2009-09-02 GnuCash IRC logs
00:22:19 *** RobotCow has joined #gnucash
01:04:20 *** dwm_lt has quit IRC
01:26:19 *** roconnor has quit IRC
01:49:33 *** warlord has joined #gnucash
01:49:33 *** gncbot sets mode: +o warlord
02:08:50 *** kling0n has joined #gnucash
02:09:52 <dbreiser> jsled: your suggestion is quieting the warnings, but I'm up to 17 files that need tweaking. I picked an abominably slow way to grind through the tree. I'll submit a patch soon.
02:10:18 <dbreiser> s/soon/tomorrow.
02:10:23 *** dbreiser has quit IRC
02:12:54 *** warlord is now known as warlord-afk
02:22:00 *** dwm_lt has joined #gnucash
02:30:00 *** XiXaQ has quit IRC
02:39:28 *** XiXaQ has joined #gnucash
02:49:56 *** kling0n has quit IRC
03:13:23 *** dwm_lt has quit IRC
03:25:34 *** XiXaQ has quit IRC
05:34:51 *** bentob0x has joined #gnucash
06:29:47 *** bentob0x has quit IRC
07:31:25 *** andyt has joined #gnucash
07:33:17 *** RobotCow has quit IRC
07:39:00 *** Jimraehl has left #gnucash
07:50:24 *** JimRaehl2 has joined #gnucash
07:52:30 *** JimRaehl2 has left #gnucash
07:52:49 *** JimRaehl2 has joined #gnucash
08:39:20 *** XiXaQ has joined #gnucash
09:03:51 *** ErKa has joined #gnucash
09:44:31 *** twunder has joined #gnucash
09:53:30 *** dbreiser has joined #gnucash
09:54:00 <dbreiser> http://pastebin.ca/1551506
09:54:14 <dbreiser> is there a simple cast fix for those complaints?
10:03:14 *** mark1 has joined #gnucash
10:08:25 *** epilot88 has joined #gnucash
10:09:17 <jsled> epilot88: welcome.
10:09:37 <epilot88> Hi
10:09:37 *** ErKa has quit IRC
10:09:45 *** warlord-afk is now known as warlord
10:10:16 <warlord> dbreiser: It's complaining because the callback function prototype doesn't match the function being passed back..
10:10:37 *** ErKa has joined #gnucash
10:11:41 <warlord> What's the function prototype for e.g. GWEN_Gui_SetProgressStartFn()? In particular, what's the prototype of the callback function argument? And does progress_start_cb() match that prototype?
10:15:27 <epilot88> My problem: My monthly scheduled transactions look ok, but when I set up my quarterly transaction (property taxes) the 12 month calender starts with November. I can no longer see September. When I then add a new September scheduled trans I can see September, but it makes my November trans appear on the September calender, and all remaining trans for "property taxes" are out of sync.
10:23:20 <dbreiser> warlord: thanks for the response. I can't figure out the answers quickly, and I have to go to the office. be back tonight.
10:23:30 *** dbreiser has quit IRC
10:31:12 *** ErKa has quit IRC
10:36:14 *** ErKa has joined #gnucash
10:39:29 *** nimrod has joined #gnucash
10:39:33 *** nimrod has left #gnucash
10:41:53 *** XiXaQ has quit IRC
11:01:58 <jsled> epilot88: out of sync on the calendar, or when you go to create the transactions?
11:02:18 <epilot88> On the calendar.
11:02:19 <jsled> I believe there are some calendar-display issues, but hopefully the actual creation will be correct.
11:02:53 <jsled> You might try setting it to "create in advance" to something large like 365 days, then run the Since Last Run dialog to see when it actually would attempt to schedule them.
11:03:04 <jsled> (you can cancel so they're not actually created)
11:04:33 <epilot88> I'll try it now..thanks.
11:13:23 <epilot88> Yep, must be a calendar issue. Created a new one called "test2". It looked fine until I created a "test3" and set it for this month, it then moved "test2" to Sept 1 in calendar view. It also changed how it looks in "since last run view", however in transactions view it is listed with a next occurrence of 11-1-09 which is correct.
