Time
S
Nick
Message
04:04 axfelix joined #dataverse
04:25 xvx_ joined #dataverse
04:35 garnett joined #dataverse
04:39 petrichor joined #dataverse
08:59 bencomp joined #dataverse
09:03 jri joined #dataverse
11:05 pdurbin
bencomp: I fixed https://github.com/IQSS/dataverse/issues/2512 for you. And everybody. :) Thanks again for the bug report.
11:05 bencomp
pdurbin: cool!
11:08 pdurbin
I mean, it hasn't been merged yet. Soon, I hope. It missed the 4.3 train. :(
11:48 xvx joined #dataverse
12:32 bencomp joined #dataverse
13:04 xvx joined #dataverse
13:06 pdurbin
bencomp: I appreciate your feedback on https://github.com/IQSS/dataverse/commit/ffd1d6ee7b6bf8d518dba4ac239008671e7ab240
13:54 bencomp
pdurbin: great :) you're welcome
13:55 bencomp
would it be possible to agree on where to ask questions about code?
13:55 bencomp
I see too many todos in the code and their number is only growing
13:56 pdurbin
bencomp: where would you like to ask questions about code?
13:56 bencomp
if you think some test is in the wrong place, how about creating an issue for it?
13:57 bencomp
s/test/check/
13:57 pdurbin
I hope you're not talking about https://github.com/IQSS/dataverse/commit/8471c02f4b662acac235d0917a869ca13fbcc63e
13:59 bencomp
well, I hadn't seen that commit yet, but it appears to me that todos in Dataverse code are not actively followed up
13:59 pdurbin
well, it may take us a while
14:00 pdurbin
bencomp: https://github.com/IQSS/dataverse/commit/ffd1d6ee7b6bf8d518dba4ac239008671e7ab240#commitcomment-16500163 :)
14:01 pdurbin
bencomp: I think you worry too much about the todos :)
14:03 pdurbin
bencomp: hey, I just set up (temporary) builds at http://guides.dataverse.org/en/2939-shib/installation/shibboleth.html so I can start improving Shib docs for https://github.com/IQSS/dataverse/issues/2953 . Is there anything in particular you'd like me to add about Shib? That issue lists some things that are on my mind.
14:05 bencomp
pdurbin: many todos could be read as "I know this is wrong, but for this secret reason I will pretend it's not my problem right now."
14:06 pdurbin
you gotta punt on some stuff so you can ship
14:08 pdurbin
bencomp: do you feel bad about opening issue or something? There's a label for "code infrastructure" for stuff like tests being in the wrong place: https://github.com/IQSS/dataverse/labels/Component%3A%20Code%20Infrastructure
14:09 bencomp
pdurbin: would you like me to open an issue? I don't understand
14:10 bencomp
fyi: there are still 650+ todos in-code. 650+ loose ends?
14:10 pdurbin
bencomp: I want you to feel like you have a place to write down your concern. You said, "would it be possible to agree on where to ask questions about code?" An issue is fine. Are you thinking of other places besides issues?
14:10 pdurbin
hmm, that's a lot of todos :)
14:11 bencomp
I think some never should have been todos, like the question I answered
14:11 bencomp
sometimes you can go over to someone's desk and ask :)
14:12 pdurbin
bencomp: maybe we should all focus on *new* todos that appear in pull requests. We know where to comment in that case... on the pull request itself.
14:13 bencomp
exactly
14:13 pdurbin
but it's an open question of where to comment on the 650+ existing todos? where would you like to comment on these?
14:15 bencomp
my suggestions for "where to ask about code" would be: look up the answer in the code, or ask a colleague for a direct decision (e.g. should we upgrade to this newer library?"), or open an issue to explain the problem and refer in the issue to the code that would be influenced
14:16 pdurbin
yeah, in the case of https://github.com/IQSS/dataverse/issues/2998 we discussed at our standup that I should open an issue
14:16 bencomp
although it would be okay to have a TODO in-code to refer to an issue
14:17 bencomp
for everyone's overview of the State of the Code/Project you don't want TODOs that have no issue
14:19 pdurbin
it makes sense that a todo should have an issue, generally speaaking
14:19 pdurbin
if we have too many open issues we'll be colored red if we ever sign up for http://www.codetriage.com :)
14:24 bencomp
I think you have too many issues ;)
14:24 pdurbin
yeah
14:24 pdurbin
bencomp: but go ahead and make another one if you want
14:25 bencomp
:D
14:25 pdurbin
I still don't know which test is in the wrong place.
