplain AAA no such environment variable: AAA % (Tested in command line mode) ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333280&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-21 10:04:34
|
Feature Requests item #1119257, was opened at 2005-02-09 05:32 Message generated for change (Comment added) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1119257&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Zoran Vasiljevic (vasiljevic) Assigned to: Stephen Deasey (sdeasey) Summary: Cache API, add Tcl interface into core Initial Comment: Those commands are in the server itself: ns_cache_flush ns_cache_stats ns_cache_size ns_cache_names ns_cache_keys I would not touch them for the compatibility reasons. However, they are pretty limited (introspection/management). What is missing is type of functionality added by the nscache module which I believe should have be done long time ago already. Suggestion: Include nscache into the core Tcl commands. Stephen, you mentioned you've benn working on alternate cache implementation. Can we use this and add better Tcl interface? ---------------------------------------------------------------------- >Comment By: Stephen Deasey (sdeasey) Date: 2005-10-21 04:04 Message: Logged In: YES user_id=87254 Attached a snapshot of development. This is obviously taking too long, so here it is. FIXME: Add explanation here... ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-02-09 10:25 Message: Logged In: YES user_id=95086 We can synthesize your changes and Stephens rewrite of the cache guts. I see no problem there. ---------------------------------------------------------------------- Comment By: Vlad Seryakov (seryakov) Date: 2005-02-09 10:20 Message: Logged In: YES user_id=184124 I added ns_cache incr command aslo some time ago, i think it is inthe CVS version of nscache. Also i used to play with core cache to add more prices size calculation incuding overhead. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1119257&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-21 07:18:45
|
Feature Requests item #1333811, was opened at 2005-10-21 01:18 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1333811&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: ns_register_url2file: url2file procs in Tcl Initial Comment: There is a C API for registering url2file callbacks (Ns_SetUrlToFile) and converting a URL to a file path (Ns_UrlToFile). There's a Tcl api for the latter (ns_url2file) but no way tom implement the callbacks. A further limitatation is that url2file callbacks are not passed any context making configuration lookups difficult. The attached patch consolidates all the url2file processing into a single file and adds a Tcl API. It also adds a new C API with the additional feature that url2file callbacks are registered for a URL hierarchy, not just per-virtual server. Ns_RegisterUrl2FileProc(CONST char *server, CONST char *url, Ns_Url2FileProc *proc, Ns_Callback *delete, void *arg, int flags) Callbacks can be registered at runtime, may have context data and opperate over some subset of the URL hierarchy, just like registered procs. Backwards compatibility with Ns_SetUrlToFile is preserved by allowing old style procs to take presedence. ns_register_url2file ?-noinherit? ?--? url script ?arg? ns_unregister_url2file ?-noinherit? ?-recurse? ?--? url ns_register_fasturl2file ?-noinherit? ?--? url ?basepath? ns_register_fasturl2file allows you to register some default url2file handler for the entire URL hierarchy, and then selectively overide some subset with the default implementation. It also optionally allows you to 'mount' some subset of the URL hierarchy onto some portion of the file system path other than the default, which is to concatenate the URL with the page root. There's a reasonably complete set of tests which demonstrate the API. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1333811&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-21 07:03:59
|
Feature Requests item #1202462, was opened at 2005-05-15 14:14 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1202462&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: C-API Group: None >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Zoran Vasiljevic (vasiljevic) Assigned to: Zoran Vasiljevic (vasiljevic) Summary: Rewrite to use Tcl VFS filesystem abstraction Initial Comment: We should really invest some time and rewrite all (most?) of the internal usages of open(), stat() (etc) so we can take advantage of Tcl virtual filesystem wappers. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-06-17 13:22 Message: Logged In: YES user_id=95086 OK. I will proceede as agreed. FYI, the fastpath code will use mmap only if fastpath.map is turned on fastpath cache is disabled or entry is too large for the cache. Only under those circumstances, the mmap is attempted. ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-06-17 13:15 Message: Logged In: YES user_id=87254 Basically, YES. I don't think you need the vfs configuration flag. The check to see if mmap is enabled comes before the vfs flag in the fastpath code, so no matter what you set vfs to, it may be overidden by mmap. You should drop the vfs flag and make vfs the default if mmap is not enabled (as that is currently the only other option). i.e. also drop the fallback to normal open() etc. if vfs is not enabled. The original concern was that the file descriptor would not be available so that more advanced IO calls could be used, not that Tcl_VFSOpen() is noticably slower than open() etc. But you've shown how to deal with that, so I think just use Tcl VFS throughout, for simplicity. The same applies to the adp code, just use Tcl VFS. Any possible minor slowdown when accessing the file sytem should be hidden by the ADP memory cache anyway. So, the instruction for folks who want to use the Tcl VFS hooks to overide normal FS processing would be to make sure mmap (currently, plus any future optimisation) is turned off. