mozilla :: #mdndev

8 Sep 2017
00:00rjohnsonin other words, if you didn't have in your system, you see no results for the expand
00:01sheppyYeah, that's not this issue at all :)
00:02rjohnsonsorry, it took me a while to get that back in my head, i don't have instant context switching like jwhitlock! :)
00:07sheppyThat's fine!
00:07sheppyI just appreciate the willingness to help :)
00:09sheppyTime to go -- night all
00:09rjohnsongood night sheppy!
00:26rjohnsonstandups: submitted and (they lay the groundwork for trying the new rds database)
00:26rjohnsongood night all!
00:37557A8I9MG[interactive-examples] wbamberg pushed 7 new commits to master:
00:37557A8I9MGinteractive-examples/master f3544b7 maboa: typed-array examples
00:37557A8I9MGinteractive-examples/master 05b4c83 maboa: changes as per review
00:37557A8I9MGinteractive-examples/master bcc90a0 maboa: changes as per review
00:37203BAALOP[interactive-examples] wbamberg closed pull request #268: Typedarray examples (master...typedarray-examples)
00:38interactive-examp[interactive-examples] schalkneethling force-pushed prod from 56ce109 to 7bc60f3:
00:38interactive-exampinteractive-examples/prod 7bc60f3 schalkneethling: Deploying latest generated pages
14:06jwhitlockgood morning #mdndev
14:55sheppyThe background color on <code> is really really distracting. Is there a reason we did that? :/
14:56sheppySeeing highlighted words strewn throughout the article is like &quot;what happened here&quot;...
14:57sheppyAll I can see are those words with the grey background; everything else, my eyes can&#39;t focus on (not even hyperbole; I really really can&#39;t read it like this)
14:59sheppyDammit, I hate how I never see any discussions anymore.
14:59sheppyI guess i need to break down and make discourse email everything to me.
15:01sheppyWhy do people like this!?
15:01sheppyI don&#39;t understand.
15:02sheppyI have never understood why anyone does this. It makes text completely unreadable when the background color changes along the way.
15:02sheppyI used to spend extra money on textbooks to be sure there were no highlighted passages.
15:02sheppy(to get new books instead of used ones)
15:04* sheppy starts working on a custom stylesheet to get rid of this.
15:04sheppyIf it makes everyone else happy then obviously I have to find a way to adapt, but I&#39;m very surprised this is a thing people like... I had no idea my eyes and brain worked this differently from others.
15:05sheppyI will point out though that the highlight blends into the background of the &quot;Optional&quot; badges, so it looks like one big blob instead of a badge next to a word or phrase
15:06interactive-examp[interactive-examples] schalkneethling opened pull request #275: There is no event.orginalTarget in Chrome (
15:06interactive-examp[interactive-examples] schalkneethling closed pull request #275: There is no event.orginalTarget in Chrome (
15:06interactive-examp[interactive-examples] schalkneethling pushed 2 new commits to master:
15:06interactive-exampinteractive-examples/master 7834a9c schalkneethling: There is no event.orginalTarget in Chrome
15:06interactive-exampinteractive-examples/master 60a1aa1 Schalk Neethling: Merge pull request #275 from schalkneethling/no-original-target-in-chrome...
15:07shobsonsheppy: I&#39;m not sure optional should be inside any code tags? Since it&#39;s not part of the code? Also, could you send me a link please?
15:08sheppyshobson: it&#39;s not inside a code tag. But its background color is the same as the color you picked for <code>
15:08interactive-examp[interactive-examples] schalkneethling force-pushed prod from 7bc60f3 to a4575de:
15:08interactive-exampinteractive-examples/prod a4575de schalkneethling: Deploying latest generated pages
15:08sheppyshobson: see thisArg in the parameters here -
15:08sheppyOops, no, never mind
15:09sheppySomeone put the optional inside <code> here.
15:09sheppyWhich is an error.
15:14rjohnsongood morning mdndev
15:39sheppyHm. I&#39;m on my local Kuma instance, and I have a page that has four tags on it. I&#39;m looking at them right now. But when I use subpagesExpand() to get a list of pages and then look at the page in question in the list, its tag list is empty.
15:42sheppyWhen I look at the page with $json, the tag list is correct...
15:43sheppyHuh. I added a fifth tag and now the script works right.
16:04jwhitlockstandups: updated bug triage agenda, other Monday agendas. Talked through AWS stuff w/ @jgmize
16:07sheppyBoy, it&#39;s still a bummer that you can&#39;t set break rules for the contents of columns in CSS.
16:08sheppyI have a list that needs to be displayed in two columns and it&#39;s a mess when the browser just breaks wherever.
16:09sheppyshobson: do you have advice on how best to display a <dl> in two columns without weird issues with the column break position?
16:13jwhitlockrjohnson: metadave: do you have a preferred channel for AWS discussions?
