The Jira Voting Charade

The blog post on improved communications, also had information regarding the Jira , they encourage people to use the Jira, they’re removing voting from the Jira. Encouraging people to use the Jira is good, although I remain unconvinced that it’s an easy to use customer facing tool, as an internal bug tracking tool it’s bloody useful, externally, not so good but it’s better than nothing.

However removing votes? This appears more and more to be a PR exercise to remove visible interest in long standing  bugs or feature requests, none of the arguments in favour of removing voting stand up to much scrutiny, especially in light of Linden Lab representatives advising people that watching an issue will be a measure of interest…. which means people will watch an issue in future instead of voting, which achieves what exactly? Well the watch feature has a defence mechanism, it emails folk everytime there’s an update, even if that update is someone correcting a typo, beyond that, it’s not really that much different to voting and already people are posting threads on forums asking for people to watch an issue (rather than vote).

The advice to avoid the emails so far has been to setup email filters …this sounds less and less convincing as it goes along. However let’s get something clear on voting on Jira issues, popularity isn’t a contributing factor in fixing a bug, nor should it be, the important factors in fixing a bug should be how big a bug is it, how many people are inconvenienced by the bug and what sort of detrimental effects the bug will have. Those are the important factors, not the number of votes.

So voting to get an issue fixed seems somewhat churlish and therefore, what’s the point in keeping it? Well one reason is that I can set myself up a very simple filter on the Jira to track all the issues I’ve voted on:

“View Voter” = currentUser()

Ah but that shows me all issues I’ve ever voted on, so I want to tweak that to show let’s say, all issues I’ve voted on that are unresolved:

“View Voter” = currentUser() AND resolution = Unresolved

The thing is, for me, with these queries I can tailor what I look for on the Jira and vote is a key marker for those queries, but I could use watch instead and set my filters using watch, as well as getting emails and filtering them, meaning my ISP, LL’s Servers and my mailbox all do work to basically trash an email I don’t want and don’t need.

However, I don’t need to watch an issue to find the issue on the Jira, and I don’t need to watch an issue to be interested in it. Whereas voting may be seen as setting false expectations, it also gives people a chance to represent their support for an issue, it maybe would be better if people could show they didn’t support an issue with a no vote, but that can be done in the comments. LL’s argument against creating the no vote is that it takes too much work, which is something I agree with, which makes me wonder why they’re bothering to tweak the system at all.

The mistake LL have made here is suggesting that watch is the new vote, this will make this a pointless exercise, except people will now get annoyed with LL over the emails they receive because plenty of people won’t go near an email filter with a bargepole, is it really so hard to change vote to watch (email unsubscribed)?

When you start using the Jira, and if you’re lucky avoid Jira wars and the closed/open/moved song and dance, you soon get used to the fact that votes don’t mean your issue will be fixed, heck I’ve had issues addressed on the Jira where I’m the only participant, people soon get used to how it works.

Really, LL have better things to do than change the Jira, votes are, if nothing else, a sign of support for an issue, it makes people feel involved and it’s a way of getting people involved in the Jira, voting should stay, the only change should be LL explaining clearly that votes don’t mean prizes, that a high vote count won’t gain priority for an issue, they should just leave it as it is and stop letting PR get in the way of the system.