Project Abandoned?

1026 views
Skip to first unread message

William Stone

unread,
Feb 10, 2015, 12:43:47 PM2/10/15
to do...@googlegroups.com
I note that the last release of dokan-sshfs was 0.6.0 on 12 Jan 2011.  It's now 10 Feb 2015.

No changes in four years sounds like an abandoned project, to me.  Is that the case?

שאול פרידמן

unread,
Feb 10, 2015, 2:05:04 PM2/10/15
to do...@googlegroups.com
There is a continue to this project here: dokanX
I think that the dev of this project went commercial here. they have the exact interface as Dokan have.

On 10 February 2015 at 19:43, William Stone <wrst...@gmail.com> wrote:
I note that the last release of dokan-sshfs was 0.6.0 on 12 Jan 2011.  It's now 10 Feb 2015.

No changes in four years sounds like an abandoned project, to me.  Is that the case?

--
You received this message because you are subscribed to the Google Groups "Dokan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dokan+un...@googlegroups.com.
To post to this group, send email to do...@googlegroups.com.
Visit this group at http://groups.google.com/group/dokan.
For more options, visit https://groups.google.com/d/optout.

psan...@codicesoftware.com

unread,
Feb 26, 2015, 3:29:36 AM2/26/15
to do...@googlegroups.com
AFAIK Eldos Cbfs existed long before Dokan. Can you confirm it is the same project as Dokan??

Glad to see there is a fork in github :-)

Sven Harazim

unread,
Feb 27, 2015, 8:53:09 AM2/27/15
to do...@googlegroups.com

Alejandro Exojo

unread,
Feb 27, 2015, 9:51:41 AM2/27/15
to do...@googlegroups.com
On 27 February 2015 at 14:53, Sven Harazim <sven.h...@gmail.com> wrote:
> fork of the fork
>
> https://github.com/Maxhy/dokany

Not really. DokanY started with a clean history, and added some
changes from DokanX, and some changes of its own.

If you want to see a real "attack of the clones^Wforks", find the
networks on GitHub of several mirrors:

* https://github.com/clone/dokan/network
* https://github.com/svn2github/dokan/network
* https://github.com/BenjaminKim/dokanx/network

I have a document where I noted a bit the state of the different
projects that I've found on GitHub, after inspecting their histories.
Unfortunately not many of them have a clean history, so this isn't a
very useful exercise. I'll be evaluating them more thoroughly in the
next days, so I can share my findings if you want.

Sven Harazim

unread,
Feb 28, 2015, 8:19:51 AM2/28/15
to do...@googlegroups.com
 I'll be evaluating them more thoroughly in the
next days, so I can share my findings if you want.

Yes, please 

Tony Gravagno

unread,
May 18, 2015, 12:47:40 PM5/18/15
to do...@googlegroups.com
With the few number of people who are a familiar with this code base it seems reasonable for everyone to organize just a bit. I can easily see all projects and responses to open inquiries pointing toward a single web page which describes each fork/version/approach so that visitors can do research from there and come to their own conclusions. I'm not asking for the code sets to merge or for any other kind of collaboration at that level. But I think the strong fragmentation of this relatively small but very valuable project is hurting efforts in the long run.

Is this worth pursuing?

Alejandro Exojo

unread,
May 18, 2015, 2:00:50 PM5/18/15
to do...@googlegroups.com
Sven, sorry, I owed you a reply here.

I don't have a lot to share, though. I jumped into implementing my
application based on DokanY. They are very kind to provide installers
with a signed driver, which also adds the necessary boilerplate to
develop (the dokan.dll, the header, the mounter, etc.). Once
installed, it's easy to start developing right away your application.

I thought I could have the application more or less started, and then
try out with some compile switch to swap DokanY with DokanX, so I
could compare the two. However, trying out DokanX seems a bit more
complex, and as I understand it, the way you have to write the
callbacks for the DokanOperations is different. DokanX uses an
std::function as parameter (fine with me, I'm compiling in C++ mode),
but those functions have to return an HRESULT instead of a Windows
error code. In conclusion, I've just used DokanY alone.

I'm not unhappy with the decision. DokanY developers where quite
friendly, answering issues very quickly, and merging pull requests,
and updating their installer with signed drivers. I think at least 2
DokanY developers work for a company which I assume is using DokanY in
production (or planning to), so this gives me some confidence.

In the usage of the software, I've found no issues. I've found no
unexplainable crash or something like that.

That's pretty much what I think I can say, but if anyone has more
specific questions, feel free to ask.

Alejandro Exojo

unread,
May 18, 2015, 2:05:48 PM5/18/15
to do...@googlegroups.com
On 18 May 2015 at 18:47, Tony Gravagno <bacj8...@snkmail.com> wrote:
> With the few number of people who are a familiar with this code base it
> seems reasonable for everyone to organize just a bit. I can easily see all
> projects and responses to open inquiries pointing toward a single web page
> which describes each fork/version/approach so that visitors can do research
> from there and come to their own conclusions. I'm not asking for the code
> sets to merge or for any other kind of collaboration at that level. But I
> think the strong fragmentation of this relatively small but very valuable
> project is hurting efforts in the long run.

Given that both DokanX and DokanY should have the same codebase for
the driver, and given that the C# bindings should be the same, I think
both projects could share that.

One could have the driver in a shared repository, and then, if they
don't agree on how the user-mode library should be compiled, put that
in different repo. Is more work to keep things compiling, and sharing
for example the dokan.h header. But that's something both projects
would have to agree, and is work that will not provide any gain in the
short or medium term.
Reply all
Reply to author
Forward
0 new messages