#dev | Logs for 2017-02-19

« return
[03:30:08] <Bytram> TheMightyBuzzard: took a quick look and it makes much more sense, now. Am utterly wiped and am calling it a [relatively early] night. Will take a more exhaustive view tomorrow. Thanks a bunch for the quick response!
[12:05:49] * TheMightyBuzzard eyes Bytram
[12:05:54] <Bytram> I just remembered a template change I made
[12:06:05] <Bytram> added "Page:" to two places on a template
[12:06:19] <Bytram> prior to whre the "(1)" stuff appears on each page.
[12:06:25] <Bytram> that was a manual edit on #dev
[12:06:38] <Bytram> don't know if you noticed or if it made it to the actual site code
[12:07:09] <TheMightyBuzzard> i don't think that bit uses a template anymore.
[12:08:01] <Bytram> I saw it appear on dev after I made the change... maybe it got dropped out subsequently...
[12:08:26] <TheMightyBuzzard> prolly. i been deploying after any changes pj requested.
[12:09:14] <TheMightyBuzzard> look, if it bugs you that much i can just have it drop the page numbers unless there's more than one
[12:09:25] <TheMightyBuzzard> Page is redundant though
[12:09:31] <Bytram> point is, and someone else noted this, the (1) does not convey much information to the user. Prefacing it with "Page:" makes that clear and also provides context to suggest that when (2) appears that these are links one can use to navigate to another page of comments
[12:09:55] <Bytram> is not just me, this was something that I saw someone else point out and suggest, too.
[12:10:43] <Bytram> just a sec
[12:12:26] <Bytram> see: linkCommentPages;misc;default
[12:12:40] <Bytram> add: \tPage:\n
[12:12:41] * TheMightyBuzzard looks at his unread feed items count in despair
[12:13:17] <Bytram> following the TWO places where there is: <div class="linkCommentPage">
[12:13:22] <Bytram> ta dah
[12:13:24] <Bytram> is on dev atm
[12:19:48] <TheMightyBuzzard> not seeing it. caching issue i guess.
[12:20:26] <Bytram> I saw it, and now I don't...
[12:21:17] * Bytram hits "Save" again
[12:21:27] * Bytram does Ctrl+F5 ona story
[12:21:39] * TheMightyBuzzard steps out for a smoke
[12:21:53] <Bytram> not seeing it atm.
[12:22:20] <Bytram> yep, break time.
[12:22:22] <Bytram> afk
[12:30:15] <TheMightyBuzzard> yeah, i hard updated the template but i gotta say i really dislike it. it puts the page numbers off center, it's redundant for a near universal convention, and it adds clutter.
[12:37:14] <Bytram> was unaware that it was a "near universal convention" -- I don't like it being a bit off-center, but that can be handled with a css fix "<div class="linkCommentPage" style="padding-right: 2em"> (or similar)
[12:39:54] <Bytram> newp
[12:40:39] <TheMightyBuzzard> check most any site that paginates. google, pornhub, yahoo...
[12:41:37] <Bytram> it already has a class... https://dev.soylentnews.org at .linkCommentPage add "padding-right: 2em;"
[12:41:38] <aqu4> No title found.
[12:41:42] <Bytram> lol
[12:42:03] <Bytram> don't use those pages, so hadn't seen it.
[12:42:44] <Bytram> oh, and "Page" s/b "Page:"
[12:44:04] <Bytram> I loaded: https://www.yahoo.com
[12:44:05] <aqu4> No title found.
[12:44:53] <Bytram> scrolled to the bottom of the page, and it auto-extended and added more stuff. Hit "End" key again, and it again auto-extended. did not see (1) or anything like it.
[12:45:12] <TheMightyBuzzard> you must have javascript enabled
[12:45:20] <TheMightyBuzzard> it only paginates if you don't
[12:45:44] <Bytram> that could be
[12:45:51] <TheMightyBuzzard> duckduckgo doesn't paginate, just has an unlabeled Next button
[12:46:34] <Bytram> yeah, but the "Next" in the button conveys what it is intended to do
[12:46:44] <Bytram> go to the Next page of results.
[12:47:18] <Bytram> http://folding.extremeoverclocking.com
[12:47:19] <aqu4>  ^ "3Folding @ Home Teams Overall Rank - EXTREME Overclocking Folding @ Home Stats"
[12:47:45] <Bytram> near upper Right-Hand corner: Pages (60): « 1 2 3 [4] 5 6 7 8 » ... Last »
[12:48:02] <TheMightyBuzzard> Yes, pages (60) is informative. pages alone would be redundant.
[12:49:28] <TheMightyBuzzard> s'why i haven't felt the need to label comment mode or sort order. the values inform you as to the function.
[12:49:43] <Bytram> let's agree that *we* disagree... seeing as another user asked the same question, what is the downside in providing syntax on the page to make clear what it is, versus assuming that people understand unstated semantics?
[12:50:59] <Bytram> and that makes it a royal PITA to describe test cases, tell you how to reproduce a bug I found, describe controls on that page in our help docs....
[12:51:26] <TheMightyBuzzard> aesthetics. a cluttered page is an unhappy page.
