Given how crufty Wordpress is, I don't even dare to imagine how bad the Tumblr backend must be that this is seen as an improvement by the developers.
this post was submitted on 29 Aug 2024
198 points (97.6% liked)
Fediverse
28248 readers
756 users here now
A community to talk about the Fediverse and all it's related services using ActivityPub (Mastodon, Lemmy, KBin, etc).
If you wanted to get help with moderating your own community then head over to !moderators@lemmy.world!
Rules
- Posts must be on topic.
- Be respectful of others.
- Cite the sources used for graphs and other statistics.
- Follow the general Lemmy.world rules.
Learn more at these websites: Join The Fediverse Wiki, Fediverse.info, Wikipedia Page, The Federation Info (Stats), FediDB (Stats), Sub Rehab (Reddit Migration), Search Lemmy
founded 1 year ago
MODERATORS
Right? At this point I'm just sticking with WordPress because I can't be bothered to migrate a bunch of sites off of it. Every year for the past decade it's felt jankier. Tumblr's backend has to be a dumpster fire for this to seem like a good idea.
My criticism aside, WP still has the convenience factor of being the open source web platform that has a plugin for just about any need. Whether those plugins are gonna break for site or introduce interesting new vulnerabilities is a different discussion.
Same boat here. I had some good times with it but these days it seems to be a bloated mess. Are there any good, lightweight alternatives these days?
Iโve been looking off and on for a few months, but it seems like there arenโt many options anymore like there were 20 years ago. A couple Iโve found are FlatPress and WriteFreely, but I havenโt tried any yet.
Depends on what exactly you want to host. If you want commercially-hosted stuff, I'd stick with wordpress or whatever your host offers, but if you're selfhosting I'd look in !selfhosted@lemmy.world or https://github.com/awesome-selfhosted/awesome-selfhosted?tab=readme-ov-file#blogging-platforms.
I suppose what I'm looking for is a lightweight, multi-user CMS, with support for both static pages and a blog. If the blog could support (at least one-way) federation that'd be a bonus. It should ideally be built to work with both desktop and mobile devices (so that I can customise the look rather than build it from scratch).
It's something I could build from scratch but if I can do it then I'm sure lots of more skilled people have done it better!
When I dug into this for myself I landed on Ghost!
I suppose at some point I should learn Node.js and other JS-related stuff. I speak vanilla JS but I've not really touched frameworks. Anyway, thank you for the recommendation.
Wordpress is all of those except lightweight, though I wouldn't really say it's a bear to manage either. I believe they have initial activitypub support as well.
You can check the selfhosted list for alternatives, but I don't think I've seen one that would be a better fit.
I mostly find the design of WP clunky as all hell. I'd like to add some features to my site and doing so feels tremendously awkward. Learning how to implement stuff in their way of doing things doesn't feel worthwhile to me, I guess.
There are Fediverse blog platforms but, as this is about Tumblr, what about a Fediverse tumbleblog? You've got:
You're not alone, I've still got clients with WP sites and it feels more and more patchworky every time I use it. The vulnerabilities may keep me up at night, but it would take a ton of effort to move them over, and my clients certainly don't want to pay for that.
I see this all the time. People complain about WP but think the alternatives are better, when they're just trading problems for others.
WP core is stable AF. I've shared in many prior comments how I spend so many more dev hours fixing other CMSes over WP.
And if you don't even need a CMS, fuck it all and switch to static hosting and markdown.
Oh yeah, for sure. Joomla still haunts my dreams.
All of my own sites are static because it's easy for me to modify. But my clients need something a bit more user friendly, unfortunately.
This smells to me like WordPress reducing their workload more than anything since they own Tumblr (unless maybe there's some sort of financial incentive to increasing the number of WordPress blogs?).
But also, considering that at one point in Tumblr's history, you could edit other people's posts, maybe it is an improvement.
But also, considering that at one point in Tumblr's history, you could edit other people's posts, maybe it is an improvement.
