[Info] Freetz-ng master SVN nun auf BoxMatrix

Shirocco88

Aktives Mitglied
Mitglied seit
4 Jan 2016
Beiträge
870
Punkte für Reaktionen
67
Punkte
28
Zuletzt bearbeitet:

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
I tried to compile an International 6.8.x and it failed on downloading images from the AVM-site.
Because I had these already downloaded in my old environment I was able to copy these files to the freetz-ng and was able to continue.
 

stoney

Moderator
Teammitglied
Mitglied seit
7 Okt 2015
Beiträge
5,147
Punkte für Reaktionen
380
Punkte
83
Ask HERE for the FW-Image and copy it into FREETZ.

Otherwise you do something wrong on copy the image from one to the other folder.
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
Unnötiges Vollzitat entfernt - HabNeFritzbox
Maybe you should read my post again.
I don't need to ask for FW-images. I have them as I'm building Freetz for more than 5 years and I have those in my old trunk.

I'm merely saying that a virgin Freetz-ng is incapable of getting the images by itself.
That's not something I need to solve.
I'm giving feedback.
 
Zuletzt bearbeitet von einem Moderator:
  • Like
Reaktionen: Micha0815

Master SaMMy

Neuer User
Mitglied seit
20 Apr 2016
Beiträge
94
Punkte für Reaktionen
11
Punkte
8
Für mich ist nun Freetz-NG gestorben ich werde nun auch gezwungen sein Assi key zu benutzen. Ich darf nicht mehr mein privaten benutzen.


ERROR: wrong password for private signature key file
 

hippie2000

Neuer User
Mitglied seit
20 Jan 2008
Beiträge
78
Punkte für Reaktionen
29
Punkte
18
@Shirocco88

Sorry für die späte Antwort, hab nur selten Zeit für Foren.

Ja es gab technische Arbeiten bei unserem Hoster.

Die dauern scheinbar noch an, aber es war jetzt 6 Tage Uptime (des Peerings), und davor gab es einen Tag an dem es dauernd Unterbrechungen gab.

Die Angelegenheit ist im Wiki oben auf jeder Seite verlinkt, ich werde da auch Entwarnung melden.

Der BoxMatrix Server schnurrt wie ein Kätzchen, nur ohne Routing/Peering für sich alleine dann :)

Uptime 150 Tage zzt.
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
What is it doing here?
Code:
[email protected]:~/trunk$ make

STEP 1: UNPACK
checking signature:


 17191 freetz    20   0   15916    716    580 R 99.9  0.1  13:15.86 tar-gnu

lsof -p17191
COMMAND   PID   USER   FD   TYPE DEVICE SIZE/OFF   NODE NAME
tar-gnu 17191 freetz  cwd    DIR   8,17     4096 540676 /home/freetz/trunk
tar-gnu 17191 freetz  rtd    DIR    8,1     4096      2 /
tar-gnu 17191 freetz  txt    REG   8,17   396216 673893 /home/freetz/trunk/tools/tar-gnu
tar-gnu 17191 freetz  mem    REG    8,1  2936176 131783 /usr/lib/locale/locale-archive
tar-gnu 17191 freetz  mem    REG    8,1    18624 132454 /lib/x86_64-linux-gnu/libattr.so.1.1.0
tar-gnu 17191 freetz  mem    REG    8,1    14592 143664 /lib/x86_64-linux-gnu/libdl-2.21.so
tar-gnu 17191 freetz  mem    REG    8,1  1869392 143681 /lib/x86_64-linux-gnu/libc-2.21.so
tar-gnu 17191 freetz  mem    REG    8,1    35264 131102 /lib/x86_64-linux-gnu/libacl.so.1.1.0
tar-gnu 17191 freetz  mem    REG   8,17    65384 673921 /home/freetz/trunk/tools/build/lib/libfakeroot-0.so
tar-gnu 17191 freetz  mem    REG    8,1   154376 143667 /lib/x86_64-linux-gnu/ld-2.21.so
tar-gnu 17191 freetz    0u   CHR  136,0      0t0      3 /dev/pts/0
tar-gnu 17191 freetz    1w   CHR    1,3      0t0   6447 /dev/null
tar-gnu 17191 freetz    2w   CHR    1,3      0t0   6447 /dev/null
tar-gnu 17191 freetz    3u   REG    8,1 26716160    781 /tmp/freetz-sig-Nu5/img
 

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
@frater:
You should provide the ".config" file, too.

