| Time |
S |
Nick |
Message |
| 00:01 |
|
iago |
sorry,anyway I was just "joking" |
| 00:02 |
|
sm |
good to check these things out, since we have different languages/dialects. No, I was just remarking that darcs convert annoys me a bit with it's silly prompt. |
| 00:03 |
|
sm |
lunch time, later all |
| 00:03 |
|
Heffalump |
hehe, I'm overdue for bed |
| 00:07 |
|
|
secorp left #darcs |
| 00:07 |
|
Heffalump |
alexsuraci: now I've got two problems. (1) there's an _darcs/lock file in my repo (ganesh/darcs-rebase-20110110) and (2) my bit-at-a-time push keeps running into problems with conflicts in the intermediate state. Might it make sense for the global pref for apply to be --allow-conflicts? |
| 00:08 |
|
|
lispy left #darcs |
| 00:13 |
|
alexsuraci |
Heffalump: have you tried specifying a bootstrap url? |
| 00:14 |
|
alexsuraci |
Heffalump: removed the lock |
| 00:17 |
|
Heffalump |
alexsuraci: I didn't, because I vaguely recalled you saying that didn't work too well either |
| 00:18 |
|
alexsuraci |
oh, i fixed it recently, shouldn't be too bad. i think. |
| 00:18 |
|
Heffalump |
is it too late to do that now I've pushed some patches anyway? |
| 00:18 |
|
alexsuraci |
uses darcs get now |
| 00:18 |
|
alexsuraci |
you can delete the repo if you want, or rename it |
| 00:18 |
|
alexsuraci |
or bootstrap as a different name and rename later |
| 00:19 |
|
Heffalump |
ok, I'll do that. But I think apply --allow-conflicts might be helpful anyway (I actually vaguely recall discussing this before, but perhaps not with you) |
| 00:22 |
|
Heffalump |
the error message when you get the boostrap URL wrong is pretty ugly |
| 00:23 |
|
Heffalump |
and it leaves a broken stub repo wwhich I now can't delete |
| 00:23 |
|
alexsuraci |
fun |
| 00:24 |
|
Heffalump |
darcs-marking-20110222 is the broken repo, if you could delete it by hand |
| 00:25 |
|
* Heffalump |
considers installing darcsden on his laptop so he can hack on it, but I'm currently running Windows on that so I can improve the darcs experience there - do you think I have any reasonably hope of getting darcsden running there? |
| 00:26 |
|
sm |
ouch |
| 00:26 |
|
sm |
there's redis, couch, snap... |
| 00:26 |
|
alexsuraci |
it's not that hard |
| 00:26 |
|
Heffalump |
well, I can but try :-) |
| 00:26 |
|
alexsuraci |
there's a readme |
| 00:26 |
|
Heffalump |
specifically for Windows? |
| 00:27 |
|
alexsuraci |
just get couch and redis running with their defaults, and the readme has the res |
| 00:27 |
|
alexsuraci |
oh, not sure |
| 00:27 |
|
sm |
I thought snap didn't support windows |
| 00:27 |
|
sm |
I would lay money there is not a readme concerning windows :) |
| 00:27 |
|
Heffalump |
screwed by mtl dependencies at the first hurdle, before even getting into Windows specific issues |
| 00:28 |
|
Heffalump |
actually, I thought you updated darcsden to darcs 2.5? Is that not on hackage yet? |
| 00:29 |
|
alexsuraci |
might not be. try from alex/darcsden |
| 00:29 |
|
Heffalump |
just doing that |
| 00:30 |
|
Heffalump |
I guess if I want to hack on it that's the right thing to do anyway |
| 00:34 |
|
|
intripoon joined #darcs |
| 00:35 |
|
Heffalump |
your dep on darcs >= 2.5.1 is too loose, it'll never work with 2.8.0 without changes. |
| 00:36 |
|
* Heffalump |
leaves cabal install running and goes to bed |
| 00:36 |
|
alexsuraci |
g'night |
| 00:36 |
|
|
idnar left #darcs |
| 00:36 |
|
|
