OpenWrt 21.02 (current TOS 6) is discontinued since May 2023, OpenWrt 22.03 (TOS 7) was released in September 2022 and will be discontinued soon, OpenWrt 23.02 (TOS 8?) will be released soon. New major TOS version development activity is below low. Quo vadis Turris?
TOS 7.0
TOS 8.0
Now, turrisOS always trailed behind OpenWrt considerably, so that fact is nothing new. And team turris by now has a relative heterogeneous hardware zoo to keep in sync… IMHO things are working as expected, folks unhappy can always switch to upstream OpenWrt, at least for the omnia (no idea about turris1 or mox).
That said, it would really be appreciated if someone of the official team could write a short post with the team’s current estimates/predictions about tos7 and tos8 and beyond? Maybe with a note what the bigest expected/experienced hurdles are?
I was expecting a similar reply and would like to point to the very important thread. It is unfortunately in Czech, but it should not be a problem for DeepL and co.
I totally agree, it’s been almost year since the last statements and it doesn’t look like anyone is increased paying attention to the release of TOS 7.
I guess your point is the stated expectation that the switch from 6 to 7 would be easier/quicker than the protracted change from 5 to 6? I think we had similar statements in English threads as well… Anyway we seem to fully agree that an update about the official plans would be really nice to have (and I understand that plans can change, be it because of unexpected issues taking longer to fix than anticipated or lack of man-power, so plans are not guarantees or even promises).
I personally agree.
Yet I think they have good reasons to focus for the given devices right now on bug fixing, because they are in the background working on finalising the announced upcoming devices (said to be ready for Q3/2023). So I don’t expect answers before end of September 2023.
Sure, but if these statements are made at a time when one transition is just completed, one would expect the relevant knowledge to be at the maximum possible level. The corresponding OpenWrt was already available at that time.
If we are talking about personnel, at the time of the release of TOS 6, someone was still working on the software side of Turris. If there was no release of TOS 7 under by the then the staffing, it’s hard to release it now that there’s hardly anyone there anymore.
The problems of the current TOS 6 compared to TOS 5 are still not solved, notwithstanding earlier promises. Of course, they were only promises.
What exactly are you missing?
IGMP issue(s)
LED issue(s)
SDIO issue(s)
For example updated U-Boot (by official distribution)
And finally…
TOS 6.2.0 unresolved
TOS 6.1.0 unresolved
TOS 6.0.1 unresolved
TOS 6.0 unresolved
This seems to be an upstream kernel issue not a TOS issue, not withstanding what affected users here keep repeating… sure team turris did not manage to fix this, but IGMP was as far as I recall never explicitly advertized as feature.
Sure the LED state retention over reboots on omnia is sub-optimal. On my personal list of grievances this is far down however as I consider that to be mainly cosmetic, but I understand others see this differently.
This I think is MOX only? No opinion on this.
This however is well explained in the wiki, took me around 5 minutes… again would be nice to have this updated more automatically (that is asking the user explicitly, the risk profile of thid makes me not want this to be applied fully automatically).
Regarding unresolved issues, any project of this magnitude is going to have those, be it because the underlying issue is (too) hard to fix, be it by under/non-responsive reporters.
Some projects routinely wipe issues inactive for too long, not sure whether that is actually better than letting them hang on.
So considering I bought my omnia in the crowfunding campaign, team turris has not gotten any money from me in more than half a decade and they still maintain the software I am hardly in a position to complain (but then with an omnia I do have the option of switching to upstream OpenWrt if things stay/get too far behind over here).
So I do not want to negate your points, they are all valid in some sort. Just personally I am still OK with team turris 
I don’t believe that is in their hands to be resolved. Yet it is by no means acceptable, agreed. I myself gave up on it - my single SDIO will only be used for non-security sensitive IoT-devices on a poor WPA2 2,4GHz SSID connected to TOR, other devices will need to use the 5GHz-SSID with WPA3.
Should be resolved by today’s 6.4.0 RC2.
It has always been common in Turris to close issues with a comment, or to move them between versions. This has stopped happening.
But solution exists.
I’d be happy for it, but a year full of one-line edits kind of indirectly speaks to the point I made at the beginning of this thread.
How do you tell this new firmware really solves our issues (maximum parallel connected devices, etc)? The thread tells NXP keeps it under NDA…
The firmware is compatible with newer kernels. In addition, it is not the last one to be released. Don’t you think the Turris team should simply get it and test it?
Don’t you think they already tested it, if it was released in 2022? The fact Turris team afterwards didn’t release it for us may have other reasons (like no benefits/solutions for our issues etc).
There is no communications manager telling us about successes / time schedules / ETAs, so for sure nobody would tell us about failed attempts with pieces of software not worth talking about… 
Hi Viktor, what are the features that you find beneficial in later openwrt releases?
Not being Victor, I am looking forward to the nftables based firewall4, as well as being able to test qosify (uses ebpf for packet classification/DSCP assignments for priority scheduling)…
I think the debate if we even need a new major version of TOS is out of place. Just as it is not appropriate to debate whether it is necessary to test new firmware for non-functioning hardware. SDIO is still on sale internationally and I don’t see anywhere in the offering at least a warning about the card’s major limitations.
The various firmware versions and SDIO behavior has been addressed in the forum, and the absence of any communication manager on the part of the Turris team has not been at all objectionable. That SDIO doesn’t work with newer kernels was known by the Turris team long before the release of TOS 6 and promised a fix. At the same time, I know very well that the firmware versions I reffered were not tested.
I have been using firewall4 and nftables on TOS7, although I have to re-install all of the packages after a pkgupdate because they don’t seem to want to use it officially yet.
However, qosify is still not in the package repositories. The turris team tends to leave out quite a lot of packages that they are not interested in, personally.
@ackstorm23 TOS6.0 is based on OpenWrt 21.02 and there is no qosify as well so thats expected. It was merged in 22.03 so probabbly you are not going to see qosify until TOS7.0. But cake algorithm for sqm is available already for TOS6.0.