1x38: Easy Being Green

Instead of just looking at Arch maybe a segment on rolling release distros? Try Arch, OpenSUSE tumbleweed, debian sid (I know sid’s a little different since it’s debians dev channel) and another one (a 4th doesn’t come to mind right now)?

On the subject of XFCE I’ve just come back to XFCE after using Gnome and it’s freaking fast… Gnome is fast but XFCE is freaking fast… Sure it doesn’t have all the bells and whistles of gnome or unity but it’s got everything you need from a DE. I do dev work so the lightweight nature of XFCE suits me because it just gets out of the way. And as for HiDPI support I don’t have a HiDPI screen so I can’t comment on that, but I’ve read that it has some support for HiDPI screens, though it’s not as east as other DE’s to configure, and like I say I haven’t actually tried it myself!

Hi, the segment about paying for fixes/development is something I’ve been interested in for some time. I’ve worked as C & Python developer writing open-source software full time and fixing bugs and development since ~2010. I’ve also funded development here and there (paid for things to be improved in other projects too).

There are a bunch of factors which make bounties for fixing issues problematic…

  • Most good developers already have jobs:
    Some extra financial incentive is fine, but to go though all the leg work, emails - agreeing on what exactly needs fixed etc… takes time to get these details sorted out - on top of an existing job, so financial incentive needs to be significant. All the communication overhead & follow up (patch review etc) adds up. Not a million dollars, but more then 1-2 days pay, else it just isn’t worth it.
  • Its risky:
    Just because you manage to fix a bug doesn’t mean that code is acceptable, sometimes I fix a glitch but end up not committing because for whatever reason, it makes things worse (unacceptable slows down, changes behavior which can break workflow elsewhere… for eg). And if you aren’t the author, upstream may reject outright. (they might have a rewrite planned that solves a slew of other problems and don’t want to loose time maintaining broken code).
  • Its hard to quote for:
    More then half the time is just finding the cause of the bug, or even just redoing the error, only to make some minor change - or to see comments in the code-noting that this is a known limitation. Either you feel like you’re over-charging or end up loosing a lot of time for something which may not have any good solutions.
  • Chances are you aren’t even the first to try fix:
    When a software project is maintained reasonably well, a lot of the low-hanging-fruit is already taken, issues that sound easy end up being cans of worms.

So if you do want to pay someone to develop, its nearly always best to hire the maintainers, or active developers who know the code at least - as Bryan pointed out.

On a simplistic level I still like the idea, but in practice fixing bugs often gets complicated.

Stuart example was a bit tricky too - errors in interaction between applications can be a can of worms… and in this case its not even clear if you need a Ubuntu or a Firefox dev. Its a valid complaint of course, just harder then normal to pin down who to ask.

Zero, of course. But isn’t anyone with an Apple monitor using thunderbolt for it?

re bug bounties: I’ve found that it’s hard enough to figure out how to contribute to the many open source projects and their many ways of doing development that, even working for free, there’s too great a barrier to entry. Also, what if I pick a bug someone’s already working on?

re thunderbolt: I have a thunderbolt to hdmi adapter cable. Does that count?

http://www.emerge-open.com/index.php/component/content/article/44-services/140-support1 may be interesting here; I didn’t know about Emerge Open, but it seems like they want to provide the service we’re talking about.

First of all, the reason I listen is the Linux and not food blendering or gun control so a big plus from me!

One thing struck me though, when you talked about giving money to bug fixers.

IMO the perfect system: http://www.ubuntu.com/download/desktop/contribute/?version=14.04.2&architecture=amd64

Greetings, Rafael

BTW: First post :wink:

Hi there, my first time here.

I was a casual listener in the past, a post by Richard Brown, openSUSE chairman, pointed me here again.

Very great show indeed - pretty unfair to challenge a new user to run E19 but it resulted in a funny nice part of the show.

I shortly resume my Linux/openSUSE experience.

At beginning 2013 I moved from Xp to Xubuntu and then I distro hopped a lot with a remarkable preference for Debian and some love for Fedora.

On June 2014 I installed Arch and kept it running till August.

Meanwhile, I spent a full day in breaking and fixing openSUSE trying to install Mate on a barebone net installed system and discovering it I kept loving it.

On September I got a new job and I got pretty busy too: I leaved Arch for openSUSE 13.1, looking for a decently up to date and maintenance effort free system. It was.

I updated to 13.2 when it was released and eventually, on Jannuary this year, to Tumbleweed.

During this openSUSE period I ran KDE, Gnome, Xfce - each and every of these DE very well deployed. At present I’m on Gnome.

