Ask HN: What sucks in the project management software you use?
I'm thinking about building a simple project management tool, since Basecamp did't really work for us.
Is there something you hate in your project management tool?
Is there something you hate in your project management tool?
So, acknowledge that. Wouldn't it be cool if our project management tools actual admitted each user has their own priorities, and just helped us communicate it clearly? Every user gets their own queue. Only that user can modify it. If the user's queue of "up next" tasks is out of sync with what a PM wants, the PM can see that... and they have to go talk to the person.
There's a core mismatch in current tools between purpose and practical use: we let people use the planning tool as a cudgel in place of real communication.
We need to have a shot at fixing that underlying issue. It's this hidden cudgel nature of issue trackers that spawns the more literal problems already mentioned in this thread (e.g. some people always want to add another field to Jira, etc -- which, sidenote, yes, oh my god). Worshipping the false idol of minimalism by stripping the ability to have custom fields won't help. Instead, redirect energy so there's simply no desire to add bureaucratic layers in the first place. If adding more custom fields won't give someone with a micromanagement complex any more power (either illusory or real) to poke your task queue around without communicating, then... why would they make custom fields? Incentives for bad behavior disappear, and so does the bad behavior.
One of the coincidental UX wins of per-user task queues is there's an obvious visual place to highlight when one person has too much on their plate, and we can use patterns in the UI to try to help. Similarly, it gives a FOSS developer a very clear way to say "yes, I care; at the same time, here's absolute transparency on the short list of issues I'm currently moving on." On the other hand, one of the commiserate challenges would be ensuring everyone can navigate to the same places in the same way; breaking that symmetry between how different people navigate the tool makes it much more complex for group use.
I've never seen a tool try this. Might be a bum idea. Definitely a little radical. The UX would be full of new challenges. I'd love to see it though, and would definitely be excited to try a system that thinks out of the box.
I haven't seen it used much, because of the way we work at Pivotal Labs. The standard work method is to pick the next story, bug or chore off the top of the backlog and start it. Not much need for "what's mine?" filtering in that case.
As for talking, I agree. That's why we colocate. I ask my PM by turning around.
It's hard to look straight at a good situation and say "why" the magic is there with any certainty. That said, from this group I've noticed our PMs are very focused on asking developers questions to help keep the issue tracking to reflect their work, and very rarely make any changes until discussing them out-of-band. Doesn't matter if it's in person or over Slack or video Hangouts; it's the pattern of involvement that seems to cause the spark. It's a habit I definitely intend to try to take with me.
I'm not sure user-oriented queues could be completely emulated with view filters. Maybe it's possible, but optional views seem to have limited impact on the overall flow of work unless everyone in the team agrees to use them.
As an aside, the abundance of optional views and filters is one of my biggest beefs with the likes of JIRA. The first thing I do if contracting within a new group that uses JIRA is ask exactly which views and filters the PM uses to navigate the backlog, because otherwise I'm hopelessly lost as far as getting on their wavelength... and this makes customizable view features quite the irony to me! Tracker mostly seems to dodge this bullet since everyone lands looking at the same lanes in the same order by default, which is cool -- that kind of informational inertia in layout and navigation is huge. (That concept of inertia also jives well with per-user queues: someone can freely organize their own tasks in a personally meaningful way without wrecking the inertia another user's view.) The trick is keeping that inertia feature intact when huge amounts of info gets dumped into the system. And I'm totally on board with the countpoint "you just shouldn't have that many tasks, because you're fibbing yourself anyway", but at the same time, I really can't recommend a big enterprise starts just haemorrhaging security tickets from the record entirely if they can't keep their todo list short enough.
Perhaps length-limited per-user queues combined with different mechanisms for "not on anyone's radar today" backlog grooming can find a new sweet spot?
It's one of the most comprehensive tools I've ever used and its also not an unwieldy beast like other tools with a similar breadth of features. I'm probably never going to try and administer Jira for a small team ever again after the grief it brought me but I'm still quite happy with my own personal Phabricator installation, it just works.
Each of its applications was 70% as good as what else is out there (especially UX-wise) and the integrations compared with Github were pitiful.
Stakeholders and lack of/ignorance of responsibility.
It really doesn't matter what piece of software you use as long as everyone uses it the same way. Often, in my experience, the best way of doing this is having a single person in control of (meaning write access to) the project. (okay, maybe everyone can comment).
A lot of businesses use this software instead of bringing on or training actual project managers. Then you have tickets being passed around between teams and stuff gets lost. People don't speak up when there's problems or bury their heads in the sand and try to deflect blame to others. Those same people will use the software, whatever piece of software, to their advantage.
There really always needs to be a captain at the wheel making sure everyone is 100% clear on what their responsibility is and where the blame will fall if things slip. Software can't do this for you, I don't care what its features are. This is 99% of the time the reason why people aren't happy with their project management software.
