=== marvelous is now known as wonderful === rickspencer3 is now known as rickspencer3-afk === wonderful is now known as vorian [11:07] MacSlow: hiya [11:08] When a defaults value is out of bounds, should i : 1) return without changing anything | 2) set to default | 3) set to nearest bound ? [11:08] (4) segfault :P) [11:52] SiDi, no crashers please :) [11:53] SiDi, just spit out a waring on stdout (e.g. g_warning()) and don't change anything [11:55] SiDi, it's debatable if you set it to the lowest possbile value if it was attempted be set lower (or the hightest possible value if it was attempted to be set higher) [11:56] SiDi, not changing anything and just spitting out a warning on stdout is the safest path [11:58] MacSlow: okey === beuno_ is now known as beuno [13:26] * MacSlow -> lunch === MacSlow is now known as MacSlow|lunch [13:48] Purely rethorical question : if i know the data given to a function is ALWAYS good (i test that it'll be good before using it), shall i remove error checking code after i called this function ? === rickspencer3-afk is now known as rickspencer3 === MacSlow|lunch is now known as MacSlow [15:52] djsiegel1: djsiegel3 https://bugs.launchpad.net/notify-osd/+bug/340996/comments/13 [15:52] Launchpad bug 340996 in hundredpapercuts "Notifications should stay on screen longer when there is a lot of text" [Undecided,Confirmed] [15:53] mac_v yes I saw [15:55] djsiegel3, mac_v: I'm pedantic there... but it's a fact... not really a bug fulfilling the papercut-requirements as the involved work is considerable. [15:56] MacSlow: really? [15:56] djsiegel3: next cycle ,how about minimizing the number of papercuts and expanding the scope? [15:56] it's not just length (message) * MILLISECONDS_PER_CHAR ? [15:56] mac_v, what do you mean? [15:56] expanding the scopr [15:56] you mean, changing the definition of a paper cut to include more things? [15:57] djsiegel1: like reducing number <100, and taking on bigger bugs? [15:57] like, why don't we just say paper cut = any bug and fix 20 of them? :) [15:57] djsiegel3: not any bug , but bug with usability issues [15:58] djsiegel3, started to work on it (preparing notify-osd with some refactoring) but some priority-shifts have me currently working on another thing in notify-osd [15:58] MacSlow, I'm curious because I wrote a notifications API in GNOME Do for plugins and did time calculations like that based on message length [15:58] it was very simple [15:58] longer messages stayed on screen longer [15:59] djsiegel3: some usability issues are not pushed since they are a little bigger to fix , like for example the click to rename [15:59] djsiegel3 you're formatting, layouting and special cases were not as complex as in the notify-osd spec I imagine [15:59] certainly [16:00] mac_v, that should be a separate project maybe/ [16:00] maybe we could do 50 paper cuts, and 10 usability bugs [16:00] djsiegel3: yeah, that would be good [16:01] djsiegel3: i fear we'll run out of *minor* things to fix in 2 cycles ;) [16:01] mac_v we can always make more buggy software [16:01] lol === beuno is now known as beuno-lunch [16:59] djsiegel3: its true you need to take into account updates + appends + priority of bubbles in the stack [16:59] + summer holidays :P [17:00] SiDi: and leap years [17:00] and angriness of kittens in your direct environment. [17:00] ..no ? === beuno-lunch is now known as beuno [17:58] MacSlow: any g_warning makes a test fail, right ? [17:58] depends on G_DEBUG [17:58] afaik [17:59] i'm running into a very stupid situation [17:59] use g_print() [17:59] i'm trying to set font size to a negative value, my setter of course doesnt modify the value, but it g_warns about the out-of-bound [18:01] SiDi, try G_DEBUG=fatal_criticals ./test-modules -p /defaults [18:01] that should not make g_warning() abort the test [18:13] apparently not [18:13] hm... [18:13] stick to g_print then === rickspencer3 is now known as rickspencer3-afk === rickspencer3-afk is now known as rickspencer3