Yes but I think that they introduced some new types of data that the old winbox just doesn't understand and they can't be bothered to patch them in.
Would you be okay if WinBox3 never got CMR? And what does it do when the WinBox3 connects to a controlled client? Is that just rejected? Or, worse, what if they want to support Unicode in CLI/etc as first class – that be a major change to WB3 which is only thing blocking Unicode.
So It's not quite just "patch it", it re-implementing the protocol client in two places. In a code base that imagine is getting long in tooth. And if engineers are working WB3, they are not working improving WB4 which the "WinBox3 forever" crowd claims have bugs.
My thought is MikroTik should set a "expiry date" for WinBox3, so it's not a surprise to anyone. They be trading engineers that could be working on bugs/features, to working on WinBox3... that's a resource choice. At their size, you're taking someone away from other tasks. I know the people complaining about WB3 also complain about "quality issues", if you have X engineers, how do you allocate them.
It's late 2026.
People have built functional clones of the Adobe and Microsoft suites from scratch without even seeing the source code.
The "engineering resources" excuse doesn't hold water anymore.
I refuse to believe that a developer or even an AI with full access to the actual repo can't patch a few new data types into WB3 and fix the bugs people are complaining about.
But sure, don't fix WB3 or WB4. Skip straight to WinBox 7. Just make sure the executable isn't the size of an entire Windows 95 installation next time or even larger than RouterOS itself without a valid reason. Idk.
Sure, but I also refuse to believe that the hardcore WinBox 3 advocates could not move on to WinBox 4 when they would want to spend some time on it.
And there in lays the problem. If it's so good, how does it have bugs? Your bug may not be what someone else considers a bug. Making it work is way different from making it "better". Better is WB4 
You're right it is less work with AI, but that not zero work. And imagine WinBox3 be easy for an agent to fix since everything is defined, and it can use WB4 for the source-of-truth.
But that's also true across the RouterOS bugs too, there is a big list I'm sure, should they spend credits on WB3 or RouterOS bugs?
That's not a matter of believe. It is a 100% factum that it can be done. No excuses. It is an "artificial" (
) decision to let WB3 die.
When the idea of WB was born it was most probably driven by "we don't have the developer knowlegde and manpower to maintain this 2 decades old codebase". Now in 2027 it would need only a proficient Mikrotik C/C++ developer (defining requirements, doing review, steer AI, etc), an (or multiple) AI agent of choice and someone taking project management. A group of 2 human + AI could fix the hell out of WB3 and stuff it with new features as well. But: there is no commitment to do so. Instead it looks like WB4 is a side-project of a part-time developer working at MikroTik.
All I know there not going to kill the WinBox4 project. That is clearly the direction. Adapt. Fill bug report on what's not working on WB4, and they might be addressed*.
Or, keep post all you want complaining, outside the release thread... but doubt the volume of complaints is going to change anything.
Whether dies at 7.25, or 7.xx...at some point it will not be support regardless of AI. Human or agents, it still resources that are limited and have to allocated. So if they decided to have "half-time engineer" (which I doubt is true), should they spend 2 hours on WB3 and 2 hours WB4 responding to agents. Or, should they focus on making the future direction better and addressing folks concerns in WB4.
* And if the bug is WB4 "too big", provide the top/PerfMon etc data showing it consuming a lot of resources when running, otherwise size does not matter
I have the doubt about the "engineer" ...
I think MikroTik has excellent engineering. That's not fair.
It's the business side that's weak. Most companies put out some EOL notices so everyone knows when something will not work anymore. Not just a RN in a beta to find out.
But @normis said it all here:
Release/notification procedures aside, it makes no sense to continue support for a platform locked tool based on a deprecated/obsolete UI. WB3 needs to die, period. Anything else is wasted development time, no matter how much (or little) may be required. Continued support makes about as much sense as MS issuing patches for Win95 in 2026 . . . Stick a fork in it, it's done (despite the modest number of fans thinking otherwise.)
@Amm0
Sure they have excellent engineers and programmers, but there must be an exception to confirm the rule.
If Winbox 4 was Winbox Mark 2, version 0.9 or 1.0, and three or four months since the beginning of its development It would be fine, but It Is v 4.5 and It took more than two years to get where we are.
I don't know what folks expected when they talked about a "cross-platform" winbox... it was going to be a little different. But geez from the commentary, you'd think they changed the whole concept and re-arranged everything. The difference are small, and never had WinBox4 "lose data" or anything really bad, it works.
I get muscle memory, but geez they actually revert a bunch of the orginal concept to actually make it more pure port from WinBox. e.g.
- 2nd level menus are one column that scrolls, just like WinBox3 (it was 3 columns)
- reverted WebFig like non-tab view, so there are tabs
- actions were always on side, and now they put them in the topbar of a list window.
That is MikroTik responding to complaints that it's not similar enough to WinBox. So they do listen, maybe someone liked the 3 columns or no tabs.... but those folks lost out in favor of familiar WinBox3 style.
My only grip in WinBox4 is they could off more options in WinBox4 to tweak more stuff to suite someones taste. My only change was using JetBrainMono which make it look very nice, so I think they did a good job on WinBox4. Now I mainly use Mac so it absolutely an improvement from wine rendering fonts like shit without customization of wine.
I wonder if the defaults were 'winbox3' defaults, would there be less complaints.
And instead of 'hey you can configure it to look almost like winbox3 if you change this and this and this' be like: we made it look as close as possible as winbox3 but now you can also customize stuff if you press here.
Also here's a legacy icon pack if you want it.
That could be helpful, but it could also be difficult to do because the look does not only depend on the version and settings of WinBox itself, but also on the environment where you run it. Native Windows is noticeably different from wine with default settings. I remember I had to tweak wine settings before it looked acceptable in Linux, and I had to tweak Winbox 4 as well. Different Linux installations have different fonts, and DPI settings are not always correct either.
I guess I was saying that's we're at now. Perhaps other than some "expand Tools to buttons" be the only thing I can think of that does not follow the winbox3 schemes. Outside the various complaint I've read about things link window boarders, colors, or other "CSS-like" things, that are not really exposed to control even.
My big grip in either WinBox is the units switching in TX/RX, so I'd like an option for "Fixed Units". That been an issue for 10+ years for me.
Seconded!
You know what? Written language with this ugly "letters" is a very old, antiquated UI.
We should just scrap this, and move to communicating trough emojis instead
We could even go the apple way, with "liquid transparent moji", because who even needs density and readability?
W3: Try understand this
W4:
And also:
Yeah, I totally agree
