I'd like to see a RouterOS GUI client like WinBox one day which works on Linux and BSD (and Mac maybe).
Do you have plans and manpower to develop it?

We can access our routers with OpenSSH, mactelnet and run WinBox in WINE of course, but native GUI would be better.

winbox works perfectly in linux with wine. Years using it.....

Wine is designed to run on x86 plaform. How about arm platform? Many new devices are based on arm. On Raspberry pi 3, i succesfully running winbox.exe after installing Exagear Desktop. Commercial from eltechs.com

winbox works perfectly in linux with wine. Years using it.....

Not really working perfectly for me when using Ubuntu 17:10 Wine,
In Ip Neighbors works only when router has DHCP running but will not list routers when removed?
No drag and drop from desktop to winbox or reverse?

I'd much rather have one really good windows app than have the winbox team spread out across multiple platforms.

Besides, there's not a single scenario where you'd absolutely need to have winbox on another platform anyway. If wine doesn't work for whatever reason, there's dual booting, VMs, or even having a dedicated windows machine(PCs are cheap..)

I'd much rather have one really good windows app than have the winbox team spread out across multiple platforms.

Besides, there's not a single scenario where you'd absolutely need to have winbox on another platform anyway. If wine doesn't work for whatever reason, there's dual booting, VMs, or even having a dedicated windows machine(PCs are cheap..)

But ROS is built on a Linux Kernel and I don't want a Winbox but a "Linbox" version and why there isn't a linux version is a mystery?

There is already linux style option - SSH CLI

If you like click with mouse in windows then it is time for Windows...

I'd much rather have one really good windows app than have the winbox team spread out across multiple platforms.

Besides, there's not a single scenario where you'd absolutely need to have winbox on another platform anyway. If wine doesn't work for whatever reason, there's dual booting, VMs, or even having a dedicated windows machine(PCs are cheap..)

But ROS is built on a Linux Kernel and I don't want a Winbox but a "Linbox" version and why there isn't a linux version is a mystery?

It's no mystery to me that mikrotik picked the most popular platform in the world for their config utility... (and not .0000...00001% of people who use linux and mikrotik)

Winbox is essentially just a graphical representation of cli so even though routeros is built on the linux kernel, there would be zero benefit to having a linbox version anyway.

I'd much rather have one really good windows app than have the winbox team spread out across multiple platforms.

Besides, there's not a single scenario where you'd absolutely need to have winbox on another platform anyway. If wine doesn't work for whatever reason, there's dual booting, VMs, or even having a dedicated windows machine(PCs are cheap..)

But ROS is built on a Linux Kernel and I don't want a Winbox but a "Linbox" version and why there isn't a linux version is a mystery?

It's no mystery to me that mikrotik picked the most popular platform in the world for their config utility... (and not .0000...00001% of people who use linux and mikrotik)

Winbox is essentially just a graphical representation of cli so even though routeros is built on the linux kernel, there would be zero benefit to having a linbox version anyway.

We have found that using windows on installation laptops that their network adapters have to be reset on a regular basis and not sure if running on Linux that the same will occur?

There is already linux style option - SSH CLI

Not every Linux user is a strange person who loves command line, edits texts in vi (or ed ), etc. LinBox would be great, because WINE is not exactly light dependency. On the other hand, there are fifty Windows users for every Linux user, so the extra work might not seem worth it for MikroTik. Especially since there's a way how to use WinBox on Linux.

I'd much rather have one really good windows app than have the winbox team spread out across multiple platforms.

Besides, there's not a single scenario where you'd absolutely need to have winbox on another platform anyway. If wine doesn't work for whatever reason, there's dual booting, VMs, or even having a dedicated windows machine(PCs are cheap..)

But ROS is built on a Linux Kernel and I don't want a Winbox but a "Linbox" version and why there isn't a linux version is a mystery?

It's no mystery to me that mikrotik picked the most popular platform in the world for their config utility... (and not .0000...00001% of people who use linux and mikrotik)

Winbox is essentially just a graphical representation of cli so even though routeros is built on the linux kernel, there would be zero benefit to having a linbox version anyway.

We have found that using windows on installation laptops that their network adapters have to be reset on a regular basis and not sure if running on Linux that the same will occur?

If you need to reset your network adapters on a regular basis, you're definitely doing something wrong..

There doesn't have to be 2 separate apps, Mikrotik can abandon WinBox, it's UI looks a little archaic anyway and release a cross-platform Qt app

Having only a console/terminal interface is less convenient for many people, same as having only GUI app. If it was OpenWrt, I'd go with SSH but RouterOS is more complicated and GUI allows for faster overview of every available feature and settings. The true way for Mikrotik is having both console and GUI.

