2013-12-15 GnuCash IRC logs

00:08:18 *** gncbot has joined #gnucash
00:42:05 *** gncbot has joined #gnucash
01:05:58 <john> @tell warlord: Derek, check out code when you get up. As of 0030R, I can't dcommit, and web access (to trac or the win32 build dir) times out. Ping shows 0 packet loss with ~140ms round trip time..
01:06:58 <gncbot> john: The operation succeeded.
01:06:59 *** john sets mode: +o gncbot
01:07:29 *** john has left #gnucash
01:07:55 *** john has joined #gnucash
01:17:28 *** gncbot has joined #gnucash
01:21:04 *** fell has quit IRC
05:01:54 *** gncbot has joined #gnucash
09:59:30 *** gncbot has joined #gnucash
10:09:47 *** gncbot has joined #gnucash
10:09:52 <warlord> And there's gncbot
10:09:55 *** warlord sets mode: +o gncbot
10:11:19 <warlord> please let me know if code isn't working now.. it should be back.
10:11:25 <warlord> LiamH: wiki should be back now
10:11:50 <LiamH> warlord: yes, looks good, thanks
10:12:58 <warlord> HtH
10:18:03 <gjanssens> warlord I am getting mailing list mails again, looks like things are working again
10:18:06 <gjanssens> thanks
10:26:23 <LiamH> OK, I seem to be getting up to date price quotes.
10:53:51 *** LiamH has quit IRC
11:04:46 <mikee> @op
11:04:47 *** gncbot sets mode: +o mikee
11:08:21 *** MrLurker has joined #gnucash
11:09:21 *** ErKa has joined #gnucash
11:12:15 *** mikee is now known as mikee-afk
11:15:10 *** jmd has joined #gnucash
11:18:08 *** jmd has quit IRC
11:34:54 *** TradeBorG111 has joined #gnucash
11:41:15 <gjanssens> warlord: some things are still not quite right
11:41:27 *** aqua___ has quit IRC
11:41:37 <gjanssens> I received an svn commit message for r23566 (by john)
11:41:58 <gjanssens> but the trac instance shows r23562 as most recent commit
11:42:09 <gjanssens> so 4 commits seem to have gone missing
11:45:13 *** aqua___ has joined #gnucash
11:46:14 *** nafg has joined #gnucash
11:48:51 <gjanssens> oops: I'm wrong
11:49:04 <gjanssens> r23562 was what I got the mail for.
11:49:24 <gjanssens> But perhaps you should trigger the git synchronization once manually
11:49:33 <gjanssens> github is still at r23461
11:50:12 *** john has joined #gnucash
11:50:26 <gjanssens> Hi john
11:50:47 <gjanssens> I was about to write a mail because your last commit is not synched to github yet
11:51:06 <gjanssens> Was it supposed to fix Christian's issue with the register as well ?
11:52:31 <john> Morning, Geert. There was a problem with Code last night that blocked the release commits. It might have messed up github as well.
11:53:55 <john> If a commit isn't synced to Github that might explain why I'm getting a weird error message from dcommit now.
11:54:00 *** amjam has joined #gnucash
11:54:48 <gjanssens> I asked warlord to check the git synch about five minutes befre you appeared online
11:54:53 <gjanssens> He didn't respond yet
11:55:52 *** nafg has quit IRC
11:58:55 <john> Looks like code's git repo isn't synced to svn, either.
12:01:07 <john> The commit that's stuck in limbo is my import problem. The one to fix Christian's entry problem is in my queue to dcommit, blocked by 23562 not being synced with git.
12:02:09 <gjanssens> Do you still have an svn working copy around somewhere ?
12:02:30 <john> Yes.
12:02:30 <gjanssens> Pushing a commit via svn directly will probably trigger the proper sync actions
12:04:49 <john> OK. I can try that.
12:06:50 *** nafg has joined #gnucash
12:12:50 *** tdrobison has joined #gnucash
12:14:16 <tdrobison> Hellp
12:14:27 <tdrobison> New to gnucash
12:15:23 <john> OK. Have you read the concept guide and looked at the wiki, especially the faq page?
12:15:48 <tdrobison> yes specifically for my issue yes
12:16:01 <john> OK, what's your problem?
12:16:57 <tdrobison> I have a liability (Credit Card) and I paid on it with asset (checking account)
12:17:34 <tdrobison> when I categorize the check acct asset to the liability I lose track of the expense category?
