Is elixir development active ?

27 views
Skip to first unread message

Romy

unread,
Jan 9, 2011, 12:20:20 AM1/9/11
to SQLElixir
Wondering because I've commented on a bug a few days ago w/out
response (http://elixir.ematia.de/trac/ticket/61) and would also like
to investigate a potentially serious bug, if it's a bug at all,
involving transactions (http://groups.google.com/group/sqlalchemy/
browse_thread/thread/f7a2e08f8be50a95).

Thanks.

Gaetan de Menten

unread,
Jan 10, 2011, 9:37:58 AM1/10/11
to sqle...@googlegroups.com
> Is elixir development active ?

Active... Not really. I have several "almost implemented" features
lying in my working copy at home, but given that I no longer use
relational databases during my day work, I only work on Elixir once in
a while (it competes with all my other hobbies). Honestly, unless
someone picks up the maintainership, it will continue to slowly slide
towards irrelevance. I'm sad to see all the effort I've put into it go
to waste, but let's face it: it's unlikely I'll ever have time to
"finish" the project (it has a quite narrow scope so finishing it
would be possible for someone with more time to devote to it than me).

On Sun, Jan 9, 2011 at 06:20, Romy <romy.m...@gmail.com> wrote:
> Wondering because I've commented on a bug a few days ago w/out
> response (http://elixir.ematia.de/trac/ticket/61)

Well that particular patch wasn't merged because it can break some
uses cases (it disallows having several entities with the same name in
the same module), which can be valid in some cases, while the correct
fix is elsewhere (as said in the bug report comments). And besides,
this wouldn't happen if migrate didn't use reload (which breaks a lot
of code).

> to investigate a potentially serious bug, if it's a bug at all,
> involving transactions (http://groups.google.com/group/sqlalchemy/
> browse_thread/thread/f7a2e08f8be50a95).

Doesn't seem to be elixir specific and your problem will probably be
soon resolved by Michael "Mr awesome support" Bayer...

--
Gaëtan de Menten

Romy

unread,
Jan 14, 2011, 6:47:53 PM1/14/11
to SQLElixir
On Jan 10, 6:37 am, Gaetan de Menten <gdemen...@gmail.com> wrote:
> > Is elixir development active ?
>
> Active... Not really. I have several "almost implemented" features
> lying in my working copy at home, but given that I no longer use
> relational databases during my day work, I only work on Elixir once in
> a while (it competes with all my other hobbies). Honestly, unless
> someone picks up the maintainership, it will continue to slowly slide
> towards irrelevance. I'm sad to see all the effort I've put into it go
> to waste, but let's face it: it's unlikely I'll ever have time to
> "finish" the project (it has a quite narrow scope so finishing it
> would be possible for someone with more time to devote to it than me).

I'm surprised nobody stepped in to work on it -- despite the extra
layer of abstraction, I would rather use Elixir than raw SQLAlchemy in
many cases. Makes me wonder what everyone else is using that's causing
it to, as you say, slide towards irrelevance ?

>
> On Sun, Jan 9, 2011 at 06:20, Romy <romy.maxw...@gmail.com> wrote:
> > Wondering because I've commented on a bug a few days ago w/out
> > response (http://elixir.ematia.de/trac/ticket/61)
>
> Well that particular patch wasn't merged because it can break some
> uses cases (it disallows having several entities with the same name in
> the same module), which can be valid in some cases, while the correct
> fix is elsewhere (as said in the bug report comments). And besides,
> this wouldn't happen if migrate didn't use reload (which breaks a lot
> of code).

Migrate's a pain in the ass to use with Elixir. This is very
unfortunate.

>
> > to investigate a potentially serious bug, if it's a bug at all,
> > involving transactions (http://groups.google.com/group/sqlalchemy/
> > browse_thread/thread/f7a2e08f8be50a95).
>
> Doesn't seem to be elixir specific and your problem will probably be
> soon resolved by Michael "Mr awesome support" Bayer...

