[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#810822: ITP: MooseFS

On 15 Jan, 2016, at 15:09, Dmitry Smirnov <onlyjob@debian.org> wrote:

> On Tue, 12 Jan 2016 04:27:06 PM Piotr Robert Konopelko wrote:
>> We're already publishing own packages repository
>> (https://moosefs.com/download/ubuntudebian.html
>> <https://moosefs.com/download/ubuntudebian.html>). We would like to make
>> MooseFS available directly in Debian! :)
> In November 2014 when I was working on introducing LizardFS (a fork of 
> MooseFS) to Debian, there were no publicly available source code for MooseFS 
> -- no tarballs or source code repository. Later it became possible to request 
> MooseFS sources by filling web form and hoping that sources will be sent over 
> email. Even though now MooseFS sources can be freely downloaded there are no 
> public bug tracker and no public VCS. Lack of VCS means one can not isolate 
> (and backport) fixes without great difficulties. 

We may make official clone of GPL'ed version on GitHub if it change anything.

> Moreover only crippled "community edition" of MooseFS is free software as 
> MooseFS developers apparently focused on proprietary "PRO" edition.

This is not true. 99.9% of my time I spend on GPL'ed version. Now there is only one feature in PRO version that is not available in GPL version. We do not plan to change that (maybe in the future exchange this feature to another).

> Debian already have MooseFS' fork -- LizardFS. To my knowledge at the moment 
> MooseFS do not offer noticeable advantages over LizardFS while the latter 
> seems to have slightly more features.

You are very wrong. We have users who switched from LizardFS. Stability, efficiency for example. Quite important advantages. Not for everyone, but at least for serious users.

> For quite a while LizardFS is developed with community using public VCS and 
> bug tracker (GitHub) as well as Gerrit code review system and continuous 
> integration system. LizardFS have more development transparency than MooseFS 
> ever had.

And in case of file system it is not good idea. 

> I believe that poor governance of MooseFS motivated forking and it appears 
> that MooseFS developers still did not learn their lesson.

No. This was financial decision.

> ***
> Based on the above I recommend to refrain from introducing MooseFS to Debian.
> ***
> Please note that IMHO MooseFS versus LizardFS situation have many 
> similarities with MySQL vs. MariaDB situation where poor Oracle's governance 
> and focus on proprietary addons discourage community from working with them.
> (MySQL is not as bad as MooseFS because MySQL have public bug tracker).

You are not right. MariaDB is developed by original author of MySQL, so the spirit of MySQL now is in MariaDB.

I invented MooseFS and I'm still developing MooseFS. Also in terms of money. We do not make money on MooseFS. The PRO version is just for us to help developing it at all. Let's say that thanks to pro version of MooseFS we are able to make GPL version of MooseFS. Thanks to that I'm able to work 8 hours a day developing MooseFS, so we made PRO version to actually improve developing of GPL version.

> As I object to introduction of MooseFS to Debian I would object to 
> introduction of MySQL if MariaDB were already available.

Why? People should have right to choose. 

> Having both is a burden and non-negligible overhead.
> With all due respect to MooseFS's former innovations and legacy I think for 
> now it would be best to refrain from debianising MooseFS and re-evaluate 
> situation in the future if there will be any development.

What kind of "development" are you talking about?

> -- 
> Cheers,
> Dmitry Smirnov.
> ---
> We often refuse to accept an idea merely because the way in which it has
> been expressed is unsympathetic to us.
>        -- Friedrich Nietzsche

Jakub Kruszona-Zawadzki
- - - - - - - - - - - - - - - -
Segmentation fault (core dumped)
Phone: +48 602 212 039

Reply to: