Koozali.org formerly Contribs.org

Contribs.org Forums => SME Server 10.x => Topic started by: ReetP on December 09, 2019, 10:51:50 PM

Title: SME10 - the clock is ticking
Post by: ReetP on December 09, 2019, 10:51:50 PM
Yes, the few of us who actually do anything here are acutely aware that time is running out on v9.

EOL is the end of 2020.

We are trying to push ahead with stuff but we are very very few, and there is a mountain of work to do, and that doesn't just mean coding.

Come and talk to us on Rocket.Chat where the idle chatter is happening.

Volunteer to do something. Anything. Don't just talk about it. Do it.

If you can't code you can document, test, or pay for a developer to help, or just chip in some cash. Right now man hours are more important than cash.

Don't wash your hands, walk away and say "I can't"' or "I don't have time".

We could do that ourselves, and then what would happen?

Koozali SME Server, like most Open Source software, is community based. Everyone in the community is responsible for it.

If v10 isn't released it isn't 'OUR' fault, and we won't have a conscience about it because we are trying.

It is EVERYONES fault, particularly that of those who did nothing.

B. Rgds
John
Title: Re: SME10 - the clock is ticking
Post by: holck on December 12, 2019, 01:54:00 PM
How do I join you on rocket.chat?

/Jesper H, Denmark
Title: Re: SME10 - the clock is ticking
Post by: ReetP on December 12, 2019, 02:09:03 PM
How do I join you on rocket.chat?

DM me name and email address and I'll set up a user account for you :-)
Title: Re: SME10 - the clock is ticking
Post by: Fumetto on December 18, 2019, 08:22:19 PM
I have a series of servers installed; at present they are just under 10, half of which are "left to themselves" while still reading administrative emails for pure "personal pleasure". I was thinking in these days that, within a year, things will have to change.

I don't have high skills and I don't think I can be very helpful (except to say "this and that don't work"). The amounts I earn thanks to SME have been insignificant (let's say I have "entered" in some companies thanks to SME and then I have done something else), however I have to give support for the systems that are still active.

A couple of servers are exclusively used as fileservers so the problem (more or less) I don't ask myself, even after the EOL date they will only have that function and should not be a problem; but I have 3 that are used as mailserver, one also very used and with about 60GB of emails inside.

Not being able to be "technically" helpful and wanting to support both my client (who pays shit but is the head of one of my family member and so I have to give him assistance (almost for free)) and the "SME project", I was then wondering if a "substantial" donation (in addition to my annual $ 50) could help. Someone talked about "paying" one or more developers to carry out the necessary preparatory steps for the release of SME 10. How much work is there to do? If you had to pay a developer how much time (indicatively) is necessary? What could be a suitable donation?

Let me be clear ... I am a "poor man" who tries to make ends meet thanks also to SME, but I can try to sensitize some customers and see if I can pull something out.
Title: Re: SME10 - the clock is ticking
Post by: TerryF on December 18, 2019, 09:47:23 PM
When a few put up their hands to do something others follow..appreciated
Title: Re: SME10 - the clock is ticking
Post by: krisden on December 20, 2019, 11:12:54 AM
If you can't code you can document, test, or pay for a developer to help, or just chip in some cash. Right now man hours are more important than cash.
Hello sme world  8)

I would not be shocked if you publish a table which would state the financial needs concerning the hours of development necessary for the success of the new version.

If the need is financial, let's stop veiling our face and express it clearly: how many hours of development should we count? How many people are needed and how much does one hour per person cost?

Could we imagine crowdfunding once the total sum of needs has been expressed? And how do you set this up if ok?

Have a nice day,

--
Christophe
Title: Re: SME10 - the clock is ticking
Post by: ReetP on January 31, 2020, 12:21:52 AM
Hello sme world  8)
Hi Christophe


Quote
I would not be shocked if you publish a table which would state the financial needs concerning the hours of development necessary for the success of the new version.

