I have spent the last 14 years hating Jira. It's terrible software. It makes me angry to use, on a regular basis.
Know what though? I've used almost every other option in the last two years. They're all worse. We currently use GitHub Projects. Holy. Crap. It sucks so bad.
This truly is a Devil You Know problem. I challenge the software industry to make something better. Please. Seriously. There's good money to be had but bug and task track is apparently an unsolvable problem, if the goal is to make software people aren't mad about using.
90% of my issues with Jira is that it's slow as balls. That's it. It's a website that cannot maintain any acceptable performance on a development machine. This applies to both UI with dialogs that open painfully slowly, and requests that take forever to arrive _and then_ keep loading megabytes of crap.
My org configured the local Jira to look like an oldschool forum: no dynamically loading content (which also means I need to refresh for newest comments), links are really just plain links, and I need to manually write markup for proper comment formatting.
It's fast and works really well compared to the guest Jiras that we sometimes use for external projects. My only gripe is that Jira's markup language is pretty bad.
Their Cloud ToS forbids sharing performance information (section 3.3i), which in its most generous interpretation is meant to avoid the type of speculation in this thread. https://www.atlassian.com/legal/cloud-terms-of-service
The theme of both your comments is that if anything is a problem with Jira, it’s all the organization’s fault because they have set it up poorly.
I would suggest that it’s 100% Atlassian’s fault that their product is so frequently poorly set up. They’ve basically abandoned their responsibility to design a good product in the name of customizability.
It's not the network that's the bottleneck, but the Javascript-bloated mess of the frontend.
edit: That isn't to say that Jira and Confluence are the only offenders. They aren't even the worst. Slack/Teams/Rocket.Chat are all just as bad and make me consider switching to a career without computers on a regular basis.
That’s kind of my point; all the tech you list here doesn’t have horrible lag, and aren’t generally even close to so bad they’d make someone want to switch careers.
So if it’s not them…
That's a lot of denial. Have you considered that Atlassian knows that these are Fortune 100 companies with contracts that would make Larry Ellison weep and therefore slap them on their biggest instances because hardware cost is negligible (as well as having dedicated employees to fix specific problems for them), and that _maybe_ the same amount of work isn't done for <generic company paying default prices for Jira Cloud>
Maybe!
But I doubt it. I run engineering at a b2b SaaS and specifically dedicating resources the way you're describing is not something I'd consider at the level Atlassian is at, especially considering there are people like you lurking out there.
I’ve wondered if the data model is wrong for bug trackers. They really are all a pain to use in the same ways, just with varying levels of intensity. Can’t find tickets, don’t know where to put some task, it's been overcustomized since the default ticket/story format was unusable, etc.
There’s probably an interesting tool to be made if anyone wants to tackle a new way to organize software products that don’t make me want to gouge my eyes out on a daily basis! Something opinionated but functional for 80% of teams, so it doesn’t get customized. And it’s not just a giant list of tickets.
Yeah, I feel like a lot of hating on Jira is just hating on bug management in general – it’s hard to know what things to work on, to deduplicate bugs, to track when they’re fixed, etc. That said, I do find Jira’s implementation pretty painful. Mostly I dislike that it is all a little bit janky/slow/inconsistent, and that it has a million features and possible ways to use it. But I’ve only used an old version and maybe the newer versions are better.
I agree with a lot of the individual gripes people raise with JIRA, but I've never heard a strong case for not using it to track the progress of a multi-discipline team to deliver projects.
By which I mean, I've heard _dozens_ of cases, but they rarely amount to more than a gripes-fueled gish gallop. Even rarer still is the offer of a functional alternative, and on the rare occasions I have seen one adopted, the same gripes have reared their heads pretty quickly.
"I just want to do X, why won't it let me" is the most common, and it is akin to "works on my machine" for me. JIRA isn't trying to solve for ten sets of workflow preferences, it's trying to give you the minimal set whereby you can reasonably track and deliver. Anything more than that is noise, not signal.
I feel like the strongest issue I have with Jira, and the one that most often trips me up, is that so many features and settings are global, and aren't the minimal set. For a sufficiently large org, you need a high degree of rigor around things like Labels and Epic naming, or you end up with labels of "production" "prod" "Prod" and "PROD" and no one knows which one to use.
It only gets worse when you're working with addons - half of the UI of any ticket is things I have to go hide because one team integrated Figma with Jira and now it shows up on every project and ticket.
The issue is Jira's permission management also seems terrible in the exact way that makes it hard to create that sort of role without making that person also responsible for a myriad of banal options that everyone needs turned on.
Better is subjective. It's less flexible now and it's not perfect. It has arbitrary constraints that make things hard (no cross Jira project tracking, wtf?) but it's as good as anything else and it's familiar.
It’s hard because there’s two very different use cases:
1. Within team, tracking work and assignments the team needs to do. They want to setup internal rituals and expectations of how work gets done on code they own.
2. External stakeholders, they largely care about the larger rocks and outcomes, why they’re not getting done, and overall status.
