News for hackers. Stuff that matters.
Re:Atlassian's new terms forbid benchmarking by jjoonathan on Saturday September 29, 2018 @07:28PM (#18103590)
Here's a benchmark for you, Atlassian: your software is slow as molasses and getting worse with each redesign.
Re:Atlassian's new terms forbid benchmarking by imglorp on Saturday September 29, 2018 @07:46PM (#18103670)
It seems like they're balancing features with speed. In order to get a bunch of extra features, every page load needs to be 200mb of extra crap that 99% of users won't use 99% of their page views.

Oh and hamburger menus are hiding the key features you DO use every page view.

Re:Atlassian's new terms forbid benchmarking by ethbro on Saturday September 29, 2018 @09:22PM (#18103980)
I always thought of Atlassian projects as inheriting all the problems of Java, exposed in UX form.
Re:Atlassian's new terms forbid benchmarking by ex_atlassian on Saturday September 29, 2018 @11:17PM (#18104282)
Its always the people/expertise, rarely the tool or language.
Re:Atlassian's new terms forbid benchmarking by ethbro on Sunday September 30, 2018 @12:47PM (#18106931)
Fair. But for purposes of language = ecosystem, I'd definitely say organizations trend to mimic language structure, and vice versa.

At least to a first order, palm-reading, internet comment approximation. ;)

Modern Java, IBM

C# .NET, Microsoft

React, Facebook

Rust, Mozilla

Re:Atlassian's new terms forbid benchmarking by anothergoogler on Sunday September 30, 2018 @03:52PM (#18108016)
I'm not sure what's more concerning, that you're calling React a programming language or that you're unaware that Facebook is written in PHP.
Re:Atlassian's new terms forbid benchmarking by ethbro on Sunday September 30, 2018 @07:37PM (#18109070)
I believe you missed the point for the trees. And for the record, I'm under no illusion that Firefox is fully written in Rust either. :p

https://4e6.github.io/firefox-lang-stats/

Re:Atlassian's new terms forbid benchmarking by beerlord on Saturday September 29, 2018 @11:53PM (#18104374)
Why doesn't JIRA offer a native desktop app? Everyone using it is on PC, Mac or Linux right? And then a cut-down iOS (iPhone and iPad) app covers the rest?
Re:Atlassian's new terms forbid benchmarking by ygra on Sunday September 30, 2018 @02:01AM (#18104716)
A few colleagues use a Visual Studio extension, although that seems to be mostly handy for time-tracking and limited interaction with issues. It's also stuck in a UI that goes back to 2008, so perhaps rather fitting ;)
Re:Atlassian's new terms forbid benchmarking by arbie on Sunday October 07, 2018 @11:08AM (#18160914)
An Electron wrapper around the web app will do wonders for productivity and RAM utilization.
Re:Atlassian's new terms forbid benchmarking by dbg31415 on Saturday September 29, 2018 @09:29PM (#18104000)
The new design for Jira really sucks, has made everything slower, and crappier to use. Page load times of 10+ seconds are common. And no clear usability gains.

I was a huge fan, but any more I just want to use GitHub.

Re:Atlassian's new terms forbid benchmarking by Karunamon on Sunday September 30, 2018 @11:15AM (#18106508)
Please talk to your admins! In my experience, Most Jira slowdown issues can be traced back to permission checks. (And you can enable tracing in the system settings to verify this)

If only your team needs to access it, set the instance to open permissions (I.e. most rights are set to “Everyone”) and then control access using the network or with a proxy in front. This took the instance I run at $dayjob from infuriatingly slow to no slower than any other web application.

—Source: 5 years and counting Jira admin

Re:Atlassian's new terms forbid benchmarking by dasmoth on Sunday September 30, 2018 @12:09PM (#18106735)
This is knowledge that could come in handy one day, so thanks.

But fine-grained permissions are a fairly regularly cited reason for using Atlassian’s stuff over other (simpler...) offerings. This sounds a lot like “don’t try the fancy stuff because it’s unusably slow”.

Re:Atlassian's new terms forbid benchmarking by Karunamon on Sunday September 30, 2018 @06:49PM (#18108880)
I'd be willing to bet part of that is too many fine-grained permissions. It's a bit of a footgun.

Basically, anytime you hit a ticket's page, Jira has to scan your account's group memberships and compare that against a litany of permissions to determine whether you can even see the ticket and what actions you can take with each of its fields. This is done to avoid showing you UI elements for things you can't do. If a given permission is set to "everyone", the check simply isn't done (kinda the equivalent of replacing a call to the user directory with a "return true;")

I'm not talking entirely about reducing security here. I mean that if everyone in your team is a member of a certain security group, and only your team touches your Jira, set that permission to "everyone" rather than using that security group - the lookup is completely unnecessary in this case.

Basically, you want as few permissions as possible to ensure the level of security you actually need. Often this isn't done, people go a bit crazy with permissions schemes trying to segregate this and that.

Re:Atlassian's new terms forbid benchmarking by nixpulvis on Saturday September 29, 2018 @08:19PM (#18103799)
Yea, they didn't say anything about UX complaints, for which there are MANY.
Re:Atlassian's new terms forbid benchmarking by ygra on Sunday September 30, 2018 @02:01AM (#18104718)
You don't like up to seven toolbars on a single page?