# 00:01:17 |
codyps |
quits : Ping timeout: 252 seconds |
# 01:48:52 |
bhundven|afk |
is now known as: bhundven |
# 03:16:20 |
bhundven |
is now known as: bhundven|afk |
# 04:59:24 |
alan_o |
quits : Quit: Leaving |
# 05:06:51 |
codyps |
joins #crosstool-ng |
# 05:39:28 |
sspiff |
quits : Ping timeout: 252 seconds |
# 06:44:38 |
smartin_ |
joins #crosstool-ng |
# 06:59:31 |
sfan5|OFF |
is now known as: sfan5 |
# 07:22:45 |
sspiff |
joins #crosstool-ng |
# 07:22:45 |
sspiff |
quits : Changing host |
# 07:22:45 |
sspiff |
joins #crosstool-ng |
# 07:22:47 |
sspiff |
quits : Remote host closed the connection |
# 07:34:33 |
sh4rm4 |
quits : Ping timeout: 276 seconds |
# 07:38:51 |
sspiff |
joins #crosstool-ng |
# 07:43:09 |
sh4rm4 |
joins #crosstool-ng |
# 08:01:38 |
sspiff |
quits : Remote host closed the connection |
# 08:10:52 |
aaribaud |
quits : Ping timeout: 245 seconds |
# 08:13:39 |
sspiff |
joins #crosstool-ng |
# 08:17:53 |
y_morin |
joins #crosstool-ng |
# 08:36:45 |
aaribaud |
joins #crosstool-ng |
# 09:21:25 |
Net147 |
joins #crosstool-ng |
# 11:16:14 |
Net147 |
quits : Quit: HydraIRC -> http://www.hydrairc.com <- Like it? Visit #hydrairc on EFNet |
# 12:05:30 |
ssspiff |
joins #crosstool-ng |
# 12:05:51 |
sspiff |
quits : Ping timeout: 246 seconds |
# 12:10:15 |
al` |
quits : Ping timeout: 256 seconds |
# 12:14:18 |
al` |
joins #crosstool-ng |
# 12:35:27 |
smartin_ |
quits : Read error: Operation timed out |
# 12:40:57 |
smartin_ |
joins #crosstool-ng |
# 13:02:12 |
ssspiff |
quits : Ping timeout: 246 seconds |
# 13:02:15 |
sspiff |
joins #crosstool-ng |
# 15:25:38 |
Kasreyn |
quits : Read error: Connection reset by peer |
# 15:32:08 |
y_morin |
quits : Quit: Going home... |
# 16:52:27 |
sfan5 |
is now known as: sfan5|OFF |
# 16:52:30 |
sfan5|OFF |
is now known as: sfan5 |
# 16:59:42 |
y_morin |
joins #crosstool-ng |
# 17:11:16 |
smartin_ |
quits : Quit: heading home |
# 17:16:18 |
sh4rm4 |
quits : Ping timeout: 276 seconds |
# 17:37:06 |
sh4rm4 |
joins #crosstool-ng |
# 21:00:08 |
y_morin |
bhundven|afk: Hello there? |
# 21:13:38 |
sfan5 |
is now known as: sfan5|OFF |
# 21:45:36 |
bhundven|afk |
y_morin: heh, just got in |
# 21:45:40 |
bhundven|afk |
is now known as: bhundven |
# 21:46:15 |
y_morin |
bhundven: Have you seen the recent "custom location" patch series for all components? |
# 21:46:29 |
bhundven |
I took quick glance, but nothing detailed |
# 21:47:01 |
y_morin |
bhundven: Recently, you've added "gcc from svn". Do you think that would be covered by "custom location" instead? |
# 21:47:08 |
bhundven |
yes |
# 21:47:30 |
bhundven |
I'm gonna take a little bit of time and work on the CT_Get stuff |
# 21:47:33 |
bhundven |
in that |
# 21:47:44 |
y_morin |
bhundven: I see that "custom location" as some sort of "I am developping that component, just use my own local changes, for ${DEITY}'s sake". |
# 21:47:47 |
bhundven |
I will remove the svn thing and move it to the new CT_Get interface |
# 21:48:19 |
y_morin |
bhundven: I was thinking of removing the "gcc form svn" code path, for the above reason. |
# 21:48:27 |
bhundven |
sure |
# 21:48:37 |
bhundven |
that totally makes sense |
# 21:48:59 |
y_morin |
bhundven: If a non-developper wants to use a gc from svn, it would be his/her responisibility to check it out first, then use "custom location" |
