[00:30:10] --- manfred furuholmen has become available [00:54:33] --- dlc has left [01:11:42] --- dev-zero@jabber.org has left [01:24:27] --- Russ has left: Disconnected [03:37:45] --- dev-zero@jabber.org has become available [04:22:30] --- matt has become available [04:37:32] --- manfred furuholmen has left [04:39:47] --- manfred furuholmen has become available [05:32:47] --- canehan has become available [06:20:14] --- dlc has become available [06:41:04] --- dmontuori has become available [08:07:35] --- dlc has left [08:07:49] --- dlc has become available [08:17:23] --- manfred furuholmen has left [08:19:23] --- manfred furuholmen has become available [08:36:10] --- canehan has left [08:36:42] --- stevenjenkins has left [08:42:03] --- SecureEndpoints has left: Replaced by new connection [09:15:17] --- Russ has become available [09:32:16] --- mmeffie has become available [09:47:50] --- reuteras has left [10:38:01] --- Russ has left: Disconnected [10:40:34] --- manfred furuholmen has left [10:50:45] --- dmontuori has left [10:57:10] --- stevenjenkins has become available [11:04:23] --- dmontuori has become available [12:06:52] --- dmontuori has left [12:06:56] --- dmontuori has become available [13:06:31] --- manfred furuholmen has become available [14:02:33] --- dev-zero@jabber.org has left [14:06:02] --- manfred furuholmen has left [14:55:07] --- dmontuori has left [15:05:40] --- dev-zero@jabber.org has become available [17:25:22] --- dev-zero@jabber.org has left [17:48:08] --- mmeffie has left [18:32:48] is jhutz? [18:33:41] is he what? [18:33:54] human? alien? male? bacon? [18:34:35] oi [18:35:41] i created tkt #123902 with no subject, can't del or take, so created #123903 with the subject line. #123902 is junk [18:41:43] looking [18:43:42] of course rt not being responsive [18:43:47] erm [18:44:04] my browser just says waiting for... forever [18:46:44] That's how long it takes to get an RT account--did you try to get one? [18:46:45] :) [18:55:59] is this the day you eat bacon? [18:56:56] tuesday [18:56:57] The Feast Day of St. Baconius of Badden-Badden? [18:57:00] yesterday [18:58:38] do you also venerate...st baconius? [19:00:31] sure, lisa, a wonderful, magical animal [19:04:15] why did the author not create the ticket? [19:04:21] and next time, we can fix the subject [19:04:27] rt eventually bothered to reply [19:04:36] author didn't create most of it, giving credit [19:04:40] taking blame [19:04:55] ok, i understand that. i have done it often enough [19:06:34] rt seems to occasionally get "unable to access the database" stupid, almost like the database is locked. [19:07:46] upgrading it soon, apparently... [19:07:47] at other times, the https port, and only the https port, becomes unresponsive and if you edit the url to http:// the page loads while the https tab continues to not load. (even if you try to reload it) [19:07:50] --- SecureEndpoints has become available [19:14:02] jeff: change 1 AFSCL... == 1500 probably is based on changes already in 1.5.55, probably by you, ne? of the other 3-4 "changes" (arent these microscopic, really?) 3 are by me--including building with WinDDK 6001, which, well, dunno if it works as posted. I decided that maybe this would save somebody an hour if they were in now, and redundant real soon... [19:14:28] er. what did i just miss? [19:14:46] (RT) >Please provide explanations for the various changes. Submit one ticket per change. mlane can submit tickets and should by sending e-mail to openafs-bugs@openafs.org so that he can take part in the dialog. [19:14:57] ah [19:15:14] kdfs branch not sync'd to 1.5.x [19:15:24] i bet RT has the capability to allow CCs on tickets. [19:16:36] anyway, given transient and modest nature of this stuff (and all build/NTMakefile only), could I request to just post a comment breaking it down? [19:16:44] there are more changes than just those. not all of them look correct or even appropriate. for example, why patch ntbuild.bat to match your environment when it is an example? [19:17:21] Isn't the AFSCL_... part not needed in the exemplar? [19:17:41] also, if as in 1.5.55 I add NSIS variable that I use later, shouldn't that be in the examplar? [19:17:55] you added it twice [19:18:07] oh, that's in error [19:20:11] anyway, not counting dups or mistakes, I think ntbuild should get certain updates--not my local config unless I add/remove something exemplary, was how I figured that [19:21:47] mlane added OutputDebugString somewhere, add him to CC for that one, if you find it bogus [19:22:08] OutputDebugString should not be in production code [19:22:27] or committed code [19:22:36] not conditionally compiled? [19:22:41] we have a trace logging system [19:22:50] check [19:26:17] but seriously, there are non-obvious changes being made in that patch. explanations for why you are making them are required. Why this change? extra: afsrdr - IF EXIST($(SRC)\WINNT\extra) && EXIST($(SRC)\WINNT\extra\NTMakefile) +! IF EXIST($(SRC)\WINNT\extra) && EXIST($(SRC)\WINNT\extra\NTMakefile) echo ***** $@ [19:26:55] since what that says is if the directory does not exist and the NTMakefile does, then .... which is clearly worng [19:27:24] hm. don't know, I take your point, I'll do as you suggest [19:27:46] and why would I even consider taking a patch that only adds an OutputDebugString? [19:28:19] and why is the afs_config.exe.manifest being altered? [19:29:26] there