IRC log for #darcs, 2012-06-22

Time	Nick	Message
00:14		preflex joined #darcs
00:35		stepcut joined #darcs
01:02		darcscommitbot joined #darcs
01:41		intripoon_ joined #darcs
05:57		sporous joined #darcs
07:02		darcscommitbot joined #darcs
07:04		kowey joined #darcs
07:04	kowey	good morning!
07:19		raichoo joined #darcs
07:26		schlaftier joined #darcs
07:45		donri joined #darcs
08:09		gal_bolle joined #darcs
09:05		owst joined #darcs
09:48		mekeor joined #darcs
13:02		darcscommitbot joined #darcs
14:42		donri_ joined #darcs
15:25		donri_ joined #darcs
15:37		bsrkaditya joined #darcs
15:45		donri joined #darcs
15:54		kowey joined #darcs
15:55	kowey	bsrkaditya: will be needing a moment to think here, be with you in a bit
15:55	kowey	sorry, not being very organised
15:55	bsrkaditya	Hi Kowey. :-)
15:57	kowey	bsrkaditya: OK, I think I'm ready. Basically I have an agenda which consists of 3 major items, each of which may take a while
15:57	kowey	as I mentioned earlier, we may want to prepare for this to be a long meeting (sorry)
15:57	kowey	[1] our 3 wk catch-up (the usual meeting stuff)
15:58	kowey	[2] overall project check-up (how is it working out for you, what to fix mentoring wise, etc)
15:58	kowey	[3] midterms
15:58	bsrkaditya	okay.
15:59	kowey	so let's started, how did the past 3 wks go?
16:00	bsrkaditya	My memory does not extend that far. :-)
16:00	kowey	that's why it's useful to blog
16:00	kowey	or keep a personal notebook if there's stuff you'd rather not blog about
16:01	kowey	now in our last meeting (wk2)
16:01	kowey	you said that you wanted to work on completing the integration of patch index into changes/annotate
16:01	kowey	the ideal being to support all of the darcs options
16:01	kowey	to support changes/annotate on multiple files
16:01	kowey	and to add non-range matchers
16:01	kowey	how did that go?
16:02	bsrkaditya	I support all options in both changes and annotate,
16:02	bsrkaditya	Multiple files in changes does work
16:02	bsrkaditya	Multiple files in annotate is not in the current work, So I did not do anything on it. :-)
16:03	bsrkaditya	Non range matchers work.
16:03	bsrkaditya	Looking through the history of my commits,
16:03	bsrkaditya	(I am starting from June 1st)
16:03	kowey	that sounds pretty good
16:04	kowey	speaking of commits, can I suggest you switch to using the rebased branch?
16:04	kowey	it will save a lot of difficulty with conflicts later on
16:05	kowey	also, is the darcsden repo current? or do you have local patches you have not yet pushed?
16:05	bsrkaditya	I am keeping the patch index repo up to date with screened. How can it still have conflicts? (Just a general question)
16:05	bsrkaditya	I just pushed all changes into the darcsden repo
16:05	kowey	bsrkaditya: see http://wiki.darcs.net/ConflictsFAQ for the full answer
16:05	kowey	short answer is that darcs conflict handling is still in pretty bad shape
16:06	kowey	and if you only resolve conflicts in one side (say your branch), without pushing the resolution over to the other side
16:06	kowey	you get into something called a “conflict fight”
16:06	kowey	which is a real headache
16:06	kowey	which is why darcs rebase is important to use as a stopgap (aside from being useful in its own right)
16:06	bsrkaditya	What is the problem with pushing the resolution over?
16:06	kowey	until we get a better theory
16:07	kowey	it could perhaps be done if you feel ready for it; would need to get it through patch review, I think
16:07	kowey	basically, I strongly recommend just using the rebased branch
16:07	kowey	it will save you trouble
16:07	bsrkaditya	I will do so. :-)
16:08	kowey	we may rebase again in the future
16:08	kowey	let's remember to talk about pushing to mainline later
16:08	kowey	so ok, you seem to have done what you set out to accomplish in weeks 3 and 4, do I understand correctly?
16:08	bsrkaditya	Yes.
16:09	kowey	is there anything else you managed to work on for week 5?
16:09	bsrkaditya	I added support for changes on directories.