# 21:49:06 |
y_morin |
gc --> gcc |
# 21:49:09 |
bhundven |
right |
# 21:50:06 |
y_morin |
bhundven: So, you agree that, if I strip it off, the "custom location" fills your use-case? |
# 21:50:29 |
bhundven |
david and I both seem to agree on a view that ct-ng has another use-case for toolchain integrators. Such that, I might have an existing source base with changes to components like gcc, binutils, etc... |
# 21:50:50 |
bhundven |
it is definitely not for non-developers |
# 21:51:17 |
bhundven |
I actually propose a CT_DEVELOPER config option under CT_EXPERIMENTAL |
# 21:51:22 |
y_morin |
bhundven: Sure. I never envioned it this way, but if users fidn it usefull for that, then all the better! :-) |
# 21:51:47 |
bhundven |
but I'm not sure your view on that |
# 21:51:53 |
bhundven |
sure |
# 21:52:55 |
bhundven |
just so that toolchain developer options are seperated from what may be considered experimental |
# 21:53:13 |
bhundven |
as a user may want experimental options, but not developer options |
# 21:53:36 |
y_morin |
bhundven: I would prefer we not have yet another hidding option. We already have OBSOLETE and EXPERIMENTAL. I guess EXPERIMENTAL conveys the proper meaning: it's experimental, either in ct-ng itself, or in a component. |
# 21:53:51 |
bhundven |
i c |
# 21:54:11 |
bhundven |
and really, a lot of what is under experimental mostly says it hasn't been tested |
# 21:54:14 |
y_morin |
bhundven: but maybe the help text for EXPERIMENTAL needs updating, then. |
# 21:54:24 |
y_morin |
yep. |
# 21:54:43 |
bhundven |
so many things I want to contribute to ct-ng, so little time... |
# 21:54:44 |
bhundven |
:p |
# 21:54:59 |
y_morin |
bhundven: have you seen Johannes' debug-shell feature? It is a neat litle trick! ;-) |
# 21:55:33 |
bhundven |
no. I just got back from talking with a recruiting/consulting agency and have been dealing with that kind of stuff for the last week |
# 21:55:37 |
bhundven |
I'll take a peek |
# 21:56:32 |
y_morin |
bhundven: when a command fails, it spawns a shell with the current environment variables, and allows you to fix up things, then {resume,re-run,abort} |
# 21:57:02 |
bhundven |
ah, nice! |
# 21:57:03 |
y_morin |
bhundven: how did things go with the startup thingy? |
# 21:57:32 |
bhundven |
still waiting for the funding to go through. but when we talk with them, they are really excited. |
# 21:57:49 |
bhundven |
sounds like it will be at most a month away |
# 21:57:55 |
bhundven |
:D |
# 21:57:58 |
y_morin |
Good thing for you! :-) (except you'll get less time for ct-ng... ;-] ) |
# 21:58:07 |
bhundven |
I hope not |
# 21:58:39 |
bhundven |
I'm actually hoping that my part for the startup should be fairly trivial |
# 21:58:44 |
y_morin |
crosses fingers for both outcomes (job + time for ct-ng! :-) ) |
# 21:59:05 |
bhundven |
"get servers up, deal with the fall-out of applications doing bad things." |
# 21:59:56 |
y_morin |
Well, that can be a challenge, if uptime is a high-risk constrain. |
# 22:00:29 |
bhundven |
hybrid cloud based, with some scaling things that need some minor tuning |
# 22:02:12 |
bhundven |
I'm hoping to get back to my personal project with the car stuff |
# 22:02:29 |
bhundven |
sooner if I can be a rockstar and pound this startup stuff out |
# 22:04:05 |
y_morin |
;-) |