Everyone tries to do use one tracking tool for both.
One of the things we struggle with is cross repo consistency. Take something like labels. They apply at the repo level and not the project level, so when you add issues from multiple repos, managing labels becomes an administrative task.
View creation is confusing and painful, due to the flexibility (which used to be a common complaint about Jira until they removed the many hundreds of options). People struggle to understand what comma does in the label list vs a space with a new label key/value.
Workflows are also per project and have to be replicated for every project and there's no global default template you can define.
Dependencies are non existent, so order of operations can't be defined in a way that's meaningful. You can order tasks but not indicate "dependency a is not complete and you may move on" in a simple way.
Task lists can't be reordered without editing the markup, which is finicky. If there's a newline it doesn't like, it errors. If there's a character wrong, it errors. Sometimes it just errors and you have to delete the whole thing and recreate it from scratch for it to render again.
There's no "move to top" or "move to bottom," which is necessary when you have long lista of items.
It's improving and if everything is a code, it's probably fine but when you're doing projects, some things are just tasks and it's not well suited for that, in my opinion.
I'm sure it will improve (as of last week you can limit WIP and export views as a CSV) but it feels so much like it's catching up, slowly, that I don't see it ever becoming a "good" alternative to systems that already exist.
I could list more but I'd have to look at my computer and there is already enough here that I'm just pulling from memory.
Not OP but also using GitHub Projects. Up until recently projects was just missing critical features like true epics and first class parent child relationships between issues. These are slowly getting added though. There is an entire product called Zenhub that exists to make GitHub issue tracking more sane. I will say the automation story in GitHub Projects is better than Jira imo
I'd say the automation is better but more complicated.
I feel bad for ZenHub. The new GitHub Projects is good enough that you can stop using ZenHub and not miss much. I think GitHub just killed their business. I love they cross project status automation they built. That's one of their killer features.
They actually have a light wrapper on GitHub actions in the projects view now that may replace a lot of the really common actions. It’s GUI based and right now it can only do very simple things but I could see it being the new way for less techy people to put basic automation in
Thinking about this a bit more (can’t edit above, so replying to myself).
The other annoying thing about the “simple” method is that it’s also outside of your repo. So that configuration isn’t versioned, it isn’t really present anywhere. Your actions code, clunky as it is, is still code.
Simple and code-driven shouldn’t be mutually exclusive.
> but bug and task track is apparently an unsolvable problem
I agree with a specific meaning of this. "Bug and task" is unsolvable. Those are very different things, usually existing for different levels of organising work. Yet, often they're forced into one system.
That solution sucks though: there's a bug, my own task may not may not relate directly to it, the team goal may or may not include that task specifically. I've enjoyed life more since those things got mostly separated for me.
The unit of tracking for how we think of things is some highly overloaded version of "project" which could mean project or team or org or repo or short lived event plan or anything else. Our terminology is probably wrong and it forces the way we think about the problem. It very much feels like a "everything is nail shaped" situation and we want everyone use the same model hammer. Jira, for all it's faults, does offer two or three different ways of doing work (with the recent addition of their DevSecOps thing that I haven't actually looked at).
I got to similar conclusions and I think I figured out why:
The ticketing software is not made to make developer life easier. It is designed to keep the rest of the company involved in the process and it's there to show the manager whatever semi-meaningless metric they want to track this week.
I don't care about the exact type of the issue, or the arbitrary priority of it. "This is security bug that needs to be fixed now" and "this goes next release" is about as much granularity as I need. What's even the difference between "trivial" "low" and "minor" ?
I do care about dependencies but in a "unless X is done don't even show Y in my list" way.
I do not care about tracking time at all
I do not care about assigning a ticket to a "component", I put the ticket number in commit anyway, obviously that already shows which component code I needed to change to fix it
Obviously many of those are useful (or are imagined to be useful) to other roles in project but as developer I just want to see the reported bugs/features and write the code for it.
I have discovered that the software is not really relevant, it's mostly the organization employing it that makes a living hell out of teams productivity. Jira can be configured to have 4 simple task states and no required fields. It can be deployed in a way that it's reasonably fast.
When you have infinite resources and a propensity for NIH, you can do a lot. GUTS is far superior to most ITIL projects but it will never see the light of day. It wasn't always good though. It was Remedy first and it evolved slowly away from it. I can 100% promise you that in 2006, bugs was actually probably the worst bug tracker I'd ever seen. In 2005 it wasn't and then it was. It took a lot to turn it into something useful.
Bugzilla is the best system I've ever used. You need a FTE dedicated to keeping it pristine, though. It's awesome for ICs but management did complain about lack of metrics.
At my day job we use ClickUp and as a contractor for another large company we use Jira. ClickUp is actually awesome. It’s a joy to use, it has more features then we need, it’s extremely flexible when it comes to linking tasks together. It has a document solution that’s very good. I would highly recommend.
I’ve just started using linear for a personal OSS project. It seems nice so far but I don’t have enough experience with it to say for sure.