16:09	bsrkaditya	And I looked over the cases where
16:10	bsrkaditya	patch index versions of the command differ from the current ones/
16:10	bsrkaditya	We already discussed the differences in annotate,
16:11	bsrkaditya	And there are differences in changes as well.
16:12	bsrkaditya	In changes (an probably annotate) the patch index version can only
16:12	bsrkaditya	miss some patches, but not introduce new ones.
16:12	bsrkaditya	This is because of how the (new) patch index changes works.
16:13	kowey	hmm
16:13	kowey	could you give an example to show you mean by missing some patches?
16:13	bsrkaditya	The key function is filterPatches,
16:13	bsrkaditya	Example, just run it over the src directory of screened.
16:14	bsrkaditya	You will see patches missing
16:14	bsrkaditya	Now, whether these patches should be there or not
16:14	bsrkaditya	is something that is disputable.
16:14	kowey	OK, I'm not going to go run this right now
16:15	kowey	how would you characterise a missing patch?
16:15	kowey	what's the difference between a patch that would be missing in the new changes/annotate from one that would be included?
16:15	bsrkaditya	The key is weather or not the patch
16:16	bsrkaditya	modifes a file that is currently in the given directory
16:17	kowey	ok, so in a simple scenario
16:17	kowey	if I create a file f
16:17	kowey	and p1 creates/modifies the content of the file
16:17	kowey	p2 deletes it
16:17	kowey	p3 creates/modifies f again
16:17	kowey	and I say darcs changes f
16:18	kowey	… is this the sort of thing you're talking about?
16:18	bsrkaditya	Yes, that is one possible scenario.
16:20	kowey	http://hpaste.org/70305 ?
16:20	kowey	this is a standard darcs
16:20	kowey	it seems to only show the second f
16:20	bsrkaditya	In this case patch index and standard darcs give the same output.
16:21	kowey	so is it possible to describe a minimal example where you would get a difference?
16:21	kowey	I'm just trying to understand what exactly you mean here
16:23	kowey	hmm, you mentioned directories
16:23	kowey	let me try one
16:23	bsrkaditya	create a dir x
16:23	bsrkaditya	create patch p1
16:23		raichoo joined #darcs
16:23	bsrkaditya	add file x/a
16:23	bsrkaditya	create patch p2
16:23	bsrkaditya	move file x/a to a
16:24	bsrkaditya	create patch p3
16:24	bsrkaditya	run darcs changes x
16:24	bsrkaditya	and darcs change --patch-index x
16:25	kowey	do you have a script that generates this in your local repo?
16:25	kowey	(not asking for now)
16:26	kowey	(am asking out of general methodological interest: having observed the difference, have you tried to make it reproducible?)
16:26	bsrkaditya	Not in the repo, but just as a script in my computer.
16:27	bsrkaditya	The basic point is that p2, and p3 will be missing.
16:27	bsrkaditya	They are patches concerning a
16:27	bsrkaditya	but they are no longer in x.
16:28	kowey	I've updated my paste
16:29	kowey	OK after running optimize --patch-index
16:29	kowey	I can see a bit what you mean now (paste updated again)
16:31	bsrkaditya	The scenario is not exactly the same is it?
16:31	bsrkaditya	(and how on earth did you get the prev error?)
16:32	bsrkaditya	You seem to have added a new file later.
16:32	kowey	I don't know, but it looks like my PI darcs is up to date
16:32	bsrkaditya	do you have the history of your commands?
16:33	bsrkaditya	it is supposed to create a patch index
16:33	bsrkaditya	if it does not already exist.
16:35	bsrkaditya	Ah! There is a bug in my functions.
16:35	bsrkaditya	:-(
16:35	kowey	updated my paste
16:35	kowey	ok, so I think I have a rough idea about this discrepency
16:36	kowey	what else did you find that was different between PI darcs and the regular version?
16:37	bsrkaditya	That is all.
16:37		favonia joined #darcs
16:37	kowey	ok, anything else to report for week 5?
16:38	bsrkaditya	I will need a clearer idea on what patches should be there and what should not be.
16:39	bsrkaditya	The current implementation is just chaotic.
16:39		schlaftier joined #darcs
16:40	* kowey	nods
16:41	kowey	so you think it would be helpful to have some sort of specification on how darcs should behave wrt filenames in the working directory and how they map on to files which exist (or have existed?) in pristine?