I've an idea, what went wrong ... but I want to have a glance at the mentioned file first.
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
@frater:
You should provide the ".config" file, too.

I've an idea, what went wrong ... but I want to have a glance at the mentioned file first.
Thanks...

I even removed the .config (saved it as well) and did a new make menuconfig and only did the necessary changes (7590) and it behaved exactly the same:
https://justpaste.it/3sp2g

The process was still running this morning. I left it running to make sure it had ample time to complete.

I changed 7590 to 7490 and it continued, even making a binary.
So far it seems to be a "7590 exclusive" (can't tell as I only did the 7490 and 7590)
 

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
The basic info - I was looking for - is as follows: 7590, international version, 06.8x

This should lead to the firmware file: FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image, don't it?

As you can see in the following console log, the last block in source image is part of the "./var/signature" file:
Code:
[email protected]:~> wget -qO- http://ftp.avm.de/fritzbox/fritzbox-7590/italy/fritz.os/FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image | hexdump -C | tail -n 27
0197a400  2e 2f 76 61 72 2f 73 69  67 6e 61 74 75 72 65 00  |./var/signature.|
0197a410  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
0197a460  00 00 00 00 30 31 30 30  36 34 34 00 30 30 30 30  |....0100644.0000|
0197a470  30 30 30 00 30 30 30 30  30 30 30 00 30 30 30 30  |000.0000000.0000|
0197a480  30 30 30 30 32 30 30 00  31 33 33 31 35 31 33 30  |0000200.13315130|
0197a490  34 30 35 00 30 31 32 30  34 31 00 20 30 00 00 00  |405.012041. 0...|
0197a4a0  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
0197a500  00 75 73 74 61 72 20 20  00 00 00 00 00 00 00 00  |.ustar  ........|
0197a510  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
0197a540  00 00 00 00 00 00 00 00  00 30 30 30 30 30 30 30  |.........0000000|
0197a550  00 30 30 30 30 30 30 30  00 00 00 00 00 00 00 00  |.0000000........|
0197a560  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
0197a600  18 36 7e a1 83 6b 20 4d  2f 3c ae 7f f7 80 b4 82  |.6~..k M/<......|
0197a610  ba f4 3b c8 80 4a 63 43  6b b5 b0 65 01 0c eb 9f  |..;..JcCk..e....|
0197a620  c8 6d 2a a6 e9 11 1d 99  4f 39 87 d6 08 52 0b bf  |.m*.....O9...R..|
0197a630  87 73 de 9c d7 7a 2e 28  4e 3c 2c 9a 46 c2 7b 59  |.s...z.(N<,.F.{Y|
0197a640  3d 0d 56 b9 3d 24 c2 d1  41 85 60 12 bd e3 bc ae  |=.V.=$..A.`.....|
0197a650  b5 33 86 bf 58 8d 86 55  ff 5e 09 31 bb 09 3c ec  |.3..X..U.^.1..<.|
0197a660  f6 29 36 cd a0 0d c5 ab  e8 c0 a3 0d 6f 52 2a e2  |.)6.........oR*.|
0197a670  a8 88 19 f2 39 c8 48 72  0e e0 fa 08 45 f3 b1 ec  |....9.Hr....E...|
0197a680  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
0197a800
[email protected]:~>
(last block starts at offset 0x0197a600 and ends - as the file at all does - at 0x0197a800).

While other solutions take the EOF-state after this block into account and continue signature checking, the too much simplified own solution from "freetz-ng" (introduced with r15284) doesn't handle such corner cases correctly (the used 'tar' command doesn't functioning here). There was a reason to implement some precautions against "not-very-well" signed images in my 'signimage' solution.

