Changes to contrib

Changes to contrib

[Kosyrev Serge]

Good day folks!

I have tried to use ox-taskjuggler and faced what appeared to be a
mountain of.. infelicities and semi-broken functionality.

A need shapes the man, so I plowed through the most direct obstacles,
and finally managed to get what I want from the beast -- mostly by
adding direct Org-property -> TJ-attribute export pathways.

I don't know what is the proper way to submit patches for the contrib/
directory, so I made a branch on github:

  https://github.com/deepfire/org-mode/commits/ox-taskjuggler-fixes

Please, do tell how you would like to proceed from there.

(One immediate nitpick, of course, is that none of the additions
 are documented..)

-- 
с уважениe / respectfully,
Косырев Сергей

Re: Changes to contrib

[Rasmus]

Hi,

Kosyrev Serge <_deepfire@feelingofgreen.ru> writes:

> I have tried to use ox-taskjuggler and faced what appeared to be a
> mountain of.. infelicities and semi-broken functionality.
>
> A need shapes the man, so I plowed through the most direct obstacles,
> and finally managed to get what I want from the beast -- mostly by
> adding direct Org-property -> TJ-attribute export pathways.
>
> I don't know what is the proper way to submit patches for the contrib/
> directory, so I made a branch on github:
>
>   https://github.com/deepfire/org-mode/commits/ox-taskjuggler-fixes
>
> Please, do tell how you would like to proceed from there.

Please see

       http://orgmode.org/worg/org-contribute.html

TL;DR: git format-patch.

Do you have signed FSF papers?  I don't know if there's any desire to move
ox-taskjuggler.el to core, nor whether it would be possible (since I don’t
know if "tj" and Baptiste have signed FSF papers).

> (One immediate nitpick, of course, is that none of the additions
>  are documented..)

Some quick comments from skimming your code (note, I have no idea what a
taskjuggler is):

I don’t know what you refer to explicitly.  But that should be fixed, I
guess.  I think you add some more taskjuggle keywords/properties for
example.

Also, you introduce a dependency on subr-x, which may or may not be an
issue since it’s in contrib.

org-export-map-special-nodes and org-element-multivalued-property are not
proper names, as they will supposedly only be used in ox-taskjuggler.

Rasmus

-- 
Don't panic!!!

Re: Changes to contrib

[Aaron Ecay]

Hi Rasmus,

2015ko azaroak 8an, Rasmus-ek idatzi zuen:

[...]

> 
> Do you have signed FSF papers?  I don't know if there's any desire to move
> ox-taskjuggler.el to core, nor whether it would be possible (since I don’t
> know if "tj" and Baptiste have signed FSF papers).
> 

[...]

> Also, you introduce a dependency on subr-x, which may or may not be an
> issue since it’s in contrib.

FWIW, my impression of the way contrib is handled is that copyright
assignment is not required, though it is strongly encouraged to leave
open the possibility of moving a package to core.  Similarly, the
restrictions on emacs version support in org core do not apply.

This is just what I’ve surmised from observation, thoguh.  Others can
speak more informatively than me.

(I do like this system since it allows org-related packages to be
effectively community-maintained, and different people take on their
maintenance at different times, as we see here.  I’d be happy if of
the org-related packages on github were migrated to contrib, so that
they don’t bitrot when their original authors move on.  One thing that
might help with that is if we could provide an ELPA repo with the
individual contrib packages, though I’ve no idea how complex that task
would be.)

-- 
Aaron Ecay

Re: Changes to contrib

[Rasmus]

Hi Aaron,

Aaron Ecay <aaronecay@gmail.com> writes:

> Hi Rasmus,
>
> 2015ko azaroak 8an, Rasmus-ek idatzi zuen:
>
> [...]
>
>> 
>> Do you have signed FSF papers?  I don't know if there's any desire to move
>> ox-taskjuggler.el to core, nor whether it would be possible (since I don’t
>> know if "tj" and Baptiste have signed FSF papers).
>> 
>
> [...]
>
>> Also, you introduce a dependency on subr-x, which may or may not be an
>> issue since it’s in contrib.
>
> FWIW, my impression of the way contrib is handled is that copyright
> assignment is not required, though it is strongly encouraged to leave
> open the possibility of moving a package to core.  Similarly, the
> restrictions on emacs version support in org core do not apply.

AFAIK: this is correct.  But we can't move a package from contrib to core
if significant parts are written by authors without assigned copyright.
As I said, I don’t know what a taskjuggler is, so I don’t know how
relevant that consideration is here.

> (I do like this system since it allows org-related packages to be
> effectively community-maintained, and different people take on their
> maintenance at different times, as we see here.

