Meeting Minutes: March 30, 2014


Announcements / General tips

  • UCOSP is wrapping up – this is the _last_ required meeting for heapify and olessia! from what m_conley can see, the two of them are in the right spot with their projects, just solidifying the last bits

  • m_conley would like them to stick around after the semester ends – and adds that others should do as well

  • m_conley: As we get closer to the end, make sure you come to us during the week with anything and remind us to look at things if you go even a day without a review.





  • question regarding to an open issue here:

  • added new url to reviewboard (r/rr_id/bug/bug_id/) that redirects to the bug tracker url to get the infobox url similarly as the user infobox constructs it (just add infobox/ to the bug url) Asks if it sounds okay.

  • m_conley and purple_cow thinks its fine

  • no other questions or blocking factors



[Absent – MIA]



  • projects are done, just fixing bugs

  • m_conley  points out she’s handling some smaller bugs as approaching final day

  • question for purple_cow: Bhushan told me you said it could be a good idea to add a check to see if the review request’s commit ID matches the revision that got pushed? (this is after the regex has matched a review request ID in the commit message) – should I add this?

  • purple_cow I think that would be something nice if the regex doesn’t match

  • oh, if the regex doesn’t match, which review request’s commit ID would we be checking the revision with?

  • purple_cow you’d use the commit ID to try to find the review request

  • m_conley heapify: awesome. Great work this semester!



[Absent – Travelling]



  • (m_conley looks at screenshots from review request)

  • Joonas_ – this point when resize the window the top navigation links move vertically aligned/one below another. now just hide them under a Menu button.

  • m_conley asks – idea is instead of sliding in a drawer from the side, to stretch down the titlebar?

  • Joonas_: yes – asks if it sounds good.

  • m_conley was personally more interested in the slide-y drawer thing, but thinks perhaps it is because he is more used to it. also curious to know what performance impact, if any, there is in increasing the vertical height of the page by stretching down that titlebar. claims sliding in the drawer over top with a CSS transform is nice because it offloads the work to the GPU and doesn’t require reflow whereas stretching an item like this will require reflowing the document

  • Joonas_ hadn’t thought about that.

  • m_conley (reflowing = making the document recalculate the heights, widths and locations of things, which is a computationally expensive thing – especially for complex documents)

  • m_conley asks how difficult it’d be to have it slide in over top. any sense of the scale of that work?

  • Joonas_ doesn’t know – just worked on the drop down menu – hasn’t done a slide out drawer before but can investigate in that

  • m_conley has done stuff like this before and suggests he talk after meeting, hash out what is required.

  • Joonas_ agrees.

  • m_conley – regarding question from status update this week – suggests sketching out something after meeting

  • Joonas_ is okay with that.

  • purple_cow  – probably also move the search box into the drop-down and hide the version number – everything that’s in the header takes space away from other stuff, so on mobile it should be super streamlined

  • * m_conley nods



  • m_conley – would like screenshots from mobile versions – to lower the barrier for the feedback-givers as much as possible – give us all of the materials we need to give you feedback quickly, and we’re more likely to give it to you quickly

  • bro1 can definitely do that

  • m_conley is really eager to see what this patch does. 🙂

  • bro1 will get screenshots up tonight

  • m_conley points out looks like Joonas_ asked you a question in your review request, that you’ll want to address

  • bro1 will take care of that as well

  • no other questions – m_conley suggests bro1 not wait for feedback but start ahead with other stuff



  • olessia is wrapping up, so really appreciate some answers to the questions on the review request

  • m_conley –  alright, can provide those

  • ChipX86 meant to get the rest of that when he got back into town, but been knocked out of commission all week from a nasty cold – going through it now

  • olessia hopes ChipX86 is feeling better

  • m_conley asks  if anything else you need from us before the term ends?

  • olessia – No – asks about performance review

  • m_conley according to Michelle, the evaluations go in on April 11 – so sometime after that

  • m_conley – olessia: great job this semester! – says we hope to see both you and heapify stick around – you’re both great contributors, and we’ve got plenty of work. 🙂

  • olessia thanks m_conley and claims she learned a lot.


  • two problems with the code in review request – if those figured out, can move forward – and also can’t get the template working

  • purple_cow asks to show the code for template

  • Audore not on review request currently

  • m_conley – :/

  • ChipX86 points out mentors want to see new code every week

  • Audore – because when I implemented I got an error right away, that the variable wasn’t declared – yea, there lot’s of other code

  • m_conley and purple_cow both want to see the errors that are generated – all of it

  • m_conley  – even if it’s not polished, we need to see where you are and what you’re doing  – impossible for us to help you if we can’t see what you’re doing. :/

  • Audore: ~~ok

  • m_conley requests to  update review request with _all_ of current code and then include a pastie of the error

  • no other questions






  • m_conley claims edwingzg identified some places where we need sandboxing.

  • edwinzg has a list on hackpad of everywhere where i found the hooks being used – noticed a lot of them weren’t really used outside of being defined in asks if thats expected. e.g. DashboardColumnsHook

  • ChipX86 claims should be at a point now where, every single week, you should be able to get us at least 2 review requests, each one sandboxing something – follow the code and note what that one does

  • ChipX86: that’s an interface around registering columns – other code then looks up from that registry – you’ll need to sandbox where the columns are being looked up – many of the hooks register into central registries like that

  • edwinzg – that part is clear, but one more question

  • purple_cow told edwinzg that for testing i should create a new extension which misbehaves and activate an extension – was wondering where the best place in the could to do that would be and how exactly i create a new extension?

  • ChipX86: we have an extensive guide on how to do just that, plus a helper script:

  • m_conley thinks it’d be excellent if this testing could be part of our automated testing suites

  • edwinzg asks how to add it to the automated testing suites.

  • ChipX86: every directory has a file where the unit tests live

  • m_conley: believes we have tests that create and register extensions somewhere

  • ChipX86 –  we already have tests for the hooks and the registries that some hooks call into you can just add new tests that simulate failures

  • m_conley:  reviewboard/extensions/ is where you’ll probably want to put your new tests

  • m_conley: so to not run the full suite each time – also



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s