Don't get me wrong ... the original file from AVM is - if you'll see it as tarball with a well-known format - nevertheless wrong. But it's accepted by AVM's own components (obviously) and by my 'check_signed_image' script, too:
Code:
[email protected]:/tmp> wget -q http://ftp.avm.de/fritzbox/fritzbox-7590/italy/fritz.os/FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image
[email protected]:/tmp> /home/GitHub/YourFritz/signimage/check_signed_image FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image -a /home/FritzBox/FB7590/keys/avm_firmware_public_key1
Found OpenSSL 1.1.0h-fips  27 Mar 2018
Check dgst command ... OK
Check rsautl command ... OK
Checking the public key from /home/FritzBox/FB7590/keys/avm_firmware_public_key1 ... OK
Checking support for the used hash algorithm md5 ... OK
Verification succeeded.
[email protected]:/tmp>
There's no obligation for AVM to create well-formed tarballs as their "firmware images" ... and it's really, really wrong under some circumstances, what's known for sure meanwhile (https://www.ip-phone-forum.de/threads/wie-funktioniert-eigentlich-das-signieren-der-avm-firmware.286213/).

The fact is, the used attempt to remove the signature file from the image to get a byte stream for hashing, hangs and has to be canceled manually:
Code:
[email protected]:/tmp> tar -f FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image --owner=0 --group=0 --mode=0755 --format=oldgnu --verbose --delete ./var/signature
^C
[email protected]:/tmp> tar -t -v -f FRITZ.Box_7590.en-de-es-it-fr-pl-nl.154.06.84.image
drwxr-xr-x 0/0               0 2018-06-28 11:57 ./var/
-rwxr-xr-x 0/0           32375 2018-06-28 11:57 ./var/install
-r-xr-x--- 0/0          284696 2016-10-31 16:49 ./var/regelex
drwxr-xr-x 0/0               0 2018-06-28 11:57 ./var/tmp/
-rw-r--r-- 0/0         4062728 2018-06-28 11:57 ./var/tmp/kernel.image
-rw-r--r-- 0/0        21057544 2018-06-28 11:57 ./var/tmp/filesystem.image
-rwxr-xr-x 0/0            2795 2018-06-28 11:57 ./var/info.txt
-r-xr-x--- 0/0          278924 2016-10-31 16:49 ./var/chksum
-rwxrwxrwx 0/0          988723 2018-06-28 11:57 ./var/urladerupdate
-rw-r--r-- 0/0             128 2018-06-28 11:57 ./var/signature
[email protected]:/tmp>
So ... the problem is an(y) original image file, which has a slightly wrong structure, if handled solely as a valid TAR output file - the FRITZ!Box model and version doesn't really matter here.
 
Zuletzt bearbeitet:

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
It isn't clear to me if you now also have provided a fix.
If so, I do not know how to properly implement it.

I had a private message ( @opto ) as well and the suggestion for a quick fix was to disable the verification of the signature.

That's an expert option when you run "make menuconfig"

That worked

It does make me wary of "freetz-ng"
 
Zuletzt bearbeitet:

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
It isn't clear to me if you now also have provided a fix.
No, there's no suggestion for a fix from my side. The "original Freetz" project uses my 'signimage' implementation, which works as shown above even for "critical images" from vendor. It's not my task to analyze the new implementation from "freetz-ng" any further (beyond of "showing the problem", that occured here) and it's certainly not my task, to provide a fix for errors with this solution.

I had a private message ( @opto ) as well
Sounds interesting ... presumably (or better: obviously) @opto is one from a bunch of accounts/pseudonyms, used by the same person - at least that's a supposition (of others and myself) based on the same behavior and not only my own "guess": https://www.ip-phone-forum.de/threads/fw-6-50-für-7490-in-trunk-rev-13490-keine-auswahlmöglichkeit.282943/page-13#post-2161302

It does make me wary of "freetz-ng"
He (or she?) is also known as @cuma on IPPF and Freetz and meanwhile as @fda77 on GitHub and the "freetz-ng" fork - due to the fact, that he's the author of the different approach, he should know, what he's writing about it. Maybe you better follow his proposals, while you're using his own fork of "freetz".