16:13metadave#meao-infra or github issues
16:13metadaverjohnson and I are 1-1ing atm
16:14metadaveyou can join if you like
16:15jwhitlockyes I would
16:16metadaveRyan&#39;s Vidyo
16:18shobsonsheppy: it&#39;s impossible as fart as I can tell, there is no way to ensure a colum break does not come between a dt and a dd
16:18sheppyshobson: Yeah
16:18sheppyI&#39;m now looking at using flexbox for this.
16:19shobsonsheppy: are you writing inline styles on a kumascript macro?
16:19sheppyshobson: I&#39;ve been trying to avoid it but likely I will have to do that then once I get the styles right, move it into the stylesheet.
16:20sheppyWe have all these two-column landing page article lists and currently building them in a macro (which we&#39;re trying to do more of) is difficult to do right.
16:20shobsonI am happy to write styles to help you achive what you are trying to do.
16:20shobsonAnd since it is not possible to include media queries in inline styles I am skeptical that inline styles can produce a good mobile user experience.
16:20sheppyAll I&#39;m trying to do is get the <dl> we use on landing pages to list article titles and summaries to draw in two columns.
16:21sheppy(when available, falling back to one if the screen is too narrow, etc)
16:21sheppySo, trying to get output kind of like
16:22shobsonI think it would be best to split it manually like
16:22shobsonOr like on /Web
16:22sheppyWriting code to do that is tricky.
16:23sheppyBecause you have to figure out how long the text is, then find the halfway point and move toward the end until you reach a </dd>, then insert your break there.
16:23sheppyAnd if you&#39;re outputting the text directly into the KumaScript results in a loop, you can&#39;t really do that.
16:24sheppyWhich means re-rewriting these macros back to pre-constructing the output and reprocessing that text to make it break, which is what it was doing before I started this refactor that was suggested. :D
16:25sheppyAh, but I have a macro already whose job it is to take a string and convert it into a two column <dl>
16:25sheppyso I should be able to just pass these over...
16:25* sheppy mutters to himself as he works.
16:28sheppyHmmm. Means moving the bulk of these existing scripts into new ones to convert the existing ones to a wrapper that does template(&quot;MakeColumnsForDL&quot;, [OldMacroWithNewName(...)])
17:01sheppyThe added highlight on <code> also adds extra spacing that is a little strange although I suppose necessary with the color added.
17:02sheppyWhy does the inspector, when I use it on MDN lately, claim that everything is unstyled? :)
17:03firebotBug 1392446 DUPLICATE, Stylo: Rules Inspector Empty
17:03sheppyOh, it&#39;s because I&#39;m using stylo.
17:03sheppyshobson: thanks :)
17:04* sheppy turns off stylo for now
17:07sheppyI know I promised not to mess with styles, but since I have to be able to work, I can&#39;t avoid at least restoring <code> to not having that background on it.
17:07* sheppy adds padding:0 to his custom stylesheet along with the extant background-color:inherit;
17:08sheppyEyes no longer bouncing around the page like those scenes in Terminator movies where the T-800 is analyzing targets nearby.
17:09sheppy(I recognize this is about me, not the design decision itself, other than whatever small concern I&#39;m not the only person in the world who has this problem) :)
17:09* sheppy works on trying to undo his toolishness.
17:14jwhitlockstandups: sync w/ @metadave, @rjohnson, and @jgmize on AWS migration work. I&#39;m up to speed on the plan now, will document the next mini-milestones w/ checklists
17:14jwhitlockbut first lunch
17:30viewsourcebot[viewsourceconf] stephaniehobson opened pull request #223: Update Inayaili&#39;s bio (master...update-bio)
17:30viewsourcebot[viewsourceconf] stephaniehobson closed pull request #223: Update Inayaili&#39;s bio (master...update-bio)
17:32rjohnsonstandups: sync w/ @metadave, @jwhitlock, and @jgmize on aws work, further discussion w/ @metadave on secrets on other details
17:35sheppyWhee! That code works.
17:54rjohnsonstandups: created (secrets for staging)
17:58sheppyWhat does it mean when git status tells something like this:
17:59sheppyHow has my branch diverged from... itself?
17:59sheppyThis happens to me sometimes and I can&#39;t figure out why or what to do to fix it.
18:01sheppyMaybe I&#39;ll figure it out after lunch. :)
18:43jwhitlock|awaysheppy: FixSubpageListMacros is the branch you get when you &quot;git checkout FixSubpageListMacros&quot;
18:44jwhitlockand a2sheppy/FixSubpageListMacros is the tracking branch for the one on github
18:45jwhitlockit is supposed to track this one -
18:45jwhitlock&quot;git fetch&quot; syncs these remote tracking branches with the remote branches they are tracking
18:46jwhitlockso as it says &quot;All conflicts fixed but you are still merging.&quot;
18:46jwhitlock&quot;use &quot;git commit&quot; to conclude merge&quot;
18:46jwhitlockI&#39;m not sure why you are in the middle of a merge, but it seems like you need to get out of it by finishing it or abandoning it
18:47rjohnsongoing to grab some lunch, back soon
18:58sheppyjwhitlock: okay. I will see if I can sort that out when I&#39;m back. Just finished lunch but have to get a test in before I resume.
