#ossasepia Logs for 02 Nov 2019

April 20th, 2020
diana_coman: whaack: ah, spaces between words in b, hm. [09:53]
diana_coman: that's why I tend to look for a single b word there; indeed, user-wise it's both unhealthy and annoying if the spaces are discarded but this might be more of an issue on browser-side really [09:57]
diana_coman: ie the browser should *not* discard the spaces I would say. [09:57]
whaack: diana_coman: i think the issue is that the links already have the spaces stripped, not that the browser is discarding them [10:05]
diana_coman: whaack: so what/where strips them? [10:06]
whaack: diana_coman: if you click on your link "How my Blog Moved North"... that was sent via pingback on the proper-html-linking post you should see that the b value is "b=andt" not "b=and t" [10:06]
diana_coman: I see; tbh that link is already dubious ie wtf "and t" did I do there, lol [10:07]
whaack: diana_coman: what do you mean what did you do? The links are generated by taking the first few characters of the excerpts. Your excerpt begins with "and the wasteful padding..." [10:10]
diana_coman: right you are. [10:13]
diana_coman: for some reason I kept thinking earlier that it was a manual select of mine instead of a pingback. [10:16]
diana_coman: whaack: as far as I understand it now, it does seem to me like an actual problem with the way pingbacks are created; just express it more clearly to MP too, as apparently I'm not the only one that had trouble figuring out exactly what problem you found. [13:14]
whaack: diana_coman: Yes, since I did not make it clear to either of you I will try to examine why I didn't articulate myself well. I made a new comment ~an hour ago attempting to better explain the problem, but it is awaiting moderation since it has an external link. [13:16]
diana_coman: ok. [13:18]
whaack: diana_coman: i kept stumbling on how to word the link i was talking about. i finally came up with "pingback linkback". you say '..an actual problem with the way pingbacks are created..' but I don't think that is the correct wording. _your_ blog correctly creates the pingback, the trilema server receives the pingback then modifies the "pingback linkback" and posts it as a comment [13:23]
diana_coman: whaack: so say it like that, even better (because yes, I kept looking at where the pingbacks are created and couldn't find why would that happen) [13:26]
whaack: diana_coman: Hmm. this makes me think that the problem is upstream. the blog that creates the pingback _should_ be the one providing the b&e values. [13:28]
diana_coman: whaack: you look and see; while I did have a look, for one thing I don't know that part inside out and for the other it was still one of those rather quick looks so don't rely on that. [13:31]
diana_coman: basically so far you have that the pingback-comments are stripped of any whitespaces they might contain in the select part; next step is to find out where & why that strip happens. [13:33]
diana_coman: maybe have a look also if you find/have any pingback from trilema to a different blog, perhaps it's again some php version clash or other such nightmare. [13:34]
whaack: diana_coman: i can't find that out because I don't know what code trilema's server is running. it is not the code he posted. I posted code that does not strip whitespace and explicitly puts +'s [13:34]
diana_coman: it runs mp-wp, the original genesis; older php version (and in general: older versions of everything) [13:34]
diana_coman: whaack: where are you otherwise with the rest of tasks for this week? [13:35]
diana_coman: you might need to focus on what's for this Sunday and then investigate this further (if needed) next week rather than right now. [13:36]
whaack: diana_coman: there is a modification to xmlrpc.php that is not vpatch'd. i'm not sure any blogs other than ztkfg/trilema are modifying the pingback linkbacks to use the new select mechanism [13:36]
diana_coman: mine isn't atm, I can confirm that. [13:37]
diana_coman: but I see what you are saying there; ok, leave it for now as it's basically waiting on clarifying it with MP and seeing his response. [13:38]
whaack: diana_coman: I am awfully behind. I am finishing up my post on examining the cause of using the link mechanism incorrectly and then I will move onto the V post. I have a rough draft for that but per our conversation it needs to be shifted quite a bit. [13:39]
diana_coman: whaack: all right; let the V post move on to next week and do it properly then. [13:40]
diana_coman: you got this not-bad investigation out of that mess, at least. [13:41]
whaack: diana_coman: yes it was fruitful. also it gave me more context to understand the bootstrapping out of dk link you recently posted again [13:43]
diana_coman: heh, well done for spotting it; yes. [13:44]
lobbes: diana_coman: in other news my first penance post on derealization is slowly morphing into a shape. Not what the shape will be at the end, but a shape nonetheless [15:17]
lobbes: has taken to keeping a notebook specifically for blog; writes down ideas as they come during the day. As days pass it tends to fall into focus [15:17]
lobbes: (ty, btw, diana_coman for original suggestion and in general) [15:17]
diana_coman: lobbes: you're welcome; and I'll read it when you publish it. [15:34]
lobbes: Hmm, so I've got a new problem that is not related to the "1, 2, 3" mp-wp problems discussed here earlier. Wondering if anyone else has run into it: [19:22]
lobbes: It would seem that trackbacks to a post url that uses the new selection mechanism do not send. I've tested this by manually sending the trackbacks with plain ol' curl, so it seems like the issue isn't on the 'sending end', but on the 'receiving end'. Here's an example paste with two curl commands to my test blog (one that sends a post url sans-selection; the other with selection): http://paste.deedbot.org/?id=dxiq [19:22]
lobbes: The ones I send with the selection syntax in the url fail 100\% of the time. And now that I'm looking, I'm also noticing this behavior on others' blogs. [19:22]
jfw: lobbes: that does sound like a new thing, though of a sort that "surely someone woulda noticed by now!!" My own blog is missing self-pingbacks which I haven't yet investigated. [20:41]
jfw: does the curl show any fault code per http://hixie.ch/specs/pingback/pingback#TOC3 (though idk to what degree that spec actually describes WP) ? [20:44]
jfw: it would appear that the inaugural link doesn't show a pingback on its target [20:46]
lobbes: jfk: fwiw here's the output of another test: "Is there no link to us?" [20:49]
lobbes: yeah, I found another Trilema self-pingback that also did not appear to send: this to that [20:51]
jfw: jfk heh. It occurs to me there's gotta be a permalink resolver in there, redundant with the apache rewrite rules since those don't apply to xmlrpc payload. Possibly whatever that thing is requires a change analogous to the "QSA," allowing it to accept query parameters [21:03]
lobbes: lol whoops. (I realize I don't bother tab-completing your nick since it is so short) [21:07]
lobbes: hmm but this may just be another 'user error' on my end actually. At least in all my test cases the source post did not have a link to the target post in the post content. Plus I just tested via the blog (making sure to actually link) and a self-pingback using the selection syntax worked [21:07]
lobbes: So false alarm I suppose. (As for why I was trackback-ing arbitrary posts that do not have links, it was part of mp-wp bot's trackback testing) [21:07]
jfw: Ahh makes sense. [21:08]