Hacker News Reader

Our org uses both Jira Server & Confluence Server - and I’m fairly sure this will make us move away from Atlassian’s products.

The split between cloud and server offerings is partly to blame here.

My thoughts go out to the businesses that have built plugins for Jira and had to endure the variances between cloud and server, or worse, only catered for the server market.

The group at tempo timesheets[1] particularly come to mind here.

Thinking ahead - I’m hoping that an org like GitLab (who have got their Saas v On-Prem offerings balance right) is able to keep building on and catering for this space.

Their planning features are not there yet, but there’s a great group of people with oodles of traction and a roadmap that aligns with this problem space[2].

[1] https://www.tempo.io

[2] https://about.gitlab.com/direction/maturity/#plan

> I’m hoping that an org like GitLab is able to keep building on and catering for this space.

That's not really an option for organizations whose main product isn't code, no?

Yeah, Jira is the standard issue tracker precisely because it works on all levels of the organization, not just the "code monkey" ones.

We (GitLab) have strong planning features but are missing workflow enforcement. GitLab also has a wiki but it is less user friendly than Confluence.

What we mainly need is cheap/free non-coding users. I'm really not comfortable paying for a gitlab seat per support staff we hire.

Thanks for the thought, I have forwarded it into an issue discussing a similar topic on issue management for specific groups not counting towards the license limit: https://gitlab.com/gitlab-org/gitlab/-/issues/2105#note_4326...

> If I was a Microsoft or another enterprise tech company I’d hire a thousand engineers tomorrow to develop a Jira/Confluence competitor before the grace period for server licenses ends.

Azure DevOps Server, née Visual Studio Team System, already exists.

TFS is the self-hosted predecessor to Azure DevOps and it absolutely sucks compared to JIRA.

>"absolutely sucks compared to JIRA"

Congratulations, according to Google you are the first person on the internet ever to use that phrase. Indeed only the second ever to string the last four of those words together.

Let me be the second person then. Jira is way ahead of TFS and I worked extensively with TFS in my last job. That could just be me though.

This quote from Bjarne Stroustroup is very relevant:

> There are two kinds of programming languages: the ones people complain about and the ones nobody uses.

JIRA is the C++ of project management software. It works. It has worked forever. It is easy to shoot yourself in the foot but a savvy operator can't be beat. It's unwieldy at first but gives you all the power to build a really sophisticated and efficient process.

"better than jira" returns 6210 results, too :)

How can you be worse than JIRA? Thats almost impressive.

My understanding of Jira is that it by itself does not suck but it bends over backwards for asinine “business processes” so it is difficult to find an implementation in the wild that does not suck.

I know a little about the inside of the jira sausage factory from second hand reports. It's got the "accrete everyone else's business processes" problem that you mentioned, but it's also got the problem that it's quite spaghetti on the inside as well, probably for the usual reason - startup founders writing lots of code in a hurry and conflating the purpose of the software with its architecture due to being too close to the problem. Compound that with the usual startup-in-growth problem of more people than necessary (with presumably very uneven capabilities) working on the codebase and you've got a recipe for a maintenance problem.

I do know that getting put in the jira team is one of the least fun parts of working for atlassian.

I’ve heard this from multiple sources as well. IMHO, atlassian has had lots of opportunity to fix this... they’ve just been too focused on the next shiny object to do the less fun work of refactoring and redesigning the engine layers into something more sustainable.

JIRA is essentially a platform. It is unopinionated and infinitely configurable. This leaves it open to abuse and many orgs will absolutely abuse it. Other products that are popular do very little by comparison. That lets them build a really tight and simple UI but any org with more than vanilla requirements will hit a wall.

Quite easy, I have am been around for 30+ years in IT and have yet to find something that beats Jira in all its integration options, possibilities of formating tickets information, and flexibility to definie project specific workflows.

Given the option, I will always push for Jira, and associated Atlassian products.

> have yet to find something that beats Jira in all its integration options, possibilities of formating tickets information, and flexibility to definie project specific workflows

I think the reason you like it is exactly the reason a lot of people hate it.

possibilities of formating tickets information, and flexibility to definie project specific workflows.

JIRA in its default state is fairly sane. People who hate it, usually with a white-hot burning passion, really hate the ham-fisted customisations their own organisation has done to it.

Indeed. Compare Jira with ServiceNow and it is like night and day. They serve different purposes, but Jira is friendly and SN almost hostile.

They're both tools to build forms. How horrible the forms are is really up to your organization. Whether they force you to fill 50 mandatory fields or not.

You would think that, until you try to use ServiceNow.