16:41	bsrkaditya	yes.
16:41	kowey	identifying how darcs behaves when files are renamed or deleted
16:41	bsrkaditya	but it also possible to cover this up.
16:42	bsrkaditya	(or do it without a spec)
16:42	kowey	interesting
16:42	bsrkaditya	Patch index stores every file id that corresponds to a file path
16:43	bsrkaditya	so, I could include every file id of every patch that is in the directory
16:43	kowey	it would be interesting to hear more about how you're thinking to approach this problem (this sort of thing is what the blog is good for)
16:44	kowey	but maybe we should try to cover a bit more ground first
16:44	bsrkaditya	cover more ground?
16:44	kowey	[more ground in the agenda] so you've identified this discrepancy, and you think you have a potential workaround that should give identical behaviour to current darcs
16:44	kowey	is that what you're saying?
16:44	kowey	but that the workaround would be essentially ad-hoc/a hack?
16:45	bsrkaditya	yes.
16:45	kowey	generally speaking, I think it's nice when you're working on a project
16:45	kowey	to leave things a little bit clearer for the next guy that comes along
16:46	kowey	eg. by producing some description of what happens in a high-level doc
16:46	kowey	but really I think beschmi (and/or Heffalump) would be better placed to help you figure out where to take that
16:46	kowey	… so any other thoughts on week 5?
16:47	bsrkaditya	patch index is best on files,
16:47	bsrkaditya	it gives the exact same output as current commands
16:48	bsrkaditya	on directories it gets slower,
16:48	bsrkaditya	and the output may differ.
16:48	kowey	slower, but still faster than non-PI?
16:49	kowey	one possibility is just to not use the patch index if on a directory
16:49	kowey	at least that way you're guaranteed identical behaviour
16:49	bsrkaditya	I yet have to find it to be slower.
16:49	kowey	how are you getting these observations?
16:50	kowey	(about what's slower/faster, etc)
16:50	kowey	is it from your understanding of how things work?
16:50	kowey	just using darcs PI on a day to day basis?
16:50	bsrkaditya	By running it on darcs screened directories?
16:50	kowey	informal testing
16:50	bsrkaditya	yes.
16:50	kowey	so something like time darcs changes --patch-index
16:50	bsrkaditya	yes.
16:51	kowey	ok, so any more about week 5?
16:51	bsrkaditya	You just found a bug. :-)
16:52	bsrkaditya	(grasping at straws)
16:52	bsrkaditya	(we should move on)
16:53	kowey	hmm, ok so for you
16:53	kowey	the last week has essentially consisted of making darcs changes --patch-index work on directories
16:53	kowey	and studying the source code (?) to figure out what differences there are between PI/non-PI in practice
16:54	kowey	one big one being the behaviour when you do changes on a directory
16:54	kowey	and files have historically been renamed, deleted, etc
16:54	kowey	you also have some informal timing tests showing that darcs PI is less of an optmisation when done on dirs
16:55	kowey	… so what did you find particularly challenging (or maybe easy?) in week 5?
16:55	kowey	any brick walls run into?
16:56	bsrkaditya	I tried to cover up the problem by including the latest file id
16:56	bsrkaditya	even if that file id no longer coressponds to that file name.
16:56	kowey	you tried to get pi darcs to behave the same way on dirs
16:57	kowey	as regular darcs
16:57	bsrkaditya	yes
16:57	kowey	but not succesfully
16:57	bsrkaditya	it will(probably) work if I include all file ids
16:57	kowey	aside from the lack of a specification on that particular problem
16:58	kowey	is there any thing that you worked on this week, where you felt like you could have used some sort of help or advice?
16:59	bsrkaditya	nothing particularly. I ran into a little problem with porting changes code
16:59	bsrkaditya	as I did with annotate
17:00	bsrkaditya	(copy paste code)
17:00	bsrkaditya	as there were a lot of functions
17:00	bsrkaditya	that work only on PatchSets
17:00	bsrkaditya	So instead, I had to let those functions
17:01	bsrkaditya	run first, and then on the resultant [Sealed (PatchInfoAnd p)]
17:01	bsrkaditya	I have used patch index to filter out patches
17:01	bsrkaditya	This made things somewhat inefficient
17:02	bsrkaditya	(for example changes on GNUMakefile got slowed down by 0.2 sec)
17:02	kowey	because of repeated traversal
17:02	bsrkaditya	sorry? (I do not understand)
17:03	kowey	I was trying to understand what you meant by inefficient
17:03	kowey	you're saying that there were lots of helper functions that “worked on” PatchSets (took PatchSets as input?)
