ever thought about ssh and tmux/winsshterm ?
40-50 winbox sessions at once? tell me how efficient that is? no really, i do not get it. as soon as i connect to more than 10-12+ routerOS sessions, winbox becomes pointless no matter v3 or v4...
I use it a lot!
Those screenshots were just to exemplify with a case with a lot of devices.
To be sincere, that customer in specific, I don't remember of having to access manually to do human tshoot.
On first take of network of a customer, that's the moment we pass a lot of time searching stupidities via winbox...
A few days of understanding... Cleaning up... Automatizing... Scripting...
And after that, rarely we access all the devices at same time.
so more a niche or a "not daily scenario where there are 30+ winbox sessions at once" rather than a everyday use-case i reckon
To me, yep... Kind of...
My colleagues deals mor directly on a day-by-day work.
So is more probably for then to be with several winbox opened.
The NOC Level2 of our customers... Those for sure are every-day with Winbox opened.
They are addicted to it!
of course NOC guys surely are - but still ... 30+ AT ONCE?
imo SSH is way more efficient here - of course there are edge-cases where it really NEEDS to be winbox.
so for multiple sessions at once upwards 20+, winbox might be not quite the right tool anyway - regardless of v3 or v4 tbh!
Well... It works with Winbox3.
Its presumable that should also work with Winbox4.
If not, should be at minimum openly declared as a not-covered characteristic.
If that were declared, and users aware, live the choice for then.
I will repeat what i Said 5 or 6 times every time Winbox4 comes to discussion(Including in this thread):
over all ... still all just speculation. why? well because a nightly build which is old (because i presume you do not have the password to current nightly builds) - so ... nothing to worry about right now imo
so we all should hakuna our tatas
EDIT: on that ... would have been a good idea to include the 7.25abXXX version in the TITLE!
It will be coming at some point (obviously) but also signaled in the Winbox 4.2 Changelog "add new field type support used by future RouterOS version;" so we are on notice that an upcoming version is going to at least need it for full functionality if not an actual block on connection.
As 7.25 is nightly so far until it appears as a changelog requirement (or enforced in released code) you don't know it is official as they may be testing new features / only "back translate" in the final stages before launch the support for the older winbox.
Regards
Alexander
I'm not a fan of v4. I'm on Linux, and I prefer using Wine Winbox3 than the native winbox v4... So I started to build my own winbox compatible client with the v3 style in mind. Still have a few tweak to do, but it usable.
That said, I'm not sure I will share my from scratch winbox compatible client... I got some dumb response on the Discord server so I might just keep it for me.
UPDATE: The following was me not realizing there is a winbox 3.43 and it caused v7.23 to crash winbox... So disregard my post, it not fixing anything regarding v7.25, and until it available, I can't see if it can be fixed. I kept it in there to show I can be stupid and smart at the same time ![]()
But for the v3 lover like me, you can patch your winbox64.exe so it work with recent router os... It will buy you some time, but at some point it won't be usable anymore...
WinBox 3.42 (64-bit) sha-256 47e83dfd0f9680d2e9623fee92c0acc4db40ea4272edeb53164304620305a24f
If your SHA-256 differs, stop! The offset below is only valid for this exact build.
Five bytes at file offset 0x90217 (589 335 decimal):
Before: 48 8b 0d a2 cd
After: 48 ff c2 eb ca
That will make winbox work again with recent version !
The issue is Address-input fields (type addr), have a allow: string saying which address form they accept. Winbox parse it and it have an 'invalid' new character which cause winbox to crash.
From the routeros change log : winbox do not accept interface without specifying IP or MAC in "Ping To" field;
Anyway, easy fix for now.
One liner on linux :
sha256sum winbox64.exe | awk '{print $1}' | grep -qx 47e83dfd0f9680d2e9623fee92c0acc4db40ea4272edeb53164304620305a24f && cp winbox64.exe winbox64-patch.exe && printf '\x48\xff\xc2\xeb\xca' | dd of=winbox64-patch.exe bs=1 seek=$((0x90217)) conv=notrunc status=none && echo patched || echo 'hash mismatch — abort'
And for 3.43 64 ( 5b41cf51fd81e8d082687f6b522159cf4aba22e6df8edc486721d4e09507b6e9 ) please?
The "Equivalent offset is 0x90227
48 8b 0d d2 cd
but i do not know if after must the same...
Thanks...
Please share the client ! I would love to check it =D
LOL ! I didn't know 3.43 existed LOL ! My fix was to resolve something caused in Winbox 3.42 with recent mikrotik firmware starting with 7.23+ ! So I wasted my time on this 3.42 fix hehe. Oh well ! Sorry for the false hope ! I don't have 7.25 so I can't see this bug and if it fixable in v3
Quick video of my custom MikroTik Winbox UI (very similar to official Winbox v3/v4):
https://www.youtube.com/watch?v=2S30ZEHigrI
I show a few of the improvements I've added:
Better Safe Mode notification
Improved keyboard control
I didn't show the connect screen, but it now remembers router details and warns you if you try to connect to a host and a different MikroTik device answers.
