Why does it take so long?

[Sim]

Even Microsoft with monolithic release cycles ships on a known / predictable schedule at least monthly.

You're seriously comparing a small company with a few employees to one of the largest software companies in the world?

The other thing not helping XenForo is lack of transparency

Look at it from their perspective. If they publish some kind of roadmap with features to be released and promise to have it released by a certain date - but technical issues (eg needing to find and implement a new editor) or just the realities of coding a massively complex system, prevent them from releasing on schedule - then everyone crucifies them for not adhering to their release promises.

The pressure to release something sooner inevitably leads to poor quality - and then they get crucified for releasing low quality software.

Estimating software development time is a very complex and invariably inaccurate task. Just ask anyone who actually works in the industry.

things go dark for a couple months, then development happens, and repeat; no one knows anything beyond what happens real-time, any sense of long-term direction or goal is undefined.

That would be because they are busy a) working on new features and b) supporting the existing software, fixing bugs. If you keep track of the bug report forum you'll see how many bugs have been closed recently - we'll be seeing another patch release soon.

As far as long-term direction, they've already announced 2.4 and we have a very clear understanding of what's coming in 3.0 - what more do you want? Version 4, 5 and 6 mapped out for you with precise dates? It doesn't work that way.

I only see Jeremy or Chris interact with the community at this point, which is scary, given some of us are building our businesses on this software for commercial purposes.

That's a commercial risk you've chosen to take. That's on you, not on them. Have you paid for some level of enterprise support package that guarantees any kind of response or performance? Have you negotiated for them to support your business? Do you know how much I would charge my business clients for promising software support for things that I write for them with service level guarantees? Hint: it's not $75 per year.

I also run several businesses based on this software, but I do so with the confidence that the software I'm using is well built and not full of bugs or rushed to release to fulfil some promise to stop people complaining.

Whether or not he is part of XenForo doesn't matter

Seriously? Expecting someone to remain at a company just to keep you happy is not employment or even volunteering - it's indentured servitude. People have to right to do what they want and if someone chooses to move on to other work or if life gets in the way or even if they've just had enough of the complaining - then who are you to demand that they stay? What right do you have to demand that XenForo insist that someone keeps posting just because it "doesn't look good"?

Things change - people move on - life happens.

Even seeing people post a photo of their garden is an update that builds a sense of a community is noticed and makes things feel like they are still being supported and there is energy behind what is going on. There is literally nothing from XF being shared now.

The last post by @Chris D was around 2 hours ago.

I count 54 bug reports being marked "resolved" in the past 3 weeks or so.

That's not nothing. That's called them working their butts off and keeping the software functioning. I'd much rather they do that than drop all support work just to release something or post pictures of their garden.

I think you need to take a deep breath and gain some perspective.

Seriously - your post says a lot more about you than it does about the developers.

 

[stromb0li]

You're seriously comparing a small company with a few employees to one of the largest software companies in the world?

The point was large companies have a ton of red tape, which leads to slow release cycles. Even many open source solutions that don't have any paid staff have established release cycles. This ties into the point that current releases are mixing major, minor, and patch releases into 3 months cycles is why the perception of slow release cycles come up.

Estimating software development time is a very complex and invariably inaccurate task. Just ask anyone who actually works in the industry.

You are mixing the release of Major and Minor features into a patch releases. Patches are to addresses issues not introduce new functionality. There is no harm in establishing a predictable release cycle. If there is nothing to release, then that's great news. But if there are changes that are completed that improve the quality of the product, I don't understand why it wouldn't be ok to release those on a predictable schedule.

As far as long-term direction, they've already announced 2.4 and we have a very clear understanding of what's coming in 3.0 - what more do you want? Version 4, 5 and 6 mapped out for you with precise dates? It doesn't work that way.

2.4 was announced to be the editor change, but seems like it is really what 3.0 was planned to be. The only announcement for 3.0 is a style change. I think this goes back to the original question of what classifies as a major, minor, and patch release. The minor releases are really major releases IMO, which is why the development cycle feels slow (not saying that is bad, but I think that leads back to OPs response on why things take longer).

That's a commercial risk you've chosen to take. That's on you, not on them. Have you paid for some level of enterprise support package that guarantees any kind of response or performance?