17:03	kowey	so instead of using them directly, you somehow had to wrap them
17:04	kowey	to get just a flattened list [Sealed (PatchInfoAnd p)]
17:04	kowey	and pi-filter said list?
17:04	bsrkaditya	Those functions in the end give out the list.
17:05	bsrkaditya	It got inefficient because, the prev pi implementation
17:06	bsrkaditya	first used the PatchSet on pi and returned [Sealed (PatchInfoAnd p)]
17:06	bsrkaditya	now I have to instead work on [Sealed (PatchInfoAnd p)]
17:06	bsrkaditya	(the inputs to the patch index changed)
17:06	bsrkaditya	because of the loss of structure,
17:07	kowey	have you discussed this issue with Benedikt?
17:07	bsrkaditya	not yet.
17:07	bsrkaditya	But I think it is inevitable.
17:07	kowey	sounds worthwhile
17:08	kowey	ok, so a point of difficulty for you was grappling with the fact that you had patchsets on the one hand
17:08	kowey	and a pi that only wanted flattened lists as input on the other hand
17:09	kowey	and that this pre-conversion was expensive
17:09		donri_ joined #darcs
17:09	bsrkaditya	pi works best if the input is PatchSet
17:09	bsrkaditya	but instead got a list
17:09	kowey	ok, well let's save this for Benedikt
17:09	bsrkaditya	(the conversion time is too small to matter)
17:10	kowey	overall then, how did you feel about how week 5 went? [:-), :-\|, :-(]
17:10	bsrkaditya	:-)
17:11	kowey	that's nice to hear, any elaboration?
17:11	bsrkaditya	because I can see the end in sight.
17:11	kowey	ah good, well then
17:12	kowey	I think we need to cover parts [ii/iii] of the meeting
17:12	kowey	so part [ii]
17:12	kowey	the project has been going on for over a month now (and a bit longer if we include the work you did during the bonding period)
17:13	kowey	and I think it's a good time for me to ask overall (as opposed to just focusing on the past week)
17:13	kowey	how you feel about it
17:13	kowey	(I may have one or two questions which are variants of this)
17:14	bsrkaditya	Can you be more specific? (I am not good with general questions like this)
17:14	kowey	ok, well first, difficulty
17:14	kowey	how challenging has been project been for you so far?
17:15	kowey	(very easy, easy, normal, difficult, very difficult)?
17:15	kowey	I'm not writing this down in a spreadsheet or anything, just poking and prodding a bit
17:16	bsrkaditya	easy-normal. I the goal were well defined, and the few problems that came up did not require too much ingenuity.
17:17	kowey	would you like the project to be more challenging?
17:17	kowey	(again, don't think there this a “right” answer here)
17:18	kowey	(I'm only as useful to you (if at all) as you can be accurate in figuring out what works for you)
17:18	bsrkaditya	Yes? I followed the path of least resistance, which may not have been the right thing.
17:19	bsrkaditya	is not the right thing
17:19	kowey	what do you mean?
17:20	bsrkaditya	Hmm, 1) I did not give enough thought how the changes I make will effect the rest of the system
17:20	bsrkaditya	2) Did not write tests to check if the functions I wrote/modified are doing what I expect
17:22	kowey	hmm, ok so you think perhaps that trying to do a bit more of the real-world-open-source stuff
17:23	kowey	(lots of projects want tests, for example, and patch reviewers can be grumpy about breaking other parts of the system)
17:23	kowey	would bring this project a bit closer to the kind of difficulty level you (maybe) think you want?
17:26	bsrkaditya	The difficulty is not something to be artificially created. But I think that I could have gone about this with more care.(Think more on things)
17:27	kowey	how about the project goals you set out initially in this project?
17:27	kowey	how do you feel you stand with respect to them?
17:27	kowey	you said earlier that you felt like the end was in sight?
17:28	bsrkaditya	We are somewhat ahead of schedule.
17:28	kowey	does that mean you think it (partly because they are well-defined), it should be easy to meet them?
