IRC logs for #openrisc Friday, 2013-04-05

--- Log opened Fri Apr 05 00:00:01 2013
stekernhah, what a silly bug, previously we have calculated the value for r9 in ctrl stage on l.jal(r).02:53
stekernappearantly that has been fine until now, but now I hit a short function where the return had time to come before the value of r9 had reached wb02:55
stekernnow there's a test that tests for that at least03:10
olofk7pm bst, is that 6pm or 8pm GMT?05:40
olofkhmm.. is there a software client needed for google hangout? Do I need a google+ account to use it? Does it work on linux?06:00
stekernhttp://www.timeanddate.com/library/abbreviations/timezones/eu/bst.html06:45
stekernthat says it's 7:45 bst now06:45
olofkI think I got it right on the Doodle page. There was no option to set the timezone to BST06:46
stekerndoes the doodle page show the time in local time?06:47
stekernbecause then, you wanted it 7pm bst, you've got it 1 hour wrong06:50
stekern*if you wanted it06:50
olofkCrap06:57
olofkShould I add or subtract one hour?06:57
stekernadd06:57
stekernjust think that *you* are always 1 hour ahead of london06:58
olofkI thought I set it to 18:00 GMT06:58
stekernthat GMT obviously follows summertime07:00
stekerni.e. BST=GMT on doodle07:00
stekernbecause it says I'm GMT+2 on the page07:00
olofkI though that GMT was one of the precious few things I could trust in this world :(07:01
stekernwhen I'm really GMT+307:01
olofkDoes hangout have any scheduling capabilities? I'm getting a bit annoyed about doodle07:03
stekernsummertime is just annoying...07:04
stekernI have never used hangout07:04
stekernin this microsoft loving environment, all we have is lync07:04
olofksame here07:05
jonibohttps://plus.google.com/events/ccdnbfrfnh4jdpa6558pvvk8lck08:45
joniboI made an attempt to set up this event on google plus08:45
joniboinvitees can invite others... please do so08:45
joniboanyone should be able to "view" the event...08:45
joniboI think only invitees can actively participate08:45
joniboso make sure everyone who needs to be invited gets invited08:45
joniboit's a bit tricky to know who's who via GPlus, so if you want to be invited, join the openrisc community and we'll find you there... and let someone know if you haven't been invited08:48
jonibohope this works... haven't tried this before either08:48
olofkWow! The bubbles move in the champagen glasses. I like google+ already08:48
jonibojust participated in someone else's conf call and it works _really well_08:48
olofkjonibo: Do I need some client, or fancy browser features to participate?08:49
joniboyes, you need the GTalk plugin08:49
jonibojust try to start a hangout for yourself and it will help you install everything08:49
joniboi've tested it on Linux 12.04, 12.10, and 13.04 and it works well08:50
olofkThat's a good idea. I found a gentoo package, so I'll try it out this weekend08:50
jonibooops... I messed up that event... it didn't get a hangout09:00
jonibowill try again09:00
stekernthat new one doesn't seem to be "on air", only private09:03
joniboyeah, this is tricky09:03
jonibodoesn't seem to be possible to do "hangout" plus "onair"09:04
stekernheh, your efforts are appreciated09:04
jonibobut we can make the "hangout" on air later09:04
stekernok, that sounds good then09:04
joniboanyway, i'm heading out of town for a couple of days... let's just go with this event for now and we can fix things up next week if need be09:06
jonibohmmm... hangouts are limited to 10 people (???).09:12
joniboare we more than 10 people?09:13
jonibooh well... we can figure this out next week.09:13
@juliusbhttp://opencores.org/or1k/OpenRISC_Project_Meeting#April_14_2013.2C_Teleconference09:57
@juliusbbasic details there, will post to the list about it later09:57
olofkRegarding the agenda... do you have anything particular in mind?10:09
@juliusbjust updates from what people have been working on10:10
olofkI would be interested in a summary of what has happened since november, and what everyone is currently working on10:10
olofkAre there any issues that should be resolved? IIRC we discussed some architectural updates, and I can't remember if everything was resolved, or if we just got too tired10:13
@juliusbput up some items on the wiki10:13
@juliusbno some items weren't sorted 10:13
@juliusblike instruction classes10:13
olofkThat's right. Let's put it on the agenda for now, and we can remove it if we have more interesting stuff to talk about10:14
@juliusbi just read your comment about the bubbles in the champagne glasses10:24
@juliusbthat is cool!10:24
@juliusbI may just have to be drinking champagne during this conference call now...10:24