[12:51:34] <Bytram> Threshold: 2 Breakthrough: 3 Mode: Threaded-TNG Sort: Oldest First
[12:51:54] <Bytram> semantic overload; syntactify your semantics.
[12:52:35] <Bytram> and on the third unlabeled drop-down box select "Threaded-TNG"
[12:52:48] <Bytram> =)
[12:53:04] <TheMightyBuzzard> we don't have a users manual. if we need one we have failed at our job.
[12:54:10] <Bytram> so what does "Threaded-TNG" *mean* and how is that different from "Threaded-TOS" ??? That is very obvious as to what their function is, right?
[12:54:17] <TheMightyBuzzard> man, i am like 30s away from removing delete comment functionality from the codebase entirely.
[12:54:49] <TheMightyBuzzard> it's extremely obvious if you select it and see.
[12:55:17] <Bytram> problem is, if delete has ever been used, then there are gaps in the DB that need to be handled, right?
[12:56:22] <TheMightyBuzzard> very likely and we do not delete comments. not ever.
[12:56:42] <Bytram> sorry, ask PJ about all those spammed messages.
[12:56:49] <TheMightyBuzzard> also fucks up parenting and a host of other things.
[12:57:15] <Bytram> It's my understanding that they were deleted -- I *could* be mistaken.
[12:57:46] <Bytram> select count(*) from comments; vs select max(cid) from comments;
[12:58:15] <TheMightyBuzzard> that was done manually i believe.
[12:58:52] <Bytram> I don't know how it was done, but if hey actually were deleted, then we need to be able to handle that when loading stories
[12:59:06] <Bytram> oh, wait
[12:59:42] <Bytram> you are talking about removing the UI elements that allow god-mode to delete a comment, instead of just how to render a story that has had comments deleted.
[12:59:46] <Bytram> two different things.
[12:59:58] <TheMightyBuzzard> right
[13:00:22] <Bytram> We can prolly do without the former (not sure, habe to think about it), but we definitely need to be able to handle the latter
[13:00:23] <TheMightyBuzzard> if we really want a comment gone, we can mysql its ass.
[13:01:01] <Bytram> set its score to -1000 == deleted
[13:01:09] <TheMightyBuzzard> can't be done
[13:01:19] <TheMightyBuzzard> you cannot have a score < -1
[13:01:33] <Bytram> update DB... =)
[13:01:40] <TheMightyBuzzard> it will bork things
[13:01:44] <Bytram> j/k
[13:02:14] <TheMightyBuzzard> much like deleting a comment that is a reply to something will fuck up the comments.count field
[13:02:24] <TheMightyBuzzard> on all parent comments
[13:02:48] <Bytram> I'm in a strange space this am... haven't slept a full night through in ages, am tired, and somehow got on an argumentative streak -- prolly partly cause I'm trying to write a reply to OO
[13:02:50] <TheMightyBuzzard> sorry, comments.children
[13:03:05] <Bytram> hrrm, yep. it would.
[13:03:50] <TheMightyBuzzard> that in and of itself i doubt would break anything too much but it could lead to a few rendering glitches
[13:04:15] <TheMightyBuzzard> now if someone replies to a deleted comment that fucks things up in exciting new ways.
[13:04:48] <Bytram> VERY exciting new ways, I'm sure
[13:05:38] <TheMightyBuzzard> hrm
[13:05:45] <Bytram> hmmmm!
[13:06:05] <Bytram> add a new column to comments: display: boolean.
[13:06:10] <Bytram> default TRUE;
[13:06:18] <Bytram> set to false if is spam
[13:06:36] <Bytram> or what would have been deleted in the past
[13:07:04] <Bytram> first rule of accounting -- NEVER remove an entry, just add a new entry that offsets the errant one.
[13:07:16] <TheMightyBuzzard> yep
[13:07:33] <TheMightyBuzzard> display, maybe.
[13:07:53] <Bytram> so, instead of actually deleting a comment, change the comment.display flag from a default:true to false.
[13:08:10] <Bytram> no gaps in db
[13:08:29] <Bytram> =) !!!
[13:13:13] <TheMightyBuzzard> gaps i'm not worried about. wouldn't change functionality though.
[13:17:43] <Bytram> okay, is beyond me at this point... just offering a different perspective/approach.
[13:28:56] <TheMightyBuzzard> yar. ima punt on this one. convince pj if ya really reckon it needs put in.
[13:30:35] <paulej72> Almost there TMB
[13:30:47] <paulej72> let me read backlog
[13:32:08] * TheMightyBuzzard puts two more sql statements in to copy flat/nocomment
[13:34:17] <TheMightyBuzzard> pushed
[13:35:32] <TheMightyBuzzard> yes, i could have done it in one statement but i don't care. this is not something that desperately needs optimized.
[13:40:53] <TheMightyBuzzard> sigh, i still need to finish updating chillax's db but i am not feeling the desire to just now.
[13:58:18] <Bytram> two sql statements vs one (optimized) statement? KISS - IF it works, don't fix it.
