I’m calm. Just being direct. Calling BS as I see it. As well as bad customer service. I’m pretty constructive, while being fed nonsence like “some bugs are typos” and “we don’t have time to explain you how you could brick your router with this release” but “upgrade to latest and you are safe”.
Must be me but I fail to see how this is being constructive ?
You’re rude, not direct. That’s a difference. Already three persons telling you that. Do something with it.
More popcorn needed…
@gdanov
And if I call you rude too, who usually say it to me…
Technically, I’m skeptical my post will have visible effect on MT’s delivery quality, so we can agree it’s not an effective post, but if they address fraction of my concerns we’ll see visible improvements so it’s constructive one. But I don’t see why you need to sway me in behavior more suitable to you. You like it how it is — great. You don’t agree with me — fine. But telling me how I should behave — no, thank you.
Maybe we'll benefit more if we focus on the content? Just suggesting.
If most people find the post rude, I'll accept it and think about that.
I think direct was actually an apt description to begin with. Rude is more of a subjective perception in this case. If you fail to see how @gdanov has genuine care and concerns for quality & service excellence then you are the one who is wrong here and that is a fact. Anyways guys let’s all be adults here and bite the bullet on this one. This is a release discussion thread, not a boxing ring or children’s playground. That will be my final 2c on this squabble. Back on topic please everyone ? :))
Haha, no more popcorn for msatter.
I think it is hilarious that gdanov thinks software companies will always provide full details of fixed bugs. I have worked for a global software company for over 20 years in a support role and found that while customers might think they are owed detailed answers on bugs/defects, it really depends on the type of fix and what the developer puts into the report. This is not something unique to mikrotik.
Complicated software usually means bugs. Some releases are better than others. Testing is your best friend and don’t assume everything you use has been tested the way you use it.
I love straw man arguments.
I wonder if gdanov is applying FOSS principles to a closed-source system. There is no public Git log to trawl for commit comments, no developer mailing list to mine, and no blogs to masticate it all for you.
What RouterOS boxes are, though, are appliances, and even modern Internet-connected appliances shouldn’t go MIA when you update their firmware. It’s hard to estimate failure rates from the moaning in threads like this, but it’s got to be higher than for, oh, let us say your everyday random smart TV. Can you imagine the hue and cry if someone like LG did this to even a tenth of a percent of their displays, just once?
This is not something unique to mikrotik.
In most cases, I don’t mind the single-line changelog entries. There’s a tl/dr problem to fight against, after all.
However, you must admit that “improved long-term filesystem stability and data integrity” makes you sit up and go, “Pray tell, may I have just a bit more detail, please?” That’s one sphincter-tightening message.
Testing is your best friend
Yes. One wonders whether MikroTik has a lab full of historical hardware and tries an upgrade to each proposed release on each one of them, and if so, whether they learn from boot-loop and other problems to improve the lab to make it test more corner cases. If not, they should. If so, one wonders why that test methodology isn’t catching these problems before release.
In case this is addressed at me, it’s excellent illustration that MT’s support attitude is unacceptable. Not surprisingly it appears to be organizational issue, not individual. Endorsed actively or passively (by tolerating it) by management.
I surely hope @mrz will clarify what’s the case?
@tangent — FOSS is a benchmark in transparency and community building, many times in support as well. Of course I make difference between open/closed development specifics and know we’ll never see bug ticket internals (if they exists, which I often doubt). That’s totally OK as long as we see some sort of tracker or redacted artifacts. Right now nobody knows (externally at least) of the known/pending bugs for the release. So I need to wait a few weeks then thrall the forums to see what to expect with each release. How should I label that without being considered rude? BTW this is not the first time I suggest this to @normis directly, nor am I the only one that requests it. But hey, “was there a suggestion” is the answer.
We surely deserve to see more than the one-liner “sphincter-tightening message” which is all we get way too often.
Just as you, I wondered if MT have proper QA lab, early in the v7 release “show”. I’m confident now they don’t. They do some limited and probably manual testing, but not the proper device farm with automated tests and release upgrade matrix. But normis tries to feed us tales of butterfly effects due to driver changes. Right. Is it going to be major investment — of course. I’ve done it for mobile app and it was expense. Is it going to be major quality improvement — absolutely.
Yes we have and it caught most of the upgrade issues when initially testing for v7 release.
That is why I suggest that the changelog be made more dynamic. I.e. you initially get a changelog with single lines, but when you see something interesting (e.g. an entry that may be referring to a problem you have been experiencing, or an enhancement you were waiting for) you can click it or hover it or whatever and get a more detailed explanation of maybe up to 3 paragraphs.
And/or a link to a place where more detail can be found, like a section of an updated manual page (on the RouterOS - RouterOS - MikroTik Documentation site).
There may be fixes that do not need this extra information, but many of them do, and result in repeated asking and discussion in the release topic.
However, you must admit that “improved long-term filesystem stability and data integrity” makes you sit up and go, “Pray tell, may I have > just a bit > more detail, please?” That’s one sphincter-tightening message.
Same for me. They could have put that explanation in the additional info section of a system like I sketched.
@tangent and @gdanov; Even if you have a lab you just can’t test al the different configurations which could cause a cascade of triggered bugs in a piece of software. The way it works is what Biomesh told you. MT is not alone in that regard. So; No, you don’t deserve to know more, even if your sphincter is close to explosion.
And one last thing for mrz. The troll sticker is not the best move. That raises a lot more questions about being professional than necessary.
I believe that to be true for the likes of Microsoft, who didn’t even manufacture most of the hardware their software must run on. Thus why we’re seeing these new initiatives from them where they tentatively roll out new releases as if each one was a beta.
That isn’t MikroTik’s position, though. They made everything but the chips, and they have NDA-level access to the creators of those.
Also, I was talking about the single-device upgrade case, where you exclude things like interactions with foreign network equipment.
That said, I’ve got to wonder why we’re seeing all these OpenVPN problems. Surely that level of interop should be tested before each release as well. Basic things like “does it connect” and “does it stay connected” and “does it connect to all the same client implementations the last release connected to”, and “does it do so under all the same profiles as before”? That sequence should pass for every release, even betas.
you don’t deserve to know more
If that’s MT’s corporate position, they deserve skeptical treatment in return.
The troll sticker is not the best move.
Indeed. gdanov has been impolitic, but never a troll.
And this on the same day one of the least moderate regulars here got a mod bit! It has me shaking my head in disbelief.
This thread is pure gold. Lots of masks going down.