It certainly was -- it didn't occur to me that I shouldn't be
expecting transactions to work under MyISAM.

>
> --
> Gaëtan de Menten

Daniel Cristian Cruz

unread,
Feb 4, 2011, 12:58:54 PM2/4/11
to SQLElixir
I would like to contribute.

But I had to study it, a lot.

On Jan 10, 12:37 pm, Gaetan de Menten <gdemen...@gmail.com> wrote:
> > Is elixir development active ?
>
> Active... Not really. I have several "almost implemented" features
> lying in my working copy at home, but given that I no longer use
> relational databases during my day work, I only work on Elixir once in
> a while (it competes with all my other hobbies). Honestly, unless
> someone picks up the maintainership, it will continue to slowly slide
> towards irrelevance. I'm sad to see all the effort I've put into it go
> to waste, but let's face it: it's unlikely I'll ever have time to
> "finish" the project (it has a quite narrow scope so finishing it
> would be possible for someone with more time to devote to it than me).
>

Gaetan de Menten

unread,
Feb 4, 2011, 2:21:50 PM2/4/11
to sqle...@googlegroups.com
On Sat, Jan 15, 2011 at 00:47, Romy <romy.m...@gmail.com> wrote:
> On Jan 10, 6:37 am, Gaetan de Menten <gdemen...@gmail.com> wrote:
>> > Is elixir development active ?
>>
>> Active... Not really. I have several "almost implemented" features
>> lying in my working copy at home, but given that I no longer use
>> relational databases during my day work, I only work on Elixir once in
>> a while (it competes with all my other hobbies). Honestly, unless
>> someone picks up the maintainership, it will continue to slowly slide
>> towards irrelevance. I'm sad to see all the effort I've put into it go
>> to waste, but let's face it: it's unlikely I'll ever have time to
>> "finish" the project (it has a quite narrow scope so finishing it
>> would be possible for someone with more time to devote to it than me).
>
> I'm surprised nobody stepped in to work on it -- despite the extra
> layer of abstraction, I would rather use Elixir than raw SQLAlchemy in
> many cases. Makes me wonder what everyone else is using that's causing
> it to, as you say, slide towards irrelevance ?

Nowadays most people use SQLAlchemy's built-in declarative module.


>> On Sun, Jan 9, 2011 at 06:20, Romy <romy.maxw...@gmail.com> wrote:
>> > Wondering because I've commented on a bug a few days ago w/out
>> > response (http://elixir.ematia.de/trac/ticket/61)
>>
>> Well that particular patch wasn't merged because it can break some
>> uses cases (it disallows having several entities with the same name in
>> the same module), which can be valid in some cases, while the correct
>> fix is elsewhere (as said in the bug report comments). And besides,
>> this wouldn't happen if migrate didn't use reload (which breaks a lot
>> of code).
>
> Migrate's a pain in the ass to use with Elixir. This is very
> unfortunate.

Could you be more specific? Have you read the simple workaround I
proposed in that bug report?

--
Gaëtan de Menten

Erik Janssens

unread,
Feb 4, 2011, 2:52:49 PM2/4/11
to sqle...@googlegroups.com
Hi,

I'm using Elixir together with sqlalchemy-migrate, and they actually
work perfect together, what are the issues you are seeing ?

As related to the state of Elixir, we are currently still using it in
several projects, and I don't see a reason not to do so or to switch
to declarative.

But as the Elixir user community we need to make a decision and
do something. I've been willing to implement some features in the
past (such as mapping to queries instead of tables), but I found it
difficult to know what exactly is happening at each stage of Elixir.

Nowadays, developing Camelot takes all my time. But if somebody
steps in to do development, I'd be willing to test new features or
commit bugfixes.

Regards,

Erik

> --
> You received this message because you are subscribed to the Google Groups "SQLElixir" group.
> To post to this group, send email to sqle...@googlegroups.com.
> To unsubscribe from this group, send email to sqlelixir+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/sqlelixir?hl=en.
>
>

Reply all
Reply to author
Forward
0 new messages