The problem with WINE as people have already said is that it requires lots of deps and many libs have to be built as 32-bit version, it's an overkill for using a small app (WinBox) not so often.
Same for dual-booting, VMs with Windows and buying "cheap PSs" (requires space and power) and Windows license.

Also, I couldn't find out how to read changelog for RouterOS updates in SSH session, I use WinBox for that.

I hope that mikrotik will not abandon the winbox. It is really useful and advantageous.

They would be crazy to abandon it, it's the killer feature, at least for me. But rewriting it as cross-platform (one code that you can compile for Windows, Linux, ...) could make sense in long term. But from a selfish perspective of Windows user who likes current WinBox, there's million of more useful things they could do.

Exactly.

There doesn't have to be 2 separate apps, Mikrotik can abandon WinBox, it's UI looks a little archaic anyway and release a cross-platform Qt app

Having only a console/terminal interface is less convenient for many people, same as having only GUI app. If it was OpenWrt, I'd go with SSH but RouterOS is more complicated and GUI allows for faster overview of every available feature and settings. The true way for Mikrotik is having both console and GUI.

The problem with WINE as people have already said is that it requires lots of deps and many libs have to be built as 32-bit version, it's an overkill for using a small app (WinBox) not so often.
Same for dual-booting, VMs with Windows and buying "cheap PSs" (requires space and power) and Windows license.

Also, I couldn't find out how to read changelog for RouterOS updates in SSH session, I use WinBox for that.

lol this isn't a calculator app.. not only is the approach to writing a cross-platform app entirely different, but winbox is deeply integrated with routeros which means they would have to rewrite that too... Do you have any idea how many additional variables and subsequent bugs that would introduce?

It's perfect as-is and the "archaic" ui is extremely functional, lightweight, no-nonsense - the kind of tool that the vast majority of mikrotik users love. They have been polishing it for over a decade so there is no way they are going to abandon/radically change it.

There doesn't have to be 2 separate apps, Mikrotik can abandon WinBox, it's UI looks a little archaic anyway and release a cross-platform Qt app

Let's hope Mikrotik don't fix the "archaic" appearance of Winbox by turning it into some low-contrast abomination with gradient blends and animations.

The fact that Winbox works very well in wine must meet some definition of "cross-platform". I have seen screenshots on Mikrotik.com of Winbox running in wine so I am reasonably certain that winbox developers are testing winbox in wine, not just on Windows.

Webfig is the real cross-platform interface. It's not far off Winbox; just needs an easier way of opening multiple windows.

It would be good if Winbox could read/write backup files and/or the output of /export

+1 - I love Mikrotik but have no Windows machine. i work with Linux and MacOS. Linux is mich closer to Milrotik wjen it comes to scripting.

Web and SSH are already as cross-platform as it is possible to be. On top of that, Mikrotik's feature-complete CLI is much easier to use and learn than competitors, afaik.

No one but some subset of power users would want anything more, and power users already have access to emulators and VMs, including entirely for free.

Those who keep asking for some native-like, cross-platform GUI solution are simultaneously severely underestimating both the high upfront development costs and long term support costs. Instead of effectively one platform on by far the most popular non-mobile, native GUI in the world - if it wasn't Windows it would be something else - now Mikrotik would have to test and support all additional platforms forever.

All the while, there would be significant inefficiencies across the board in having at least two well-developed GUIs, assuming a perfect overlap with a cross-platform library (!). Adding or improving features to all these advanced user interfaces would also suffer. There is also no guarantee of being able to replicate the responsiveness or all the GUI features of the current Winbox. So, now you risk a trade-off ...

Such enormous use of resources that could be better spent right now to bring real features to you makes zero sense.

You could be asking Mikrotik for significant features compared to the market, features reducing wasted time, features broadening access to more customers, features making your purchases inherently more valuable to you, features bringing the product up to date, or enabling you to recommend them wholeheartedly to others.

Even the most fundamental things like a public bug tracker and public feature request tracker - where you could have put these requests for visibility or voting - are missing ...

What are you thinking?

Mikrotik is no small company. If they don‘t have the recources to develop for multiple operating systems it would be seriously conceirn me.
The most used system on the road is Android an iOS - not Windows.

Webfig is the real cross-platform interface. It's not far off Winbox; just needs an easier way of opening multiple windows.

I think the "way to go" is to merge WebFig and WinBox. I.e. make all WinBox functionality available in WebFig.
There could be some separate tool that provides a MACtelnet/MACwinbox gateway to HTTP that can be started
on the local computer to have MAC-level access, and it could also perform tasks like netinstall controlled from the
browser.
This should preferably be available on the common desktop OSes, and probably be open-source.
It does not need a GUI so it should be "quite easy" to write it as a Posix compatible program.

I think the "way to go" is to merge WebFig and WinBox.