It also saves the MAC of every router you connect to. When you see a neighbor's MAC, it can match it to a saved entry and let you connect with those credentials (even if the entry was originally saved by IP).
Other features:
Automatic config backup + auto-backup during the session (I use /export to limit disk writes, I wish we could do an in-memory backup download)
Full audit recording of all actions
On recent RouterOS, it can try all your saved passwords against a single host (useful for devices I had lying around when I didn't know which password/entry to use). With the new recording of identity/MAC/serial this should be less needed in the future.
Here's another sample of the UI:
It shows the tweaks I've made to the Bridge VLAN Interfaces view so it's much quicker to control and manage a large number of interfaces at once, including better sorting and a clear view of what's already added versus what isn't.
Improvements like this are what I'm focusing on to make the overall client experience better.
I've done other things, but I just can't remember it all. I'm still not 100% compatible with everything yet, but nothing hard left to do, just tweaking.
Other stuff I did, I have a ssh type terminal via winbox protocol, so I don't need to open 2 port(ssh+winbox). Also I tried to tweak it so it usable for "blind" user, but I don't know much about what they need, but I still tried to make it usable with better keyboard control... That said it easy to make a text command line interface using the lib I made which would probably be better for them.
I don't fully like the UI yet... It template base, so it easy to tweak, but it really hard to make it exactly how I feel it right. There is also a mix of Usable <-> Pretty which I'm not a designer... I just work by feeling here, which is not great, but I don't have the knowledge to say what I do like, and what I don't... So be able to tweak small things at a time is nice.
No idea what will come out of it.. But for now, I had fun working on it and it useful to me. One things I didn't test is full routing table from BGP... I'm sure I have a lot of optimization to do there to make it usable ! Maybe implement a Dual winbox connection mode, where I got 1 that handle long transfer list, and one which answer quicker (though I didn't test much if it an actual issue, or it just an UI issue...). Anyway soo much stuff to do !
Has anyone of the complainers actually asked the question: why did Mikrotik require newer Winbox version for 7.25? I can't believe it is a "artificial" limitation. There must be a technical reason. Some improvements or fundamental changes to Winbox protocol, they can't (yet) port to Winbox 3? Something like that.
If I were them, I wouldn't want to support their v3 neither, and update it with new features. But if I were them, I would also address the things that hold back people from using v4 and prefer remaining with v3. It rare people stick to older version, and when it happen, there are good reason for it.
There are reason for forcing people into a user experience that worst... I'm not saying it good reason for the users though... Microsoft is a great example of doing this over the years and still actively doing it. Car maker are probably the worst at this though, knowingly removing sometime people like to put it back a decade later to make it feel like it fresh new things (This is where Tesla flip the switch a little bit on this and part of their success).
I'm not going to comment on whether Winbox 4 is an improvement, but things, especially around wifi seem to have become progressively broken in Winbox 3.
If they have decided to end development on Winbox 3, then this is actually reasonable, if for no other reason than to decrease the number of support tickets and reduce the avalable footguns for newcomers.
I would just hope that if that is the case, that they would officially make the statement that they have decided to "pull the chain" and abandon the deprecated, single platform code base of WinBox3. (Or make a statement otherwise, a timeline, something . . . . ) It's just not rational to make the user base constantly guess and/or wonder if "this is it".
That's how it works nowadays: Every random dude on the internet feels to know it all and to be entitled to berate others without having knowledge about the topic, related technologies and future strategies. They feel their ignorance is as good as other peoples knowledge. And most important they think buying Mikrotik products entitles them to have a saying on what Mikrotik should or should not do.
They just should grow up: Take it or leave it.
End of rant.
In a more "normal" world, they would have killed WB3 with ROS7 (not sure how the timelines aligned though.
A totally new OS with a totally new UI would have made lot of sense, although a level of rearward compatibility to allow migration of installed devices does as well.
The only 100% true statement is that it has to die at some point. Two radically different code based being maintained and developed to do the same thing is close to the definition of insanity. (ROS6 and ROS7 is another pretty much identical case . . . )
At the end of the day, some folks will always whine and complain, but if the new version can perform all the tasks the old one can (and likely more), at some point, personal preferences become nothing more than noise (especially if the number of folks griping is minimal).
IMHO . . .
In that scenerio... MikroTik would have started with a new version of "The Dude" and/or "new controller", tried the UI there etc. That was the need, IMO.
And work through the UI kinks there with Qt framework in a "New Dude"... and then come back to a WinBox refresh using same framework... Instead... we're left with two winbox clients, but no tools for managing multiple MikroTik's than the decades old Win32 dude.exe.
But ship sailed long ago. So I'd imagine WinBox3 will be dead for new versions at some point... since improvements to WinBox4 (like better errors/help/validation) likely require changes to the WinBox Protocol itself so it router can offer new UI-based data/properties/help/etc.