11:14:43 <warlord> Sounds like a cal display issue.
11:15:07 <jsled> Yeah, I know I saw some weirdness, and I believe there's even a bug filed for it.
11:15:09 <jsled> Sorry about that.
11:18:13 <epilot88> Please forgive me, but I just started using this program yesterday. Issue #2. I set up a "test1" scheduled transaction yesterday to start running today. When I turned on the computer today there was no pop-up to alert me. Should there have been?
11:18:41 <epilot88> Btw, thanks Jsled and Warlord.
11:19:51 <jsled> epilot88: so long as the scheduled transactions are set to "run on startup", then yes, when you start gnucash, the "Since Last Run" dialog should appear (if there are new scheduled transactions to create)
11:19:59 <jsled> That setting is in the Preferences.
11:20:08 <jsled> I believe it's enabled by default, but I could be wrong.
11:20:32 <jsled> And, just to be clear, it's when you start *gnucash*, not when you start the computer.
11:21:09 <epilot88> Ok, thanks for all the help.
11:22:24 *** epilot88 has left #gnucash
11:32:30 *** warlord has quit IRC
11:54:01 *** mib_y9641xzy has joined #gnucash
11:54:14 <mib_y9641xzy> hi
12:00:42 *** twunder-konv has joined #gnucash
12:03:58 *** warlord has joined #gnucash
12:03:58 *** gncbot sets mode: +o warlord
12:05:36 *** twunder has quit IRC
12:08:32 *** mib_y9641xzy has quit IRC
12:14:12 *** ErKa has quit IRC
12:15:24 *** ErKa has joined #gnucash
12:36:36 *** bentob0x has joined #gnucash
13:23:51 *** Freak_NL has joined #gnucash
13:24:26 *** hvx has joined #gnucash
13:24:29 <Freak_NL> Hello, there is a bug in the webserver configuration for the on-line documentation. Is there someone here who maintains this?
13:24:37 <Freak_NL> (character-encoding issue)
13:25:38 <warlord> What's the issue?
13:27:08 <Freak_NL> The documentation HTML pages are written in UTF-8, and specify UTF-8 as encoding in the meta tags too, but the webserver overrides this by setting the character encoding to iso-8859-1 in the HTTP headers sent
13:27:23 <Freak_NL> this makes the documentation pages look odd
13:27:25 <warlord> LOL
13:27:38 <warlord> What URL are you going to?
13:28:20 <Freak_NL> http://www.gnucash.org/docs/v2.0/C/gnucash-guide/oview-features1.html
13:28:21 <Freak_NL> etc.
13:28:39 <Freak_NL> I just followed the links on the main page
13:29:05 <warlord> Right.
13:29:06 <Freak_NL> See what I mean?
13:29:07 <warlord> linas?
13:29:24 <warlord> No, I'm not looking... Just wanted to know which server needed to get tweaked.
13:29:42 <Freak_NL> You can set the HTTP headers character encoding in .htaccess if the server allows it
13:30:43 <Freak_NL> It's a minor issue because of the overlap between the two encodings, but I recognised it as an issue I have dealt with before
13:32:51 <Freak_NL> (AddDefaultCharset UTF-8 in the .htaccess for the docs)
13:34:52 <jsled> we don't have access to www
13:34:53 <jsled> linas: ^
13:35:56 <jsled> Hrm. But the .htaccess in that dir is in SVN.
13:40:14 <jsled> Huh. In fact, there's already an «AddType text/html;charset="UTF-8" .html» in the .htaccess there.
13:40:26 <jsled> But the server seems to be overriding it.
13:40:38 <jsled> Or not respecting the .htaccess. Or something.
13:50:31 <Freak_NL> Shall I consider this bug reported?
13:51:58 <jsled> i suppose so, yeah.
13:59:35 <Freak_NL> Good luck with the .htaccess issue
13:59:40 *** Freak_NL has quit IRC
14:31:52 <linas> ola
14:32:37 <warlord> hiya linas.