17:28	bsrkaditya	Yes.
17:28	kowey	ok, so we may need to pile more work on then if we finish ahead of time
17:28	bsrkaditya	:-)
17:28	kowey	(it's all about finding that sweet spot of productivity = happiness, IMHO)
17:29	kowey	what about the blogging?
17:29	kowey	what's that like for you?
17:29	bsrkaditya	Mixed feelings. It was quite tough for me.
17:30	kowey	is it more the act of writing? or perhaps the extra complication of writing in English? (I'd pretty much die if asked to blog in French)
17:31	kowey	or perhaps something else?
17:32	bsrkaditya	The act of writing is hard. It probably says something unflattering about me, but I write best in English. (and speak as well, I suppose)
17:32	Igloo	Also, are other people reading the blog? Do they find it useful/interesting?
17:32	Igloo	It's worth doing something hard if it's useful, but not if it's not :-)
17:33	kowey	well sometimes the hardness is inherently useful
17:33	kowey	although that may not be pertinent here
17:33	kowey	ok, so I'll make a mental note that you're finding the writing hard
17:33	Igloo	You mean writing about the design may force you to think about it more? That's true
17:34	kowey	(and forgive me) continue to expect that we keep up the blog-by-Sunday commitment
17:34	Igloo	Anyway, sorry for butting in :-)
17:34	bsrkaditya	I will keep it. I hope.
17:34	kowey	because I think the practice (always good to see somebody else is participating in this discussion; always afraid of saying something silly and it not being caught)
17:34	kowey	will be useful for you in the long run
17:35		amgarchIn9 joined #darcs
17:35	kowey	and also because it's helpful for us as a project
17:35	kowey	to have the blogs as a sort of long term archive
17:35	kowey	and something we can use to eg. show other gsoc students what gsoc is like
17:35	kowey	or google how we use our GSoC time, etc
17:35	kowey	(no pressure!)
17:36	kowey	well, a commitment is a commitment; you don't hope to keep it; you keep it (somewhat just teasing, but half-serious)
17:36		donri_ joined #darcs
17:36	kowey	sorry, being moralistic
17:36	kowey	OK what about our current bi-mentor setup
17:36	kowey	and the meetings?
17:36	kowey	how is this format working for you?
17:37	bsrkaditya	I missed a meeting with you, and 2(I think) with Benedikt.
17:37	bsrkaditya	They were quite helpful.
17:38	kowey	I'm sorry they take so long!
17:38	kowey	tend to be a bit long-winded :-/
17:38	kowey	but am reassured to hear you're still finding them helpful
17:38	bsrkaditya	I just do not notice. :-)
17:38	kowey	so it is still Sundays that you two are meeting?
17:39	bsrkaditya	I am meeting him tomorrow actually.
17:39	kowey	ah right, he had mentioned travel plans
17:39	kowey	OK, two more questions for part [ii]
17:39	kowey	so generally, is there anything about how we're mentoring the project that you think we could maybe try doing differently?
17:40	kowey	(because it's only the uhm, 4th time, and each one is a learning process)
17:40	bsrkaditya	It seems to be working. I think I am fine with it.
17:41	kowey	(hoping that with enough practice/feedback, darcs can get better at mentoring projects)
17:41	kowey	ok so final question on my mind
17:41	kowey	what's your typical (for part ii), work day like on the project?
17:41	kowey	what sorts of things do you find yourself tending to spend more time on than others?
17:41	kowey	(eg. hacking, testing, staring at code, just thinking…)
17:42	bsrkaditya	I have no idea. :-)
17:42	kowey	really?
17:42	kowey	well ok, tell me about your last two days
17:42	bsrkaditya	It generally goes like I want to accomplish something specific
17:43	bsrkaditya	I read some code
17:43	bsrkaditya	I think of some approach
17:43	bsrkaditya	I refactor/copy some code
17:44	bsrkaditya	I write up a few functions(I found a few times that there are better ways, only after finishing)
17:44	bsrkaditya	The code does not compile
17:44	bsrkaditya	I do crazy things, and mess up the code
17:45	bsrkaditya	The errors vanish one by one
17:45	kowey	so would “damn it, work!” adequately capture the spirit of a typical day?
17:46	bsrkaditya	it captures the ending part of the day. :-)
17:46	kowey	and when you say errors, what do you mean? how do you find them?