Indeed.  But it has a risk if contrib is thought of as a staging area.
Again, I guess this should be done on a case-by-case basis.

> I’d be happy if of the org-related packages on github were migrated to
> contrib, so that they don’t bitrot when their original authors move on.

I agree with you, but others want to move stuff out of contrib.

Rasmus

-- 
m-mm-mmm-mmmm bacon!

Re: Changes to contrib

[Serge Kosyrev]

Rasmus <rasmus@gmx.us> writes:
> Kosyrev Serge <_deepfire@feelingofgreen.ru> writes:
[..]
>> I don't know what is the proper way to submit patches for the contrib/
>> directory, so I made a branch on github:
>>
>>   https://github.com/deepfire/org-mode/commits/ox-taskjuggler-fixes
>>
>> Please, do tell how you would like to proceed from there.
>
> Please see
>
>        http://orgmode.org/worg/org-contribute.html
>
> TL;DR: git format-patch.
>
> Do you have signed FSF papers?

No, didn't.

> I don't know if there's any desire to move ox-taskjuggler.el to core,
> nor whether it would be possible (since I don’t know if "tj" and
> Baptiste have signed FSF papers).

I'm not sure how wise it would be to raise barriers for contribution,
given the current state of the thing..

> Some quick comments from skimming your code (note, I have no idea what a
> taskjuggler is):
>
> I don’t know what you refer to explicitly.  But that should be fixed, I
> guess.

I'm sorry, what should be?

> I think you add some more taskjuggle keywords/properties for
> example.

[...]

> Also, you introduce a dependency on subr-x, which may or may not be an
> issue since it’s in contrib.

I'm sorry, I'm not sure what do you mean?  I looked at subr-x.el, and
I didn't find any reference to any of those functions..

> org-export-map-special-nodes and org-element-multivalued-property are not
> proper names, as they will supposedly only be used in ox-taskjuggler.

Well, the functions themselves are supposed to be generic.

There are two ways to name a function, in my mind:

  - by intended application
  - by what it does

..it's just that I chose the second.. in the futile hope that someone,
sometime will move/use them. : -)

If you deem it not to be an appropriate course of thought, I will
change the name, sure.

-- 
с уважениeм / respectfully,
Косырев Сергей

Re: Changes to contrib

[Rasmus]

Hi Serge,

Serge Kosyrev <_deepfire@feelingofgreen.ru> writes:

> I'm not sure how wise it would be to raise barriers for contribution,
> given the current state of the thing..

I don't know what state you refer to.

>> Some quick comments from skimming your code (note, I have no idea what a
>> taskjuggler is):
>>
>> I don’t know what you refer to explicitly.  But that should be fixed, I
>> guess.
>
> I'm sorry, what should be?

Sorry, I was referring to this quote by you, which seems to have
disappeared:

>>> (One immediate nitpick, of course, is that none of the additions
>>>  are documented..)


>> I think you add some more taskjuggle keywords/properties for
>> example.
>
> [...]
>
>> Also, you introduce a dependency on subr-x, which may or may not be an
>> issue since it’s in contrib.
>
> I'm sorry, I'm not sure what do you mean?  I looked at subr-x.el, and
> I didn't find any reference to any of those functions..

You use string-join, which is in subr-x.  I think subr-x was not a
dependency before.

>> org-export-map-special-nodes and org-element-multivalued-property are not
>> proper names, as they will supposedly only be used in ox-taskjuggler.
>
> Well, the functions themselves are supposed to be generic.
>
> There are two ways to name a function, in my mind:
>
>   - by intended application
>   - by what it does
>
> ..it's just that I chose the second.. in the futile hope that someone,
> sometime will move/use them. : -)

Then they are in the wrong library.

> If you deem it not to be an appropriate course of thought, I will
> change the name, sure.

At this point, they lack appropriate names.

Rasmus

-- 
Summon the Mothership!

Re: Changes to contrib

[Kosyrev Serge]

Rasmus <rasmus@gmx.us> writes:
> Serge Kosyrev <_deepfire@feelingofgreen.ru> writes:
>> I'm not sure how wise it would be to raise barriers for contribution,
>> given the current state of the thing..
>
> I don't know what state you refer to.

Well, I couldn't use it, without significant tweaking.

It feels as if the thing doesn't have a maintainer and slowly decomposes.

>>> Some quick comments from skimming your code (note, I have no idea what a
>>> taskjuggler is):
>>>
>>> I don’t know what you refer to explicitly.  But that should be fixed, I
>>> guess.
>>
>> I'm sorry, what should be?
>
> Sorry, I was referring to this quote by you, which seems to have
> disappeared:
>
>>>> (One immediate nitpick, of course, is that none of the additions
>>>>  are documented..)