# 22:04:55 |
bhundven |
I've been looking at directfb, buildroot and yocto project, and ct-ng... and twittling my fingers... chanting... "soon..." |
# 22:05:07 |
bhundven |
:D |
# 22:05:35 |
bhundven |
but I'm also looking at wayland |
# 22:05:38 |
bhundven |
instead of directfb |
# 22:06:22 |
bhundven |
http://wayland.freedesktop.org/ |
# 22:11:11 |
bhundven |
I really need a timemachine |
# 22:11:17 |
bhundven |
so i can just stop time and code |
# 22:11:44 |
bhundven |
forget about going forwards/backwards in time. Just stop. |
# 22:15:59 |
y_morin |
bhundven: I know the feeling! |
# 22:18:02 |
y_morin |
wrt wayland: I'm a bit skeptic. The idea is good. I didn't see the code, but there's no reson it can't be good. The issue I see is about migrating all the existing (even if there's an Xserver on-top). |
# 22:18:13 |
y_morin |
reson --> reason |
# 22:18:56 |
y_morin |
There is a lot of inertia all over the place, so much that the migration can take years, if not decades. |
# 22:19:25 |
bhundven |
I don't think that is the goal. I think it's to cut the xserver api/abi in half. The technology of X11 is mostly wasted on desktops and laptops. I would never have multiple xdmc sessions on my laptop for instance. |
# 22:19:40 |
bhundven |
xdmcp |
# 22:19:41 |
bhundven |
eh |
# 22:19:46 |
bhundven |
so many acronyms |
# 22:19:59 |
y_morin |
bhundven: Ah, the network-transparency! I for one tend to have a few of them. |
# 22:20:08 |
bhundven |
for a server, it makes sense |
# 22:20:16 |
bhundven |
and that's why you'd run x11 on wayland |
# 22:20:26 |
bhundven |
but for an embedded project |
# 22:20:41 |
bhundven |
just wayland and a custom gtk/kde interface could be sufficient |
# 22:20:57 |
bhundven |
no x11 overhead |
# 22:21:08 |
bhundven |
just straight kms/fb access |
# 22:21:17 |
y_morin |
bhundven: not only for servers. I have a bunch of desktop machines which I do not want to attach a display to each, so I login via XDMCP from my main machine. XDMCP in Xnest, for that matters. |
# 22:21:30 |
bhundven |
ah, yea. |
# 22:21:34 |
y_morin |
bhundven: yep, on specialised setups such as embedded, it makes sense. |
# 22:21:40 |
bhundven |
yes. |
# 22:21:58 |
bhundven |
I don't think many people with your use-case or servers would move to this |
# 22:22:14 |
bhundven |
embedded/mobile otoh |
# 22:22:29 |
bhundven |
and directfb is a direct competitor |
# 22:22:37 |
bhundven |
...directly... :p |
# 22:22:48 |
bhundven |
lol |
# 22:23:14 |
y_morin |
bhundven: for Joe Random User on his laptop to edit spreadsheets, show pictures to Grandma, send mails and play games, yes, wayland is the way to go. |
# 22:23:21 |
y_morin |
bhundven: lol! ;-) |
# 22:23:32 |
bhundven |
yup |
# 22:23:51 |
bhundven |
yea, all the need for user-space server drivers |
# 22:24:08 |
bhundven |
could be elimnated by just having applications draw to the framebuffer |
# 22:24:18 |
bhundven |
through wayland or directfb |
# 22:24:59 |
bhundven |
and with the nouveau and open source ati stuff getting squared away |
# 22:25:02 |
bhundven |
it's a good time |
# 22:25:16 |
bhundven |
to get away from that driver model |
# 22:25:31 |
bhundven |
less package dependencies |
# 22:25:35 |
y_morin |
I haven't followed wayland too much, but would it be possible to 'tunnel' a remote frambuffer through a VNC-like connection (eg. with comporession, and stuff like that) ? |