If the need is financial, let's stop veiling our face and express it clearly: how many hours of development should we count? How many people are needed and how much does one hour per person cost?

I have absolutely no idea how many man hours or what the cost might be.

We are chiselling away. But we are very few, and we all have other lives. No one currently working on SME really derives an income from it..... we all use it, but we are not really 'IT Specialists'. It is something we do in our spare time between wives, and families, and jobs, and life!!

Quote
Could we imagine crowdfunding once the total sum of needs has been expressed? And how do you set this up if ok?

The problem is there are several thousand servers out there, but most are so reliable that the admins rarely come here, so we have no real way to communicate with them, until they need something....

So I am not sure most would see an appeal for money. Or at least they won't see it until too late.

It also means we would have to spend a lot of time trying to find and brief a developer to help us.

It just isn't so simple (I wish it was!)

If anyone has any bright ideas then do let us know!
Title: Re: SME10 - the clock is ticking
Post by: ReetP on January 31, 2020, 12:26:49 AM
If you have ANY interest at all in helping with development please come and talk to us.

Even if it is just running a test VM and trying things out to see what breaks. The more people test, the better it will be.

We are currently looking at:

Fixing Horde to work with the latest versions of PHP
Proper certificate support (if I can make it work !!)
Then another Alpha ISO

I am not sure what is next. We need to figure something with Samba 4.

PM me for a login to our Rocket.Chat instance and come and chat. We don't bite!

Title: Re: SME10 - the clock is ticking
Post by: Fumetto on January 31, 2020, 12:52:40 AM
An idea that should not be underestimated, which I had already expressed years ago, is to "impose" an annual payment for updates. It might be interesting to add it as "features" to version 10.
For each server instance, access to the smeupdates repositories is tied to a code, which would be obtained every year (or one-off) by paying a "donation" of a few dollars ... let's say 50 or 100.

I know that many will turn up their noses but it would be a good way to raise money to pay one or more full-time developers; in the end, precisely because of the fact that the installations are very numerous, this would allow to recover a good portion of economic resources.

... and sorry for the english, google translate is probably not the best but for long concepts I need it ... ^ _ ^
Title: Re: SME10 - the clock is ticking
Post by: ReetP on January 31, 2020, 01:15:06 PM
An idea that should not be underestimated, which I had already expressed years ago, is to "impose" an annual payment for updates. It might be interesting to add it as "features" to version 10.

The funny thing about open source is people don't expect to have to pay for it even when they are profiting from it... :-)

The number of times I have heard people wanting support and help 'here and now', but don't want to pay or donate, and won't help because they are 'too busy'. Yes - earning from SME....

It is 'frustrating'.

Like most open source projects, the way to go is probably with paid for support. However, we don't really have enough people to be able to support it 24/7 now (we did talk about this a number of years ago).

There has always been a case for 'paid for' contribs, but I know that people who did make them struggled to make any money.

In any event, none of of us are really keen on 'paid for' updates... we like it being open source, and we would need infrastructure to handle payments, licensing it etc etc etc so it isn't that simple.

Please donate to keep the infrastructure running, and PLEASE come and help. Even just chatting gives us some moral support! Testing, reporting bugs etc helps even more.

Quote
... and sorry for the english, google translate is probably not the best but for long concepts I need it ... ^ _ ^

You are doing fine so please don't worry!

Come and join us at chat.reetspetit.info and you can practice your English ;-) Just ask me for a login.
Title: Re: SME10 - the clock is ticking
Post by: gieres on January 31, 2020, 09:57:10 PM

Even if it is just running a test VM and trying things out to see what breaks. The more people test, the better it will be.

Hello,
I would like to help but I don't even know how to install a virtual machine!
And I don't know how to test, either ...
If you want, we can make a real machine available, but since I am not a computer scientist, you would have to write procedures to say how to test. Because I don't think it exists in the wiki, does it?
Would the cost / benefit ratio still be interesting?
Good night.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on February 01, 2020, 02:44:21 AM
Hello,
I would like to help but I don't even know how to install a virtual machine!