14:26 pdurbin
bencomp: also, I know you have tools that you use to get a sense of The State of the Code but it would be great if you could help us get this set up in some sort of central location so we could all look at the same overview.
14:26 pdurbin
we don't see some of the things you see
14:28 bencomp
regarding the "test": I meant the checking of trimmed email addresses instead of untrimmed ones
14:29 pdurbin
oh
14:29 bencomp
counting TODOs can be done from NetBeans
14:29 pdurbin
so you seem to agree that there's something weird there. you're glad that I at least added a todo.
14:32 bencomp
to be honest I thought you shouldn't have added a todo, but you should have done the research to answer that question
14:32 pdurbin
hopefully you're glad that I added tests for those methods too :)
14:33 pdurbin
bencomp: do keep in mind that that todo is in a feature branch. it hasn't been merged yet
14:33 pdurbin
I haven't even made a pull request.
14:35 bencomp
right
14:37 pdurbin
in this case I think adding that todo is starting to spur a useful discussion
14:37 pdurbin
it will appear in the pull request
14:44 bencomp1 joined #dataverse
14:45 pdurbin
A todo is much lighter weight than a GitHub issue. I wouldn't want to discourage developers from adding todo's. I'd want them to have an outlet for whatever is bothering them.
14:50 bencomp
sure, but I have seen a lot of laziness and loose ends
14:51 pdurbin
bencomp: we would probably benefit from the "overview" you mentioned
14:52 bencomp
I would want developers to actively ask for opinions and not use TODOs to "throw their concerns just out there"
14:53 pdurbin
we do. I could forward a thread from yesteday to you if you like
14:54 pdurbin
bencomp: please check your email for "DatasetFieldServiceBean - findByName vs findByNameOpt"
15:02 bencomp
thanks, got it
15:03 pdurbin
sure
15:05 bencomp
pdurbin: the general problem with my "overview" is that it isn't an overview, it is just a very long list of issues. in NetBeans/Eclipse/IntelliJ IDEA you can group the issues by type, but in no way will any tool that I looked at successfully link an issue on GitHub to todos in the code
15:06 bencomp
so todos get buried in a big pile of code
15:08 pameyer joined #dataverse
15:09 pdurbin
the todos are right in your face when you're actually in the code. it's nice to read how developers think the code can be improved
15:11 bencomp
I don't think we can settle this discussion today ;)
15:11 pdurbin
bencomp: if you want, try to get it on the agenda of the next community call :)
15:13 bencomp
not sure anyone but us is interested, but I might be wrong
15:14 bencomp
maybe on the dataverse-dev list? :)
15:14 pdurbin
what dataverse-dev list? ;)
15:15 bencomp
TODO set up dataverse-dev list
15:15 pdurbin
I created that todo a while ago: https://trello.com/c/aEtEaWia/267-dataverse-dev-list
15:16 bencomp
gotta do some other things now before the weekend
15:23 pdurbin
ok
15:23 pdurbin
pameyer: mornin'! we're still meeting at 1pm at IQSS, right?
15:25 pameyer
pdurbin: morning - that's the plan
15:26 pdurbin
pameyer: cool. If we want we could grab lunch together (and maybe with Bill) before, after, during, or none of the above. :)
15:28 pameyer
not a bad plan - I've got a meeting in a few minutes, and was planning on heading over after that
15:28 pdurbin
ok
15:28 pameyer
Bill's not in the office at the moment; but we can ping him
15:29 pdurbin
I just texted him. I think he still plans to work from IQSS today.
15:29 pameyer
yup
16:26 axfelix joined #dataverse
19:06 cnk joined #dataverse
20:46 axfelix joined #dataverse
21:01 pameyer joined #dataverse
21:26 garnett joined #dataverse
21:38 garnett joined #dataverse