You know ServiceNow is the de-facto standard in ITSM, don't you?

This really does not say any thing outside you have 30 years+ in IT. What did Jira fix? I find what you said fully false given the context of the statements.

Why has this had adds playing every time I turn my radio on for the last 2 years as well..?

It means I have seen a lot in what concerns bug tracking and planning systems, from home grown cgi scripts with a basic form submit to enterprise deployment platforms like DOORS and ClearQuest, with dedicated IT support teams to just keep them running, and anything else that you can think of in between.

What do you think is better than Jira, and why?

I'm not the previous commenter, but I also have a similar 30+ years experience, and I've used quite a lot of project management software. I've seen Jira used poorly, and used quite efficiently for managing projects. My current job uses Jira and nobody in the company has ever expressed anything negative about Jira, and it's become quite nice to use. But again, I have seen some companies absolutely fail at Jira. YMMV.

What's wrong with JIRA? If it is about the usage of it within some orgs, I don't know if it is the tool or culture.

I have seen some people fawning over tools like Asana but in my usage I just found it fancier and slower than Jira.

Personally, as a developer I find GitHub issues to be good enough.

I feel like I've had two experiences with it. Probably 10 years ago I was at a medium sized facility and we migrated 10 different ticketing systems to Jira hosted on-site. It was a little awkward, not quite lightweight, but served all the needs and it was amazing to move tickets between departments while maintaining history and the customization of presentation and fields. Confluence was a much nicer wiki system compared to others at the time and it was nice to integrate Jira/Confluence. Search and Filters are something many other systems suck that for both wiki and ticketing.

I left that job and for awhile working at places using different ticketing systems, hearing complaints on places like HN about Jira, and wishing for features those other systems were missing. A few years ago I started at a place that used the cloud hosted Jira at a much smaller company (with an enthusiastic lead who likely tweaked the setup himself). It was a much, much worse experience. It was horribly slow and I guess they had focused on "management" features that made a normal ticket process annoying and confusing (it may have been how it was configured). Reading around I think the horrible performance was because of scaling architecture decisions that Atlassian made, which I heard they've at least acknowledged and are addressing. I gasped when I saw this headline they plan to drop local hosting before getting cloud performance issues behind them.

Man Asana is a real mystery. It has fewer features and worse UX than JIRA. I constantly have trouble clicking on the right thing or having some auto action just do something unexpected. It's coasting on having a nice stylesheet.

Oh, it is very possible. Look at most ITIL software, although it might be hard, they usually don't have screenshots on their websites so the only way to see the product is in the company of a sales person.

I wish we had something even half as good as Jira.

People are spoiled.

JIRA sucks, but it sucks compared to more lightweight tools and the modern aesthetic

It outperforms previous tracking tools like bugzilla, trac, or those proprietary piles of crap like the "Rational" tools

Can you give me some examples of tools that are better? Genuinely curious.

I've used JIRA at multiple workplaces and really enjoy it. I like Github issues and Trello for my personal task/issue tracking but I don't think they're suitable for more complex projects.

I don't think it's possible to say that one project management project is universally better than anything else. What a team needs out of project management tools varies immensely.

In this thread there's people saying that inability to host on prem is an absolute blocker for them, for others they have zero problems with their data being SaaS-hosted. Some people need deep analytics to measure how their team is doing and they can meet deliverables, and some people could care less. Some people have 10+ workflow stages with tons of rules, some people are fine with Todo / in progress / done.

If you have a small team and dont really care about reporting, Trello is probably a great option for you.

If you want reporting, but are willing to work with a somewhat prescribed workflow / process, across a few teams at most, clubhouse or pivotal tracker are great.

If you need full control and flexibility, and are willing to give up some ease of use and speed, JIRA probably is your best bet.

ClearQuest wants a word.

Now that brings back some bad memories from around 2001.

HP Quality Center.

OMG.

In the place I worked at with quality there was a multi page document on how to file a ticket. It was a nightmare.

Polarion clears it's throat...

Thanks. So true. How a UI can damage a product is never as obvious as with ticket management systems.

This disagrees with my experience in both systems, especially if you’re using TFS for your build system, too and Visual Studio for programming. TFS was far better.

If I was going to switch anyway, I wouldn't switch to Azure DevOps Server, I'd switch to GitHub Enterprise. I don't think there's much of a future for Azure DevOps.

> I don't think there's much of a future for Azure DevOps.

This is just... wow.

Every Microsoft shop I've ever worked at defaults to Azure. You need a really good reason to use GCP or AWS for anything, and it's an uphill battle. They're already paying licensing fees to Microsoft, there are incentives and discounts to stay within the ecosystem. My current employer has GitHub Enterprise and we still use DevOps for a lot of stuff, especially anything that business stakeholders have to touch.