That is true, yes.

The question is.. the only documentation for ox-taskjuggler that exists
in Worg, seems to only cover the surface aspects of export -- none of
the existing documentation touches on the multitude of details that the
existing code does involve itself with.

The point I'm trying to make, is that producing some documentation that
would cover the details (that one actually expects to be covered, coming
from the TaskJuggler background) seems like a separate task.

That is, not in scope of my little "tweak it to work for myself" series..

>>> I think you add some more taskjuggle keywords/properties for
>>> example.

I do, indeed -- and they are undocumented in a matter that is similar to
the 80% of pre-existing properties.

>> [...]
>>> Also, you introduce a dependency on subr-x, which may or may not be an
>>> issue since it’s in contrib.
>>
>> I'm sorry, I'm not sure what do you mean?  I looked at subr-x.el, and
>> I didn't find any reference to any of those functions..
>
> You use string-join, which is in subr-x.  I think subr-x was not a
> dependency before.

Oh, indeed -- missed that!

What should I do about it?

>>> org-export-map-special-nodes and org-element-multivalued-property are not
>>> proper names, as they will supposedly only be used in ox-taskjuggler.
>>
>> Well, the functions themselves are supposed to be generic.
>>
>> There are two ways to name a function, in my mind:
>>
>>   - by intended application
>>   - by what it does
>>
>> ..it's just that I chose the second.. in the futile hope that someone,
>> sometime will move/use them. : -)
>
> Then they are in the wrong library.
>
>> If you deem it not to be an appropriate course of thought, I will
>> change the name, sure.
>
> At this point, they lack appropriate names.

Understood, will rename, then.

-- 
с уважениeм / respectfully,
Косырев Сергей

Re: Changes to contrib

[Rasmus]

Hi,

Kosyrev Serge <_deepfire@feelingofgreen.ru> writes:

> Rasmus <rasmus@gmx.us> writes:
>> Serge Kosyrev <_deepfire@feelingofgreen.ru> writes:
>>> I'm not sure how wise it would be to raise barriers for contribution,
>>> given the current state of the thing..
>>
>> I don't know what state you refer to.
>
> Well, I couldn't use it, without significant tweaking.
>
> It feels as if the thing doesn't have a maintainer and slowly decomposes.

The question is whether there's any desire down the road to move this
backend to core?  Given that I don’t know that paper status of two former
contributors, and nobody chimed in, it’s likely that that race is already
done.

> The question is.. the only documentation for ox-taskjuggler that exists
> in Worg, seems to only cover the surface aspects of export -- none of
> the existing documentation touches on the multitude of details that the
> existing code does involve itself with.

OK.  There should be docstring and potentially comments, if necessary.
Worg is a separate issue.

>>>> I think you add some more taskjuggle keywords/properties for
>>>> example.
>
> I do, indeed -- and they are undocumented in a matter that is similar to
> the 80% of pre-existing properties.

OK.  It would be better to have some sort of documentation there, but
that’s a separate bug.

>> You use string-join, which is in subr-x.  I think subr-x was not a
>> dependency before.
>
> Oh, indeed -- missed that!
>
> What should I do about it?

I’d just use mapconcat TBH, since it’s just one place as I recall.

    (equal (string-join '("a" "b") " ")
           (mapconcat 'identity '("a" "b") " "))

> Understood, will rename, then.

Thanks.  And thanks for working on this!

Rasmus

-- 
This message is brought to you by the department of redundant departments

Re: Changes to contrib

[Kosyrev Serge]

Hi,

Rasmus <rasmus@gmx.us> writes:
> Kosyrev Serge <_deepfire@feelingofgreen.ru> writes:
>> The question is.. the only documentation for ox-taskjuggler that exists
>> in Worg, seems to only cover the surface aspects of export -- none of
>> the existing documentation touches on the multitude of details that the
>> existing code does involve itself with.
>
> OK.  There should be docstring and potentially comments, if necessary.
> Worg is a separate issue.

OK, will do.

>>> You use string-join, which is in subr-x.  I think subr-x was not a
>>> dependency before.
>>
>> Oh, indeed -- missed that!
>>
>> What should I do about it?
>
> I’d just use mapconcat TBH, since it’s just one place as I recall.
>
>     (equal (string-join '("a" "b") " ")
>            (mapconcat 'identity '("a" "b") " "))
>
>> Understood, will rename, then.

All done, sending mails..

-- 
с уважениeм / respectfully,
Косырев Сергей