Great!! First requirement is wanting to help. We can help with the rest!!

You don't need to be a scientist. Just interested in learning.

We are going to add some stuff on creating a test machine and what to do.

A lot of of us run Proxmox which is good.

You can even do basic testing even with virtual box.

I also have some SMEs on digital ocean and vultr.

The benefit with VMs is that you can snapshot, test, revert. Makes testing much easier

Give us a few days and we will try and sort something out.

Thank you for volunteering!!
Title: Re: SME10 - the clock is ticking
Post by: Jean-Philippe Pialasse on February 01, 2020, 05:11:02 AM

If the need is financial, let's stop veiling our face and express it clearly: how many hours of development should we count? How many people are needed and how much does one hour per person cost?

Could we imagine crowdfunding once the total sum of needs has been expressed? And how do you set this up if ok?
Christophe,
Here is a late answer.
I would say just in term of dev time without the testing and without implementation of samba 4 AD it would represent 3 months of one full time perl developper with experience with all the services implicated in SME and systemd.
It is both not that much and a lot depending on the perspective.

For the cost per hour, well it will greatly depends on where is living the developper, but also we are asking for a very particular profile. In clear words we are asking for a perl/bash developper who is also a system administrator... not quite the same jobs and trainings.
Title: Re: SME10 - the clock is ticking
Post by: krisden on February 01, 2020, 10:10:13 AM
As an example, I use Dolibarr ERP/CRM to manage my shop and I need to use accountancy advanced module.
For three years, the Dolibarr association has called on financial participation for the development of this module (luck, the main dev is accountable  8-) ).
Here is the link wich reveals the spreadsheet for the course per year and Dolibarr versions (sorry, in french) :
https://www.open-dsi.fr/financement-participatif-2019-bilan-et-quelques-nouvelles-du-module-comptabilite-dolibarr/

The progression is well structured and organized in stages over a year. As long as level 1 is not exceeded in financial terms, the development concerning requests for improvements, various corrections, etc., does not start and remain as it is. When a level is crossed, the dev begins and we pass to level 2, etc ...
In fact, a survey is set up to collect the opinion of the participants on the need or not to use the funds according to the importance of the functionality concerned :
https://nextcloud.open-dsi.fr/index.php/apps/polls/poll/BP1ATFnmSEpMKHWx
There is a report published at least once a week which reports on the progress of development, what has been achieved, what is missing in the objective, it is quite reassuring and good for confidence.

If we look on the forum, the 4 levels were crossed in not even a month with around twenty contributors :
https://www.dolibarr.fr/forum/t/financement-participatif-compta-2019-go/29289

So I am aware that we are talking about a developed operating system but finally, there are models, why not take inspiration from them ? It's a total success  :-o

Note: I am not a developer and my knowledge is very limited in this area. I am not a project manager either and I have no experience in this area.
But I can participate in translation, tests and of course financially.

Have a nice week-end
Title: Re: SME10 - the clock is ticking
Post by: stephdl on February 03, 2020, 07:28:41 PM
I would say just in term of dev time without the testing and without implementation of samba 4 AD it would represent 3 months of one full time perl developper with experience with all the services implicated in SME and systemd.

Don't put too much effort in Samba AD, RHEL has dropped the support for rhel8, openldap is obsoleted, so now it is Free IPA the next solution. Moreover if you read the Samba documentation, the implementation needs two servers, one for the file server, the other for the authentication (samba AD), for sure this can be used with a container, or with a remote account provider but no one is easy (of course hard should  never be a blocking point)

Obviously systemd is a good development goal from my point view and it worths really an implementation.
Title: Re: SME10 - the clock is ticking
Post by: Stefano on February 04, 2020, 12:14:50 PM
Don't put too much effort in Samba AD, RHEL has dropped the support for rhel8, openldap is obsoleted, so now it is Free IPA the next solution. Moreover if you read the Samba documentation, the implementation needs two servers, one for the file server, the other for the authentication (samba AD), for sure this can be used with a container, or with a remote account provider but no one is easy (of course hard should  never be a blocking point)