He decided to remove the signature checks from the original "freetz" project (the reasons are "unpublished") and he has replaced it with an own implementation ... but that's no obvious reason to wary from it. But it may happen, that the simplified new implementation does not handle the above mentioned "corner cases" as needed - you were run into such a pitfall.

Anyway ... it's an interesting (and new) info, that he's still using the other identity to communicate privately with IPPF members ... while he wants the identity @cuma to get erased from all IPPF history. This occured, after he tried to remove all posts from the @cuma identity. He had already removed - some times ago - all posts himself, that he had written with his @opto identity. Therefore you'll find some remaining quotations in posts from other authors as the only hint, that he ever had exist here.
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
As merely an end user it is not really clear to me why there is a fork at all.

The only reason why I downloaded this freetz-ng is to be able to keep up with newer firmwares and newer models (7590).

Am I mistaken that switching to freetz-ng is my only route ?

I have around 70 boxes that I need to keep running. The only reason I am using Freetz is that I can put a zabbix agent on it and can use the SSH-server as a proxy tunnel to be able to use a browser in our client's network.

If it dies now I can keep loading an August 2018 firmware on future boxes. It also means that I can't go beyond the 7490 nor version 6.83.

Hopefully things will improve.
I'm not in a hurry
 

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
The only reason why I downloaded this freetz-ng is to be able to keep up with newer firmwares and newer models (7590).

Am I mistaken that switching to freetz-ng is my only route ?
Sure ... who or what has led you to this - by the way very wrong - judgement?

You may continue to use the "original Freetz" without any problem.

There's no automatic for the latest beta versions yet, they still must be included by a collaborator first.

But as far as I understand, that's not a 'no-go' for your intentions, because you're interested in the latest stable versions, aren't you?

If you've used the SVN server until it wasn't available anymore (the "git" option was present for years already and the SVN server was not very reliable in its last months), you need (once) to checkout a fresh version from GitHub (https://github.com/freetz/freetz.git). Afterwards copy your (original) ".config" file into this checkout and run "make oldconfig", followed by "make".
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
I started to use Freetz some 8 years ago. Before I used DD-WRT.
I'm not following the development closely and after I noticed that the SVN update stopped working I went to this forum and noticed there was a fork.

This led me to believe that the development stopped and I was now forced to use this fork.
As I'm using International versions I'm already forced to use older versions. I also need to investigate how I can create port forwards to the router itself.

I even never found the time to investigate how I can upgrade a new 7490 router.
I'm always downgrading it to version 6.30 before and then use the AVM-interface to flash Freetz on it.
I know there is a better and maybe even an easier way to do this. I just don't have the time to investigate.

This also stops me from using Freetz on a 7590 as there is no tool to downgrade it to version 6.30.
All this is my own fault and I'm not asking any help here for that. I need to, at least, try this to do myself with the info available here....

I am, probably just like you, always busy with much more than I can cope each day.

So yes.... I'm sure that I'm taking some wrong decisions.
I will probably switch to git.
I never knew that Freetz was there.
 

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
I will probably switch to git.
Fine ... it's much easier than it sounds.

But you should really invest 30 minutes and try to understand (afaik, you're able to read in German?), how the firmware is installing itself, if loaded into RAM via boot-loader. The needed hardware setup is the same as for any recovery program and doing it "the right way" should save you much more time than the invested half of an hour.

EDIT: And the best news at all ... a 7590 uses exactly (at least in most parts) the same way to install a firmware via execution from RAM - your investment will open you a new world, where even the new GRX models are "freetzable" again.
 

frater

Mitglied
Mitglied seit
23 Nov 2008
Beiträge
440
Punkte für Reaktionen
3
Punkte
18
Fine ... it's much easier than it sounds.
Oh, I know...
I already did a git clone and will be adapting that environment so I'm at the same point I was when I was using svn

But you should really invest 30 minutes and try to understand (afaik, you're able to read in German?), how the firmware is installing itself, if loaded into RAM via boot-loader. The needed hardware setup is the same as for any recovery program and doing it "the right way" should save you much more time than the invested half of an hour.
I always thought so, but was always afraid it would take me a complete afternoon and when I need to create a new firmware that part has already taken up most of the time.