idnar joined #darcs |
| 00:36 |
|
|
intripoon__ joined #darcs |
| 00:36 |
|
|
intripoon_ left #darcs |
| 00:37 |
|
|
jutaro left #darcs |
| 00:38 |
|
|
intripoon left #darcs |
| 00:40 |
|
|
kolmodin_ joined #darcs |
| 00:58 |
|
sm |
snap is rumoured to not be impossible to install on windows, due to libev flag being false by default |
| 01:11 |
|
|
iago left #darcs |
| 01:50 |
|
|
copumpkin left #darcs |
| 02:21 |
|
|
gwern joined #darcs |
| 02:21 |
|
|
gwern left #darcs |
| 02:21 |
|
|
gwern joined #darcs |
| 02:56 |
|
|
copumpkin joined #darcs |
| 03:43 |
|
|
owst left #darcs |
| 03:55 |
|
|
intripoon joined #darcs |
| 03:58 |
|
|
intripoon__ left #darcs |
| 04:47 |
|
|
gwern left #darcs |
| 05:37 |
|
|
teratorn left #darcs |
| 05:37 |
|
|
teratorn joined #darcs |
| 05:45 |
|
|
gwern joined #darcs |
| 05:45 |
|
|
gwern left #darcs |
| 05:45 |
|
|
gwern joined #darcs |
| 06:20 |
|
|
secorp joined #darcs |
| 06:38 |
|
|
copumpkin left #darcs |
| 06:39 |
|
|
copumpkin joined #darcs |
| 07:26 |
|
|
lelit joined #darcs |
| 07:34 |
|
|
sm left #darcs |
| 08:01 |
|
|
darcscommitbot left #darcs |
| 08:02 |
|
|
darcswikibot left #darcs |
| 08:02 |
|
|
darcscommitbot joined #darcs |
| 08:03 |
|
|
darcswikibot joined #darcs |
| 08:10 |
|
|
jutaro joined #darcs |
| 09:32 |
|
|
raichoo joined #darcs |
| 09:55 |
|
|
jeltsch joined #darcs |
| 10:29 |
|
|
teratorn left #darcs |
| 10:29 |
|
|
teratorn joined #darcs |
| 11:05 |
|
|
iago joined #darcs |
| 11:37 |
|
|
jeltsch left #darcs |
| 12:00 |
|
|
jeltsch joined #darcs |
| 12:01 |
|
|
jeltsch left #darcs |
| 12:21 |
|
iago |
hi |
| 12:33 |
|
Heffalump |
hi |
| 12:46 |
|
|
raichoo left #darcs |
| 12:49 |
|
|
copumpkin left #darcs |
| 12:50 |
|
|
copumpkin joined #darcs |
| 13:21 |
|
|
McManiaC joined #darcs |
| 13:21 |
|
McManiaC |
darcs failed: error opening _darcs\pristine.hashed\0000003820-fd58c08ae78a7e6575589eb300a2c1a255287e61501c2f8360b72a87a9807193 |
| 13:21 |
|
McManiaC |
how can I fix that err? |
| 13:21 |
|
Heffalump |
was that at the end of a get? |
| 13:22 |
|
McManiaC |
get that whenever I try to run a cmd |
| 13:23 |
|
Heffalump |
try darcs repair then (back up your repo first) |
| 13:24 |
|
McManiaC |
hmk |
| 13:24 |
|
McManiaC |
thx |
| 13:25 |
|
McManiaC |
Replaying patch 21 done, 23323 queued. Tue Feb 6 15:32:22 Mitteleurop[_\e4_]isc |
| 13:25 |
|
McManiaC |
gonna take a while :> |
| 13:26 |
|
Heffalump |
do you know what triggered the problem? |
| 13:27 |
|
|
raichoo joined #darcs |
| 13:28 |
|
McManiaC |
well, we mixed darcs and git, so there are a lot of reasons why this might have happened |
| 13:28 |
|
McManiaC |
^^ |
| 13:37 |
|
Heffalump |
how did you mix them? |
| 13:55 |
|
|
jutaro left #darcs |
| 14:00 |
|
|
jeltsch joined #darcs |
| 14:20 |
|
McManiaC |
GHC was/is using darcs, we were using git for our working repo, so yeah… |
| 14:20 |
|
McManiaC |
^^ |
| 14:33 |
|
Heffalump |
ah |
| 14:50 |
|
Heffalump |
all hell would probably break loose if git started trying to VC the files under _darcs. That's the only reason I can think of for it to go wrong, though. |
| 14:54 |
|
McManiaC |
yeah |
| 14:55 |
|
McManiaC |
I might just revert those changes |
| 15:03 |
|
|
owst joined #darcs |
| 15:15 |
|
|
gh_ joined #darcs |
| 15:26 |
|
|
intripoon left #darcs |
| 15:27 |
|
|
intripoon joined #darcs |
| 15:29 |
|
|
intripoon left #darcs |
| 15:30 |
|
|
intripoon joined #darcs |
| 15:33 |
|
|
