I don’t know about you but I am getting d*&% tired of having to change my scripts EVERYTIME Mt puts out a new ROS version. In fact I have spent over 8 hours just trying to get this one little line to work - later in the post.
How about it MT - can you stablize the scripting parameters NOW instead of 5 versions from now? I didn’t mind the original change from ros2 to ros 3 - that was to be expected with the new capabilities of ros3. But the simpler script commands should not have changed much from 2 to 3. But in ros3, every version has issues with scripts - and ‘you’, (MT), tell us LITTLE to NOTHING about the changes. Nor does it seem that you have tried any scripts before posting the latest ROS.
I don’t know if ‘you’ (MT) realize it or not but most of us are trying to run reliable business models out here in ‘userland’, I have over 5000 wireless clients, backup lines and links, how the H#$% am i supposed to be able to sleep at night (or my support staff) if we have to constantly monitor everything because a SIMPLE script does not work??
Ok - done ranting.
First run (and in console as well) :global gw $gw 0; global gw1 $gw1 0
:environment print gives, “gw” =0 “gw1” =0
Here is the simple line; :global gw; global gw1; :if (($gw =0) && ($gw1 =0)) do= {:put done}
I get; conditional is not boolean
But if I just do; :global gw; :if ($gw =0) do= {:put done} OR :global gw1; :if ($gw1 =0) do= {:put done}
Works just fine. I even tried changing out " && " for the keyword " and ", same issue.
i agree MT please respond to this. It is not correct for someone to change the way the damn script works and not tell the people that are writing them . i have spent numeros hours debugging stuff that used to work fine on previous versions . && is completly broken in 3.4 and a host of other problems that we are haveing to work around . i see a host of post on this 3.4 not doing script correctly and MT just sets back and says nothing not even a hint of help here.
anyway I love MT but this scripting issue has just made for a couple really bad weeks.
MT people, I hope you read this. I have upgraded most of my routers to 3.4 because one problem I had was solved in this.
but now I find my most important script is not working any more.
As I see in this forum it is the result of the new ros3.4 package. Are upgrades not mend to solve problems, instead of creating new?
My auto fail over system depends on it and thus the reliable working of my production network.
Can we expect a new 3.5 soon with corrections or do I go back to the working 3.2 environment?
Is it possible to document the new script and have it also test against the most commonly used ones to see they are not broken in new ros releases?
And put a notice that differences do exists in between ros 2.9 and 3.x version family. The Wiki now has several scripts that probably bring more users in problems then they would help them only because the scripts might not always working under the different ros versions. The question becomes now; ´When to use which script language?¨ regarding the several ros versions.
I think MT should do us, and themselves a big favour to solve the problem and to warn users for the differences.
On this moment I am very disappointed in this while the MT ros product and hardware otherwise is performing very well and to all my wishes!
PLEASE LET US KNOW WHAT IS HAPPENING!
OK, there is a serious bug in scripting in v3.4, we just tested 3.5 and the original poster issue is solved, also this (http://forum.mikrotik.com/t/3-4-script-weirdness-put-string1-string2/19627/1) is solved. so this is not us changing scripting in every revision, it’s just a bug which has been fixed. Sorry for not replying earlier.
It is very frustrating to have a group of rather large scripts (the snipet above was the crux of our issue) that either tell me they are running, or the scheduler telling me it has run the required item - yet nothing shows as having been run - and then having an actual failure show and failover that you depend on to watchdog your network does not work…
I wanted to make the point that this can’t be the ‘norm’. I understand about writing code - I was a programmer and a software project leader starting in 1974. So I do have an appreciation of what it takes to make software like this work. A little more through testing on MT’s part would be prudent in the future. I can supply some scripts that use several key scripting elements, Sam (ChangeIP) also regularly submits scripts (Hi Sam), he could be another source and there is the fellow over at Cable Free Solutions - sorry I don’t remember his alias right now - he also regularly submits/corrects scripts.
Anyway - I did want to thank you for answering the mail, it is good to know you guys will ‘take it on the chin’ for issues like this and not take it personally.