12:17:55 <tdrobison> the accounts balance but I still would like to categorize it as clothing expense
12:18:13 <tdrobison> does that make sense?
12:19:09 *** ErKa has quit IRC
12:19:14 <john> Gnucash doesn't do "categories". It does double-entry accounting. Go back and read the *whole* concepts guide. Gnucash is *very* different from other money management products.
12:19:38 <tdrobison> sure thanks
12:19:52 <gjanssens> john: your svn commit seemed to have done the git sync trick
12:20:05 <gjanssens> my local repo is now up to svn r23563
12:20:37 <john> It did indeed, and the dcommit is proceeding. I'll be able to tag the release in a few minutes. Could you fire off the Win32 build when I do?
12:23:12 <gjanssens> I still got a little time left before our guests arrive, so that should be possible
12:24:23 <gjanssens> I'm rebuilding trunk with your most recent commit
12:24:53 <gjanssens> Thanks for reverting my unwanted changes
12:25:00 <john> Beware the directory name changes in src/import-export
12:25:23 <john> You're welcome. OK, the release is tagged.
12:25:26 <gjanssens> I forgot about them in all the experiments I had to do to figure out the guile locale issues
12:26:28 <john> ;-). Reverse of what happened to me last time.
12:29:34 *** amjam has quit IRC
12:33:31 * gjanssens triggered the windows build
12:33:58 <warlord> john, gjanssens, most likely the issue is that the trigger file is in /tmp (or /var/run) and that may have got cleared as part of the reboot.
12:34:54 <john> Derek, do you mean the trigger for the build? Didn't get that far last night.
12:35:38 <warlord> No, I mean the trigger to push from svn -> git
12:36:02 <warlord> (or, rather, pull from svn into git)
12:36:30 <warlord> The windows build probably got stuck because code was unavailable..
12:36:50 <warlord> The VM was spinning; load avg of 53 once I finally got a shell.
12:37:23 <warlord> I didn't do any investigation into the cause (although I did see an "Out of memory" from "umount" in the log..
12:37:41 <john> The windows build never had a chance. I couldn't even commit the release, never mind tag it.
12:37:42 <warlord> I was still receiving emails so it was a VM guest issue, not a VM server issue.
12:38:14 <warlord> Yeah, sorry about that.. NOt sure what happened, but code's OS is definitely getting long in the tooth.
12:38:26 <warlord> (and due to the way it's partitioned I cannot upgrade it directly)
12:39:01 <john> Which is why you want to replace it over the holiday. Roger that.
12:39:20 <warlord> Right.. F20 is coming out in 2 days.. :)
12:39:25 <warlord> (code is on F15)
12:39:52 <john> F15 isn't that old, is it? 2 years or so?
12:41:29 <warlord> 2 1/2
12:41:50 <warlord> But there have been 0 updates in 1 1/2 years..
12:43:01 <warlord> @op fell john
12:43:01 <gncbot> warlord: Error: You don't have the #gnucash,op capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified.
12:43:12 <warlord> @op fell john
12:43:14 *** gncbot sets mode: +oo fell john
12:43:58 <warlord> I would've just tried to upgrade-in-place if the partitioning would've allowed it. code was last installed-from-scratch using F10, so 5 years ago.
12:44:23 <warlord> (on 2009-01-28 to be exact)
12:44:29 <warlord> ... so ALMOST 5 years ago ..
12:49:34 <gjanssens> gtg
12:49:37 *** gjanssens is now known as gjanssens_
12:49:42 <john> Heh. That's about the same time I last upgraded my server, to Debian Lenny. I'm planning a shuffle as soon as Apple gets of its arse and ships the new Mac Pro.
12:51:56 <warlord> see ya, gjanssens_
12:52:08 <warlord> john: :)
12:52:37 <warlord> Yeah, I usually do an upgrade every 2ish years if I can.
12:54:20 <john> I'm too lazy for that, but then, I'm not presenting anywhere near as much to the public.
12:54:54 <john> Like nothing. I used to run a mail server, but Comcast got mad at me for backscatter and blocked it.
12:56:13 <warlord> Yeah, it's tough to protect against backscatter.
