Latest comment: 28 days ago2 comments2 people in discussion
Hello I am interested in reviving the Wikiproject Report. I am very active within wikiprojects and most of my edits can be traced to either wikiproject television or wikiproject Doctor Who. I feel that WikiProjects are key to wikipedias growth and deserve more recognition. What steps would be required in order for me to join the Signpost Team. Some of the projects I wish to cover include WP:COMICS, WP:Wikipedia, WP:DISNEY, and WP:AUSTRALIA. Questions?fourOlifanofmrtennant (she/her) 00:04, 28 February 2024 (UTC)Reply
@OlifanofmrTennant: seems nobody has gotten back to you on this yet, apologies on behalf of the Signpost team!
The landing page has been very broken since the introduction of the new format a few issues ago, and it has not been fixed or reverted. I am in awe of the improvements that have been made to the Signpost infrastructure over the last year or so, but this change is definitely still in early beta. I don't think I have the skills to fix it. – Jonesey95 (talk) 16:26, 3 March 2024 (UTC)Reply
Can we please revert to the previous layout for Wikipedia:Wikipedia Signpost? We have black text on black backgrounds (accessibility problem), overlapping images, and images that exceed the cell-like spaces that they are being put in. The large images are not aligned vertically when the captions are different lengths, and the images overlap some of the text. It has looked bad for a few months now. I tried to insert a {{clear}} template after the first row of images to vertically align the next row, but it looks like the images are not set using divs or a table. – Jonesey95 (talk) 14:36, 25 April 2024 (UTC)Reply
One answer to your question is simply that there was not enough volunteer labor to write the article. Anyone can still cover the story. Bluerasberry (talk) 15:49, 27 April 2024 (UTC)Reply
As someone who participated in the case, I don't believe I should be writing about it. ~~ Jessintime (talk) 03:22, 29 April 2024 (UTC)Reply
Latest comment: 9 hours ago13 comments8 people in discussion
As long as everyone is complaining, I might as well voice my own gripe. In the good old days (yeah, former glory, wooden ships and iron men) the front page was a list of headlines, or rather sections and subsections. It was easy to read. This year, half the text is superimposed on a background photo. This degrades both the photo and the words, making them less easy and pleasant to read. It was a brave try, but it didn't work, at least for me and I'd like to click an option to show the old version, without pictures. Or words all below, above, or beside the picture; whatever. Jim.henderson (talk) 23:13, 1 May 2024 (UTC)Reply
I made a similar note above. The two pages that Bri linked to are fine. It's Wikipedia:Wikipedia Signpost that has been a problem for a few months. It can't possibly comply with our accessibility policies, for one thing. – Jonesey95 (talk) 00:53, 2 May 2024 (UTC)Reply
I'm trying to improve the layout/readability. I got something decent ish here I think, but I can't for the life of my find how the hell to adjust padding/margins so that things don't overlap with each other. I'd get rid of text shadows on the In the media etc. sections, but that would mean playing with the live version since there's no CSS sandbox. Headbomb {t · c · p · b} 03:01, 2 May 2024 (UTC)Reply
I think the only way to make the text accessible with arbitrary images is to separate it the text from the images. Black on translucent gray on an unknown image background is probably not accessible. – Jonesey95 (talk) 17:03, 2 May 2024 (UTC)Reply
The latter version with more details as given by Bri is my preference. But either of these no-picture front page versions is better than what we've been getting the past couple months. Jim.henderson (talk) 09:32, 2 May 2024 (UTC)Reply
I too prefer the pictureless version. But only because the picture-full version is badly aligned and legibility is compromised. I'm not, in principle, opposed to pictures. But the layout needs improvement. Headbomb {t · c · p · b} 20:06, 2 May 2024 (UTC)Reply
agreed; i think pictures are good and well, but i also don't like that the text goes over them. ... sawyer * he/they * talk 22:14, 2 May 2024 (UTC)Reply
The Big Pictures are a distracting, less than good (trying to say that gently) use of space requiring extra scrolling and making skimming very difficult, so the amt of time I care to spend on it is less- to un- informative, imo. Alanscottwalker (talk) 22:47, 2 May 2024 (UTC)Reply
Can you post a screenshot of this? I tested it in a couple browsers when I wrote the styles for this page and it doesn't look like they overlap to me. Here is what I am seeing: jp×g🗯️ 06:57, 13 May 2024 (UTC)Reply
See File:Screen Shot 2024-05-13 at 5.51.36 .png. Note text obscured by images, images that are not aligned, black-on-black text, images that overlap. This is in Firefox for Mac, the latest version. – Jonesey95 (talk) 12:54, 13 May 2024 (UTC)Reply
Over the course of the last year or so I put substantial effort into repairing some of the technical debt in the Signpost codebase, and fixing decade-old broken features, updating the indices, et cetera. I think I did have done pretty good job of most of it (e.g. the module actually works, articles have images, there's a frontend that shows working preview cards and allows Signpost articles to be linked on other websites, the scripts integrate a wider variety of information into the module which can be used freely, author and subheading metadata is saved and accessible, et cetera). I also wrote Wegweiser, which allowed us to actually use the tagging system and module which had been developed most of the way to completion in 2014-15 and then abandoned. I was basically fine with nobody caring about most of this, because I was not doing it for the glory, just because it had to be done at some point. The newsroom display and the next-issue mockup is pretty much completely automated at this point, which eliminates the previously difficult-and-unreliable process of manually updating it. There are a lot of things like this, which I am pretty happy about.