20:17rjohnsonstandups: created and
20:32sheppyjwhitlock: I was doing a rebase because there&#39;d been some changes elsewhere since I started working on my patch, and while working on the rebase there was a conflict, so I fixed that, and then it claims I have no rebase in progress but there&#39;s this situation with the diverged branches and I&#39;m stumped. :)
20:35jwhitlockwithout more context (which may be hard to give over IRC), it&#39;s hard to say what situation you&#39;re in
20:35jwhitlockis still your situation?
20:35sheppyjwhitlock: yeah -- I&#39;m checking a couple things then may have to get into this more deeply, hang on a sec.
20:41sheppyjwhitlock: So, here&#39;s my current situation... git log tells me this:
20:42sheppyAnd git status says:
20:42sheppyEvery time I follow the instructions and get things so they look right in the code again, the list of things it claims need fixing in git status is longer. :)
20:43sheppySo I figure I should stop following its recommendations and ask for help now.
20:43jwhitlockyou are in the middle of a rebase - &quot;You are currently editing a commit while rebasing branch&quot;
20:43sheppyI thought I was done. I&#39;ve done all the stuff that I thought was needed to finish the rebase so many times now...
20:44sheppyIt&#39;s like an inception-like rebase where every rebase attempt makes it want to do the same set of changes yet again.
20:44jwhitlockso next is probably &quot;git rebase --continue&quot;
20:44sheppyOK... I just did that three times. :)
20:45sheppyWhen I tried it just now, I got that: ^
20:45sheppygit reset as it says?
20:46jwhitlockprobably git rebase --skip
20:47sheppyGuess it didn&#39;t work...
20:47jwhitlockno it worked
20:48jwhitlockNext commands to do (5 remaining commands)
20:48sheppyOh? That looks like an error message to me :)
20:48jwhitlockit was Next commands to do (6 remaining commands): before
20:48jwhitlockyeah it looks like you picked poorly for the branch to rebase on top of
20:48jwhitlockgit rebase --skip again
20:48sheppyI was rebasing vs master to bring in the latest stuff, after doing a git fetch --all
20:49jwhitlockalso commit c444a2a has the worst commit message I&#39;ve seen in a while
20:49sheppyThought I was doing the usual thing to be sure I&#39;m up to date before doing a PR.
20:49sheppyjwhitlock: really?
20:50jwhitlockWAY over 50 characters in that one
20:51sheppyAh. Well, I knew I was looking at merging commits, so I was going to reformat when I did so.
20:52jwhitlockstill on ?
20:52metadaverjohnson: awesome work getting secrets integrated so quickly :-)
20:52sheppyjwhitlock: Did another rebase --skip, have a conflict to resolve this time, oddly.
20:52rjohnsonmetadave: thanks man!
20:53* sheppy runs his test page on this stuff again to be sure nothing is broken
20:53jwhitlockanother option is to give up, git rebase --abort, and we can figure out what was wrong with the rebase command
20:53sheppyjwhitlock: Yeah, maybe that&#39;s a good idea...
20:55sheppyAnd git log:
20:55sheppyjwhitlock: I&#39;m bummed I got into this mess -- I&#39;d been getting on pretty well with git recently. Guess I was due. :)
20:57jwhitlocklooking at the log
20:58jwhitlockit is weird that there&#39;s a merge commit and what appear to be the same commit twice
20:58sheppyI have no idea how that happened. It&#39;s really strange.
21:01jwhitlockcommit 7bd5390 is the actual latest commit
21:02jwhitlockoh hi commit 39a24f11b60 and commit 0e6b6229da5 also look like the same commit
21:04jwhitlockok and it is up-to-date with a2sheppy/FixSubpageListMacros
21:04sheppyHm, I think my mistake may have been that when trying to update my branch to have the latest changes on master, I didn&#39;t do pull --ff-only.
21:05jwhitlockok so there are currently 6 commits that are new. How many do you want?
21:06jwhitlockI think it is easier to see them on github than git log -
21:10sheppyjwhitlock: sorry, got pulled off on something, back
21:10sheppyThis is weird.
21:10jwhitlockhow many commits do you want in your PR?
21:11sheppyThese commits are a mess.
21:11sheppyI mean, I&#39;m looking at one commit that fixes one bug and unfixes another one. WTF?
21:13sheppyOK... I only want one commit in my PR.
