Just read the article at Forbes and I love this idea. I’d love for it to have a real impact and I just assumed something like this existed already. I could chip my hat into the ring as a coordinator/PM and tester. John C Z. - Project Manager - Systems Engineering | LinkedIn
1 Like
Here’s a tough question to contemplate:
I’ve actually worked with professional software testers in the past; the kind that have Computer Science degrees, and make a rather large paycheck. When I conversationally asked them to explain some of the challenges and frustrations they faced in their jobs, they brought up the thorny issue of “test coverage”. Basically, software developers can easily fall into an assumption which goes like this: there is an infinite amount of tester labor out there, so the sky is the limit as to how many tests can be performed. So there is no harm in supporting many different operating systems, many different database backends, many different architectures (ISAs), many different languages, etc. There’s all these different dimensions that testing permutations explode out of.
But then the testers have to push back, saying, sorry, but there is a very finite amount of time to do tests. The more tests you’d like to be performed, the more thinly spread all the testing becomes. So consolidating around the fewest permutations of possible testing combinations is the smartest, from their point of view. Less things that need testing in the first place, will make for higher quality in the end.
I submit for your consideration this dilemma: will your testing platform be wide-open to all Open Source software, all distros, etc (leading to the unhappy outcome of thinly-spread labor, and some rather nasty bugs still have good odds of slipping by uncaught), or will it be “opinionated”, focusing on maybe just a smaller number of the most popular distros (leading to the unhappiness of some people’s favorite software and distros getting excluded)?
Which is the lesser of two evils? Less “test coverage” (but wide open testing choices), or less permutations tested (and more “opinionatedness”, meaning fewer testing choices are presented to testers)?
I guess we could call this the “wide-open vs. opinionatedness” dilemma. 
1 Like
Good point @esbeeb
It brings a question to my mind though. Who are the testers?
If the testers are anyone and everyone, i.e. the general public, then wouldn’t restricting the available systems open for testing also reduce the available number of testers?
But if we are selective with who can test, then their numbers are more finite and which makes your point of focusing on fewer systems sensible.
I’m assuming this project is going to be open to the general public for testing input, so the quantity of testers may not be a big concern. However, this would pose another bottleneck. Being open and accessible to the general public would mean there would be a drop in the quality of testing. What I mean is that we’d have submissions from people who do not understand how to test and report their issues in a clear and effective manner. Resources would need to go into helping these people effectively test… in someway, somehow.
It’s kind of like a different flavour of the same problem you described @esbeeb
Or am I off somewhere?
My sense is that when you have (hopefully) large numbers of the general public approaching such a system, most of whom are new to Linux, then “opinionatedness” is the right way to go: reducing the number of choices down to a more psychologically manageable number.
The newbies have much better odds of actually wrapping their heads around a smaller number of (already bewildering) choices. If newbies are thrown into a wide-open arena, with effectively infinite choices, they will likely make more scattered, random contributions that don’t make all that much impact at the end of the day (for the larger community as a whole). Note that there have been entire podcasts, focused just on trying to help newbies make sensible choices, amidst all this bewildering choice (like Linux4Everyone, and Choose Linux). These praiseworthy efforts are a testament to how hard all this distro choosing can be for Newbies.
I agree that a small minority of potential testers will see that their fringy favorites are not there, wipe a tear from their cheek, then hopefully shuffle on to the QA systems already in place with the less popular distros, making their valuable contributions there instead. So their labor hopefully won’t just fall into a wastebasket. It will just move out to the margins.
Remember Bryan Lunduke’s great slide about Darth Vader, in his latest Linux Sucks 2020 speech?
I personally think this project should go full Darth Vader, in whittling down the choices.
Hey all! I found about about this from Jason Evangelho’s twitter and came to say hi. I’m all for more Linux testing and it’ll be interesting to see what y’all come up with. You might find it interesting to look the Fedora QA team’s Fedora Test Days, both as example of what distros are doing now and as something that a community like this could bring wider exposure to.
2 Likes
Thanks for mentioning your credentials here. Your experience will be highly relevant, I’m sure
Welcome!
The initial key objectives are here: dasgeekcommunity.com - This website is for sale! - dasgeekcommunity Resources and Information.
The aim is a minimal viable product that’s extensible for later expansion.