You are correct that the terms do not state a specific SLA, however this is paid software and a license grants access to support. I'm not saying support is bad by any means, Jeremy and Chris have been awesome developer advocates. But, if I submit a bug on the latest stable release and it is never addressed to even confirm if it is something simple as user error I guess I don't understand that. I'm using a stable release of their software and have run into financial loss using the features in 2.3, which is said to be stable (look at PayPal). I don't expect software to be bug free, but if you can't open a case and even get confirmation if the issue is user error or a bug, and it's been over 3 months, am I in the wrong to leak out a little bit of venting?

I think you need to take a deep breath and gain some perspective.

I'm not mad by any means, I only responded with an opinion of why I agreed with OP, instead you made personal assumptions of what my career is, my development experience, and why I can't express an opinion.

 

[Sim]

2.4 was announced to be the editor change, but seems like it is really what 3.0 was planned to be

They've already announced that 2.4 will be the new editor plus quality-of-life enhancements.

The only announcement for 3.0 is a style change

The style change will have a huge impact and will require pretty much all customised themes and the majority of addons to be updated to work with it - so that's a pretty big deal. That definitely needs to be a major release - and even if there's not much else new in that release, they can start adding new functionality as of 3.1.0

I think this goes back to the original question of what classifies as a major, minor, and patch release. The minor releases are really major releases IMO, which is why the development cycle feels slow

Standard semantic versioning has:

Major release (X.y.z) - major new functionality, potentially breaking changes, possibly rewrites of underlying code

Minor release (x.Y.z) - new functionality, ideally no breaking changes

Patch (x.y.Z) - bug fixes, no new functionality, no breaking changes

There's nothing wrong with the release cycle they've been on - minor release versions can have new functionality - basically you can throw anything in that doesn't introduce backwards compatibility issues.

FWIW, I wasn't that impressed with the breaking changes introduced in 2.3.0 (mostly wide-spread changes to the templates which broke many template modifications, plus the change from Guzzle v6 to v7) - I think those changes should have been released as 3.0.0 - but that would have meant either delaying the release until the new theme was ready - or else releasing the new theme as 4.0.0 ... and two major releases that close together can be problematic. Also, 2.3.0 wasn't a "major" release of new functionality - so most non-technical people would no doubt have been disappointed with the lack of new features befitting a 3.0.0 release - or if they had waited until the new editor or new theme was ready - they would have complained about how long it's taking to release new versions, so I get why they released it as 2.3.0. Either way, It is largely "semantic" what version you call it and the changes needed to be released at some point - so the backwards compatibility issues were going to need to be addressed at some point in any case.

if I submit a bug on the latest stable release and it is never addressed to even confirm if it is something simple as user error I guess I don't understand that. I'm using a stable release of their software and have run into financial loss using the features in 2.3, which is said to be stable

Have you submitted a support request or just a bug report?

Customer area login

xenforo.com

 

[stromb0li]

There's nothing wrong with the release cycle they've been on - minor release versions can have new functionality - basically you can throw anything in that doesn't introduce backwards compatibility issues.

This is where I respectfully disagree with you. In November, a significant number of bugs were fixed that could have been released as a patch to address some of the current problems in the latest stable release, including one I reported myself. The release date for 2.3.5 is unknown—it could be tomorrow or in a month; it's speculative to say that just because a new batch of bugs were solved, that a new release is around the corner. I had the same thoughts after observing the bug fixes in late November.

Going back to the original question about why releases take so long, this is why I believe the development cycles are perceived as long. As I mentioned earlier, major and minor releases are different from patch releases. Your response supports my point—there’s currently a mixture of major, minor, and patch releases in each update, which is why we see such significant changes in release cycles between versions.

Have you submitted a support request or just a bug report?

Originally opened via bug, but I needed help urgently due to some external factors going on. The response was then that it would be answered in the forum and to wait. Bumped up the post after a month since no response was made and was told bumping was not allowed. Since then, it's still unanswered after 3 months. It's the only "support" ticket I opened as a customer in all of 2024. Given no response, our workaround was to use a variant of the feature, which goes against security best practices for our use-case.

 

[Chris D]

The ticket is closed and will likely not receive a response. You posted it to direct us to the bug report which was already there and we were already aware of.

Once the bug has been triaged, it will either be confirmed, fixed, or we’ll ask for more information.

I have directed some resources towards it to see if we can get a response to the bug report by 2.3.5.