[13:58:50] <TheMightyBuzzard> Bytram, for updating the database during install. run exactly once ever.
[13:59:15] <Bytram> zactly - I fully agree... premature optimization ... causes trouble.
[14:04:19] * Bytram rereads TMBs release notes story
[14:37:54] <paulej72> Bytram, TheMightyBuzzard I cleaned up the comments pref page. I got rid of some unused options and moved a few things around and made some edits on the comments
[15:05:06] <paulej72> Bytram: TheMightyBuzzard do we need to have a comments FAQ?
[19:14:32] <Bytram> paulej72: just saw your message... THANKS!!! On my way to have a look!
[19:33:09] <Bytram> found a few things, some that are not new to this release, but while we are here... this is wrt: editComm;users;default
[19:33:26] <Bytram> (1) "Do not display an e-mail address."
[19:34:17] <Bytram> this has always irked me... it's not clear at the outset which email address it is talking about, the current user's e-mail or the e-mail addresses of all the other users on the site.
[19:34:21] <Bytram> suggest changing it to:
[19:34:31] <Bytram> ==> "Do not display my e-mail address."
[19:35:49] <Bytram> and, to continue with parallel construction, change the following lines to be:
[19:36:05] <Bytram> ==> Show my email address with random SPAM-armoring applied.
[19:36:17] <Bytram> ==> Show my real email address without cowering behind childish anonymity or obfuscation.
[19:37:25] <Bytram> (2) Order of the comment page controls should match the appearance of the matching controls when reading a story's comments:
[19:37:30] <Bytram> (a) Threshold
[19:37:35] <Bytram> (b) Breakthrough
[19:38:25] <Bytram> (c) Mode
[19:38:31] <Bytram> (d) Sort
[19:39:53] <Bytram> IOW, move "Display Mode" and "Sort Order" to follow the drop-down list selector for "Breakthrough"
[19:42:36] <Bytram> (3) "Reason Modifier" displays as an underlined link on my display, but when I mouse over it, the underline disappears.
[19:42:55] <Bytram> current code: <p><b><a name="reason_bonus">Reason Modifier</a></b> (assign a modifier to different moderation reasons)<br>
[19:45:19] <Bytram> This code is historical in that it appears to be the same on production.
[19:50:06] <Bytram> One approach is to add CSS to make it not look like a link, possibly by adding an ID on a parent, encompassing block control (maybe on the table?) OR add a DIV around the whole set of controls with an ID that can be referred to.
[19:50:55] <Bytram> then have css of the form: text-decoration: none; to apply to the whole TABLE or DIV
[19:57:12] <Bytram> ==OR== replace the use of an HTML named anchor with using, in this case, ID="reason_bonus" on, say, a <LABEL> tag.
[19:57:16] <Bytram> ...
[19:58:00] <Bytram> whichever of the above approaches, or even a different one, needs to also contend with these entries which also exhibit the same problem:
[19:58:21] <Bytram> >> Reason Modifier
[19:58:30] <Bytram> >> Anonymous Modifier
[19:58:38] <Bytram> >> Karma Bonus
[19:58:46] <Bytram> >> New User Modifier
[19:58:57] <Bytram> >> Small Comment Modifier
[19:59:08] <Bytram> >> Long Comment Modifier
[20:00:16] <Bytram> ^^^ all of these have named anchors, with underlines that disappear on hover, but no other 'labels' on the page has a named anchor.
[20:02:52] <Bytram> NOTE: I am under the impression that old browsers have not have as good support for ID= as they have for NAME= but am not at all certain on that one.
[20:05:41] <Bytram> (4) Do we need a comments FAQ? Good question!! I can go either way on this, but I'm leaning towards it being a good idea as we could use that page to show some actual examples of how the page would layout differently depending on what choices a user selected.
[20:07:33] <Bytram> I do believe we need something to explain what "Threaded-TOS" and "Threaded-TNG" actually *ARE*. Someone new to the site should not be hit with meta-babble based on what our site or the old site used to do.
[20:17:10] <Bytram> SUMMARY: That's a lotsa words, I really appreciate the cleanup, there's really only a couple/three *kinds* of things that I see that could use some changes -- well done.
[20:17:46] <Bytram> paulej72++ for taking the initiative
[20:17:46] <Bender> karma - paulej72: 8
[23:04:04] <paulej72> Bytram made all of the fixes. we did not need the anchors so they were eliminated
[23:05:58] <paulej72> I also reorg the whole page into sections and grouped functions based on display, posting, points modification, or admin together.
[23:15:20] <paulej72> I am inclined to bang on this a bit before deploying. I think next weekend should be acceptable
[23:15:49] <paulej72> I want to make sure NCommander is OK with trheadtos
[23:23:08] <cmn32480> ermagerd.. we might actually roll this out?
[23:23:16] * cmn32480 slinks back inot his hole
[23:24:37] <NCommander> paulej72, it looked good when I looked at it earlier
[23:24:52] <NCommander> I'll re-review and give it another sign-off, I was waiting for TheMightyBuzzard or you to give me a sign for final review
[23:25:19] <paulej72> I think we have it ready for review.