olofkThat would be great. If google nose works with hangout, I could probably smell your champagne10:31
@juliusbgoogle nose all10:42
@juliusbso I'm getting a new laptop (after using my various work laptops for my open source hackery for years)10:42
@juliusbwhat OS should I install? I'm thinking just basic Gnome-based Ubuntu 10:42
@juliusbapparently there's some way of getting the new gnome ontop of Ubuntu10:43
@juliusboh ok it looks like they're (the Ubuntu folks) are just going to do a GNOME-based Ubuntu release now, officially10:44
@juliusbsounds like what I want!10:44
stekernIMO, xubuntu is the nicest ubuntu flavour10:45
stekernthat is, ubuntu with xfce10:46
olofkI've been using xfce (both xubuntu and on gentoo) for some time now and like it a lot10:49
olofkOnly thing is that I just got a new laptop with touch screen, and gtk2 isn't very multitouch friendly10:49
olofkIs g++ supposed to inline functions declared with static inline when I use -O1 ?10:52
@juliusbhmm, OK will check it out maybe. Why xfce over GNOME, though? Just less crap?10:58
stekernolofk: you can print what optimisation passes are used with: g++ -O1 -fverbose-asm -S empty.c -o -10:59
olofkAH.. good to know11:01
olofkjuliusb: Less crap would be my main reason11:01
stekernjuliusb: yeah, less crap11:01
stekernI have never liked gnome11:01
olofkgnome is for gardens!11:02
stekernfluxbox ftw! ;)11:02
olofkYeah, I'm a big fluxbox fan too11:12
stekernmy desktop 10 years ago: http://oompa.chokladfabriken.org/skarmskjut/harvaretransparent.jpg11:14
olofkAhh... fake transparent terminal, xmms and conky11:21
olofkBrings back memories11:21
olofkI assume there is a reason why you didn't put you11:21
olofkr terminal over your fluxbox menu.11:22
stekernheh, yeah, I guess so too ;)11:22
stekernit's gkrellm, not conky, btw11:27
_franck_I'm not a big fan of ubuntu and friends, you always have tons of update until it breaks everything....11:49
_franck_centos or scientific linux is what you want :)11:49
_franck_it's based on RHEL11:49
olofkI usually run Jesux. It makes me feel better about myself. http://pudge.net/jesux12:00
stekernI haven't actually had much trouble with the ubuntu updates, up until the last one12:42
stekernI for some reason like apt-get better than rpm12:43
stekernbut the last time I tried redhat was around 95, so I guess my experience with that is somewhat outdated12:43
stekernjuliusb: 2 cosmetic questions about mor1kx:13:44
stekern1) I have called some signals in execute stage 'exec_' because 'execute_' is so long, I reckon it's bad that I have deviated from the norm in that regard, so we should probably rename stuff consistantly. do you have a preference of one over the other?13:46
stekern2) I have used /* */ block comments in some places and // block comments in some, those should probably be consistent too, but what do you prefer?13:47
@juliusbYes, true, probably should be consistent. In general I think longer names makes things easier to read, but I've probably contravened that myself.13:57
@juliusbI prefer fairly descriptive names for signals, as you might have seen13:58
@juliusbI reckon it makes life easier13:58
@juliusb2) dunno13:58
stekernyeah, I agree on the descriptive names, but it can go overboard (not that I think you have): http://thedailywtf.com/Articles/CodeThatDocumentsItselfSoWellItDoesNotNeedComments.aspx14:01
stekern;)14:01
stekernbu I'm fine with execute_, not sure where I started on that path, probably at a line that annoyingly just went over 80 columns =P14:03
stekern+t14:03
olofkGreat article. I get a lot ofhttp://thedailywtf.com/Articles/CodeThatDocumentsItselfSoWellItDoesNotNeedComments.aspx14:05
olofk14:01 < stekern> ;)14:05
olofk14:03 < stekern> bu I'm fine with execute_, not sure where I started on that path, probably at a line that annoyingly just went over 80  columns =P14:05
olofk14:03 < stekern> +t [14:05] [olofk(+i)] [3:freenode/#openrisc(+cnt)] [Act: 2]14:05
olofkStupid paste14:06
olofkSorry about that14:06
stekernon 2, perhaps // should be preferred, because those can be nested into a /* */, which is useful if you quickly want to comment out a large block of code temporarily14:13
stekernI'm not going to do a big bulky change of those anyways, I'll do it gradually while changing something related14:16
stekernjuliusb: you're just not keeping me on a short enough leash, letting me get away with stuff like that =)14:21
@juliusbstekern: I agree, keeping to // is nice14:44
@juliusbno tight leash for you stekern 15:05
--- Log closed Sat Apr 06 00:00:03 2013

Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!