I think I would hate that. WebFig is nice, it allows you to set everything (at least I didn't notice anything missing), but compared to WinBox, as was already mentioned, current version lacks multiple windows. It makes it much less convenient to use, because it means endless switching between different views. It's not impossible to fix, I've seen different attempts to have windowed user interface in browser, but not a single one was pleasant to use, it's always somehow slow, laggy, ...

As I wrote, with a merge of WebFig and WinBox I mean to put all functionality of WinBox into WebFig.
Of course that includes the way of handling separate windows etc.
This is not at all impossible. Environments do exist that implement functions similar to WinBox in a browser window.
Of course it requires way more processing power on the client side, but that is what everyone is expecting already.
(see the state of the art in websites like Youtube or Twitter, that require multi-GHz processors to render a simple page
within 10 seconds)

As I wrote, with a merge of WebFig and WinBox I mean to put all functionality of WinBox into WebFig.
Of course that includes the way of handling separate windows etc.
This is not at all impossible. Environments do exist that implement functions similar to WinBox in a browser window.
Of course it requires way more processing power on the client side, but that is what everyone is expecting already.
(see the state of the art in websites like Youtube or Twitter, that require multi-GHz processors to render a simple page
within 10 seconds)

You are obviously not coming from an web developer field. Javascript gets basically compiled in the browser and is blazingly fast when it comes to simple tasks like a Winbox. Think of Google Docs or even Designer Apps like Figma. Html and Javascript have come a long way in the recent years. Compared to some Webapps Winbox is just a tiny application that uses RouterOSs APIs - Router OS is the real deal.

You are obviously not coming from an web developer field.

No, I come from the user-that-is-tired-from-having-to-upgrade-computer-all-the-time-to-get-decent-performance field.
I detest the "developers" that always get the fastest computers and declare their developments fast and usable, while
leaving the world with almost unusable slowness. That was already my opninion when Java was introduced.

However, I don't understand why you send this reply quoting my text, because I already confirmed that it would be
possible to make something like winbox entirely as a web interface. It was the others above that questioned this.
I think it is really possible, I only fear the slowness. But that could mean I finally need to update my machine because
Intel(R) Core(TM)2 Duo CPU E6850 @ 3.00GHz with 8GB RAM apparently is not enough anymore to run a browser.

I also remembered Java. It too was supposed to be cross-platform and fast, and it sort of was. As long as you had the right Java, if you didn't mind ten times longer startup times to load million libraries, double used memory, and wasn't too picky about how things look, because it was wrong on all platforms. Once it got going, it did work, but the whole impression wasn't great.

For modern web-based applications, you also need the right browser, it takes time to load, browsers are not exactly memory savers and look of each UI is original, but nobody seems to care about that anymore.

My objection is that some stuff just doesn't belong in web browser, but I admit that it may be more of a philosophical point. That's why we have operating systems, to allow us to run different programs, optimized for given purpose. It's all those tiny things like having own window, instead of being burried somewhere between other browser tabs (I know that browsers can have multiple windows too, but that's just not it), standard consistent and predictable GUI (too bad that "make your own GUI" seems to be the trend everywhere), etc. Web UIs are great too, but as a simpler extra option.

I agree that a Browser might not be the best choice. But the current state of Winbox ain‘t it either. I got multiple Windows 10 machines where dialog boxes get stuck and i have to kill Winbox via the task manager. It feels like Winbox is in an emulator even on Windows.

However, I don't understand why you send this reply quoting my text, because I already confirmed that it would be
possible to make something like winbox entirely as a web interface.
Intel(R) Core(TM)2 Duo CPU E6850 @ 3.00GHz with 8GB RAM apparently is not enough anymore to run a browser.

I just wanted to tell you that performance is not going to be an issue. Your system is perfectly fine and will do a decent job - we develop web apps in our company and we support systems with way less power under the hood.

My objection is that some stuff just doesn't belong in web browser, but I admit that it may be more of a philosophical point. That's why we have operating systems, to allow us to run different programs, optimized for given purpose.

That is the old-fashioned way. My laptop is a Chromebook, and it has absolutely nothing installed besides the browser and a Citrix receiver.
When looking at work (more representative), the vast majority of what used to be installed programs has been replaced by cloud services accessed via the browser.
The advantage of this solution is that you are always ready to go, no matter what computer you work on, no need to install software or find a USB stick with portable application.
(and of course a well managed computer would not allow you to run software from a USB stick or the download directory...)

I find parts of this discussion funny since Webfig is practically a web version of Winbox as it is. You can even call up a windowed terminal in Webfig!

The main thing that could not be implemented in a web version is the "MAC-Winbox" protocol, which is a killer feature that I use all the time. For this reason, it seems unlikely that MT is going to give up on Winbox anytime soon.