> Every Microsoft shop I've ever worked at defaults to Azure. You need a really good reason to use GCP or AWS for anything, and it's an uphill battle

(1) Github Enterprise, the alternative suggested, is a Microsoft product (Github is Microsoft), not GCP or AWS. Given the way Microsoft seems to be investing and promoting products, seeing Github Enterprise as having more legs than Azure DevOps is completely reasonable. (And I'm the one who introduced Azure DevOps into the discussion.)

(2) I work for what has historically been an enterprise Microsoft shop; our cloud transition, except for some use of Github and Azure AD, been primarily AWS-based.

> been primarily AWS-based.

Same here for the MS shops I work with; actually no Azure at all.

I think GP is more referring to the fact that GitHub Actions et al are a better reimplementation of Azure DevOps. I suspect Azure DevOps will be deprecated in favour of GitHub.

This shouldn’t be downvoted, I’m certain it’s true.

If you talk to a Microsoft rep, they “recommend github enterprise for new projects”.

Microsoft obviously cares about backwards compatibility, and there is no way to currently migrate your stuff.

So... it’s not going to happen tomorrow, but it is going to happen.

...and no, it doesn’t have the nice clickops UI from devops: but that’s legacy now in favour of the multistage yaml pipelines.

You can close your eyes and cross your fingers and say Azure Devops will Be around forever... but, you’d be wrong.

I work in a Microsoft shop and we're all in on AWS. Seen a few transition from Azure to AWS also.

I built a pipeline for testing purposes / for an upcoming presentation with github actions.

The amount of bugs i encountered is.. Bad. From not starting build agents, to stopped agents to hanging etc etc etc..

That said: i looooove the docker container concept for builders :)

Azure devops was the first to offer Windows based self-hosted runners - if you’re stuck in the drudgery of building Windows software, because your customers only run Windows servers, it was pretty much the only modern devops option for a while. I don’t think Devops is going anywhere anytime soon, I’ve heard utilization is quite high.

One of my customers uses Azure DevOps Server, self hosted. And it's missing a lot of features that the cloud version offers.

I think this is on purpose to push people to the clould offering.

It already exists. JetBrains YouTrack has an on-site version with similar features to Jira, and the latest version comes with a wiki feature called Knowledge Base. [1]

There's also an automatic import process from Jira [2].

YouTrack has a cloud version now. Unfortunately they used the introduction of this to raise the prices for their self-hosted version. However they do allow migration both to and from the cloud version, and they plan to introduce a self-hosted version of their newest Space product, so it seems JetBrains is more DIY friendly than Atlassian.

JB seem like a better company anyway. It's hard to forget Atlassian's decision to stop rewarding employees based on merit, a system they claimed was created by "white men" [3]. The new system rewards employees largely for following the "company values". Under the company values page [4], they claim one of their values is "Don’t #@!% the customer". Interesting way to show it.

[1] https://www.jetbrains.com/help/youtrack/standalone/Getting-S...

[2] https://www.jetbrains.com/help/youtrack/standalone/importing...

[3] https://www.news.com.au/finance/work/at-work/atlassian-ditch...

[4] https://www.atlassian.com/company/values

In our case we're ditching our office and that probably means we'll be using the Atlassian cloud instance (since I imagine our server racks will be scrapped eventually), which might be a trend they're seeing altogether.

I agree. this is a jump-the-shark move.

I've worked with confluence and jira and they are not as good as other products - but we can put any kind of confidential data in them. Everybody likes sharepoint/google docs/slack way better.

I feel like SharePoint and Google Docs are fundamentally different products that Confluence.

Looking for a Confluence competitor with permissions, version tracking, templates, and search but don't feel like I have found a good alternative.

I don't think we use version tracking, templates are not really used (but they should be!) and search could be better.

We basically use confluence like a wiki, but it's abnormally hard to put info into it.

Hard in what way?

by spending a lot of time in the confluence editor doing things the confluence way.

Yeah, it can definitely feel like a time-suck sometimes. What "confluence way" things would you do differently? And how?

XWiki is the closest thing I've found to Confluence.

Sharepoint? <spits out glass of water>

ok i take that one back personally, but somehow microsoft stuff keeps getting chosen by the decision makers.

This sounds like a terrible move and will probably force all medium-size organizations I work with to ditch Atlassian.

Me too. How awesome is that! Its absolutely fantastic news.

Edit: This is great for those of us that have to suffer Atlassian products that we didn't choose. It is also good news for companies such as Gitlab that do offer self-hosted solutions.