14:32:50 <warlord> There seems to be a problem on www about charsets?
14:33:10 <linas> let me look ...
14:36:59 <linas> default is is8859
14:38:43 *** hvx has quit IRC
14:45:42 <warlord> linas: right, but the data is utf8, and the .htaccess says it's utf8.
14:48:18 <linas> I changed default to utf8 but the page http://www.gnucash.org/docs/v2.0/C/gnucash-guide/oview-features1.html still shows carat A
14:48:45 <linas> but if you look at page source, you'll see the source has carat-A in it (for the foooter)
14:49:05 <linas> and for the header, its incorrectly placed in the html head instead of html body
14:49:26 <linas> so that particular page is broken.
14:50:10 <linas> btw, jsled, fyi I have a number of dead links that people click on and get 404 not found. who would want to clean that up?
14:50:41 <jsled> There is no 'Â' in the page.
14:50:54 <jsled> There's a U+00A0 NO-BREAK SPACE ( )
14:50:58 <jsled> which is being mis-encoded.
14:51:34 <jsled> Because the server is still saying Content-Type: text/html; charset=ISO-8859-1
14:51:47 <linas> hmm
15:01:50 <warlord> linas, can you compile a list of dead links?
15:02:00 <linas> yes
15:02:21 <linas> where should I send it?
15:02:24 <warlord> -deve?
15:02:26 <warlord> devel
15:02:36 <warlord> anyways, lunchtime.
15:02:42 *** warlord is now known as warlord-afk
15:11:47 <linas> gnucash.org web pages are being served without the usual http headers ... is this normal?
15:13:35 <Simon> erm, I don't think you can specify charset with AddType, only ForceType
15:18:10 *** twunder-konv has quit IRC
15:19:12 <jsled> linas: what do you mean "without"?
15:20:25 <linas> well, if I telnet into port 80 and say GET /bogus-url that I get the normal http header. But if I give a valid url, there's no header
15:20:38 <linas> try it you'll see what I mean.
15:20:50 <linas> the header contains info like page encoding, etc.
15:20:59 <jsled> the Content-Type header?
15:21:30 <jsled> Do it as "GET /bogus-url HTTP/1.1"
15:21:44 <jsled> The header *is* there; that's what I pasted.
15:21:54 <jsled> Use live http headers or firebug or wireshark or something.
15:23:56 <linas> header is there when a bogus url is specified, but not when a good url is specified
15:25:25 <jsled> ah.
15:25:27 <jsled> now it's working.
15:26:29 <jsled> As it's sending the correct Contnet-Type back now http://pastebin.ca/1551837
15:29:02 <jsled> for some reason, the server refuses to accept a *very* bare bones HTTP request.
15:29:18 <linas> try a different URL
15:29:28 <linas> since I did a touch of that file.
15:29:29 <jsled> In particular, it requires a pretty valid Accept: header.
15:29:36 *** cortana has joined #gnucash
15:30:00 <jsled> http://www.gnucash.org/docs/v2.0/C/gnucash-guide/oview-about1.html looks good too, after a force-refresh.
15:30:02 <linas> one thing that was happening was that apache was returning 304 not modified for that page
15:30:08 <jsled> right.
15:30:20 <jsled> Thoguh I was careful to do a force-refresh before, too.
15:30:20 <linas> and so when I did the touch, it resent the page (and thus the new encoding
15:33:23 <linas> ah whatever. The missing headers may be due to my sloppy use of telnet
15:34:20 <jsled> linas: if you don't specify HTTP/1.0 or HTTP/1.1 on the GET, you won't get response headers back, it appears.
15:34:30 <linas> yeah.
15:35:00 <linas> but when I did specify the HTTP/1.1 by hand, I kept getting an erorr page about malformed request.
15:35:27 <linas> maybe telnet wasn't sending crnl which is what's expected
15:35:33 <jsled> right … that's what I was saying about the Accept header
15:36:26 <jsled> I dunno why, but your server is picky about that one.