We were using Trello but it was too simple and went looking for a more advanced tool. We tried ClickUp, it looked awesome, flexible, easy to use and had good range of features, but it was buggy as hell. During testing period we ran into multiple issues that we decided we can not trust it.
What do you not like about GitHub Projects? I haven’t used the new version of it that came out last year, but prior to that I really liked it. Does the new one still have an option for a Kanban interface? I hope it does, as that’s a must for me.
It's hard to say, as Jira is so configurable that some people that dislike Jira don't really dislike Jira itself. They dislike customization their company put in, or didn't put in.
I used to feel this way but the cloud version seems to have gone the opposite direction in recent years. Don't get me wrong, there's still some complicated config you can do but my new gripe is that there are parts that are too limited.
I too would be thrilled for someone to eat Jira’s lunch, and am happy to rattle off a litany of what I dislike about it.
But I understand why it’s used everywhere, and there’s more to that than mere inertia. It’s the least worst option, like democracy and (historically) capitalism. There’s other things out there that are better for a specific subset of users (engineers, management, PMs, whatever), but none that I know of that successfully manage the balance of giving everyone enough of what they need to be serviceable.
Caveats: assuming a semi-traditional software company of moderate size using something Agileish. Places that are more process-innovative or -anarchistic or composed of all technical people or anything like that probably _can_ find a better fit with something else.
I've been using jira-cmd for my personal use and i maintain it. It has saved me a lot of time like shortcut for transition, jira creation templates for default config when creating jira etc
Could you please quote where I shallowly dismissed the CLI tool as it relates to making Jira easier to use? I don't think I've done so, as I haven't talked about the CLI tool in any way, I only ever talked about contributions to the CLI tool.
> Anyone thinking about contributing to this project should keep in mind that it is under Reddits control. You will effectively be donating your time to a company that treats its users and partners very poorly.
This undermines the submission by reducing all tech to "Does Reddit control it? Then it's bad."
Do you notice how the part you quoted doesn't contain me shallowly dismissing the CLI tool as it relates to making Jira easier to use? How it contains a completely different point? You did just switch your complaint (from me dismissing a concept to me undermining the submission), so I think you noticed.
> This undermines the submission by reducing all tech to "Does Reddit control it? Then it's bad."
I also did not reduce it to that, as I didn't state anything about the tech. I only stated something about contributions to the tech. It's fine for me if you want to interpret this in the way you did, but trying to silence me because you don't like what you interpreted into what I wrote is a pretty strange thing to do. Just move on, and if you really want to, downvote me.
Otherwise your comment chain is much more non-productive than my original comment could be, since it doesn't even relate to what I wrote, only to what you think I meant.
Eh, agree to disagree then. In my view, calling out Reddit as the owner and claiming they’re “bad” is a shallow dismissal. You clearly disagree, but that doesn’t make you right.
And “you calling me out is the real problem” just seems not well thought out. So in your view nobody should disagree with you, most of all when you’re acting poorly? That’s convenient.
> In my view, calling out Reddit as the owner and claiming they’re “bad” is a shallow dismissal.
If this were in any way an accurate representation of what I wrote, I might agree with you. But it's another misrepresentation. Could you please stop with that? I did not claim Reddit is bad, and I didn't call out Reddit as the owner to dismiss the project, I specifically and only talked about contributions to the project.
> And “you calling me out is the real problem” just seems not well thought out. So in your view nobody should disagree with you, most of all when you’re acting poorly? That’s convenient.
Aaaand the next mis-representation (if I counted correctly, this is your sixth! Wow!). I did not claim at any point that nobody should disagree with me. But if you disagree with me, please engage with what I write and say, not with what you want me to have said. You keep talking about how I am shallowly dismissing this thing while completely and utterly misunderstanding or misrepresenting my position. This means that this whole discussion chain, as well as every following attempt of yours to misrepresent my position, is off topic. Even if someone were to agree with you that my initial position is a shallow dismissal, they'd have to admit that it's at least on topic.
Just take what I write in good faith instead of repeatedly putting words into my mouth.
Sorry but no, I don't think it was a relevant or insightful comment about the quality of the cli tool or even Jira in general.
One reason I have for thinking this way is that you call me disagreeing with you as "misinterpretation" rather than what it is: someone who has all of the facts, an understanding of the situation, and who is arriving at a different conclusion than you.
It's clearly important to you that I don't understand what you said, rather than disagree with it.
And now we've arrived at you misrepresenting my claims about you misrepresenting my claims. It's clear this is going nowhere and I will stop wasting my time now. In the future, please try to engage with what people write, or don't engage (at least with me). If you read back on our conversation, you'll see that you did not lose a single word about "contributions", which was my whole point the whole time.
Know what though? I've used almost every other option in the last two years. They're all worse. We currently use GitHub Projects. Holy. Crap. It sucks so bad.
This truly is a Devil You Know problem. I challenge the software industry to make something better. Please. Seriously. There's good money to be had but bug and task track is apparently an unsolvable problem, if the goal is to make software people aren't mad about using.