21:13sheppyI just need to get the right stuff merged together into that one.
21:13jwhitlockok lets first try &quot;git rebase origin/master&quot;
21:14jwhitlockand tell me what it shows
21:14jwhitlockgit rebase -i origin/master
21:14sheppyWow... it says a lot :)
21:14sheppyLOL - yeah
21:14* sheppy aborts that first one and does the second
21:15sheppyDo you want what comes up in my editor after doing the git rebase -i origin/master, or the result after I save the message?
21:15jwhitlockwhat comes up in your editor
21:17jwhitlockok trying to figure out the order
21:20jwhitlockthe order is bizarre, jumps around
21:20jwhitlockbut I think what you want is
21:20sheppyYeah, when I realized that is when I realized I better ask for help.
21:20jwhitlockkeep &quot;pick&quot; for the first one
21:20jwhitlockchange lines 2, 3, and 4 from &quot;pick&quot; to &quot;fixup&quot;
21:21jwhitlocksave and close the editor
21:22jwhitlockok git rebase --abort
21:22jwhitlocktoo screwed up to mess with, sorry
21:23jwhitlocknew plan
21:23sheppyThe weird thing is that I know nobody touched these files but me, so how the heck did they wind up with conflicts to resolve in the first place?
21:23jwhitlockbecause you did something like &quot;git pull&quot; when you shouldn&#39;t have
21:23sheppyThought I had the order down pat for stuff like this but clearly I was mistaken. :)
21:23jwhitlockif the upstream changed, &quot;git pull&quot; defaults to creating a merge commit
21:24jwhitlockand then you end up in a situation like this
21:24jwhitlock&quot;git pull&quot; is a great shortcut when you know what you are doing
21:24jwhitlock&quot;git pull --rebase&quot; should be the default
21:24sheppyI&#39;ve been using git fetch instead lately... I must have slipped and did a pull by mistake.
21:25sheppyIs it time to save my work aside and nuke this all from orbit? :D
21:27jwhitlockoh wow you copied a file too
21:27jwhitlockok here&#39;s what you type
21:27jwhitlockgit checkout origin/master
21:28sheppyActually I just copied and pasted content from one file into a new one, but apparently that counts. :)
21:28jwhitlockgit checkout -b fix_with_jwhitlock
21:29jwhitlockgit checkout a2sheppy/FixSubpageListMacros macros
21:29jwhitlockgit commit
21:30jwhitlock(add your commit message. It would be awesome if the first line was 50 char or less, and followed by a blank line)
21:30sheppy50 characters isn&#39;t very much... gosh
21:31jwhitlockyeah it&#39;s just a summary, like for an email
21:31jwhitlockbecause it literally is copied as the subject line in kernel workflows
21:34jwhitlock(and then save and close your editor to finish the commit)
21:35jwhitlock(and then say something so we can continue onto the next step)
21:35sheppyjwhitlock: almost done
21:38jwhitlockok check that it looks ok
21:38jwhitlockgit log
21:38jwhitlockgit show
21:38rjohnsonstandups: created/merged
21:40* sheppy looks over show&#39;s output
21:41jwhitlocklog looks good
21:44sheppyshow does too
21:45sheppyWhat next?
21:46jwhitlockgit push --force a2sheppy fix_with_jwhitlock:FixSubpageListMacros
21:48jwhitlockok now
21:48jwhitlockgit checkout FixSubpageListMacros
21:49jwhitlock(just a sec)
21:49jwhitlockgit reset @{u} --hard
21:50jwhitlockok you&#39;re done
21:52jwhitlockthat&#39;s it for me - have a good weekend #mdndev
21:53jwhitlockstandups: helped @sheppy recover from bad merges
21:53rjohnsonhave a great weekend jwhitlock!
21:57sheppyjwhitlock|away: thank you!
21:59sheppyHuh. How do I make it update my PR though
22:01sheppyOh, it already is
22:16sheppystandups: Finished a more thorough overhaul of the macros SubpagesWithSummaries, SubpagesWithTags, ArticlesByTag, and MakeColumnsForDL. Added new macro SubpagesWithTagsOneCol to do a 1-column form of SubpagesWithTags (which calls the one-column version then delivers the results to MakeColumnsForDL to build a 2-col form).
22:17sheppystandups: These edits resolve potential security flaws in these macros, improve code readiability, and improve the resulting output to make it cleaner and more responsive.
23:05rjohnsonstandups: dumped non-custom-collated local sample db and loaded into aws rds stage instance
23:21rjohnsonstandups: created/merged, deployed secrets into staging k8s namespace
23:29rjohnsonstandups: tried deploying latest k8s work (including rds mysql!) to staging namespace (not too bad, some connection kinks to sort out on monday)
23:29rjohnsongood night all!
9 Sep 2017
No messages
Last message: 12 days and 6 hours ago