17:46		favonia joined #darcs
17:46	bsrkaditya	compiler errors.
17:46	kowey	ah
17:46	kowey	so much of your time is spent trying to make it build?
17:46	bsrkaditya	There are a few errors that are not compiler errors,
17:47	bsrkaditya	which are hard to find/notice
17:47	kowey	like just now
17:47	kowey	with the auto update
17:47	bsrkaditya	Yes
17:47	kowey	so could I ask again (just to make sure I understood)
17:47	bsrkaditya	I don't think that I spend too much time fighting the compiler
17:48	kowey	ah ok
17:48	bsrkaditya	But I do not have any numbers
17:48	kowey	that's fine
17:48	kowey	just trying to see how you feel about things really
17:48	kowey	what it's like
17:48	kowey	general impressions
17:48	kowey	sometimes the struggle is code
17:48	kowey	sometimes it's about “gosh, I don't know what the right design is for this”
17:48	kowey	and getting a feel for what your day is like
17:48	bsrkaditya	I find it hard to measure time when programming.
17:49	kowey	helps us as mentors to figure out where to focus our energy
17:49	kowey	yeah, not really talking about time so much, well a bit
17:49	kowey	more like what's easy vs hard for you
17:49	kowey	it's ok if you don't really have a good answer, just poking
17:49	kowey	i poke and prod a lot, can be a bit annoying
17:49	bsrkaditya	It was generally easy. :-)
17:50	kowey	OK phew, part [iii] of the meeting
17:50	kowey	MIDTERMS!
17:51	kowey	we have another 3 weeks to go
17:51	kowey	before we have to assign a Pass/Fail on the project
17:51	bsrkaditya	there is a pass fail at midterm?
17:52	bsrkaditya	I thought it was at the end.
17:52	kowey	yes (it allows projects to terminate early if things are really just not working out, or at least be half-done)
17:52	bsrkaditya	So, what should be done by the next three weeks? :-)
17:53	kowey	well, we've been talking a bit about it
17:53	kowey	and in a way we're thinking that it's a bit more about how than what
17:54	kowey	for the what (check with Benedikt to make sure)
17:54	kowey	it seems like it would be good to make sure we really get #1 (automatic update) polished and ready
17:54	kowey	http://wiki.darcs.net/GSoC/2012-PatchIndex#goals
17:54	kowey	there seem to be a lot of dones there, but also some things missing
17:55	kowey	and perhaps it would be good to finish the job
17:55	kowey	make sense so far?
17:55	bsrkaditya	Yes.
17:55	kowey	but really the how is the important part here
17:55	kowey	and you've alluded to this in your comments about how things wetn
17:55	kowey	which is that this is a real project with real (believe it or not) users
17:56	kowey	so we kind of want to make sure we are doing our best to ship people some quality code
17:56	kowey	and to this end, we've devised 4 criteria to help maybe get a better idea what this means
17:57	kowey	1. that does what is intended
17:58	kowey	2. that it supports all of the options, ie. that real users can actually use it in production, so all flags, etc do what they're expected to (may be less applicable to our automatic update goal here)
17:58	kowey	3. that the code itself is of push-to-mainline quality
17:58	kowey	4. and there are adequate tests
17:58	kowey	does that make sense?
18:00	bsrkaditya	Yes.
18:00	kowey	#3 may seem tautological, but what we mean is “a darcs patch reviewer would be happy to accept it”
18:00	kowey	but here we'll use Benedikt as our reviewer
18:01	kowey	OK now we're going to have to change some aspects of the project to make this work
18:01	kowey	I tend to prefer a loose highly autonomous student-led style, but hopefully we can do still run the project with that sort of flavour
18:01	kowey	… but I think we're going to have to crank up the formality a bit
18:02	kowey	because actually getting a nice prototype with promising numbers
18:02	kowey	*getting from
18:02	kowey	to actually-used code is a terrifyingly large gap
18:02	kowey	and you may need the extra support
18:02	kowey	so far so good?
18:02	bsrkaditya	yes
18:03	kowey	so what we'd like you to do
18:03	kowey	and this ties into the criteria for midterms
18:03	kowey	is have your work signed off by Benedikt according to the 4 criteria
18:03	kowey	I think we all need an outside perspective
18:03	kowey	and while it's easy to think something is done
18:04	kowey	it's sometimes helpful to have somebody remind us that it's not done till it's… done
18:04	kowey	do you think you can do that?
