Where have the old posts gone?

Maybe because we are a Free Software community, so we’d like to use Free/Open source tools? :wink:

Askbot works quite well and we hosted it on our infrastructure and initially tried to contribute to the development process as we do in the Free software community. However, the reality was that it was not sustainable with the resources at the disposal of the community—it’d be great if some of you could join in and help out here, of course. So, we moved to Askbot being hosted by Askbot. Here, it worked a lot better, and when Discourse came about, we moved because we felt it fitted our needs better.

Most users like Discourse—it’s a great platform. I do not know why you’ve, in your very first post, labelled it as a “poor tool”.

Um… gonna have to disagree on that one. It was buggy for me when making posts or comments, had a really dated look and feel, and generally was a more frustrating experience than other platforms. My first exposure to Askbot (which as you may have noticed, most people don’t even know what it is) was through Ask Fedora. By extension, this to me also reflected poorly on the state of the Fedora community.

FOSS is great and preferable but I think a focus on usability usually takes priority over ideology. A sad example is the proprietary nvidia driver versus nouveau.

I’ll say it again: I think the move to Discourse is a very good one, because it allows discussion and more engagement than cut and dry questions and answers. The interface and experience are modern and far more enjoyable.

But that isn’t to say there isn’t a place for curt questions and answers. Not everyone wants a discussion or learning exercise when they have a problem to solve. Not to mention, in the age of data, it seems really odd and backwards to let go of such a database built up over the years. It’s understandable that the value of that information may decrease over time and there are also the realities of limited resources. But the willingness to let it suddenly vanish, and break links, seems quite disconnected with the value most people place on those assets and generally on consistency.

Again, if there was a way to migrate the data with our resources, we would’ve done it. We did not decide to let the data vanish on a whim. We worked on moving information to quick docs too. And by we, i mean a set of less than 10 volunteers. Maybe take a minute to read through the work that went into it? Someone shared a link to the pagure tickets at some point—thatd be a start.

I really wish the practicalities of the process would be taken into account when all of this is being discussed. “It should be done” does not imply “it can be done with the available resources” :slightly_frowning_face:

I personally disagree, but thats OK. In my book, if the point is merely to have a working system, a finished product with paid support is the way to go. Youll have a product that is tested on specific hardware and so everything works, and a support forum where professionals will fix bugs for you ASAP.

The whole point of FOSS is to the ideology. The software, the communities, are all tools to achieve it.

1 Like

I sympathize with this important detail. I think most people would be understanding too.

The frustration is with what appears to be a downplaying of the importance of the Q&A resources and link consistency by the decision makers.

The phenomenon is quite measurable and observable, rather than personal.

This is where things get philosophical and off topic. Succinctly, there are many pragmatic reasons and arguments in favor of FOSS. I do not see ideology as necessary, even if that is also an element for some.

On the contrary, I would suggest that expressions of ideology can even be counter-productive to their stated goals, since they tend to cause (strong) disagreement.

Tendentious, but arguably the case. I would also say the whole point of the web is the ‘ideology’ of a globally-linked commons of documents and information, to which link-breaking is damaging (& not only to the purely derivative search indices)

Anyway I’ll say no more on this topic: clearly it’s a done deal. Thanks for your engagement, and my apologies if my tone was unduly sharp.

@FranciscoD It looks like the uri structures for askbot and discord sites don’t overlap.

One can setup a permanent (301) redirect by regexp from ask.fedoraproject.org/en/question/(.*) to askbot.fedoraproject.org/en/question/$1 and this won’t affect the discord site.

This way old links will be preserved, and reindexing in search engines also going to be much easier.

2 Likes

A ticket was filed here about this already. Please take a look there to see what the infrastructure team said: https://pagure.io/fedora-infrastructure/issue/7759

(Another case of “should be done” but “cannot be done” from the looks of it).

https://docs.fedoraproject.org/en-US/project/#_what_is_fedora_all_about

We are dedicated to free software and content.

Advancing software and content freedom is a central community goal, which we accomplish through the software and content we promote. We choose free alternatives to proprietary code and content and limit the effects of proprietary or patent encumbered code on the Project.

Sometimes this goal prevents us from taking the easy way out by including proprietary or patent encumbered software in Fedora. But by concentrating on the free software and content we provide and promote, the end result is that we are able to provide:

  • releases that are predictable and 100% legally redistributable for everyone;
  • innovation in free and open source software that can equal or exceed closed source or proprietary solutions;
  • and, a completely free project that anyone can emulate or copy in whole or in part for their own purposes.
1 Like

The community has multiple channels of support, as listed here: Communicating and getting help - Fedora Project Wiki

We have 40+ channels on IRC, many of which are bridged now to Telegram. We have an equal if not more number of mailing lists (powered by hyperkitty so one can use the web application to participate). We have presence on Social media where people ask queries. At the end, we also have fedoraforum and Askfedora along with the many regional support web-applications. All of these have dedicated channels for Q&A that users can use.