Obviously systemd is a good development goal from my point view and it worths really an implementation.

leaving lurking mode just to ask to you: how did you manage the AD stuff "on the other side" (if you know what I mean)?
TIA
Title: Re: SME10 - the clock is ticking
Post by: stephdl on February 04, 2020, 12:55:43 PM
Nethserver uses a systemd container to run the samba AD but the next version could see some changes

No samba AD at all, windows system administrator expects a full ad directory replacement and it is clearly not

Use a second vm  to authenticate the users, but it could complicate  the vps installation

The container solution is hard to maintain
Title: Re: SME10 - the clock is ticking
Post by: ReetP on February 04, 2020, 01:42:13 PM
I get the feeling there is the usual crazy dance for control going on here some place.

Somewhere between RH, M$, Samba and systemDisaster (ooooooohh whoopee I can now make my home directory portable. I wonder if it includes my Mounts as well??)

https://www.theregister.co.uk/2020/02/03/linux_home_directories_merged_into_systemd/

If you are a Windows admin how are you meant to manage linux servers/desktops? Just use Windows? If you can't use Samba for AD them what can you use (just curious)?

Might be fun with my desktops and their LDAP/SSSD logins :-(

Wish they'd stop screwing things around.
Title: Re: SME10 - the clock is ticking
Post by: Jean-Philippe Pialasse on February 04, 2020, 06:33:44 PM
I feel the future will tends more toward cloud solutions for file storage and remote software.
Traditional cifs shares are behind us except for few cases.

You can see nextcloud with onlyoffice integration where the whole office can interact on the same document at the same time.
Nextcloud has some limitations in term of access performance, but if you check seafile you have something more robust, while less modular.

Of course there is stil’ the question for SSO, and administration of all windows client and security policy.
Title: Re: SME10 - the clock is ticking
Post by: nicolatiana on February 05, 2020, 05:57:32 PM
I feel the future will tends more toward cloud solutions for file storage and remote software.
Traditional cifs shares are behind us except for few cases.

You can see nextcloud with onlyoffice integration where the whole office can interact on the same document at the same time.
Nextcloud has some limitations in term of access performance, but if you check seafile you have something more robust, while less modular.

Of course there is stil’ the question for SSO, and administration of all windows client and security policy.
Maybe so for people dealing with CRM solutions, accounting, small documents.
My customers deal with heavy graphics, CAD or BIM with files sizing 50/100 Mb or BIM projects sizing a few Gbytes .... it's quite difficult for me to think those stuffs residing in cloud for technical (speed) and economical reasons.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on February 05, 2020, 06:20:07 PM
Maybe so for people dealing with CRM solutions, accounting, small documents.
My customers deal with heavy graphics, CAD or BIM with files sizing 50/100 Mb or BIM projects sizing a few Gbytes .... it's quite difficult for me to think those stuffs residing in cloud for technical (speed) and economical reasons.

Yes we have exactly the same issue, and slow ADSL (not everyone has 10s of Mbs connections, not will have for some while yet, especially in developing countries.).

Try opening a cloud stored artwork file and editing it on your desktop...... ain't going to happen.
Title: Re: SME10 - the clock is ticking
Post by: Jean-Philippe Pialasse on February 05, 2020, 07:01:13 PM
Cloud is not equal to someone else computer.
Cloud could be your SME exactly at the same spot. It is just shifting protocol. 
However, you point the situation i pointed also out with « few cases », where you have huge files to handle and where i feel CIFS is better in term of performance, except with a solution offering a local cache on the client machine.