intripoon_ joined #darcs |
| 15:33 |
|
|
sm joined #darcs |
| 15:35 |
|
|
intripoon left #darcs |
| 15:38 |
|
|
intripoon_ left #darcs |
| 15:40 |
|
|
jutaro joined #darcs |
| 15:42 |
|
|
intripoon joined #darcs |
| 15:45 |
|
|
intripoon_ joined #darcs |
| 15:47 |
|
|
intripoon left #darcs |
| 15:49 |
|
|
raichoo left #darcs |
| 15:51 |
|
|
intripoon_ left #darcs |
| 15:55 |
|
|
intripoon joined #darcs |
| 15:57 |
|
|
intripoon_ joined #darcs |
| 15:57 |
|
|
gh_ left #darcs |
| 15:59 |
|
|
intripoon left #darcs |
| 16:05 |
|
|
intripoon joined #darcs |
| 16:06 |
|
|
jeltsch left #darcs |
| 16:07 |
|
|
intripoon_ left #darcs |
| 16:09 |
|
|
intripoon left #darcs |
| 16:11 |
|
|
intripoon joined #darcs |
| 16:14 |
|
|
intripoon left #darcs |
| 16:16 |
|
|
intripoon joined #darcs |
| 16:18 |
|
|
Heffalump left #darcs |
| 16:55 |
|
|
raichoo joined #darcs |
| 17:00 |
|
|
sm_ joined #darcs |
| 17:03 |
|
|
sm left #darcs |
| 17:03 |
|
|
sm_ is now known as sm |
| 17:13 |
|
|
amoe left #darcs |
| 17:17 |
|
iago |
does some Darcs developer think that Darcs is a easy to maintain piece of software? |
| 17:23 |
|
|
owst left #darcs |
| 17:33 |
|
|
jeltsch joined #darcs |
| 17:38 |
|
sm |
no |
| 17:40 |
|
|
intripoon left #darcs |
| 17:41 |
|
|
intripoon joined #darcs |
| 17:44 |
|
|
bonobo joined #darcs |
| 17:51 |
|
|
jeltsch left #darcs |
| 17:56 |
|
iago |
sm, noone or simply you no |
| 17:57 |
|
iago |
and if someone is able to explain why, I would like to ask why too |
| 17:58 |
|
|
jutaro left #darcs |
| 18:00 |
|
|
exlevan left #darcs |
| 18:02 |
|
sm |
iago: just giving a broad answer to your broad question :) but I would guess none of the darcs devs typically think "man, this is easy to maintain" |
| 18:03 |
|
sm |
though I guess it's relative. Coming from something larger/worse, many people might think so |
| 18:03 |
|
sm |
and we here might not fully appreciate it's goodness, being so close to it |
| 18:04 |
|
iago |
well, you can change the question |
| 18:04 |
|
iago |
does Darcs developers they produce software easy to maintain ? |
| 18:04 |
|
iago |
it's hard to critic yourself |
| 18:04 |
|
iago |
if that helps |
| 18:04 |
|
iago |
(does Darcs developers think that ...) |
| 18:04 |
|
sm |
I too am interested in how to make it better. But have to go walking with gf, sorry |
| 18:05 |
|
iago |
nothing |
| 18:05 |
|
iago |
just asking for how you see this |
| 18:05 |
|
iago |
not saying my opinion |
| 18:08 |
|
|
jaredj joined #darcs |
| 18:11 |
|
jaredj |
Q. I developed four separate programs for my AVR microcontroller in separate dirs, then pulled all of them into one dir, and fixed up the inevitable Makefile conflicts. Now I've changed something from R3. I want to pull that change back into R3 but I don't want all the changes I pulled from other dirs to come with it, making R3 into something that it's not. Is there a way to do that? |
| 18:12 |
|
jaredj |
(oops, separate dirs R1, R2, R3, R4) |
| 18:14 |
|
jaredj |
if it were svn i'd svn merge -C27327 ../together/the-file.c; under darcs i don't suppose i can split patches apart so willy-nillily |
| 18:15 |
|
jaredj |
both repos have a file called main.c; the patch i'm trying to pull changes both files |
| 18:15 |
|
jaredj |
ergo, unrecord, record two separate patches, pull the one |
| 18:17 |
|
jaredj |
oo. that would probably work if i hadn't accidentally trashed R3...? agh |
| 18:27 |
|
|
bonobo left #darcs |
