The QA Tracker status

Hello everyone!

Once in a while, I see threads here and there about testing, and there’s a (not very hidden) subtext that some folks have… feelings… about the ISO tracker website itself.

I’m keen to understand what the current state of play is, in simple terms, so we can chart a sensible path forward for volunteer ISO testing.

I may be completely out of date with my understanding of where things stand and any existing plans. Please tell me if that’s the case - I can take it! :smiley:

From where I’m sitting in the community, it looks like the site is perceived as some combination of:

  • An important resource with genuinely valuable features for those who want to contribute to the various flavours of Ubuntu through testing and feedback
  • A somewhat frustrating and challenging piece of infrastructure to actually use and develop on
  • A lower-priority bit of kit with few (if any) developers actively working on the codebase

I’m not saying any or all of these are gospel truth, but there seems to be at least some sentiment floating about along these lines.

So here’s my question: are there some specific, atomic, actionable things we could tackle to improve this situation?

What’s on your wishlist for making the site better? Whether that’s user experience improvements, technical debt cleanup, or even bigger architectural changes - I’d love to hear what people think would move the needle.

A few starter thoughts that have been rattling around in my head:

  • Could we modernise the existing codebase incrementally? (It’s currently Drupal 7, which is… well, let’s just say it’s been around the block a few times)
  • Would it make sense to explore a complete rewrite using more modern frameworks and tools?
  • Are there specific pain points in the current workflow that could be addressed without major surgery?

Obviously with limited developer resources, we’d need to be pragmatic about what’s actually achievable, but I reckon it’s worth having the conversation!

Please keep suggestions respectful, positive, and forward-looking! The goal here is to figure out how we can make ISO testing a more pleasant experience for everyone involved.

What do you think? What would make you more likely to jump in and help test those ISOs?

Cheers!

4 Likes

I often test the ISO of the development version and I don’t find any problems. However, I believe it would be necessary to adjust the instructions on the first page regarding the package to be cited in case of errors

since the package relating to the error is different depending on the ‘flavor’ tested, perhaps it would be necessary to report those instructions at the top of the product page: Edubuntu, Kubuntu, Lubuntu, Ubuntu…

2 Likes

I’m not using the site much; if I’m lacking time I’m not doing much QA testing…

If you look at the top100 though I’m there, so I’m somewhat familiar with iso.qa.ubuntu.com

My major issue has always been that it can be slow at times.

I’m not just making an entry at the time of my testing, but I enter the page and enter some entries (list of box(es)) I’m going to test in the testcase, eg. if I was using this current box in the comments, ie. copy/pasting dell [optiplex] 7050 (i5-6500, 16gb, intel hd530/i915, amd cedar radeon) from a list of my hardware onto testcase page & saving with status in progress.

If bugs are discovered; or actual comments need to be made, I’ll return & edit that page & add bug links, comments etc… so another edit (or even 2-3)

Finally the test is completed & I’ll edit the entry again, making any additions or changes as necessary & finally PASS or FAIL the test.

ie. multiple edits of the single QA test item… which has at times cost my 20 minutes waiting for the page because of timeouts/50x errors; 20 minutes I could have spent actually testing (or ~18 minutes given the multiple edits would require ~1-2 mins of entry & save if super responsive)…

Historically that’s been my biggest annoyance with it, and that’s infrastructure related.

I don’t have anything beyond that though sorry; as I’ve been using the site for awhile and have adapted what I do based on how it’s worked last ~decade.

( Please also note the 20 min example, whilst real!! isn’t that recent; as stated my QA is way down & due to lack of time, I thus can’t use recent timings in my example as I have none; loss of 32mins on site I think the maximum I recall; but I’d encountered a bug I wanted to record which I why I didn’t just give up! )