15:37:00 <jsled> In a way I haven't seen before. Mabye I don't do enough HTTP by hand against httpd, though. :)
15:37:29 <jsled> (though I've done a fair bit of non-browser HTTP spidering, and don't recall Accept being so critical.
15:39:52 <linas> hmpf. well, no matter what I can't get a telnet-based fretch to work
15:39:55 <linas> ohhh!
15:39:59 <linas> wait ...
15:41:36 <linas> the Host: www.gnucash.org is critical, because apache is configured to be a virtual host
15:41:40 <jsled> Host: is probably … right,.
15:42:10 <linas> so none of those urls exist for the default host ...
15:42:17 <linas> I wonder why I configured it that way ...
15:42:48 <linas> yeah, I guess I uses to have other servers hosted there once upon a time ...
15:43:36 <linas> anyway ... all problems solved now, right?
15:43:44 <linas> oh, and the bad url list
15:49:36 *** cortana has quit IRC
15:55:16 *** ErKa has quit IRC
16:01:17 *** cortana has joined #gnucash
16:02:51 *** ErKa has joined #gnucash
16:09:22 *** ScislaC has joined #gnucash
16:09:50 *** warlord-afk is now known as warlord
16:20:15 *** ErKa has quit IRC
16:48:23 *** bentob0x has quit IRC
16:48:50 *** twunder has joined #gnucash
16:58:40 *** vitoc has joined #gnucash
17:28:42 *** josh has joined #gnucash
17:36:29 *** MiladZ has joined #gnucash
17:38:09 *** MiladZ has quit IRC
17:42:13 <josh> when i try to 'get quotes' from the price editor, i get an error message saying 'Unable to retrieve quotes for these items: CURRENCY:CRC'
17:42:55 <josh> i also tried some other currencies with the same results
17:44:56 <warlord> josh: you need to patch F::Q
17:45:00 <warlord> see the user mailing list.
17:46:54 *** twunder has quit IRC
17:56:12 *** andyt has quit IRC
18:25:22 *** motin_0 has joined #gnucash
18:52:40 *** twunder has joined #gnucash
18:56:14 *** twunder has quit IRC
19:06:48 *** vitoc has quit IRC
19:12:53 *** twunder has joined #gnucash
19:13:06 *** ScislaC has left #gnucash
19:29:51 *** cortana has quit IRC
21:02:45 *** dbreiser has joined #gnucash
21:04:24 <dbreiser> does the information in the comment to http://pastebin.ca/1551506 indicate the function is expecting a uint32_t and is getting a guint32 from progress_start_cb()?
21:05:24 <dbreiser> obviously I'm over my head programming-wise...
21:18:36 <warlord> Yes, that's the probably the complaint, guint32 v. uint32_t, guint64 v uint64_t
21:19:17 <dbreiser> can one cast it directly (if that's even a coherent question)? how?
21:19:24 <warlord> Or... more likely, it might be the char v. gchar
21:19:49 <warlord> Yes, you could just cast it to (GWEN_GUI_PROGRESS_START_FN)
21:20:01 <warlord> But then if the API changes the compiler wont catch it.
21:20:25 <warlord> First, try changing the gchar to char and see if that fixes it.
21:20:49 <dbreiser> ok
21:21:40 <dbreiser> in progress_start_cb? both of them?
21:21:49 *** Michael1 has joined #gnucash
21:22:21 <warlord> Well, let's try one first to see if it helps..
21:23:26 <dbreiser> ah. a cautious man. (Whereas I, the clueless one, seem ready to make multiple changes all at once...)
21:25:47 <warlord> heh. well, no point in making changes everywhere blindly. Let's see if it works, then if it does let's try the other. Then commit :)
21:27:05 <dbreiser> oh. I meant both of the gchars in progress_start_cb(), not both functions
21:30:09 <dbreiser> going to do this from a fresh checkout of 2.2.9. could take a bit while I fiddle
21:30:39 <warlord> Yes, you need to change both the gchars
21:30:48 <dbreiser> thanks
21:36:15 *** JimRaehl2 has left #gnucash
21:36:32 *** Jimraehl has joined #gnucash
21:38:56 <warlord> Did that help?