For larger projects, especially in software development, I usually recommend JIRA. It might seem a bit intimidating and process-heavy but it's very flexible and works well with agile processes, too.
My general recommendation would be keeping it simple for as long as you can (i.e. Trello, GitHub or the like). If your organization has grown beyond those simple means both Basecamp and JIRA are good options (the former being more on the opinionated side, the latter offering a lot more flexibility).
That said, given the myriad of project management options I can't think of any reason to build yet another project management tool. You say Basecamp doesn't exactly work for you. What's so wrong with it that makes you want to develop another tool (Which is no small endeavour by the way. It took 37signals literally a decade to get Basecamp to where it's right now.)?
I tried to leave Jira once and after a 6 month experiment realized how much better Jira was than everything else out there.
But, yeah, at the same time, it's sort of 'too flexible' up front, and requires a lot of decision making (which feels like you can make the 'wrong' decision). I think they've done some better guided setup stuff in smaller chunks in the past couple years, but need to go further.
Choose from:
"I'm a small team with 3 people"
"We're a dev shop with 12 people over 2 teams"
"We're 75 folks with 5 teams and 8 projects"
And have the system autoconfigure a recommended setup.
I wish more tools actually forced managers to put one thing in front of another - like in the real world
A project management tool should have the ability to Monte Carlo estimations.
The "solution" to every "problem"* seems to be more fields to some people. Or perhaps you want another report? Madness.
* "Problem" being something with people's wetware, so just add baby-sitting.
I'm frontend dev and I often work on images/designs/psd's made by designers and I find it cumbersome to manually find image each time I need and to download manually to my project folder via "Save as" browser option and then copy from Downloads folder to project's folder. I think It should be automatic. Click on image and it lands in my project's asset folder (BTW this could be hard in pure web based apps. Such app would need to be desktop based to have access to hard drive).
And it should display path image was saved, to allow me to copy and paste this path to my code.
And it should display me notifications where designers put a new version of image.
I'm also disappointed when a tool doesn't offer to make comments on graphic assets or drawing arrows.
We get annoyed at these limitations of Trello: cant assign checklist items, can't have discussion on checklist items.
We get annoyed at these limitations of Basecamp: projects quickly turn into checklist hell, cannot connect ANYTHING to it's user story, poor calendaring.
We've looked at JIRA, but it's just not client friendly, and we really don't want another tool in the mix.
Things I hate seems like a good starting point for automating an existing manual task, but interestingly, a lot of contemporary project management is moving toward manual/physical processes or hybrids where manual process play a significant role: e.g. physical Kanban boards.
One thing that sucks about a lot of software is that it takes up screen space.
Good luck.
We think our free, agile, open source tool has satisfied people. We're 1 year into it, gotten 100,000 registered users, and have been highlighted as one of the top 10 open source tools of 2014 and won the 2015 most valuable agile tool from the agile awards.
Hope it can help some of you.
The thing that sucks about it is that it was a home-built thing. It's really rough around the edges and in order for it to get any better, someone on the team would have to stop working on product, so it rarely improves. I'd love to switch to something out there, so that I'm paying for a team whose sole focus, full time, is to constantly make it better.
For us, the pain was that the project changed too quickly for todo lists, so most of the tasks wouldn't get completed and the whole thing represented what we needed a week before, rather than today. Done lists vs. todo lists would solve this (not sure if you're familiar with done lists, but they're awesome).
I guess everyone needs their own features, even on a per-project basic.
I guess it would be cool if you had plugins (like Slack/Google Drive integration, etc.) and one could compose their own project management environment by mixing these plugins together..?
We've loved this way of sharing tasks to others so much that, we have even did write about this way of creating todo-list here: http://www.insalgo.com/blog/2016/01/03/management-use-cases-...
Basecamp 2 is excellent though.
Also, accommodation for 'casual' users. It really is painful to onboard a marketer that needs to watch a few tickets.
1) I find the search clunky and rarely helpful.
2) Of the applications—all web-apps or "native" apps that are actually still JS+HTML (so, Electron apps), surprising no-one—whose memory-hogging and generally poor performance have rendered <8GB RAM machines unusable in our company, Asana is the worst.[1]
3) I think part of this is due to the way we use it, but if I look for information in Asana and don't find it I have very low confidence that it is not there. The aforementioned unhelpful search function contributes to this, but it's not the only reason.
[1] I just had an idea: we should rate productivity software based on the last year in which leading PC video games' average recommended (not minimum) RAM specs roughly matched that of the software in question. Lower is better. This would be the last year in which that application would send anything else running on a decent gaming machine into swap. I'm guessing Asana would land in the 2003-2005 range.
[EDIT] If you're just using it as a fancy TODO list it's probably fine, but then its enormous memory footprint is really unjustified.
And you can create tasks per story.
Incidentally, "Pivotal" is the company. "Tracker" is the hosted product.
Basecamp: Not enough complexity.
Both: Pricing that isn't sensitive to smaller orgs.