Dedoimedo: the end of a long, beautiful run?

[JayM]

I had installed Plasma on MX-19 running in a VM this morning, then logged out and selected it as my default DE. It just now took me 15 minutes to find out where to disable single-clicking to open stuff. It's not an option within Dolphin anymore, it's a system-wide setting. What I wonder is, how do people who prefer single-clicks accomplish common tasks such as selecting multiple files?

It's funny, but up until two years ago I panned any distro that didn't come in a KDE/Plasma version. Now that I've gotten used to Xfce I can't stand all the bling, gewgaws and floopydoopies. It reminds me of Windows before I turn off all those animations and junk, which along with changing the theme and menu to Classic style was one of the first things I always did. If that's what Dedo prefers, good on him. Some of us prefer an OS that stays in the background and lets us get things done instead of trying to entertain us with unessential eye candy.

That being said, he raised several very valid points about MX-19: the known glitch in the panel's notification area that some people have that shows artifacts of another icon, the fact that samba and bluetooth connectivity could be easier. Some of the things he complained about are rather odd though: the plank docker? Come on, that's an add-on. It doesn't ship with MX. If he's going to review distros he should stick with what comes with 'em, not with things he's installed. If he wants to do that he could have easily installed and logged in with Plasma, then no more complaints that MX looks "so '90s." (It's not called patito feo for nothing.)

I suspect that he wrote what he did partially due to having a deadline to make. It's partially clickbait. He should know by now that every major version release of any OS is going to have a few bugs during the first couple of months or so. That's why the wise hold off on upgrading to the latest versions until after the first dot-release (or the first service pack in the case of Windows.)

[BitJam]

I wonder - why did you decide to upgrade xfce version together with Debian version, since Debian changes alone suck up so many developer resources? This could have been an incremental process. For example, keeping xfce4.14 in MX test repo till the next point release of MX. (??)

Just to be clear, that wasn't my decision. I remember conversations about it but I did not pay attention to them. I'm not saying it was a bad decision, just saying it wasn't mine.

I imagine there was probably a clamor for it and after testing it seemed good enough or the benefits seemed to outweigh the detriments. And that could very well be true. Our "about" page used to describe one thing that makes us different from many other distros. Our "thing" is we port foreign packages and backport unstable packages to Debian stable in our community repo. We have a forum for it: Package Requests / Status. So it's in our nature to be more cutting-edge than Debian stable. That's what we do.

We started out as a small distro. Just a few years ago (2016, I believe) we were hoping to break into the top 20 at DistroWatch. Now we are famous for being famous and our very small dev team is, in some ways, struggling to keep up with all the new input from our greatly expanded user base. In addition, there could be a psychological factor that more perfection and polish is expected from us because we are now so popular. I could give you some very specific examples of this. For years I said we were blessed by our relative obscurity because it gave us a chance to develop our systems and work out the kinks without being overwhelmed by user input. IMHO, we've had more of a struggle with MX-18 and MX-19 due to our higher profile and greater popularity.

Yet another way to see it is that, especially with our antiX roots, we've been a very R&D oriented shop. There was often a frantic rush and Herculean efforts to add finishing touches and finishing features to our releases. We went from conception of the distro to the first release (MX-14) in just a few months! That's insanely fast. In some ways our beta-testing is more like alpha-testing (in terms of the number of people) so there is now a huge jump in users between beta and the final release. With our larger user base, we may need to switch to be a more production oriented shop. The main difference between production and R&D is the quality control step which is what I was talking about with more formalized testing. The downside of becoming a production oriented shop is we become less nimble. Changes will need to go through a quality control phase before they're released. I don't know exactly how this will play out.

[ChrisUK]

If you ignore the XFCE and app stuff, it's a pretty good review - although it does remind me of something that occurred to me when 19 was released...

I was surprised how early the release date was, especially considering the new Debian base plus a new XFCE version. IIRC, didn't December used to be the month for major releases? And if so, would an extra month of development/testing made any difference to any of the complaints in the review? (I suspect not)

[Beratung]

I used to enjoy reading his reviews but eventually felt that by spending superficial amounts of time with distros, those reviews obviously aren't the same as reviews written by someone with hands-on long-term professional working experience. With that said ...