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-06-17 03:58 Message: Logged In: YES user_id=95086 This is how I plan to proceede on this topic. 1. Rewrite all filesystem-specific calls to use Tcl pendants except thoser related to ADP and fastpath processing. This would induce no speed penalties to the request processing code. 2. Add fastpath.vfs flag which will, if set, use Tcl VFS path *after* exhausting all other possiblities (mmaping, direct access to file). This way one can override a page from the container file by providing the page on the filesystem. The vfs toggle will bet set to false by default. 3. Change ADP to use TclVFS in the same manner: first trying the file on the filesystem (as done now) and if it fails, using Tcl VFS wrappers. This way the VFS is a second-class citizen and would only be used as a fallback if the file cannot be otherwise found on the filesystem and if the VFS toggle has been flipped. Please give YES or NO (if NO, tell why) ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-05-16 05:43 Message: Logged In: YES user_id=95086 Oh yes. The packaging is what I'm (mostly) after. The only parts where mmap is used is fastpath.c. I do not know what/where is sendfile... So, effectively, the mmap has to be *enabled* to be used, right? Well, there is nothing wrong in leaving the mmap inplace. Actually, it should be also done in the way that Windows platform can use it (nees some wrapping but its easy done). But, it mmap is not enabled? Then the server uses standard open/read/close stuff from the OS. Now, what's wrong in using Tcl API equivalents of that? In that case, it is transparent where the file actually comes from, filesystem, zip/tarfile, etc... Do you think this could hit us performance-wise. I do not know. Maybe, but it's worth trying it out. Also, there are whole lotta places where the server uses OS calls to the filesystem which has nothing to do with serving pages (logfiles, access files, config files, etc...). These are not sensitive performance-wise and those places should/could be rewritten so we can bootstrap from zipfile if needed. Or be entirely wrapped in a startkit. What do you think? ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-05-16 05:30 Message: Logged In: YES user_id=87254 I'd be reluctant to let go of mmap and sendfile, considering our niche as a performace server. You can already use the vfs to read zip files etc., it's the single file packaging your after, right? ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-05-16 00:55 Message: Logged In: YES user_id=95086 The advantage is that you can for example, assemble all of the server distro in a starkit and distribute it as a wrapped application. Also, all sorts of filesystem wrappers (tar, zip, metakit, http, ...) are already implemeted in TclVFS so you can serve pages out of the tar/zip files directly, for example. No, mmap kind of things would be not possible since all would be going thru Tcl API. But, where is the mmap used? I believe only in fastpath and only if you explicitly enable it. And, only for Unix port. ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-05-15 17:25 Message: Logged In: YES user_id=87254 What would be the advantage of using the Tcl VFS? Would it be just as fast? Could we still use performance enhancemts such as mmap() and sendfile()? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1202462&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-21 06:59:10
|
Feature Requests item #1244177, was opened at 2005-07-24 20:06 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1244177&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Load binary modules from init.tcl Initial Comment: This patch add an ns_moduleload command, which calls down to the existing Ns_ModuleLoad(), to load binary modules. It is called from the init.tcl boot-strap file to load the global and per-server modules specified in the config file. It also makes pre-startup callbacks available to Tcl via ns_atprestartup, and adds ns_runonce which makes sure the given script is run only once either globaly or on a per-server basis. It cleans the startup logic a bit by putting more under the control of the init.tcl file. It also paves the way for moving the location and directory structure of modules in the future. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1244177&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-20 17:01:50
|
Bugs item #1333116, was opened at 2005-10-20 13:41 Message generated for change (Settings changed) made by vasiljevic You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333116&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: trying to uuencode single character @ kills server Initial Comment: Giving the '@' character as only character argument to ns_uuencode or ns_base64encode kills the server. It does not affect the decode variants of the commands. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333116&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-20 14:21:06
|
Bugs item #1333280, was opened at 2005-10-20 14:21 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333280&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_env get -nocomplain complains; ns_env get will not Initial Comment: Using 'ns_env get' behaves like 'ns_env get -nocomplain' while the latter complains: % % ns_env get AAA % ns_env get -nocomplain AAA no such environment variable: AAA % (Tested in command line mode) ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333280&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-20 14:17:12