So, pardon me but I don’t understand where the impression that we’re downlplaying it comes from.

If anything, we struggle because we constantly try to have channels everywhere and that spreads us rather thin. So I would encourage everyone here to please be present on multiple channels to help the sub-communities there.

1 Like

Maybe I am missing smth, but from

Fedora Infrastructure does not have a way to do either of these tasks. We do not run either service and have no way to put in redirects in either place.

it looks like Fedora Infra team doesn’t manage the ask site. Thus it cannot be done by them, but it doesn’t mean it cannot be done at all.

You still run the ask site behind smth, which resolves https certificates. Who manages nginx- or the whatever-proxy config for it? How is it deployed, Openshift?

1 Like

Discourse hosts it for us. We can ask them, but our business plan seems to have limited features.

It’s a real shame that Ask Fedora decided to delete every user contribution. Some of us had put a lot of time and effort into that! The community has lost a great resource.

Could an admin please get in touch with the Internet Archive to archive everything on the old URLs so there will be a way to still get to them in the future. The old questions are linked to form all over the web.

1 Like

Have you looked to see if they’re there now?

In any case, I have an additional suggestion, half-baked because I don’t know how feasible it is, but it has potential to be a good user experience. Could the 404 handler be amended to check the IA for the nonexistent url automatically, and if it’s found, offer this up to the user as a clickable link?

[ Edit: I just checked a random couple of old ask urls, and they were already in the archive. There’s also this which may be of use: https://blog.archive.org/2013/10/24/web-archive-404-handler-for-webmasters/ ]

1 Like

Please read the replies on this post to learn of the reasons, and current status.

@FranciscoD Now it explains a lot…

I would still suggest contact discourse team and get it fixed. They have some tools to do it, for example permalinks https://meta.discourse.org/t/custom-permalinks/56867 Even though these tools are a bit limited, they can at least suggest something.

At the very list, all /questions/ url could be redirected to a post which explains how to find the old data on the askbot-site.

2 Likes

It’s on our list, will hopefully have the time to investigate it in detail soonish.

This is absolutely, 100%, without a doubt, what must be done, and fast. Host Discourse on a dedicated system somewhere if necessary to get it done. There is massive amount of critical content that must be preserved! Example: This morning I couldn’t boot my Fedora system due to a broken initramfs. I needed access to content on the old site to find the solution and ended up at a 404 error on this site with no way to get to the content on the “new” URL.

Sadly, it seems the transition was not planned out well, and the Fedora project is going to suffer irrepubable harm as a result. People will surely leave for other distributions because they can’t find the answers they need. These important links are already falling outof search indexes because the URL’s are now all 404 errors. Just ask the W3C, the leader of the World Wide Web standard: “Cool URI’s don’t change.”

https://www.w3.org/Provider/Style/URI

Did you ask a question about your query, which is what this platform is for?

Thanks for your comments. The transition was planned out—please read the other posts to see the limitations that needed to be dealt with.

Or they’ll ask new questions, get them solved, and happily proceed with their lives :slight_smile:

The important links are on fedora quick-docs. The rest are individual cases which while do apply to multiple users sometimes, tend to rely heavily on softawre and fedora versions. No solutions that were provided for Fedora 29 or earlier versions are guaranteed to work for Fedora 30 and future versions. So please, you, like everyone else, are better off asking a fresh question.

This is the #site-feedback category, so my reply to the topic continued the feedback about the site. My question had already been answered on the old site, so no, I’m not going to ask it again. I just wanted to be able to get to my answer that was already there.

In other words, there were limitations that weren’t dealt with. How exactly is that not lack of planning before executing the transition?

Forcing people to ask new questions, and wait for resposnes, rather than have access to the knowledge already created is presumptuous and rude, as seems to be the tone in all of your replies to the constructive criticism offered here.

People need to solve problems quickly and not wait around, and the tools for them to solve those questions are on the old site that is quickly being wiped off the face of the Internet.

“Important” as judged by whom? And by what process?

Maybe, maybe not. That doesn’t reduce the usefulness of having the old material. In my case, once I was able to dig up my answer on the old site, I just had to edit the kernel version and it worked. Again, such a presumptious attitude suggests poor planning and lack of understand of the use cases for people looking for content on Ask Fedora.

Again, this is the #site-feedback category. Asking new questions aren’t on topic in this category.

1 Like