12:56:45 <warlord> (one key is to refuse the messages during the incoming SMTP request instead of accepting them and then bouncing them back)
12:57:01 <warlord> Alas, I do issue a bunch of crap when spammers sign up for the wiki.
12:59:35 <john> I know. I even knew then. I'd just upgraded spam assassin and it had some new settings that screwed up my config. By the time I figured out the problem, Comcast was already peeved.
13:01:13 <warlord> I wonder -- was there a leap second last night?
13:01:30 <warlord> I only ask because it looks like stuff failed around midnight.
13:03:16 <john> Leap seconds happen at 0000Z. My commit that didn't sync went in at 00:47R.
13:04:34 <warlord> Clearly something happened around 00:00 EST..
13:05:21 <warlord> It's also possible that it was an upstream network issue, although that wouldn't explain the 53 load avg
13:06:29 <john> Like I said in my email, ping times were around 140ms with no packet loss. I don't think it was a network problem. Code was just so slow that everything was timing out.
13:07:06 <john> Even gncbot got logged off because of timeouts. Twice.
13:07:43 <warlord> Yeah, I noticed that... Which of course Iwouldn
13:07:49 <warlord> 't have if I was having network issues...
13:08:10 <warlord> (I just looked at mrtg and my tunnel traffic was flat from 00:00 -> 10:00
13:09:04 <john> Tunnel traffic?
13:09:33 <warlord> The fact that my VM server is 5 years old is also going to be an issue in the not-too-distant future. I've already replaced all the disks at least once.
13:09:51 <warlord> Yeah, I have a tunnel to serve my class-C network (where code resides)
13:10:46 <john> What kind of router do you use?
13:11:39 <warlord> Right now it's a linksys runnig dd-wrt
13:13:14 <john> That can handle a tunnel? I'm assuming you mean an ssh tunnel, and that the router is doing the auth.
13:14:01 <warlord> No, it's a GRE tunnel
13:14:39 <john> Huh. That's a new one for me.
13:15:34 *** tdrobison has left #gnucash
13:19:00 <warlord> It is like an IP-in-IP tunnel.. sort of. Basically I am tunneling ipv4 and ipv6 in the gre tunnel; the upstream is a shared router in Boston that advertises my public networks.
13:19:43 <john> Yeah, I looked it up on wikipedia. ;-)
13:21:02 <warlord> :)
13:21:23 <warlord> Converting from GRE to IPsec would be nice, but I dont think the upstream router supports it.
13:25:40 *** jmd has joined #gnucash
14:17:20 <john> warlord: Derek, it looks like the Win32 nightlies have been failing since Monday because of the error "failed to load external entity "gnucash-guide.xml" in the German translation. It looks OK to me in Git, and it built fine on the Mac just now. Can you have a look and see if there's something wrong with the Win32 checkout?
14:25:38 <john> Oh, I think I see the problem. Frank renamed the language subdirectories the other day. de_DE is now de. Just before the error it says "Processing guide (de_DE); it should say "Processing guide (de)". The nightly needs to autogen.
14:27:38 <john> Nope, I'm wrong, it's hardcoded into the stupid win32 packaging commands.
14:31:26 <john> Fixed, r23570
14:34:56 *** MrLurker has quit IRC
14:35:10 <john> And retagged. Would you please kill the running build and restart it?
14:43:15 *** aqua___ has quit IRC
14:55:03 *** TradeBorG111 has quit IRC
14:58:22 *** GNUCashUserBrazil has joined #gnucash
15:31:06 <fell> Hello GNUCashUserBrazil, we need an update of our brazilian translation.
15:33:45 <fell> If you are interested, see https://bugzilla.gnome.org/show_bug.cgi?id=646585 and http://wiki.gnucash.org/wiki/Translation
16:05:38 *** aqua___ has joined #gnucash
16:11:02 <GNUCashUserBrazil> Yes, I'm interested.
16:11:39 <fell> Fine!