... I've experienced strictly snapshot and iso creation problems with MX 19 until just a couple of days ago. Finally, for the first time in a couple of months, a snapshot was created properly which was then possible to be installed on two other laptops (one of them 14 years old with 1 GB of RAM) without presenting me a black GRUB> screen. For awhile there, there was something wrong with either the snapshot creation (which actually worked) or the live .iso from the snapshot creation (which failed strictly during boot, afterwards).

Aside from that I believe that MX 19 is the nicest, most polished MX yet. We're not living in an age of strictly "power & performance" anymore. A lot of people like to actually feel good about what they're seeing and doing on a machine while having to look at it 5+ days out of the week for 8+ hours out of the day. There's absolutely nothing wrong with combining power, speed, performance, and attractiveness. At this very moment, the only MX Linux "gripe" (a ridiculous one I'm sure) that I have is with the dark themes. I've been looking for that perfect dark theme for the past couple of years now. The best one out there (IMO) is Blackmate by the developers of the Mate desktop. Works on MX 19 too. It's just a tad too light in a couple of places though, but aside from that it's perfect.
So aside from superficial stuff, I'm extremely pleased with MX 19 and feel that the developers did a magnificant job. :)
Dedoimedo also did a magnificant job, considering the little amount of time that was spent on a review.

PS: If you look at the number 1 slot on distrowatch for the past 12 months, that speaks for itself as well.

.

[BitJam]

I was surprised how early the release date was, especially considering the new Debian base plus a new XFCE version. IIRC, didn't December used to be the month for major releases?

Yes, that was partly/mainly because we tried to release antiX before we released MX due to the developer overlap and initially the overlap of having the same lead dev for both distros. We developed them in parallel this year which might have put a strain on devs who work on both.

And if so, would an extra month of development/testing made any difference to any of the complaints in the review? (I suspect not)

TBH, I think we were getting diminishing returns from our small but extremely dedicated group of testers. I discuss the reasons for this and a possible solution in my post above yours. TL;DR: we have a very low ratio of beta-testers to users.

I agree with you it's a good review. And honest and useful. Dedoimedo's reviews have always been extremely useful.

[junoluna]

if MX has gnarly bugs, i haven't managed to find any

suffice to say, i absolutely love this distro and wish i had found it earlier ... after sticking it on 4 laptops without hitch, i am on the verge of becoming evangelical about it

[asqwerth]

Regarding XFCE 4.14, I personally didn't have any objections to introducing it in MX19 myself.

I'd been on the gtk3, pre-4.14 version on Manjaro for quite some time, as it slowly developed and rolled towards 4.14-final. And on another Manjaro install, I was on gtk2 XFCE 4.12 until 4.14 went final, at which point an update just rolled Manjaro over to 4.14 automatically.

In both installs, 4.14 seemed ok and very usable to me, although my use case is admittedly very simple and run of the mill. I also didn't have icon-artifact issues in the panel in Manjaro**. So this was and still is a puzzling issue to me.

As for whether Dedo's reviews are clickbait, I don't think so. He is just very fixated on aesthetics, and also likes to review distros from the viewpoint of a newcomer who would not know how to use the command line or edit config files in order to get things to work. If he has to resort to that to get something to work, that is a minus point in his review. He can also be overly dramatic in his prose at times. Accept the relevant stuff and don't bother about the rest.


** NB. although Manjaro has a different update notification icon problem even now, namely that the tooltip for the update icon shows up below the icon and panel, so if you place the panel horizontally on the bottom edge, the tooltip message can't be seen since it's cut off by the bottom of your screen. It's fine if your panel is at the top.

[Friedrich]

One area where I think we could improve a lot is having more formalized testing procedures. This boils down to better communication between the devs and the testers. ISTM our base of testers has not kept pace with our base of users. I hope that with more formalized procedures (like checklists) we will be able to make the best use of our testers and also make it easier for users to become testers.

Very much agree.
Do you think of improvements inside the forum?

[Jerry3904]

Dedoimedo also did a magnificent job, considering the little amount of time that was spent on a review.

Yeah, that's the point. Very few reviewers actually take the time to look carefully at the distro before popping off. This review must have taken Dedo many, many hours and I know I speak for the MX Devs in saying we are grateful for that careful attention.

He always gives us a lot to think about.

[richb]

The bottom line for me is that the review is very useful. Consider if Dedoimedo was a beta tester. The issues he raised would be evaluated by the devs and solutions pursued, Obviously it is a review on a release but the same process can be pursued. The general opinion of the devs that I gather from this thread is that they see this as another opportunity to make improvements not only in MX as released but in the release testing process as well.