EDIT: And the best news at all ... a 7590 uses exactly (at least in most parts) the same way to install a firmware via execution from RAM - your investment will open you a new world, where even the new GRX models are "freetzable" again.
No more flashing to 6.30
That's a promise.
 

hippie2000

Neuer User
Mitglied seit
20 Jan 2008
Beiträge
78
Punkte für Reaktionen
29
Punkte
18
@frater: there's no "the freetz" any more. It splitted in multiple forks.

The tool used to checkout isn't the issue, it also says which fork you follow, and both are highly different. The curent SVN fork will also be accessible via git, as well as opposite. That's just a matter of time.

There's one fork which enforces githup over nearly everything the community ever contributed, and disposed the trac and svn and owns the domain to forward all accesses to github (most of which are 404). This fork is mainly controlled by one person, and even prior contributors do not have write access any more.

And there's the second fork which attempts to preserve every single byte which has been contributed by nearly 1000 people which currently offers svn access (like freetz did until the split).

You should choose your fork by its content and not by the tool to checkout. Freetz-ng will offer both, Freetz-org (the highly restricted fork) will not offer this freedom (there's an svn bridge at github but its pretty limited and you can not upgrade existing builds).

The question never was SVN vs. git, the question is which fork to follow.

For an overview of all Forks (which you will not see on the restricted fork) see the freetz related sections here:

> https://boxmatrix.info/wiki/Development

They are fair for all forks unlike the org fork which claims to be the one and only and excludes a lot of manpower which has to use own ways to release. Compare yourself.

Choose by your needs and see the comments (read them) vs commits ratio to see which form of development fits your needs.

There is no The Freetz any more, and it likely will not return unless there are new maintainers with more community spirit.

I will provide a way to switch between forks at any time, regardles of the chheckout tool one uses, however, its not finished yet.

If you prefer new ideas and solutions while preserving all the past you are best served using the ng fork (currently svn), if you prefer max stalled and controlled development choose org (reduced to github abilities).

Both are maintained by core developers, which know freetz, going different pathes of restriction.

> https://boxmatrix.info/wiki/Freetz-Contributors

Cleaning out the zero (white space cleaning) or negative commits (deletion of other people's work) there's likely no big difference left between er13 (org fork) and cuma (ng fork).

If you are a coder willing to contribute in a way similar to the development before the split you may prefer ng, if you are willing to disuss every single step of your work org is your friend.

The prior way how freetz was developed and maintained was the ng way of collaborative development with write access for all main contributors. This has changed since the Github fork which is pretty much a one man show of a gatekeeper where you have to pray to not get your work rejected.

Even people who believe they are team members have to "pray in" their code. There's no other project in the open source world being more restrictive against nearly ALL contributors...

Of course this habit yields in slower development. And quality obviously does not increase at all (which could have been a reason for the slowdown).

See the Commit counters (Com column) of both ways to work:

> https://boxmatrix.info/wiki/SVN-Server

Choose yourself...
 
Zuletzt bearbeitet:

PeterPawn

IPPF-Urgestein
Mitglied seit
10 Mai 2006
Beiträge
12,475
Punkte für Reaktionen
839
Punkte
113
Wow, what a fair, objective and never polemically post in #38 ... and I'm really astonished, how it could stand in common line with earlier rumors, assumptions and statements, like those: https://www.ip-phone-forum.de/threads/boxmatrix-svn-server-und-freetz-svn-backup.302398/

@frater: You should really try to get an own impression - especially the arguments regarding the involved people and their count are very funny, if you try to look behind the scenes.