Ryan (@dasgeek) came up with the idea and DLN is kindly bootstraping the project start to finish with their platform, servers and reach. They’re a valuable neutral party so we’re not reliant on the whims any one-developer’s account or hosting and if we can make something DLN can be proud of hosting they’ll keep putting the wind behind us.
The key here is going to be humility, this is about serving software freedom.
Think about how you can best serve the people who’ll make this a reality as i’ll be doing. Our job is to save them time; it’s not a training opportunity, it’s a chance to be proud of contributing, to give someone else a leg up.
For those worried about missing the Sat meeting, it’ll likely be weekly and aimed at producing proposals which should go to the forum for discussion prior to being finalized. Timing rarely works for everyone so it’ll make sure everyone’s tied in. Future scheduling will also be tweaked to not conflict with the DL live show @thatComputerKid
I’m currently working with @MichaelTunnell to insure everything’s ready so I can just hand things over to the people that need them and support them as best I can.
There’s no sign up form for offering support though there may be a call for particular skillsets soon when we have a better idea of what we’ll be using. That announcement will be pointed to from this thread. I’ll log what’s been offered here and make it an open document prior to the meeting.
1 Like
Hey,
Heard about this project idea on the Linux For Everyone podcast and thought I’d drop by and check out what the ‘plan’ was.
I have experience as a full-stack enterprise developer working in Java/Kotlin on the back-end with various database tech experience and Angular/React on the front-end. Also got PM experience but happy to lean on others there, can’t do everything!
Be interested in joining the conversation to see what stage this is at and what stakeholders there might be if this is something which specific developers have approached DLN about previously.
Don’t suppose there is a Jira/Trello/Github or anything yet where these product ideas are being captured outside of this thread? Be good to just collect everything in one place and begin to asses what would provide most value and what is feasible given the expertise of everyone willing to get involved?
Might be good to get somewhere a little more organised for people to record their availability and skills as a starting point.
Anyway, @Ulfnic you seem to have been doing a lot of the legwork so far. What are your thoughts?
3 Likes
My apologies on saying it coincided with DL, I was confused and off by a day. I have since deleted that message and I hope you will forgive me 
I actually have no idea when the DL live show is, I was going to find out lol. Hopefully I can forgive myself 
Anyone joining is encouraged to read the thread to catch up to speed. After the meeting there’ll be more formal instructions and proposals based on what’s discussed.
DL is providing a GitLab/GitHub space (contributor choice), likewise with coms and testing environments that’ll be ready by Saturday.
Good suggestion.
Just checking, @Ulfnic is there another meeting tomorrow (29/08)? Will try and attend if possible, be nice to get an idea of the goals and ideas from others involved so far. Happy to just sit and take notes too if that’s beneficial and no one already is 
GMT+1 timezone here, dunno what timezones the previous meetings were in…
Hopefully see you there.
Meeting stands tomorrow: 2020-08-29T18:00:00Z
I’ll trust the forum schedule object to do time conversion as I scarce trust myself.
I’ll get the instructions out to everyone on how/where to join shortly, just making sure the comms subdomain is locked in.
Everyone’s free to take notes and we could use a collaborative document tool to collectivize them for a forum post near the end.
Looking forward to it!
2 Likes
Are you all still looking for contributors?
I’m still pretty inexperienced, but I work with Oracle SQL a lot at work. I’ve developed a Django application before and am fairly familiar with HTML, CSS, JS. I’m definitely interested in helping any way I can.
1 Like
Hello @jakesco.
I’m sure we’re always looking for contributors.
If you have strong opinions on the feature set, I suggest you join us in the chat, but otherwise I would recommend you wait a couple of days for a list of what needs to be done. It will be easier to grab a task and contribute, once we know a bit more.
Nothing is set in stone yet, but I doubt you will get to flex your Oracle SQL muscles on this project 
Is there something posted for the meeting invite? Zoom or some collaborative tool? Planning on being there, so maybe I missed the link or instructions.
1 Like
The information will be posted here 1hr before the meeting and i’ll describe why in that post 
One of the reasons is not jump starting things before everyone’s in.
If anyone would like to update their skillsets and see who’s interested prior to the meeting, do so here:
edit: changed link to new link
https://devops.destinationlinux.network/index.php/s/mMcJDzpbRECoT9W
No login required, just find/add your name and start editing.
Information will be moved to a better system sometime post discussion.