18:05	bsrkaditya	It seems my part is small. :-)
18:05	bsrkaditya	I can do it.
18:05	kowey	OK so (and this scratches my autonomy itch a bit)
18:05	kowey	I'm going to ask to you keep in mind that
18:06	kowey	you're working on this fulltime, whereas the rest of us are trying to fit this into a regular work week (not complaining!)
18:06	kowey	which means we're going to need you to take on quite a lot of responsibility here
18:06	kowey	ie. it's going to be up to you to make sure this signing off stuff happens
18:06	kowey	that Benedikt does get a chance to look and review the code, etc
18:07	kowey	cos otherwise he may forget and go off and do some more work writing his PhD thesis
18:07	kowey	and no PhD thesis write-up is complete without a lot of procrastination in the middle
18:07	kowey	can you also manage the sign-off process? ie. the part of about making sure it happens, chasing up on B, etc?
18:08	bsrkaditya	procrastination = darcs work? :-)
18:08	kowey	got me through my thesis
18:08	bsrkaditya	I will do it.
18:08	kowey	basically, just make sure it happens, thanks!
18:08	kowey	ok, one final thing for part 3
18:09	kowey	so you have a potentially enormous amount of work on your hands (eep!)
18:09	kowey	but never fear!
18:09	kowey	you also have help
18:09	kowey	I think it would be good if you could have a chat with Heffalump sometime
18:09	kowey	about the technical aspects of this project
18:09	bsrkaditya	Tomorrow?
18:09	kowey	thereby getting a second tech perspective from somebody more steeped into the darcs work
18:10	kowey	and also more of a design big-picture perspective
18:10	kowey	I'll leave it up to you (ah! more responsibility!)
18:10	kowey	to make sure a meeting gets arranged that works for the two of you
18:10	kowey	good?
18:11	bsrkaditya	okay.
18:11	kowey	ok so to sum up part 3
18:11	kowey	A. the focus for midterms is to make sure the automatic index stuff is completed
18:12	kowey	B. by complete, we mean 1. does what is intended 2. supports all options, 3. passes patch review 4. has proper tests
18:12	kowey	C. we're adding mentor-sign-off to this the project, upping the formality a little, but you have to make sure the sign-off happens
18:12	kowey	D. do meet with Ganesh
18:13	kowey	phew!
18:13	kowey	I think that's all from me
18:13	kowey	any final remarks?
18:13	bsrkaditya	Is Ganesh free now?
18:13	kowey	send him an email if not
18:13	Heffalump	not really, train should get home soon
18:14	kowey	right, same time, same place, bsrkaditya ?
18:14	Heffalump	weekends and evenings are actually bad for me (family/travelling)
18:14	kowey	(next week, that is)
18:14	bsrkaditya	Heffalump: What time is convenient for you?
18:14	bsrkaditya	kowey: bye.
18:14	kowey	good luck with the Sunday blog, and the pre-midterms push!
18:16	Heffalump	does sometime between 5:30am-6:30am on a weekday work for you?
18:16	Heffalump	UK tmie that is
18:16	Heffalump	(what timezone are you in?)
18:16	bsrkaditya	+05:30
18:16	Heffalump	ok, so that'd be middle of the day for you
18:17	bsrkaditya	So monday 05:30 am utc?
18:18	Heffalump	Tuesday ok? I think this particular Monday may be busy (a final part of our ClearCase to Perforce migration is happening this weekend!)
18:18	bsrkaditya	Tuesday then.
18:18	Heffalump	cool, see you then
18:47		donri joined #darcs
18:57		bsrkaditya left #darcs
19:02		darcscommitbot joined #darcs
19:06		donri joined #darcs
19:45		rui joined #darcs
20:13		babalone joined #darcs
20:32	* darcscommitbot	move Darcs.Witnesses to Darcs.Patch.Witnesses (Guillaume Hoffmann)
21:08		donri joined #darcs
21:13		favonia joined #darcs
21:31		babalone joined #darcs
22:10		babalone left #darcs

← Previous day | Channels | #darcs index | Today | Next day → | Search | Google Search | Plain-Text | plain, newest first | summary