21:41:39 <dbreiser> haven't got that far. I've been sequentially patching the format spec problems and running full packaging attempts. For this I'm going to a sane environment and building incrementally, but I'm not fully set up for that yet
21:42:04 <warlord> oh
21:42:06 <warlord> ok
21:43:19 <dbreiser> for as little disruption as Snow Leopard had on the standard mac realm, it's causing some heartburn in the open source realm...
21:44:44 <warlord> Heh
22:04:30 *** warlord has quit IRC
22:09:04 *** warlord has joined #gnucash
22:09:04 *** gncbot sets mode: +o warlord
22:09:25 <dbreiser> warlord: no the gchar -> char did not help
22:09:34 <warlord> :(
22:09:53 <warlord> Okay, what about changing the guintXX to uintXX_t as well?
22:10:13 <dbreiser> will do.
22:12:51 <dbreiser> hmm. broke it worse.
22:14:00 <dbreiser> http://pastebin.ca/1552240
22:14:26 <dbreiser> though by now, it's probably my misunderstanding some piece of this operation
22:20:11 <warlord> Well, sorta.. The callback function is declared in one place and then implemented in another. You changed one place (probably the declaration) but not the other.
22:20:36 <warlord> So you need to make sure that the declaration at line 120 matches the implementation at line 1027
22:20:49 <warlord> But clearly changing the functions made the callback setting work.
22:21:25 <dbreiser> and the stuff beyond 1027 also needs tweaking: gnc-gwen-gui.c:1035: warning: format ‘%lu’ expects type ‘long unsigned int’, but argument 9 has type ‘uint64_t’
22:21:53 <dbreiser> after I changed the stuff at 1027 to match 120.
22:22:14 <warlord> Yeah, for something like that you can change it from lu to llu and cast it to (unsigned long long) for the printf
22:24:34 <dbreiser> the %lu seems to be coming from a G_GUINT64_FORMAT. what's the right text to get llu instead?
22:25:49 <warlord> Oh... Then cast it to a guint64
22:26:01 <warlord> You must be on a 64-bit machine.
22:26:34 <dbreiser> yep. the default for snow leopard. reportedly fewer problems than trying to force it to 32
22:26:44 <warlord> Right.
22:27:00 <warlord> So cast the uint64_t to guint64. I'm surprised they are not equivalent types.
22:27:22 <dbreiser> so, being totally in the dark, what's the syntax for the cast?
22:27:30 <warlord> (guint64)
22:29:29 <dbreiser> in the prototype or in the implementation?
22:29:39 <warlord> in the printf
22:29:44 *** motin_0 has quit IRC
22:29:59 <warlord> so.. printf(".....", (guint64)intArg)
22:30:44 <dbreiser> That statement is: ENTER("gui=%p, flags=%d, title=%s, total=%" G_GUINT64_FORMAT, gui, progressFlags, title ? title : "(null)", total);
22:31:59 <warlord> so (guint64)total
22:32:36 <dbreiser> oh. I was off in outer space...
22:33:19 <dbreiser> that doesn't look too portable. am I missing something else?
22:33:35 <warlord> What do you mean?
22:33:52 <warlord> That's perfectly portable, if total was a uint64_t
22:34:42 <dbreiser> OK. And that solved the first incompatible pointer type. I can probably handle the second. Let's see...
22:35:49 <warlord> Anyways, I need to get outta here and get dinner and head back to my hotel room.
22:36:04 <dbreiser> thanks for the help and patience
22:38:34 <warlord> you're welcome.
22:38:39 <warlord> I'll be back online in an hour or two.
22:39:30 *** warlord has quit IRC
23:21:22 *** dbreiser has quit IRC
23:44:38 *** dbreiser has joined #gnucash
23:45:34 <dbreiser> grumble, grumble. gnucash 2.2.9 built on OS X 10.6, but it won't run because dlopen is trying to find .so files instead of .dylib files.