16:13:00 <GNUCashUserBrazil> I have naver done any action in open source projects. I don't know the tools used, but I'm interested in helping the comunity
16:14:49 <fell> wiki:Translation has a section Tools
16:15:29 <GNUCashUserBrazil> I would like to know more about creating and modifying the reports and graphics
16:16:15 <GNUCashUserBrazil> Got it. I'm studying the translation wiki
16:21:01 *** StuM has joined #gnucash
16:29:57 *** nafg has quit IRC
16:50:45 *** nafg has joined #gnucash
17:14:38 <GNUCashUserBrazil> C you soon
17:14:47 *** GNUCashUserBrazil has left #gnucash
17:22:17 *** jmd has quit IRC
17:43:06 *** aqua___ has quit IRC
17:57:29 *** gjanssens_ has quit IRC
18:03:48 *** fell has quit IRC
18:11:50 *** PaulFertser has quit IRC
18:21:14 *** puck has quit IRC
18:22:39 *** puck has joined #gnucash
18:28:10 *** kpreid has quit IRC
18:28:37 *** kpreid has joined #gnucash
18:47:41 *** amjam has joined #gnucash
19:00:16 *** puck has quit IRC
19:01:50 *** puck has joined #gnucash
19:35:18 <warlord> john: I just kicked the win32 builder (after killing the previously in-process build)
19:37:24 <john> Derek, thanks. Any idea what was hanging up the previous build? Geert fired it off 7 hours ago!
19:38:12 <warlord> john: it wasn't hung -- it was still running.
19:38:37 <warlord> tag builds take a very long time because they rebuild all the dependencies.
19:38:41 <john> Wow. That machine is *slow*.
19:38:47 <warlord> (whereas the nightly builds do not)
19:38:58 <warlord> It's not just the machine, it's mingw/windows..
19:39:22 <warlord> ... and the fact that tag builds rebuild all the dependencies as if it's a fresh build from scratch.
19:39:45 <warlord> (whereas, as I said, the nightly build doesn't do that, it reuses the deps already built so builds much quicker)
19:39:55 <warlord> I think it was on the 'gnucash' build of the tag build when I killed it.
19:40:20 <john> Yeah, but a Win7 VM on a Mac mini can build from scratch in a couple of hours, while another build is going on either on another VM or in OSX.
19:40:29 *** amjam has quit IRC
19:40:42 <warlord> How much RAM and how many CPUs assigned to the VM?
19:41:03 <john> 4G and 2 CPUs.
19:41:24 <john> Well, cores. There's only one i7 CPU.
19:41:25 <warlord> Right, 1G and 1 CPU here.
19:41:57 <john> Memory is cheap. ;-)
19:41:58 <warlord> (ditto, cores... I've got 2 AMD quad-core CPUs in the VM host...2.3GHz)
19:42:18 <warlord> Not really.. DDR2 is actually quite expensive right now.
19:42:33 <john> That's 'cause it's so *old*.
19:42:51 <john> Won't DDR3 work in the box?
19:43:09 <warlord> Unlikely, because it's so old! :)
19:43:39 <warlord> It's a Supermico H8DA3-2
19:44:22 <john> That doesn't mean much to me, I'm afraid.
19:45:23 <warlord> No, but you could google it to figure out if it would take DDR3 ram ;)
19:48:10 <john> Nope. From Wikipedia: "DDR3 SDRAM is neither forward nor backward compatible with any earlier type of random access memory (RAM) due to different signaling voltages, timings, and other factors."
19:50:25 <warlord> It currently has 8 2GB sticks, using up all the slots..
19:51:04 <john> And doing a lot of work besides Gnucash's windows builds, I'm sure.
19:51:59 <warlord> right, it's the vm server also hosting code and a bunch of other servers, too.
19:55:09 <warlord> I could double the ram via 4 8G (4Gx2) kits @ $390 each, so $1560. Or I can max out the RAM at 64G via 4 16G (8Gx2) kits @ $740ea, so $2960. But at those prices it's almost worth just buying a new system.
19:56:33 <john> Yeah. The H8DCL series looks nice. Bet they're expensive, too, even with DDR3 instead of DDR2.
19:57:50 <warlord> a single 32GB ECC DDR3 stick is ~$600
20:00:08 <warlord> Looks like ~$300-$360 for the mobo, but I don't think I can re-use my existing CPUs. (AMD Opteron 2378s)
20:00:17 <john> So you could max out a H8DCL for only $4800. ;-)
20:01:44 <warlord> the IF-O or 6F-O model?
20:02:02 <warlord> (I guess both have 8 slots..
20:03:09 <warlord> 8x32G would max out the mobo, so I guess that's good. Then again I don't have $6k to spend on a new system right now.
20:03:23 <warlord> (or possibly more?)