What ๐ญ
So the way Tumblr works is that your account is basically a blog, with your home page on the site being populated with posts from the accounts that you follow. You can reblog posts onto your own account and comment on them to create individual conversation threads like this one. At one point, there was a bug in the edit post system that let you edit the entirety of a post when you reblogged it, including what other people had said previously, and even the original post. This would only affect your specific reblog of it, of course, but you could edit a post to say something completely different from the original and create a completely unrelated comment chain.
John green sends his regards
I don't remember that. He had his posts edited, or he edited someone else's posts?
Doesn't he have a funny username, like "FishingBoatProceeds" or something?
It makes sense.
Supporting Tumblr backend with patches vs building on top of stable WP and improving it seems like a win win.
Tumblr's backend has been passed between several companies* for several decades, it's a miracle it still works and can be updated
*some of which don't even know what tumblr is
Could you elaborate?
Wordpress is a pile of decades old php code* that is held together with string and tape pretty much.
*php isn't the problem itself, modern php is actually pretty nice.
I'll take your word for it. Apparently tumblr and WordPress (and WordPress.com) are owned by the same company, so this change would make sense to reduce maintenance workload.
WordPress core is pretty wild. But modern WordPress isn't working purely in that. The latest WP uses PHP primarily as a backend, and modern JS as a frontend and passing data through filters->DB.
I won't call it elegant. But it's not the PHP experience from five years ago.
And despite all that it's likely that the Tumblr backend might be even worse.
Remember when Verizon paid a billion dollars to ruin Tumblr and get a fraction of that back for it?
Ruin it by removing porn?
This could indeed potentially lead to a future with tumblr on the fediverse. I don't see why not ๐คท
Give it to meee
Tumblr announced they were gonna integrate with ActivityPub years ago. Itโs been silence on the topic since then, hopefully this move to Wordpress is a step towards joining the Fediverse.
The link to the original TechCrunch article doesnโt seem to be working, but hereโs a mirror on Slashdot: https://m.slashdot.org/story/407458
The question is whether or not Tumblr users would want such a thing.
I feel like the same thing will happen like when WordPress introduced a (bad) TikTok/streaming clone called Tumblr Live. I think less than 10% of the userbase ever interacted with it, most of the community openly hated it, and the people who did use it largely didn't use Tumblr themselves.
I could see Tumblr users actively finding a way to defederate their blogs from everything Fediverse related.
Wordpress federation is pretty one directional, you can follow a blog from mastodon, but you can't use your blog to follow other people.
I understood you could, with the friends plugin:
If you also have the ActivityPub plugin installed, you can follow people on Mastodon and other ActivityPub-compatible social networks.
Interesting, I didn't realise that.
However I assume they will be migrating them to wordpress.com, which is their proprietary hosted solution, as opposed to wordpress.org, which is the open source software. Plugins don't work on wordpress.com free accounts, only paid ones. I believe outward federation is integrated into .com though.
they're just moving the tumblr backend into the WordPress software. No cause for WordPress.com to be involved
I guess this means Tumblr is coming to the fediverse?
There has been some signs and rumours to that effect since Mastodon started taking off; I believe Automatttic specifically advertised for ActivityPub developers to work on Tumblr?
But Tumblr moving on to a Wordpress base only opens for the option of federation, if that is even accessible in the Wordpress installation that will drive Tumblr โ as I understand it. Or has the WP-ActivityPub plugin been rolled into core WP without me noticing?
The AP plugin isn't in core but it's developed by Automattic.
True. All I'm saying is, it's not a given that a WPified Tumblr will use it as a default.
Oh boy my tumblr will be now 518% more hackable!
How so?
By being moved to the main Wordpress branch, where everything has been known to be hackable since 1999, rather than staying in Tumbr's however-modified branch where probably some exploits don't work or have unexpected results.
I believe there is already plans in the timeline to bring Tumblr to the fediverse.