If anybody truly wants a native, compiled GUI management app for $PLATFORM, MT has given you the ability to craft one yourself with the API. Unfortunately, it also isn't possible (as far as I know) to direct API calls to a MAC address, so you would still be unable to write a "MAC-Winbox" substitute with the API.

-- Nathan

I find parts of this discussion funny since Webfig is practically a web version of Winbox as it is.

-- Nathan

Webfig can‘t even copy entries.
The lack of Mac support renders it a supplemental tool that can never replace Winbox. The API i good, I used it myself, but it uses a different syntax than ssh and imho you shouldn‘t rely on third party tools when you wan‘t rely on functionality.

The main thing that could not be implemented in a web version is the "MAC-Winbox" protocol, which is a killer feature that I use all the time.

That is why I suggested the development of a separate tool that you can run on the local machine, which serves as a relay between MAC based
access and IP based access. Just makes available a local IP socket on the machine, and sends everything back and forth between that IP socket
and a raw socket that uses the appropriate ethernet protocol. You could then start this tool when you require MAC based access (of course it has
to support RoMON as well).

Then, all other tasks can be done by WebFig. It could be convenient when the tool also can do netinstall with the user interface in a browser window,
but of course it would be a bit more complicated (would have to include a small webserver). But such code is available as libraries.
The tool should be written in some widely available "scripting" language like Go, Python, Perl, etc. So it is easy to run it on many platforms without
having to use workarounds like wine.

Webfig can‘t even copy entries.

Of course, as I wrote, WebFig has to implement everything that WinBox currently has. And there is not really much left:
- multiple windows
- select columns in lists (persistent, preferably also dragging column order)
- filter
- Inline comments
- copy records

I find parts of this discussion funny since Webfig is practically a web version of Winbox as it is. You can even call up a windowed terminal in Webfig!

Can you paste code snippets into WebFig's terminal? It doesn't work for me. I'm working in an all Mac shop, has been for decades. There is some configuration that needs added to new devices which takes a while to click around in WebFig to add. I suggested, "Just copy and paste this into the WebFig terminal." Lots of egg on my face. I tried with Firefox, Chrome, and Safari all on MacOS X. I just tried Chrome on Ubuntu 16.04. Apple took away Ethernet ports so I got a Dell w/Ubuntu yesterday. I don't have a Windows machine, or license to load in a VM, to test with.

I use winbox under wine and also know how to delete 192.168.88.1 from my ~/.ssh/known_hosts. The installers think I've added an hour to their installs vs UBNT AirRouters with four tabs of settings pages. We can't use quickset for everything that needs to be set. Going back to quickset after the other stuff has been set is dangerous. Quickset will undo the management rules for the firewall, and re-enable the 5GHz radio on the hAP AC lites. The installers often accidentally switch the quickset mode to CPE or something other than Home AP dual, then things get hosed and they are locked out, mostly because they don't have the MAC access method without Winbox. I think they may just be using the scroll function of the trackpad when the mode drop down ends up under the cursor. I'm having that issue with system preference dialogs on Ubuntu. I haven't spent enough time in webfig to replicate the issue for myself.

I've put a bottled winbox on their laptops but they don't want to use it because it's not a MacOS program. They also think it's chewing up RAM and making their laptop slow worse than I see on my MBP, which is older than their laptops. Some of it is just the techs whining about the new procedure. Nobody likes change. Our techs happen to be over 50, so there is more bias against change.

I would feel better about WebFig, despite it's shortcomings, if the default routeros config would create a self-signed cert and use it for https access to WebFig. If the admin want's his own cert, he can delete the auto generated cert later, but secure our access by default, please!

I understand how difficult it is to build a web app with the functionality that Winbox has. Each browser has it's own quirks and sometimes it feels like it would be no more difficult to build a standalone app for multiple OSes. Browsers and code libraries can probably abstract away 70 to 80% of the cross device issues whether you're doing a web app or standalone. The rest is where it gets "interesting."

In the OS/2 days, some developers with Windows source code were using the Wine source to build OS/2 apps. The distributed binaries weren't that much bigger than the windows app. When they compiled in only the APIs they needed it didn't bloat as much as the wine bottler approach. The developers could also add in small chunks of OS specific code where the wine APIs weren't good enough. I wish MikroTik could / would do that for the MacOS X / Linux binaries. Maybe they've tried and found it unworkable already.

The developers could also add in small chunks of OS specific code where the wine APIs weren't good enough. I wish MikroTik could / would do that for the MacOS X / Linux binaries. Maybe they've tried and found it unworkable already.

Couldn't agree more.

Winbox works fine most of the time under wine, but not with some quirks, .e.g. winbox window losing "focus" and being impossible to type anything on any opened winbox until all wine processes are killed, which usually happens exactly the time you're most busy with an urgency...