Nevertheless I want to state, that I earlier did not try to advise against using the "freetz-ng" fork:
but that's no obvious reason to wary from it.
This fork is mainly controlled by one person, and even prior contributors do not have write access any more.
Let's have a look first at the "freetz-ng" fork and its contributors/collaborators:
Code:
vidar:/tmp/trunk $ svn log | grep "2019)" | sed -n -e "s#r[0-9]* | \([^ ]*\) *.*#\1#p" | sort | uniq -c
      5 administrator
    318 fda77
vidar:/tmp/trunk $
This means ... in 2019 (the fork started on 2019-01-30) there were 318 commits by fda77 and 5 commits by administrator to the SVN based fork named "freetz-ng" - whether there were other contributors of code and what they did contribute in detail, is not ascertainable as easy.

The start of this fork was - in contrast to text from #38 - not from another or new SVN server. No, the fork has its roots on GitHub, too - and it only moved later to SVN, while co-operations (with low efforts) and synergy effects for Freetz users got blocked at the same time.

The original Freetz project had its GitHub mirror for years, before it has moved to GitHub overall. The "freetz-ng" fork had its GitHub mirror (or its GitHub content) for 'round about two weeks and there's (now since 6 weeks) no further activity visible, to give the users the "freedom" to select the used SCM system themselves.

It looks to me, as if these words:
Freetz-ng will offer both, Freetz-org (the highly restricted fork) will not offer this freedom.
are nothing more than rhetoric - providing ONLY a SVN repository to Freetz users isn't really the promised "freedom" ... it's more or less only an attempt to fool them. At least, if not the words, but the facts count.

But I'm a little bit uncertain, too ... especially while facing the unsettling, insisted question resulting from the quote above: Where and why is the "freetz" fork (let's call it a fork for now and pay our attention first to more important things) "highly restricted"? Especially, if you take the own words of @hippie2000, that it's unimportant, which tool is needed/used to checkout the code, into account?

Best is still, you compare the count of contributors - by yourself - with the "fork" of this project on github.com:
Code:
vidar:/home/GitHub/YourFreetz $ git log --since=2019-01-30 | sed -n -e "s|^Author: \([^<]*\).*|\1|p" | sort | uniq -c
     22 Eugene Rudoy
      2 PeterPawn
      7 PrinzVonBillAir
      9 WileC
     12 YourFritz
     14 f-666
      2 f666
      3 fda77
vidar:/home/GitHub/YourFreetz $
Looks like a dissent to the quoted text from above to me - there's not only a single person as contributor.

And last, but not least, let's have a look at this quote:
there's no "the freetz" any more. It splitted in multiple forks.
OK, what's a fork? There are two different kinds of "forks" ... one of them describes the existence of an own copy of the master repository for each developer, where (s)he may create and test new code, until it's ready to get merged into the master repository. Such "forks" are part of the daily work and are a normal step in a workflow ... obviously not that, what was meant above. Do you agree?

The other meaning is more organizational - a project splits (ah ... there's the same word as above) into groups, where each of these groups has different ideas regarding the further development of this project. So they want a break-off of collaboration ... that's the "split" for the project.

With these aspects of distinction let us now count the forks first, which claim to have different directions of development. Even using the table on boxmatrix.info, I get a result of ... <drum-roll> ... 2.

The repository of @DHU is still hosted on GitHub and was last synchronized with freetz:master on 03-24-2019. It contains some additions and extensions to use a FRITZ!Box device with home-automation platform FHEM. I would not see this as a "fork" with an idea of different "directions of development" and "contributors" in mind - it's a "normal" fork of the "original freetz" repo, following the "GitFlow".

The next one is "freetz-ng", which is only available from the new SVN repository. It contains - more or less - the Freetz content from beginning of 2019, with some new patches by fda77 ... but this fork lacks most of other contributors, if you'll not take @hippie2000 into account - looks like he's the person behind the "administrator" identity.

And the remaining "fork" is not really a fork (https://github.com/Freetz/freetz/network/members), it's still the root and a continuation of the project from the last years on GitHub (exclusively now). Hopefully now with a more open handling, 'cause there were earlier really some reasons for disputes, but that's another question.