is a reason the control panel is required to run under highest privileges, so that under UAC the user gets prompted to run as Administrator so that the changes the user makes to the registry actually occur to the registry instead of the shadow copy created for the unprivileged application [19:30:00] I take a look at a patch like this and ask myself "why should I waste my time thinking hard about this?" [19:30:02] Not something that should be committed, the latter, just flags an issue with vs2008 build, probably shouldn't have sent that. [19:30:37] Is afs_config.exe in fact part of redirector dist, in fact? [19:31:10] its not part of the distribution that is used by REDACTED [19:31:28] and the drive mapping functionality has SMB dependencies [19:31:40] so it won't work with the redirector as is [19:31:48] and replacing it is not part of the contract [19:31:56] or even modifying it [19:32:25] That's what I thought. Would it be preferable not to link it? [19:32:35] huh? [19:33:02] this is a patch against a branch [19:33:21] which has to be merged against 1.5 [19:33:40] ok, will send alternate change [19:33:44] I don't understand why you are touching afs_config.exe? [19:34:00] the manifest change you sent breaks it worse that it already is [19:34:12] Not, per se. Sorry. [19:34:40] the manifest change disables the ability for Windows to detect that the process requires Admin privs [19:35:02] anyone that uses it will think the process is working but in reality it won't [19:35:28] Yes, the binary was not intended to be used, as such. I am not proposing any change to any shipping afs_config.exe. [19:36:12] I don't understand what benefit would come from that change being applied to the afs_config.exe on kdfs [19:36:37] None. It builds to completion. It meant nothing else. I assumed you would find it convenient to pick off anything useful from diff--evidently, this was confusing. [19:37:42] you submitted a patch to RT. That requires that I spend my time evaluating it. If you want to ask whether something might be useful, ask the question. [19:38:17] Point taken. If it's useful, I'll pick this apparent and send separate diffs--is it? [19:38:24] not really [19:38:27] sorry, apparent == apart [19:38:36] it won't help me any [19:38:56] it won't help REDACTED [19:39:32] what REDACTED needs is someone to write a new MSI for them that is not based on the OpenAFS.org distribution but based upon the way they distribute the code internally. [19:39:42] REDACTED does not use the OpenAFS.org build. [19:39:51] they have their own build process [19:42:22] Aware of that. I think, though, that if some set of build changes is obviously correct, it might help people other than you, build src. [19:43:45] we don't support building openafs with 2008 yet. we build openafs for windows with 2005 because we still need to build for 2000 [19:44:23] Right, but we will at some point, no? [19:45:02] I guess I can just keep patches, and discover whenever you've made the equivalent changes, but avoiding duplication is what RT and such are for, I thought. If a patch doesn't revert preferred build configurations, I'd think we'd want to merge it. [19:45:04] when we give up win2000 support [19:45:32] not on the kdfs branch. if you want to submit 2008 changes, submit them for the head [19:45:58] Ok, thats helpful. [19:45:58] if you want to improve the build documentation, submit it for the head [19:46:32] Is 1.5.x useful? [19:46:41] there are only three directories that will be merged from kdfs to the 1.5-something branch [19:46:57] everything must go to the head first. [19:47:09] if you give me for 1.5, I have to rework for the head [19:47:09] ok [19:47:39] don't create unnecessary work [19:47:51] I'll send head, I wasn't sure how good head was. (Not trying to, obviously) [19:48:46] the kdfs branch is a temporary work branch. its not up to date with the rest of the repository. putting stuff there will only result in it getting lost [19:50:39] GONE [19:50:47] --- matt has left [20:03:14] regardless of how good the head is, everything starts there. this was a huge problem for rees, for reasons i never entirely grasped. [20:03:48] the head was broken long term once. we then reverted the rx tcp work and moved it to a branch [20:04:07] which is where it should have been in the first place, but we were foolish. well, ken and i were, at least [20:08:03] speaking of branches..how is the cvs -> git work going? [20:10:25] i have the git blob from max. i am still working through the conflicts. most look like just a filure to detect merged changes in the automated process. there are a couple i don't get [20:11:13] max was concerned that roughly half the deltas created merge conflicts. do those mostly look solvable in an automated way? [20:15:25] he has an idea yesterday that he was going to try which sounded plausible [20:50:09] --- matt has become available [20:52:20] I think Jim was concerned that Unix and Windows stable branches, are, not the same branch. I don't know what he thought about head. [21:16:51] --- matt has left [21:49:09] jim wasn't interesting in committing his changes to the unix devel branch, regardless of what branch of windows it was, in spite of other gatekeepers telling him it was important. but it's water under the bridge now. [22:57:56] --- reuteras has become available [23:59:30] --- manfred furuholmen has become available