Tumbleweed is stable and pretty up to date: I have always the latest and greatest packages with no maintenance effort.

openSUSE selling points to me are:

  • the installer, the best one in the Linux world

  • yast, a powerful sysadmin GUI tool

  • zypper, very powerful package manager

  • the OBS, together with the 1-click install

  • Tumbleweed

TBH I spent some time in collecting resources and in understanding how openSUSE and its tools work, I’m still learning things about it today, however once I got used to it I started finding it amazing indeed.

If someone is interested in, I can post my new comer resources list.

Cheers,

gabriel

Which is fair, but remember that Bad Voltage is about technology generally – we all rock the Linux so it’ll show up a lot, but it’s not everything :slight_smile: And everybody needs to eat, whether it’s Soylent or perfectly _sous vide_d steak :slight_smile:

Ubuntu inviting donations for downloads does work – the community team have some budget now to donate to worthy initiatives! What makes that seem like the perfect solution for you? Was elementary doing the same thing a good idea?

Also, welcome to the forum, @pi_falera! Nice to have you here.

Yes. Yes it was.

That aside, you’re identifying one of openSUSE’s leading points as yast? I found it terribly confusing, and ready to configure a bunch of stuff that I think that real people ought not to fiddle with. Am I wrong here? Do tell me if I am!

Also, welcome to the forum, @gabriel_3!

re: the openSUSE challenge:

I’ve been an intermittent user of SuSE and openSUSE for well over a
decade, starting back in the 6.x days if I remember correctly. I’ve
always admired the project for its professionalism and broad scope. I
recently switched back to openSUSE thanks to the rolling Factory (now
Tumbleweed) announcement, and I also have 13.2 installed on other
machines. I’m really enjoying the experience, and this time, I’d like to
stay.

I really want to see openSUSE succeed, but unfortunately I tend to see
it dismissed by potential new users as “a nice distro, but…” followed
by certain annoyances or “paper-cut” issues. The openSUSE 13.2
review by Jesse on Distrowatch is a recent example of such comments. I
feel that openSUSE is fundamentally extremely well designed and
executed, but I have also noticed the same recurring “paper-cut” issues
year after year, release after release. So I’m wondering if anything
could be done about them, or if the developers have any interest in
fixing them.

Without going into detail, here are a few general areas of concern that
I often hear and/or have personally noticed throughout many openSUSE
releases:

  1. Ugly font rendering
  2. Automatic installation of extraneous/unrequested packages, especially after initial installation. Even after un-installing, they are later automatically re-installed due to the pre-installed patterns.
  3. Crippled packages meant to prevent compatibility with proprietary multimedia formats.
  4. Difficult to install multimedia and/or proprietary formats and drivers.
  5. Breakage of YaST modules.
  6. Not everyone wants Gnome and/or KDE, and it requires to much effort to remove the undesired desktop and install an alternate one. More officially blessed custom spins would be very nice.

All of the above issues do have workarounds, and think it would be best for this thread to not get into the technical details of these aforementioned issues. But in general terms, these issues come up time
after time in reviews and forum posts about openSUSE. So my question is: Would the openSUSE project be interested in working to resolve any or all of these issues? And if so, where would be the best place for me to
bring up these issues and work with developers to make improvements? Some of the issues are not related to any one specific package, but are rather of a more systemic nature. In some cases legal/patent issues are
probably involved. (Again, let’s please avoid legal discussions in this thread.) Although I am not a coder or developer at all, I do have an eye for detail and polish, and I have a good idea of what typical users
expect out of a Linux distro on the desktop. I’d like to help to identify and test solutions to these papercuts if any developers are willing to look into some of these long-standing issues with a fresh eye.

Thanks in advance for your time and consideration.

And so does the Juju team and the QA team. Perhaps you saw that the web team got a new 60inch monitor! We should ask popey why…
In all honesty, I think that if a community-listening project like the Ubuntu project(lately) gets money for certain categories, they are the ones who can best decide which bugs or features need work.
Elementary’s implementation was pretty bad for obvious reasons. Not because it was hard to bypass the paywall, but because it was possible but not discoverable! Also, they didn’t provide categories right?

@sil

Thank you for welcoming me!

IMO your point of view on yast is wrong:

  • it is like the HUD of a starship: you can drive a complex system, as every and each operating system is, by it;

  • if you want to configure something you have simply to open up yast and look for it: this is by far easier than search for it on the docs or web, find the right file and edit it.

An example: I like fish as shell environment: on openSUSE I open yast and set it as my default from a drop down list; on other distros I have to change the user config by CLI. It is simple in both ways but the second one requires me to remember the right command.

Also consider yast from the Linux newcomer point of view: on yast you look around on a GUI instead of trying to guess which in the dozens tips googled is the right one for you.