|
Bugs item #1333276, was opened at 2005-10-20 14:17 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333276&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: 'ns_env unset' removes value, not key Initial Comment: As seen in command line mode: % % ns_env set LALALA 123 % ns_env exists LALALA 1 % ns_env get LALALA 123 % % ns_env unset LALALA % ns_env get LALALA % ns_env exists LALALA 1 % This is contrary to the usual use of variables (and arrays) where an 'unset' command will remove the variable (not only value) and 'exists' will return 0. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333276&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-10-20 11:41:42
|
Bugs item #1333116, was opened at 2005-10-20 11:41 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333116&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: trying to uuencode single character @ kills server Initial Comment: Giving the '@' character as only character argument to ns_uuencode or ns_base64encode kills the server. It does not affect the decode variants of the commands. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1333116&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-08-02 05:09:14
|
Bugs item #1230150, was opened at 2005-06-29 23:42 Message generated for change (Comment added) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1230150&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Tcl-API Group: Current Status: Open Resolution: None Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Nobody/Anonymous (nobody) Summary: ns_log Fatal does not shutdown server Initial Comment: In nsd/log.c one noted side effect of the severity flag "fatal" given to ns_log is "WILL CAUSE AOLSERVER TO EXIT!". AFAIK this does not work, e.g. in command mode NaviServer happily logs but does not exit. ---------------------------------------------------------------------- >Comment By: Stephen Deasey (sdeasey) Date: 2005-08-01 23:09 Message: Logged In: YES user_id=87254 Is anyone buying this, can I close this bug? I recently added some code to make Tcl log Tcl_Panic messages to the server log, then abort(). Which is unlike Ns_Fatal, which logs the message and then calls _exit()... ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-08 04:10 Message: Logged In: YES user_id=87254 The Fatal log level is used only twice in the source code: once in the Ns_Fatal command which then calls _exit(), and once in the fatal signal handler which then calls abort(). I don't think you can have Ns_Log exit the process. Tcl code should not, in general, be logging fatal errors because, in general, it's hard to get Tcl fataly wrong. I was saying it would be a bad design to ns_log fatal, not that there aren't other ways to shutdown the server and we need to hide them... :-) The only thing I'd change is the documentation: don't mention 'fatal' as a valid level in the Tcl docs. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-07-07 01:03 Message: Logged In: YES user_id=95086 The "fatal" severity would/should trigger server to go down. Even from Tcl scripts. Otherwise the "fatal" has no real meaning. ---------------------------------------------------------------------- Comment By: Bernd Eidenschink (eide) Date: 2005-07-06 04:31 Message: Logged In: YES user_id=197910 >I'm not sure whether it's a good idea for Tcl procs to >shutdown the server like that. It should be prety difficult >to get into a 'Fatal' situation with Tcl... But they can already do 'ns_shutdown (-restart)'. And many, many more nasty things ;-) IMHO, without a general concept to disarm the developers and their scripts, it should shutdown the server. At least it was always documented that way (in the ancient docs) and people believed it :-) Or we make it a config option: ns_param ShutdownAtFatal true ? ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-06 03:26 Message: Logged In: YES user_id=87254 This comment goes with the function Ns_Fatal(), which does indeed call _exit(1) which should shutdown the server :-) I'm not sure whether it's a good idea for Tcl procs to shutdown the server like that. It should be prety difficult to get into a 'Fatal' situation with Tcl... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1230150&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-08-02 05:03:16
|
Bugs item #1231853, was opened at 2005-07-03 11:14 Message generated for change (Comment added) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1231853&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: Current >Status: Closed >Resolution: Fixed Priority: 5 Submitted By: Bernd Eidenschink (eide) Assigned to: Stephen Deasey (sdeasey) Summary: Incomplete config file leading server to unusable state Initial Comment: I'm not 100% sure if this behaviour should be classified as Bug, lets see. I found that at a bare minimum the server needs something like this to start: --------------------------------------------------- set servername "server1" set serverdesc "My first server" set homedir [file dirname [ns_info config]] ns_section "ns/servers" ns_param $servername $serverdesc ns_section "ns/parameters" ns_param home $homedir --------------------------------------------------- when starting the server with: ./nsd -u nsadmin -g www -f -t ../smallest-config.tcl it says (timestamps and threadinfo deleted): [-main-] Notice: nsmain: NaviServer/4.99.0 starting [-main-] Notice: nsmain: security info: uid=1003, euid=1003, gid=8, egid=8 [-main-] Notice: nsmain: max files: FD_SETSIZE = 1024, rl_cur = 1024, rl_max = 1024 [-main-] Notice: conf: [ns/server/server1]enabletclpages = 0 [-main-] Notice: nsmain: NaviServer/4.99.0 running [-main-] Notice: nsmain: security info: uid=1003, euid=1003, gid=8, egid=8 and waits. It's visible in the process list (ps shows 2 threads), but not binding to a port ('lsof -i' shows no nsd). When stopping the Server with Ctrl-C: [-main-] Notice: nsmain: NaviServer/4.99.0 stopping [-main-] Notice: driver: triggering shutdown [-main-] Fatal: driver: trigger send() failed: Socket operation on non-socket If it is not considered to be a bug there should at least be some log notice like: "Hint: You may have to specify... " or "Incomplete Configuration...." or not start at all. ---------------------------------------------------------------------- >Comment By: Stephen Deasey (sdeasey) Date: 2005-08-01 23:03 Message: Logged In: YES user_id=87254 A server without any configured drivers now logs a warning. It will still shutdown uncleanly, but I don't want to sprinkle checks for zero drivers through the trigger code. AOLS seems cleaner, but alot has changed. TODO... ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-08 06:01 Message: Logged In: YES user_id=87254 That's what I was wondering, does anyone have a use for this? 4 it is then, the server will log a warning if no drivers are configured. I'll add this. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-07-07 01:07 Message: Logged In: YES user_id=95086 The option of 4 is actually OK. I can cook up an application with Tcl using NS as MT-app server having no visible network connections. I would not stop the server if no drivers are loaded, but a warning log seems OK. ---------------------------------------------------------------------- Comment By: Bernd Eidenschink (eide) Date: 2005-07-06 04:23 Message: Logged In: YES user_id=197910 Hm, I tend to Option (3); But just because I can't imagine a use case for (4), is there? Or is it then just a tclsh gone wild? ---------------------------------------------------------------------- Comment By: Stephen Deasey (sdeasey) Date: 2005-07-06 03:45 Message: Logged In: YES user_id=87254 If the only configured driver fails to bind an error is logged but the server starts successfully. I'm not sure how useful a naviserver with no listening drivers is..? But you're right, not specifying any drivers should not leave the server in limbo -- it should either start or it should fail. So here's some options: 1) not specifying any drivers is a fatal error 2) not specifying any drivers is ok (log warning?) 3) no listening drivers for any reason is a fatal error 4) no listening drivers is ok (log warning?) I think maybe option 4, with a warning. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719006&aid=1231853&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-30 03:18:18
|
Feature Requests item #1241432, was opened at 2005-07-20 03:09 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241432&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Tcl error info, with connection info Initial Comment: Here's a patch which moves the Ns_TclLogError and Ns_TclLogErrorRequest routines from nsd/tclinit.c into nsd/tclmisc.c. I've deprecated Ns_TclLogErrorRequest (which isn't used) in favour of detecting whether there is a current connection, and if so, logging the method, url and remote IP. There's a new config setting, errorlogheaders, which is a list of connection headers to log, in addition to the standard info. Just like the nslog module. I've also added a new function Ns_TclLogErrorInfo(Tcl_Interp *, char *) which takes an extra argument and adds it to the Tcl errorInfo stack trace, as a convenience. Ns_TclLogError becomes a simple wrapper of this function. ---------------------------------------------------------------------- >Comment By: Stephen Deasey (sdeasey) Date: 2005-07-29 21:18 Message: Logged In: YES user_id=87254 Commited to CVS. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241432&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-30 03:12:25
|
Feature Requests item #1241351, was opened at 2005-07-20 00:02 Message generated for change (Settings changed) made by sdeasey You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Allocate single interp per-server per-thread Initial Comment: When some C code want to run some Tcl code it first calls Ns_TclAllocateInterp(server) to get an interp. Interps are de-allocated and cached per-thread (per-server) for the next allocation. If there is no cached interp, a new one is created. This can also happen when there was a cached interp but it has already been allocated. Recursive allocation can happen e.g. when a script registered with ns_register_proc (which allocates an interp to run the script) calls the command ns_pagepath (which may allocate an interp to run a registered Tcl callback). Then end result is that more than one interps per-thread per-server can be allocated, the number depending on the ammount of recursion due to Tcl callbacks. This is bad from a memory point of view and a speed point of view as interp traces run each time an interp is allocated. Attached is a patch which implements reference counting on NsInterp structures. Interps are only allocated fresh if no interp has yet been allocated for that thread. If an interp has already been allocated, the reference count is incremented. The refcnt is decremented on de-allocation untill all references are released, and then interp traces and cleanups etc. are run. ---------------------------------------------------------------------- >Comment By: Stephen Deasey (sdeasey) Date: 2005-07-29 21:12 Message: Logged In: YES user_id=87254 Commited to CVS. ---------------------------------------------------------------------- Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-07-25 10:07 Message: Logged In: YES user_id=95086 Go ahead.... The previous one seems like a big waste and memory hog. Why they made it so in the first place? I would also use refcounting on the interp as you did... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-25 16:07:06