The issue is that both us and Samba people think they have the only solution and protocol to access shared files.
With this situation they can just start to do whatever they want and everybody just follow.
Title: Re: SME10 - the clock is ticking
Post by: gzartman on February 05, 2020, 08:11:03 PM
Don't put too much effort in Samba AD, RHEL has dropped the support for rhel8, openldap is obsoleted, so now it is Free IPA the next solution. Moreover if you read the Samba documentation, the implementation needs two servers, one for the file server, the other for the authentication (samba AD), for sure this can be used with a container, or with a remote account provider but no one is easy (of course hard should  never be a blocking point)

Obviously systemd is a good development goal from my point view and it worths really an implementation.

I don't believe Free IPA replaces Samba 4, but runs along side it.  My understanding is that Free IPA is more or less a client for talking to Samba4/Windows Domains and provides a front end for LDAP, and also other types of auth.   If you don't give a rip about supporting windows clients, then you can just use Free IPA and not even install Samba 4.  Free IPA lets you play in a windows domain, but not visa versa.

"IPA's proposed solution calls for using Samba 4 together with IPA, working from one LDAP back-end and sharing a kerberos server. Samba 4 will present IPA to the AD world as a separate domain forest and will be responsible for establishing a cross-forest domain trust between the IPA/Samba domain and the Windows part of the enterprise."

Also, there is no need to run 2 instances of Samba with one being in a container.  Both the AD part of Samba and the file server part of Samba can run side by side.

Title: Re: SME10 - the clock is ticking
Post by: gzartman on February 05, 2020, 08:14:15 PM
Nethserver uses a systemd container to run the samba AD but the next version could see some changes

No samba AD at all, windows system administrator expects a full ad directory replacement and it is clearly not

Use a second vm  to authenticate the users, but it could complicate  the vps installation

The container solution is hard to maintain

They only did this because they didn't want to figure out the proper way to run Samba4 on RHEL/COS.  At the time they did that, it was very difficult to find Samba 4 packages with the AD bits included for RHEL/COS.  Sernet was the only source.  I believe Netserver used a compiled from sources version of Samba, or something like this, in their container.
Title: Re: SME10 - the clock is ticking
Post by: Jean-Philippe Pialasse on February 05, 2020, 10:26:04 PM
Greg,
Thank you for your input. It is really appreciated as you are the far mor knowledgeable here about samba.
Title: Re: SME10 - the clock is ticking
Post by: bunkobugsy on February 07, 2020, 10:34:07 PM
I'll always favor a private cloud type of environment provided by (local) network file sharing.
Samba 3.6 still works great in NT4 Style PDC, but SMB1 days are numbered (and unsafe).
With the EOS of W7 I fear that the day is near when W10 finally breaks down support for it.

I know Koozali builds its own samba 4 packages, but I just came across Tranquil IT's repos
of latest AD enabled samba 4 packages for EL7 (and even EL8) and great documentation.
http://samba.tranquil.it/centos7/       https://dev.tranquil.it/samba/en/index.html
Title: Re: SME10 - the clock is ticking
Post by: calisun on March 01, 2020, 12:59:47 AM
There are sites out there that provide financial support for Open Source projects. I have just nominated this project for a grant at: Mozilla Open Source Support (MOSS) "Secure Open Source" Nomination.
I am not sure if it helps if more than one person nominates, but it can't heart to try more of you to nominate (Choose Track III): https://www.mozilla.org/en-US/moss/

I am sure there are other organizations out there that give out funding, lets see if anyone else can find some other donation sites and nominate this project.
If other organizations are found, share here so more people can nominate this project.

Plus,
There are also developers out there that are willing to donate their coding skills for free to worthy project.
How about promoting this project on:
https://www.codetriage.com/

It is unfair to few developers here to shoulder all this work by themselves. That is why we need to spread the word and get more people involved by advertising project on codetriage.com and many other similar sites.
Title: Re: SME10 - the clock is ticking
Post by: csn on March 01, 2020, 08:43:02 PM
Slightly late to the party, but keen to help however I can.