# 22:25:45 |
bhundven |
:/ idk |
# 22:25:51 |
y_morin |
bhundven: yep, kill the NVidia binary blob! ;-) |
# 22:25:58 |
y_morin |
hides... |
# 22:26:00 |
bhundven |
hehe |
# 22:26:29 |
y_morin |
Well, back to some coding... |
# 22:26:34 |
bhundven |
indeed |
# 22:26:50 |
bhundven |
I'll take some time tonight to get back to the ct_get |
# 22:26:58 |
bhundven |
and clean it up |
# 22:27:01 |
y_morin |
Just one before: do you find the new patchwork usefull |
# 22:27:02 |
y_morin |
? |
# 22:27:08 |
bhundven |
if you want to kill off gcc/svn thing |
# 22:27:19 |
y_morin |
bhundven: Yep, I'm eager to see that CT-Get! |
# 22:27:21 |
bhundven |
didn't even notice |
# 22:28:09 |
y_morin |
bhundven: http://patchwork.ozlabs.org/project/crosstool-ng/ |
# 22:28:46 |
y_morin |
Yep, not hosted on ct-ng.org: needs Django and a DB. I'm not too much fond of running those heavy-weights on the machine. |
# 22:29:25 |
bhundven |
nice |
# 22:29:35 |
bhundven |
that gets us a bit away from hg email |
# 22:29:51 |
bhundven |
well, to the mailing list anyways |
# 22:29:52 |
bhundven |
? |
# 22:30:00 |
y_morin |
bhundven: well, you still need to send the patches so pathwork gets them. |
# 22:30:06 |
bhundven |
ah, ok |
# 22:30:20 |
y_morin |
bhundven: patchwork is only a mean to see the pending stuff. |
# 22:30:34 |
bhundven |
I see |
# 22:30:43 |
bhundven |
ok, I'll have to read up on that when I do my stuff |
# 22:31:44 |
y_morin |
bhundven: that should be transparent to you: send the patches as usual,patchwork is subscribed to the list, and when a patch is sent, it stores it. |
# 22:31:59 |
bhundven |
ah, ok |
# 22:32:14 |
bhundven |
new process: everything is the same. |
# 22:32:15 |
bhundven |
:D |
# 22:32:23 |
y_morin |
bhundven: then, anyone can run: pwclient list -p crosstool-ng to get the list of pending patches |
# 22:32:35 |
bhundven |
nice |
# 22:32:40 |
y_morin |
bhundven: and run: pwclient view to see the patch. |
# 22:33:17 |
y_morin |
bhundven: and I have a script here that gets a patch and applies it to my local tree. Then when I push, patchwork is updated, and a mail is sent to the author. |
# 22:33:37 |
bhundven |
very cool |
# 22:33:49 |
y_morin |
bhundven: I've recently learnt how to use 'git rebase -i' (thanks to kos_tom!). It rocks! :-) |
# 22:33:55 |
bhundven |
now if we could automate supporting people on irc/email |
# 22:33:58 |
bhundven |
:p |
# 22:34:10 |
bhundven |
yea.... |
# 22:34:15 |
y_morin |
bhundven: but I don't think I'm ready to switch over to git. |
# 22:34:30 |
bhundven |
for this project, I'm rather fond of hg |
# 22:34:41 |
y_morin |
bhundven: yep. an bot that replies to questions with something like : RTFM. :-p |
# 22:34:48 |
bhundven |
for other projects I work on, git is a better solution |
# 22:35:04 |
bhundven |
I just dislike svn |
# 22:35:06 |
bhundven |
and perforce |
# 22:35:08 |
bhundven |
and cvs |
# 22:35:16 |
y_morin |
bhundven: perfarce, you mean? ;-) |
# 22:35:22 |
bhundven |
lol |
# 22:35:34 |
bhundven |
don't even get me started on clearcase |
# 22:35:49 |
y_morin |
bhundven: ditto. I;ve suffered enough on it... |
# 22:50:45 |
bhundven |
is now known as: bhundven|afk |
# 23:07:57 |
devcoder |
joins #crosstool-ng |
# 23:08:44 |
y_morin |
quits : Quit: Nighty Night! |