Nevertheless the "freetz/freetz" project on GitHub is still the same project and I can't see any reason, why someone (with an objective view) should call a (current) count of two forks "multiple". And why should he try to deny, that "the old freetz project" (or call it 'genuine', 'authentic', 'real', 'legit' and so on :)) still exists and - even more important - is meanwhile with more contributors as in the last two years more than alive?

There's one (means: a single) fork of "freetz" with a different attitude and (probably) different targets... but even for this fork the further decisions of directions aren't really clear. OK, removing full-functioning parts from the fork and replacing them with a worser solution (as shown above) or removing partial functions completely, is obviously also an attitude and at least a sign of the selected direction of development - I didn't want to embezzle this fact.

But even then there's no (plausible) reason to give someone the impression, that Freetz has been splitted into various smaller forks with different intentions and teams. The collaborators from "freetz-ng" (as the only real fork with an organizational aspect) did not contribute to Freetz in the last years (let's select a concrete number: the last three years). Therefore the original project did not even lose a single active contributor due to the existence of the "freetz-ng" fork. Only @Whoopie has left the Freetz team, but afaik he did not enter the "freetz-ng" team instead.

fair for all forks unlike the org fork which claims to be the one and only and excludes a lot of manpower which has to use own ways to release.
Yes ... these famous lots of manpower. But definitely the owner of the "freetz-ng" fork is not willing to contribute to Freetz the usual way - (s)he wants to get some "special permissions". Otherwise (s)he requests to ignore his/her changes to his/her own fork (https://www.ip-phone-forum.de/threads/freetz-ng.302327/page-5#post-2317571) - this desire was ignored in some cases, that's why his/her name "fda77" is in the above list of contributors to "freetz/freetz" (his/her earlier synonym "cuma" was an active contributor until February 2014). If you're interested in the discussions regarding this topic on GitHub, you may read about it here: https://github.com/Freetz/freetz/issues/89

And where did the "original project" ever claim, that it wants to be the only one? It's a demand from Freetz users, that there should be a common project, containing all the "goodies" from all collaborators. But collaboration requires a common understanding and the acceptance of and compliance with common rules, too. And if someone explicitly states, that (s)he is not ready to use the available (and free) tools, which were and are in many project the "best practice", how could anybody help him/her or collaborate with him/her? There's a suggestion, how it should and may work (see the link from above) ... if the consignee doesn't want to collaborate this way, others aren't any longer in charge here, to offer him/her other options. If the offered "manpower" leads in the end to more tasks and efforts for others, it's not worth any further thinking - and that's probably the core from earlier statements of the "developer community" of Freetz, regarding the activities of former developer "cuma".
 
Zuletzt bearbeitet:

hippie2000

Neuer User
Mitglied seit
20 Jan 2008
Beiträge
78
Punkte für Reaktionen
29
Punkte
18
Of course a lot of noise will appear here to disturb other people's work. Nothing new. Always the same disturbers.

But rhetorics can not fix the numerous disadvantages the github fork and the new restrictive way to work has introduced.

Many try it, and many return. And even if some are satisfied since they don't know the state before the split I'm happy that they found their home. This is about giving a present to users and not about who is the winner.

Please try it all and compare, evolution and manpower can not be restricted down to zero by rethorics, it will happen elsewhere if not needed by the regents who believe they are kings or gods.

Currently there are 100 checkouts/updates per day in april, 50% using the lab automatic.

This number will increase regardless of which text will be spawned to stop it. BoxMatrix is an international project, and it will fill all the gaps which are left (which increased horribly since the reduction of the project to the github subset of abilities, correctness and integrity).

I will continue to promote all forks on BoxMatrix, and accept if there's not the same fairness on the other side.

Time will dispose one fork into redundancy (or much better into the return of the cooperation it had before the takeover). I doubt it will be the restrictive fork which wins...
 
Zuletzt bearbeitet:

Zurzeit aktive Besucher

3CX PBX - GRATIS
Linux / Win / Cloud

Statistik des Forums

Themen
233,769
Beiträge
2,038,214
Mitglieder
352,702
Neuestes Mitglied
Mieze09