So yes yast is a leading feature of openSUSE.

Cheers,

gabriel

If you’re worried about that - you can generally comment in the tracker that you’ve started look into it (include hints for anyone else - if you find the file/function at fault). I don’t think its such a problem.

… getting involved in a new (complex) software project generally isn’t easy, probably you should have at least a bit of experience developing the software (really basic stuff) before attempting to fix more complicated bugs. Some projects have a Easy Hacks page, (LibreOffice for eg), this tends to work better then bug fixing for developers who want to get involved.

1 Like

Close enough! Though its rather enterprise focused - cant blame them, it’s where the business is.

I’m curious if they are doing all the development themselves, or if they have some contact with existing developers of the projects they support .

Would be good if they could branch out by making contacts with other popular projects (a few active devs in each project may be enough), then they can handle the customer, basic triage - then forward the issues to the active developers to quote on.


Edit: would be great to try and get these guys on the show.

@sb56637

If I’m not mistaken, we meet also elsewhere :smile:

Another point of view on the very same core issues you listed is that openSUSE requires the user to learn about it.

I never faced your points 3 4 5, because I use VLC for media playing and open driver for my pretty old GPU.

As you wrote, there are some aspects that need settings while on other distros they are defaults.

However, while being my observing horizon by far shorter than yours, some improvements happen from release to release: I gave up the first time I installed openSUSE 12.2 because of networking not working - later on I discoverd that it was not enabled by default - today networking is working out of the box.

As another example of improvements, font rendering: on Tumbleweed there’s a nice new package, yast2-fonts, that eases the systemwide font rendering configuration, I guess it will be added in the next release.

The real openSUSE problem IMHO is that it is a hidden gem: there’s very limited media coverage, so it is hard to enlarge the user base which drives to reduce the number of prospect devs, and this results in your list of required improvements.

Cheers,

gabriel

Hi there @gabriel_3, good to hear from you!

Well, since I posted these points in the mailing list (#6 is new), I’ve continued to use openSUSE on all my systems, and I’ve also switched several other very new Linux users to it. Unfortunately in my experience, these points continue to be a pretty major roadblock to adoption of openSUSE by users that are newer to Linux or impatient. I appreciate openSUSE for the flexibility that it offers me, but I’m not a typical user. I think we all need to take a look at what we do to make openSUSE work the way we want it to, and then ask ourselves if others will be able to figure it out with relative ease. Most of us already know the idiosyncrasies of openSUSE, but I really think that new users shouldn’t have to:

  • tweak font rendering styles (most will just simply switch to another distro with prettier fonts out of the box
  • install, re-install, un-install, and un-install again loads of packages with confusing (for them) names. It makes the distro feel slow and bloated when they try to install a small program and see a massive list of completely unrelated packages ready to be installed, some of which they might have already tried to un-install but which are being automatically re-installed. Yes, there are ways around, but the user shouldn’t have to learn this.
  • learn the names of the repos and the package names and/or the unofficial openSUSE guide that contains a “1-click” install (which is not 1 click) for all the confusing multimedia file names. And to add to the confusion, you can’t just install additional packages, you have to replace already existing packages with identically named packages from another repo. This is very counter-intuitive for new users (“why should I have to install VLC again if it’s already installed? I just want to add proprietary media support.”)
  • search around for the names of alternative desktop environments that are not KDE or Gnome, which act more or less like a traditional Windows computer. Cinnamon and Mate continue to be the best received by new users; there’s a reason why Mate is wildly popular in this regard.

I’d definitely agree with this. But I don’t think it’s going to make news until it becomes more approachable for new users. Unfortunately it looks like a super easy to use distro, but then after installing the user has to learn a lot of little openSUSE “tricks” to make things work the way he wants it to. I don’t mind learning these “tricks”, but pretty much everybody else does mind. :wink:

I see your points @sb56637.

Possibly I’m an unusual user too: I like the process of making openSUSE up as I like it. And obviously the more I use it the easier it is.

Maybe an external tool like Fedy for Fedora could help, or better a two mode yast, one with basic settings and one with advanced ones.

Hm… The display works out of the box, the sound works out of the box. Isn’t that good enough?

As a general rule I avoid derivative distros. Too many downstream projects (not just distros) have slammed into a concrete wall when the upstream made some incompatible decision (either technological or political). So I figure why take the risk, if Arch is what I want then I’ll use Arch.

1 Like

I thought it was all about @bryanlunduke. The description of the show should be something to the effect that ‘Bryan Lunduke and some blokes talk about things.’ Or, ‘Bryan and his entourage talks tech.’ :smile:

1 Like