20:04:01 <warlord> Honestly, if DDR2 ram were less expensive I'd seriously consider the upgrade, but even $1560 for the upgrade feels like I'm over-paying.
20:04:06 <john> I was going to say it would probably be more once you got the "optimized chassis", the cpus, the disk drives...
20:04:48 <warlord> Right..
20:04:53 <john> Well, yeah, at 2.5x the price of DDR3 for less performance.
20:05:10 <john> It's all about volume.
20:05:14 <warlord> Right.
20:11:34 <warlord> Hmm, I wonder if this would work: http://www.ebay.com/itm/Server-RAM-8GB-2-4GB-PC2-5300P-ECC-Registered-DDR2-667MHz-PC5300-2Rx4-Memory-LOT-/291034212574?pt=US_Memory_RAM_&hash=item43c2fda8de#viTabs_0
20:12:24 <warlord> 8GB (2@4G) for $38, so $152 for 4 sets, 32GB. *THAT* would be worth it, assuming these modules would work.
20:13:09 <john> And that you get them for $38.
20:13:20 <john> But he's only got 5 sticks.
20:13:40 <john> No, i misread it, he had 5 pairs.
20:13:41 <warlord> No, he has 5 sets of 2 sticks.
20:13:46 <warlord> Right, I need 4 pairs.
20:24:21 <warlord> Looks like that memory *would* work.. I wonder if I could get approval from -core to spend the $150+ on the upgrade? ;)
20:24:42 <warlord> (then I could easily give 4G + 2CPUs to the win32 build machine)
20:25:35 <john> I'd think so... but if Gnucash pays for the memory, we should get all 16G assigned to the gnucash VMs. ;-)
20:25:43 <warlord> :-P
20:29:51 <warlord> oooh, and it's free shipping, too!
20:30:17 <john> I sense a sniper-bot being fired up...
20:30:38 <warlord> ??? Nah, I'm a 'buy-it-now' guy, especially for that price.
20:31:42 <warlord> Although I do wonder if I should get the 5th pair, *just in case*
20:31:49 <john> Besides, he's not even running an auction.
20:32:51 <warlord> Yeah...
20:32:58 <john> Well, he's offering a 30 day warranty.
20:33:23 <warlord> True..
20:37:05 <warlord> I'm thinking I should bite the bullet and just get it. It would give the VM server a few more years of life, and it's a great price.. Let me see if I can find any 8GB sticks.
20:39:29 <john> You're getting greedy...
20:39:41 <warlord> Hey, it's only a few searches..
20:40:00 <warlord> And I'd rather only spend the upgrade money once.. ;)
20:40:54 <warlord> But I'm not seeing any.
20:42:46 <warlord> I should probably just get them now so I have them on the way...
20:51:14 <warlord> Should be here Thursday..
20:53:23 <john> Christmas comes early!
20:53:41 <warlord> They I have the joy of taking the server down to update the RAM.. Whee!
20:54:05 <warlord> s/They/Then
20:54:27 <warlord> I'll send out email to schedule it for this weekend (21st/22nd)
20:55:54 <john> That's good. No releases!
20:58:49 <warlord> LOL
20:59:05 <warlord> What's the schedule for the next release?
20:59:48 <warlord> (so I can think about timing for system downtime and server migration)
21:03:28 <john> The 29th, which will be 2.6.0. I'd expect 2.6.1 about a month later, but it depends on what bugs come in, how bad they are, and how long it takes to fix them.
21:04:54 <warlord> Okay..so I'll need to do the migration either before or after that weekend.
21:05:10 <warlord> We'll see how much time I actually get to work on it.
21:05:58 <john> We'll want to turn off SVN the 1st week of Jan. After that we can enable pushing to Github if Code is going to be down for a while, and we'd only need the Windows build server for a release. That's an emergency fallback.
21:06:08 <john> Dinner time. TTYL.
21:06:18 *** john is now known as john_afk
21:07:19 <warlord> I would hope we wouldn't need lots of downtime. Most of the downtime would be the actual data migration for the services we're going to keep..
21:08:04 <warlord> ttyl.
21:09:50 *** amjam has joined #gnucash
21:43:19 *** ErKa has joined #gnucash
22:37:09 *** PaulFertser has joined #gnucash
23:47:14 *** Jimraehl has joined #gnucash
23:52:18 *** Jimraehl has quit IRC