| 18:49 |
|
|
Heffalump joined #darcs |
| 18:51 |
|
Heffalump |
jaredj: you can reconstruct R3 by pulling the relevant patches out of the merged repo |
| 18:52 |
|
Heffalump |
iago: I agree that darcs is not easy to maintain. I think that is mainly (but not entirely) a legacy issue. |
| 18:52 |
|
jaredj |
Heffalump: good point >.< |
| 18:52 |
|
jaredj |
i don't know if i care about R3 independently anymore though |
| 18:53 |
|
Heffalump |
well, in that case your entire question is moot :-) |
| 18:53 |
|
jaredj |
things are coming 'together' quite nicely :) |
| 18:54 |
|
jaredj |
academically though, if i pulled everything from R1, then R2, then R3, then i want to get the R3 patches, are they certainly last, or in among the other patches? |
| 18:54 |
|
Heffalump |
certainly last |
| 18:55 |
|
Heffalump |
but you can also use things like --match 'touch filename' to help identify the relevant changes (you might have to repeat that for lots of filenames though) |
| 18:55 |
|
|
jeltsch joined #darcs |
| 18:55 |
|
jaredj |
yeah i just found out about --match, pretty cool |
| 18:56 |
|
jaredj |
unfortunately the whole problem is that i have a main.c in R3 and a main.c in together |
| 18:56 |
|
jaredj |
i think that's why i got so many wrong patches back in R3 in the first place |
| 18:56 |
|
Heffalump |
yeah, that's a pretty fundamental conflict |
| 18:56 |
|
Heffalump |
we have a vague plan to make that be less messy but that's very much "future work" |
| 18:58 |
|
iago |
Heffalump, thanks for your opinion, though I would appreciate something more concrete than "a legacy problem". I'm asking for example if you consider that the size of Darcs is the main maintainability problem and that it would be solved once you split darcs-core as a lib (just to say something) |
| 18:59 |
|
iago |
perhaps you consider the main problem is the lack of documentation, or whatever |
| 19:00 |
|
Heffalump |
I think the main problem is the lack of proper modularity in the code. |
| 19:01 |
|
Heffalump |
It's not well factored into separate pieces with a well-defined specification. |
| 19:01 |
|
Heffalump |
darcs is not particularly large, so that's not the problem itself. I don't think making a library will help in itself, but the work needed to make a good library will necessarily also improve the design of the code as a whole, and perhaps being forced to maintain the library will help keep us honest. |
| 19:06 |
|
iago |
thanks |
| 19:25 |
|
|
McManiaC left #darcs |
| 20:19 |
|
|
jaredj left #darcs |
| 21:45 |
|
iago |
Heffalump, simple curiosity, there is some idea to replace conflictors in the future Darcs3 ? |
| 21:46 |
|
iago |
some idea you really think it deserve more thinking |
| 21:48 |
|
Heffalump |
I have a plan. I still need to write it down and convince everyone else it works :-) |
| 21:48 |
|
iago |
what is the general high-level idea? |
| 21:48 |
|
iago |
(I mean, what you can express in few simple phrases every mortal will understand) |
| 21:49 |
|
Heffalump |
using a graph to represent the patches in conflict, to avoid ever having to have duplicates stored |
| 21:50 |
|
iago |
so, the same idea of conflictors but refined |
| 21:52 |
|
Heffalump |
yeah |
| 21:55 |
|
|
jutaro joined #darcs |
| 21:55 |
|
|
gwern left #darcs |
| 23:13 |
|
|
iago left #darcs |
| 23:26 |
|
|
iago joined #darcs |
| 23:42 |
|
|
jeltsch left #darcs |
| 23:59 |
|
|
jutaro left #darcs |