However, in the last few months, I have become convinced that fixing technical issues at the Signpost is not a really good use of my time, primarily because a gigantic proportion of said time is wasted on bureaucratic issues seemingly for no reason. For example, over the course of 2023, I cleaned up a bunch of old accidental redirects from typos, unused shortcuts and the like; in January of this year I deleted one with no incoming links and no pageviews (the PrefixIndex being crammed with unused junk pages makes it harder for scripts and queries to work properly). But the speedy deletion was contested, and it was demanded that I go through a formal RfD discussion, which took an excruciating three months and was finally closed at the end of March. The discussion was eventually closed with a consensus to delete, but three months of time that I could have spent fixing stuff, I spent sitting on my butt doing nothing, because I had to wait for a formal committee process to approve it.
This wasn't even the first time this had happened: in January 2023 I had also been forced to stop my cleanup efforts for several weeks because somebody objected to a couple speedy deletions, and demanded I file individual MfDs for each of the seventeen unused redirects and misnamed templates. So I did, and I waited, and they were all deleted: but I still had to stop working for a month. There is nobody I can send an invoice to for this. The time is just lost.
In general, the impression I get is that the time I spend working on technical issues is considered worthless -- in fact, the more of it I do, the more worthless it's considered, because that means the amount of work I'm willing to do for free is even higher. This means that people feel comfortable doing things that break stuff, or intiating gigantic timesink processes, and cause dozens of hours of pointless busywork, and they don't care, because my labor is worth $0 -- and I can't raise the price, or get people to stop buying it, so the only option I have is to stop selling it entirely, so I have mostly stopped doing technical Signpost work unless it's strictly necessary to prevent things from breaking. I would very much appreciate if anybody was willing to help with this stuff -- I would be glad to give some guidance to anyone who wants to lend a hand, and there's a lot of difficult and tedious work that needs to get done.
I'm aware that the front page design has some issues, e.g. it uses shadow instead of stroke or outline (because of browser compatibility issues iirc but maybe that's changed); opacity and padding need some work; I have a big page of notes on improvements I was planning to make some months ago. Most crucially, it looks kind of ass on mobile. The thing Headbomb posted looks pretty good and I think it would be nice.
When I was trying to figure out how to format images for article cards, I had a significant challenge with aspect ratios. Basically: if you go to a bunch of randomly selected news websites (wapo, sacramento bee, the telegraph, le monde etc) you will see that they have a bunch of images for their articles. These have some landscape ratio (2:1, 3:2, 16:9, etc). That's all good and cool -- but if you look at, for example, an article or two you'll see that there isn't just one image for each article: there's a header image, which has the flattened landscape ratio, but also a sidebar thumbnail, which for each article is a 1:1 square. Adding any kind of metadata to articles requires a ton of extra code in four separate programs (Wegweiser, in the Signpost Publishing Script, in Module:Signpost, and also in the SignpostTagger) all of which interact with the modules. For article pictures, since every image we want to use for an article doesn't come in a pretty correct-aspect-ratio version, this means we have to have a scaling factor, x offset, and y offset (in addition to image link, author name, and license text). Since I didn't want to do this gigantic ton of extra code twice (see, allowing different aspect ratios for images would require a scaling factor/X and Y offset for the 2:3 and also for the 1:1 and God forbid somebody wants a 3:2 portrait version!) -- I just said, you know, whatever, I will just have there be a single set of params for 1:1 images -- then on the main page for the Signpost we can just have text overlaying the bottom part of it so that, in actual practice, we are still dealing with a 3:2 or 16:9 or other landscape-ratio image (because this is what all of the news sites I looked at seem to have, and it looks pretty good) but we just have the lower part of the image less visible because of the overlay. Anyway, this was the idea at the time -- you can see the implementation is not perfect. jp×g🗯️ 06:51, 13 May 2024 (UTC)Reply