And what on-premise substitute do you think will be superior? Asking seriously.

I have no answer to that, but if one is forced to go off of on-prem Jira and go into the cloud, it may not follow that people will go to in-cloud Jira. Once an organization is forced to do work, it may be the opportunity re-evaluate things and find something "better" (local or in-cloud) instead of just renewing annually due to inertial.

Our org uses both Jira Server & Confluence Server - and I’m fairly sure this will make us move away from Atlassian’s products

The split between cloud and server offerings is partly to blame here

My thoughts go out to the businesses that have built plugins for Jira and had to endure the variances between cloud and server, or worse, only catered for the server market.

The group at tempo timesheets[1] particularly come to mind here

Thinking ahead - I’m hoping that an org like GitLab (who have got their Saas v On-Prem offerings balance right) is able to keep building on and catering for this space.

Their planning features are not there yet, but there’s a great group of people with oodles of traction and a roadmap that aligns with this problem space[2].

[1] www.tempo.up [2] https://about.gitlab.com/direction/maturity/#plan

Yes, there still is no easy way to migrate Tempo data from Server to Cloud. Their advice is "there's an API - good luck" [1].

Gitlab is great for their developer niche, but the relatively weak issue tracker and wiki lets them down for more general use. Still, Gitlab is a great company with momentum and a roadmap. Feel free to join us [2] in looking at such alternatives.

[1] https://tempo-io.atlassian.net/wiki/spaces/THC/pages/8969584...

[2] https://www.goodbyeserver.org

Hi, Developer Evangelist at GitLab here.

"Relatively weak issue tracker and wiki" got me interested - mind elaborating in more detail what brings you to this conclusion? :)

Thanks!

Please stop doing this. Fishing for product insight on such broad opinions in an open forum does a disservice for Gitlab the product and the team. If it was a natural question... But this is so canned, every gitlab employee here does it, sounds flaky and impersonal. Almost troll-like, maybe even arrogant tbh.

You really don't know why Gitlab is not an issue tracker like Jira is? I'm sure you do, or at least you should know. Being an issue tracker like Jira is not even a bad thing! So why try to look like you think you can be as bad as Jira?

/rant (from a Gitlab fan)

Hi,

sorry that it came around this way. I was really interested in your honest opinion, as my personal experience with Jira is limited at this point with 7 months into my new role. I have been asked about it during past GitLab trainings in my old job, but never used it in production myself. May sound weird, but I learn the most from users sharing their experiences :) Hence my question, it would help my research.

https://github.com/oldboyxx/jira_clone

Not just "medium-sized".

We're really larger than "medium-sized", but we're not the kind of company that can or will force everyone to use Jira.

Jira adoption could have happened, but it would only have been possible in a bottom-up, team-by-team basis. And certainly not with a $40000 starting price tag.

As it is, yeah, we're planning on migrating off Jira now.

> If I was a Microsoft or another enterprise tech company I’d hire a thousand engineers tomorrow to develop a Jira/Confluence competitor before the grace period for server licenses ends.

Microsoft already has GitHub that works fine as a Jira/Confluence substitute for technical people. Maybe this will be their call to make it more accessable.

> Microsoft already has GitHub that works fine as a Jira/Confluence substitute for technical people.

But if you want to replace Confluence as the knowledge base for the business people, there's Sharepoint, which is better than massaging your genitals with broken glass, but not by much.

Microsoft also has Azure Devops, which essentially competes with Jira + Teamcity in one. Confluence is covered by either Azure Devops wikis or SharePoint.

Microsoft is busy also pretending OnPrem does not exist, They cancelled all of the OnPrem cert programs, and everything is "Cloud First", with Onprem products being treated as second class with price increases to push people to subscriptions

Microsoft is doing the same with Azure DevOps and big orgs with sensitive data are moving there as well. I don't think you can even have Azure DevOps on premise.

So Atlassian is doing it because they know there is no competitor that would jump in. Because everyone else also goes that direction.

> I don't think you can even have Azure DevOps on premise.

https://azure.microsoft.com/en-us/services/devops/server/

“Previously known as Team Foundation Server (TFS), Azure DevOps Server is a set of collaborative software development tools, hosted on-premises.”

> So Atlassian is doing it because they know there is no competitor that would jump in. Because everyone else also goes that direction.

New Confluence licenses will no longer be available in 2021 and support ends in 2024 (though it seems the Data Centre version will still be available, but it's expensive). That's potentially 3-4 years lead time to get something going. Either totally on-prem or an either-or option (à lab GitLab).