With the Dude being officially retired now, I would love some clarity from MikroTik if they plan on developing an updated solution like they did with Winbox, or are they throwing their support behind a different solution like NetXMS, Zabbix, etc? What are they recommending?
Thanks!
MikroTik Devices Controller soon?
I think they plan to open app store (Docker apps running in container on routers). Or why they added /app? Maybe paid apps, maybe not, maybe some
When can we expect the Dude source code to be released for community development?
Something similar happened when they stopped updating the x86 version of the server… later they ported it to RouterOS package, but the improvements were minimal.
It would be fantastic if they released the source code so the community could continue its development: many of us here love "THE DUDE."
This is like having a funeral years after the person died. Glad MikroTik formalized what was already true.
They've been talking about a controller for years. I'm sure they are working on something. MikroTik likes their surprises. I can see the UI could be troublesome in 2026 (e.g. looks like a souped up MFC-app from 1990s). Yet my only complaint was you could not move one device from a map to another map - that is limiting (but after years of using it, you know to carefully think about the maps you want).
But what I find powerful in The Dude is its data model. I hope the don't lost that part, that's what make the Dude great. You could "slap another UI" over the existing schemes and be done, IMO. Many things in the Dude are quite sophisticated (e.g. "CAPsMAN-like inheritance"), and everything was rationalized. What I liked was it was just give it a disk, enable, and a few clicks it does actually monitor a network (with some understanding of the complex options of auto-scan). No heavy weight Postgres/mySQL/Promesthis/etc/etc, just SQLite tunable to mange historical stats vs data size. I have one Dude instance that go back ~10 years now.
But what does that announcement actually mean in terms of functionality?
Will the Dude package no longer be available in new versions?
According the announcement, yes
“no further updates, enhancements are planned for the foreseeable future.”
Would it be too much to ask to Open Source it now that we are in the AI era?
Now that we are in the AI era, it would probably make more sense to reverse engineer the communication between The Dude and ROS using AI, and then implement a new “The Dude” lookalike.
The source code of this very old application would likely need to be heavily refactored or even partially rewritten by AI anyway, so investing effort into the existing codebase may not be very practical.
I am not aware that MikroTik has ever open-sourced anything in its entire history. Not even parts of ROS are open source, even though it is based on Linux, the epitome of open source. So why would MikroTik start doing that for such an old software product?
Most likely, the source code even contains secrets or other sensitive information, which would create a security problem that did not exist before.
Just to not start from scratch. What communication between the dude and ros could be needed to reverse engineer? Isnt it polling information like any other monitoring tool would?
agree with @infabo - It's not needed.
And there is CoreBit – Network topology & monitoring tool inspired by “The Dude”, that I'd have to guess was built with some assistance of AI, and seem to have good reviews from forum. Whatever anyone think of AI, it very good at generating web pages and SQL, or reverse-engineering stuff too. So don't let the fact /interface/bridge is also a mystery to LLMs fool you in what they can do ;). CoreBit opted to "standard" monitoring things, thus the "inspired" by
Well, at least my AI (with instructions, prompting, access GPL code, previous code I wrote, etc) has a good handle /console/inspect, /app, and CHR upside-and-down, see my tikoci.github.io which I cleaned up with CoPilot/Claude with some effort on my part for sure, but trust me the robots worked hard (e.g. hours on QEMU "guest tools" on X86, /app editor, OpenAPI3, and cleaning up webpages, etc.). Now I have a background in software engineering, so I know what I want pretty specifically and fill in gaps in RouterOS like is someone in "Beginner Basics" here, so that helps.
But we have not had the clunkers take a look at the Dude yet. It's on the list. I want to unwind the Dude SQLite database, since I've long wanted to extract information, and likely from getting to bottom of the SQLite schema used by The Dude, it have a good sense of the protocol.
Now, AFAIK, The Dude uses same winbox protocol, just with different "TLV IDs"... This would be a bit more work and more effort. Especially to get right, since it likely figure out the basic protocol (and likely already in training for NSA/other leaks/security reports/etc long ago). The problem is it's a packed binary protocols using IDs for everything, which is very efficient, but the mapping of those IDs packed in structures to path/commands/attributes, likely be much more work. And The Dude fits in the "winbox" protocol, some how. See Open-Source Python Reimplementation of WinBox Terminal Access who has implementation of "terminal-over-winbox" so I suspect The Dude is just different set if some TLV ID in winbox that have to be map (for dude likely a bunch more than opening the terminal which does not attribute mapping)
And that is the key question here... If by "retired", they mean the Win32 dude.exe app only... then I'd imagine that some AI could re-create the Dude (without the source, only screenshots, and direction if you really wanted a Win64 version of it, or something web-based...) - over the course of many hours, many tokens, and good prompting.
But if "retired" means the dude.npk you install... that's a different story since there really be no dude. Nothing to reverse engineer then. Since I think it's Dude data model and architecture are excellent, and give works by only package install, not local hardware needed...it fills an important need (e.g. I have a lot of LTE devices, so some local dude that just collects the signal other data from a set of LTE devices at a site, is just so simple - no container, no X86, one package install). And the dude.npk and efficiency of winbox protocol that still has legs...it's packed binary so less on wire than SNMP or even native API.
The Win32 client can go away IMO... but the core architecture embodied in the how The Dude works is hard to replicate. In fact, it more sophisticated in some way than many people think since it even has some notion of hierarchical dude server, that the Win32 API did not really help with, and I still not sure how to use... but allow rolling up data in UI without a central database.
The Dude abides. Maybe?
@Amm0 thats an insane amount of information and knowledge, give me a couple of days, ill get to it en opensource it
Do we really need dude?
There’s actually a really beautiful monitoring application, I’m not trying to advertise it,but it’s from Latvia and starts with “N.” Those who know, know 
New WinBox?
No, NotThatOne
There’s actually a really beautiful monitoring application, I’m not trying to advertise it,but it’s from Latvia and starts with “N.”
Come on 
, give us some hints, how many letters?
netxms
I would say slightly more expensive than The Dude.
It somewhat reminds me the time I was suggested to buy a Lamborghini Gallardo when I had to replace my Fiat Panda and the new Fiat Panda wasn't available...
@Amm0 send me your github account so i can add you to the repo , i need some feedback
Side note here on migration from Dude to something. I did find this project: GitHub - german77/TheDudeToHuman: Converts "The Dude" database from Mikrotik into something more usable than binary blobs · GitHub project at some point that extracts a dude.db into something more usable. I have not extensive testing but it works.
I have a research project, donny GitHub - tikoci/donny: Experimental: TypeScript library and CLI for MikroTik The Dude network monitoring database (dude.db) · GitHub ), which has more lower-level details in TypeScript on The Dude internals, but it's not really designed to be user-facing since it feeds into another project trying to dissect winbox protocol ("m2ir"). But the TheDudeToHuman is worth a look for migration.