I can write passable PHP and Python, I'm great at SQL and database architecture, and I've been adminning SME Servers on small networks since it was branded Mitel, so feel free to throw some testing at me.

(https://media.giphy.com/media/xT8qBmSnYDXS21ZvHO/giphy.gif)
Title: Re: SME10 - the clock is ticking
Post by: TerryF on March 01, 2020, 09:06:42 PM
More than enough room in this house :-) to start, have a look in bugzilla currently efforts by a couple of hard workers on a new server manager framework - https://bugs.contribs.org/show_bug.cgi?id=7819

More than welcome to join those doing what they can, see Rocket Chat hosted by user ReetP, contact him with name and email, his details are here he can then setup a login.

The knowledgeable there will be able to  point you in the right direction
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 01, 2020, 10:02:05 PM
I'll always favor a private cloud type of environment provided by (local) network file sharing.
Samba 3.6 still works great in NT4 Style PDC, but SMB1 days are numbered (and unsafe).
With the EOS of W7 I fear that the day is near when W10 finally breaks down support for it.

It isn't clever but depends on how exposed SMB is to the wider world. YMMV.

But yes support will be forcibly removed  at some stage for sure.

Quote
I know Koozali builds its own samba 4 packages, but I just came across Tranquil IT's repos
of latest AD enabled samba 4 packages for EL7 (and even EL8) and great documentation.
http://samba.tranquil.it/centos7/       https://dev.tranquil.it/samba/en/index.html

No, we built it for testing but nothing for production as yet.

The whole field is shifting and we need to have a think about what to do with this.

We try to align with RH/Centos as much as possible to save ourselves work.

If they have dropped S4 + AD we need to decide which route to pursue and whether ro rely on a 3rd party.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 05, 2020, 06:49:22 PM
It is unfair to few developers here to shoulder all this work by themselves. That is why we need to spread the word and get more people involved by advertising project on codetriage.com and many other similar sites.

Yes, and we can only do so much.

However, to get new developers on board it means we will have to stop developing and spend time hunting. And then teaching/mentoring. No dev is going to be able to just 'leap in' and do things-  they need to understand the underlying systems, what needs fixing etc etc. It is not as simple as as it sounds, and pretty time consuming.

What we really need are people who use SME and have a reasonable grasp of the basics to step up to the plate and come and learn some more. Quite frequently they are the people earning money from SME as well..... (or at the least saved themselves a lot of money)

If every admin who used SME donated 2 hours a week of their time we'd have something like 15,000+ man hours a week. Let alone the number of users that are also supported who could possibly help.

It really is that simple.

It is time that is the most valuable commodity right now.

if you really can't do that then donate to keep the hardware running. Or follow up on some of your ideas yourself and encourage others to do so - I just haven't got the time myself between a full time job and trying to code here.

There are really only 4 of us writing any code at the minute.

Terry is Mr Test  - he doesn't really code much but and spends hours on testing and verification. Boring shitty rinse and repeat, but vital, and the reason your SME is usually so stable. He is a God.

I have spent the last week or more with him updating the PHPKi contrib for certs which need to be updated to to go with my ipsec contrib because that has been updated, and also trying to get a replacement for PPTP which is horribly insecure, all of which I use at work - my justification for working in SME.

So you don't all need to be whizz bang coders - there are plenty of other things to do like testing and verification and bug hunting.

It just needs enthusiasm, willingness to learn, and a bit of motivation to get off your butts and volunteer.

Simples :-)

If anyone wants to chat then do like 'csn' did above and come and talk to us on Rocket.Chat

We are just ordinary people, and we don't bite!! (we actually share some really bad jokes too.....!!)
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 05, 2020, 07:00:04 PM
How about promoting this project on:
https://www.codetriage.com/

Unfortunately that is github related and currently we don't use github.

Why? Because it is a huge amount of work and we just don't the time or people right now.