|
Feature Requests item #1241351, was opened at 2005-07-20 08:02 Message generated for change (Comment added) made by vasiljevic You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Allocate single interp per-server per-thread Initial Comment: When some C code want to run some Tcl code it first calls Ns_TclAllocateInterp(server) to get an interp. Interps are de-allocated and cached per-thread (per-server) for the next allocation. If there is no cached interp, a new one is created. This can also happen when there was a cached interp but it has already been allocated. Recursive allocation can happen e.g. when a script registered with ns_register_proc (which allocates an interp to run the script) calls the command ns_pagepath (which may allocate an interp to run a registered Tcl callback). Then end result is that more than one interps per-thread per-server can be allocated, the number depending on the ammount of recursion due to Tcl callbacks. This is bad from a memory point of view and a speed point of view as interp traces run each time an interp is allocated. Attached is a patch which implements reference counting on NsInterp structures. Interps are only allocated fresh if no interp has yet been allocated for that thread. If an interp has already been allocated, the reference count is incremented. The refcnt is decremented on de-allocation untill all references are released, and then interp traces and cleanups etc. are run. ---------------------------------------------------------------------- >Comment By: Zoran Vasiljevic (vasiljevic) Date: 2005-07-25 18:07 Message: Logged In: YES user_id=95086 Go ahead.... The previous one seems like a big waste and memory hog. Why they made it so in the first place? I would also use refcounting on the interp as you did... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-25 02:06:29
|
Feature Requests item #1244177, was opened at 2005-07-24 20:06 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1244177&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Load binary modules from init.tcl Initial Comment: This patch add an ns_moduleload command, which calls down to the existing Ns_ModuleLoad(), to load binary modules. It is called from the init.tcl boot-strap file to load the global and per-server modules specified in the config file. It also makes pre-startup callbacks available to Tcl via ns_atprestartup, and adds ns_runonce which makes sure the given script is run only once either globaly or on a per-server basis. It cleans the startup logic a bit by putting more under the control of the init.tcl file. It also paves the way for moving the location and directory structure of modules in the future. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1244177&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-20 09:09:03
|
Feature Requests item #1241432, was opened at 2005-07-20 03:09 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241432&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Tcl error info, with connection info Initial Comment: Here's a patch which moves the Ns_TclLogError and Ns_TclLogErrorRequest routines from nsd/tclinit.c into nsd/tclmisc.c. I've deprecated Ns_TclLogErrorRequest (which isn't used) in favour of detecting whether there is a current connection, and if so, logging the method, url and remote IP. There's a new config setting, errorlogheaders, which is a list of connection headers to log, in addition to the standard info. Just like the nslog module. I've also added a new function Ns_TclLogErrorInfo(Tcl_Interp *, char *) which takes an extra argument and adds it to the Tcl errorInfo stack trace, as a convenience. Ns_TclLogError becomes a simple wrapper of this function. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241432&group_id=130646 |
|
From: SourceForge.net <no...@so...> - 2005-07-20 06:02:27
|
Feature Requests item #1241351, was opened at 2005-07-20 00:02 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Stephen Deasey (sdeasey) Assigned to: Stephen Deasey (sdeasey) Summary: Allocate single interp per-server per-thread Initial Comment: When some C code want to run some Tcl code it first calls Ns_TclAllocateInterp(server) to get an interp. Interps are de-allocated and cached per-thread (per-server) for the next allocation. If there is no cached interp, a new one is created. This can also happen when there was a cached interp but it has already been allocated. Recursive allocation can happen e.g. when a script registered with ns_register_proc (which allocates an interp to run the script) calls the command ns_pagepath (which may allocate an interp to run a registered Tcl callback). Then end result is that more than one interps per-thread per-server can be allocated, the number depending on the ammount of recursion due to Tcl callbacks. This is bad from a memory point of view and a speed point of view as interp traces run each time an interp is allocated. Attached is a patch which implements reference counting on NsInterp structures. Interps are only allocated fresh if no interp has yet been allocated for that thread. If an interp has already been allocated, the reference count is incremented. The refcnt is decremented on de-allocation untill all references are released, and then interp traces and cleanups etc. are run. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=719009&aid=1241351&group_id=130646 |