Many people don't realise but we have an automated build system for packages

I grab a copy of the code from CVS
Modify the package
Test build it locally to check for failures
<Here comes the trick bit>
Submit it to Buildsys
Buildsys then automatically pickups up the code, queues, builds, checks for failures and then signs the packages
<^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^>
You can then install from a test repo
Once tested and verified we manually it to a release repo

The tricky bit is the automated building.
Buildsys is currently wired to talk with CVS. The backend code is scary!!!!!

Converting that all to git will take a lot of work. It may happen at some stage, but it is just not that easy.....

We need to get v10 done log before we worry about that.
Title: Re: SME10 - the clock is ticking
Post by: Jean-Philippe Pialasse on March 06, 2020, 06:15:21 AM
To add pootle translation interface is wired with cvs repo to generated missing translation to fill and also patch to apply on new translation available.


This code IS the scariest ;)
It ´s full of pythons.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 06, 2020, 02:24:52 PM
To add pootle translation interface is wired with cvs repo to generated missing translation to fill and also patch to apply on new translation available.

This code IS the scariest ;)
It´s full of pythons.

Ugly !!!!

I don't like snakes....
Title: Re: SME10 - the clock is ticking
Post by: bunkobugsy on March 06, 2020, 06:14:27 PM

No, we built it for testing but nothing for production as yet.

The whole field is shifting and we need to have a think about what to do with this.

We try to align with RH/Centos as much as possible to save ourselves work.

If they have dropped S4 + AD we need to decide which route to pursue and whether ro rely on a 3rd party.

This may sound silly, but wouldn't releasing SME 10 with RHEL/COS non AD/DC samba 4.9 make thing much easier? I gues there shouldn't be so many differences between SME 9 an 10. PDC type domains should work for the next 4 years and SMB1 could be easily turned off. Nobody should use Win XP anymore. W7, W8 and hopefully Win10 should be just fine.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 06, 2020, 08:44:13 PM
Quote
This may sound silly, but wouldn't releasing SME 10 with RHEL/COS non AD/DC samba 4.9 make thing much easier?

If you came and talked to us on Rocket.Chat or followed the bugs you'd realise that is not going to happen in a hurry.

Getting v10 finished will be a huge task even without Samba AD, and so far we have a lot of people making suggestions and pretty well no one actually doing any work.

Everyone thinks it is someone elses problem or that they don't know enough to help.

This is bullshit.

EVERYONE is responsible for producing a distro. Not just me or JP or Terrry.

If v10 doesn't get finished in time you are as guilty as anyone else.

So, less talk and more action please.

Stop making excuses.

Get INVOLVED.

DO something.

Don't tell us you haven't time or knowledge. What makes you think I have??

Start your testing here:

https://wiki.contribs.org/SME10.0_QA

Ask me for a rocket account.
Title: Re: SME10 - the clock is ticking
Post by: bunkobugsy on March 07, 2020, 03:51:15 AM
Getting v10 finished will be a huge task even without Samba AD

Then jumping directly to COS8 would make so much more sense...
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 08, 2020, 01:55:00 AM
Then jumping directly to COS8 would make so much more sense...

You clearly have no idea about the complexities involved here. If you actually got involved and did something to help instead of sniping from your armchair you might understand a bit more.

If it COS8 'makes sense' to you then go right ahead and build it. I look forward to seeing your work on it. Let us know when it is done and we can all migrate.

Thanks.

I'm going for beer and popcorn and waiting to see your handiwork.

Title: Re: SME10 - the clock is ticking
Post by: bunkobugsy on March 08, 2020, 05:46:07 AM
You clearly have no idea about the complexities involved here.

You're absolutely right!
However I did some reading yesterday and I have a feeling that the future with Samba4 and Win10 is going to be a constant headache. It seems that SMB1 is the only way and they want to force everyone to move to windows server.

https://forums.contribs.org/index.php/topic,53587.msg278605.html#msg278605

https://bugs.contribs.org/show_bug.cgi?id=10429#c9

https://wiki.samba.org/index.php/Required_Settings_for_Samba_NT4_Domains#Windows_10_and_Windows_Server_2016:_There_Are_Currently_No_Logon_Servers_Available_to_Service_the_Logon_Request

Will get back to work in a month or so and then I'll be able to do some more testing with the latest SME10 and Win10. I'll also test what SMB level is used in ClearOS/NehtServer 7, maybe in an AD/DC domain Win10 behaves differently and SMB2/3 works with Samba4.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 08, 2020, 02:23:12 PM
You're absolutely right!

Probably.


Quote
However I did some reading yesterday and I have a feeling that the future with Samba4 and Win10 is going to be a constant headache. It seems that SMB1 is the only way and they want to force everyone to move to windows server.

Seriously save yourself some time & trouble unless it is purely for educational purposes.

We are well aware of the problems and complexities involved. Some of us have been looking at this for several years.

Greg Zartman, who wrote some of the test S4 AD code here and who is one of my best friends, knows more about this than anyone.

Yes, M$ wants you to use their cloud/servers and/or other online type authent systems which are a pain. But equally SMB1 is old hat.

There are solutions, but it is a constantly moving feast. Yes, going Samba 4 and full AD looked the way ahead, but that seems now to have fallen by the wayside. It is also extremely complicated, and any one who has built packages from source usually then hides them behind a pay wall - like Sernet did. RH are clearly moving away from S4 AD  - they never really committed due to the encryption/authent systems they prefer to use (see kerberos et al).

If you are on Linux or Mac none of this is really a drama. The only people who have issues are Windows users.

The problems for them particularly are network share browsing, and domain authentication.

Either way, building a full S4 AD implementation is NOT going to happen in the short term unless someone wants to build it. And in the medium to long term it may not be the direction required anyway.

Quote
Will get back to work in a month or so and then I'll be able to do some more testing with the latest SME10 and Win10. I'll also test what SMB level is used in ClearOS/NehtServer 7, maybe in an AD/DC domain Win10 behaves differently and SMB2/3 works with Samba4.

Well, we know what levels are used already and the issues involved so save yourself that.

But straight forward testing, verification and bug fixing in SME v10 will be the best way to contribute. There is plenty to do.

And as I have said many times, if you talk to us on Rocket.Chat you will get answers to a lot of these things, and save me the time writing it all out again.

And again, don't just sit there procrastinating. Go and look at the jobs we actually need doing and do something.
Title: Re: SME10 - the clock is ticking
Post by: bunkobugsy on March 08, 2020, 08:02:43 PM

Either way, building a full S4 AD implementation is NOT going to happen in the short term


Then I realy don't understand why not just go with RHEL/COS Samba 4.9 in PDC mode (for the next 3,5 years).

https://wiki.samba.org/index.php/FAQ#Samba_as_NT4_Primary_Domain_Controller


Well, we know what levels are used already and the issues involved so save yourself that.


I'm just trying to figure out which way to go next year: PDC, Samba4 AD/DC or throw money at MS for CALs.

Sorry for annoying you, it was not my intention.
Title: Re: SME10 - the clock is ticking
Post by: ReetP on March 08, 2020, 11:28:13 PM
Quote
Then I realy don't understand why not just go with RHEL/COS Samba 4.9 in PDC mode (for the next 3,5 years).

Who said we wouldn't?

You are chasing your own tail here and doing nothing but confuse yourself.

You don't understand much because you have chosen to not get involved in the conversations as to what we will do, despite saying several times to get involved. You have half read bits here & there and not really understood anything at all so you are now needlessly asking pointless questions and wasting time.

Please, either join in properly with development and help build the system how you want it, or sit and wait quietly whilst we see if we have time to finish v10.

This thread was to point out that if you want to see v10 then stop talking, and do something to help.

That is still the case. I for one don't have time for any more of this waffle.

Please ask for a login at our rocket.chat and ask us there, or follow with your testing via bugs etc.

Thank you.