Re: [elpa] master 0d619ea 10/19: counsel.el (couns-clj): Add with-no-warnings

Re: [elpa] master 0d619ea 10/19: counsel.el (couns-clj): Add with-no-warnings

[Stefan Monnier]

> -      (cider-sync-request:complete str ":same")))))
> +      (with-no-warnings
> +        (cider-sync-request:complete str ":same"))))))

To silence the "unknown function" warnings, we recommended to use
`declare-function' rather than with-no-warnings.


        Stefan

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Stefan Monnier]

> --- a/swiper.el
> +++ b/swiper.el
> @@ -4,8 +4,8 @@
 
>  ;; Author: Oleh Krehel <ohwoeowho@gmail.com>
>  ;; URL: https://github.com/abo-abo/swiper
> -;; Version: 0.2.0
> -;; Package-Requires: ((emacs "24.1") (ivy "0.2.0"))
> +;; Version: 0.2.1
> +;; Package-Requires: ((emacs "24.1") (ivy "0.2.1"))

The `swiper' GNU ELPA package can't be installed because it requires the
GNU ELPA package `ivy' which doesn't exist.


        Stefan

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Oleh Krehel]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>> --- a/swiper.el
>> +++ b/swiper.el
>> @@ -4,8 +4,8 @@
>  
>>  ;; Author: Oleh Krehel <ohwoeowho@gmail.com>
>>  ;; URL: https://github.com/abo-abo/swiper
>> -;; Version: 0.2.0
>> -;; Package-Requires: ((emacs "24.1") (ivy "0.2.0"))
>> +;; Version: 0.2.1
>> +;; Package-Requires: ((emacs "24.1") (ivy "0.2.1"))
>
> The `swiper' GNU ELPA package can't be installed because it requires the
> GNU ELPA package `ivy' which doesn't exist.

Is there a work-around? I'd prefer to keep them in a single repository.
It's possible to publish separate files from one repo as separate
packages in MELPA. That's what I've been doing so far.

Oleh

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Stefan Monnier]

>> The `swiper' GNU ELPA package can't be installed because it requires the
>> GNU ELPA package `ivy' which doesn't exist.
> Is there a work-around?

For what problem?
The fix is simple: just use

   ;; Package-Requires: ((emacs "24.1"))


-- Stefan

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Oleh Krehel]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> The `swiper' GNU ELPA package can't be installed because it requires the
>>> GNU ELPA package `ivy' which doesn't exist.
>> Is there a work-around?
>
> For what problem?
> The fix is simple: just use
>
>    ;; Package-Requires: ((emacs "24.1"))

But it needs to require ivy for MELPA purposes, since the "swiper" MELPA
package is only swiper.el.

Oleh

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Stefan Monnier]

>> ;; Package-Requires: ((emacs "24.1"))
> But it needs to require ivy for MELPA purposes, since the "swiper" MELPA
> package is only swiper.el.

The problem here is that you have two different "it"s, then.
The "it" in GNU ELPA and the "it" in MELPA are not the same yet you give
them the same name and describe them in the same file, hence the problem.
Currently the one in GNU ELPA is broken.

I recommend you use the same packaging for both, or you drop one of the
two archives (why do you need to distribute it in two different
places?).


        Stefan

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Oleh Krehel]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> ;; Package-Requires: ((emacs "24.1"))
>> But it needs to require ivy for MELPA purposes, since the "swiper" MELPA
>> package is only swiper.el.
>
> The problem here is that you have two different "it"s, then.
> The "it" in GNU ELPA and the "it" in MELPA are not the same yet you give
> them the same name and describe them in the same file, hence the problem.
> Currently the one in GNU ELPA is broken.
>
> I recommend you use the same packaging for both, or you drop one of the
> two archives (why do you need to distribute it in two different
> places?).

OK, I'll merge swiper and ivy together in MELPA. It was MELPA mainterner's idea
to split them, since he likes lightweight packages.

Oleh

Re: [elpa] master 3b78e0e 18/19: swiper.el: Bump version

[Artur Malabarba]

[-- Attachment #1: Type: text/plain, Size: 653 bytes --]

Can you make them into separate Gelpa packages? Instead of 'git pull
subtree' you'll have to do 'cp git-repo-dir/ivy.el Elpa-dir/packages/ivy'.
On Apr 16, 2015 7:53 PM, "Oleh Krehel" <ohwoeowho@gmail.com> wrote:

> Stefan Monnier <monnier@iro.umontreal.ca> writes:
>
> >>> The `swiper' GNU ELPA package can't be installed because it requires
> the
> >>> GNU ELPA package `ivy' which doesn't exist.
> >> Is there a work-around?
> >
> > For what problem?
> > The fix is simple: just use
> >
> >    ;; Package-Requires: ((emacs "24.1"))
>
> But it needs to require ivy for MELPA purposes, since the "swiper" MELPA
> package is only swiper.el.
>
> Oleh
>
>

[-- Attachment #2: Type: text/html, Size: 1094 bytes --]

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

> +  (cond ((functionp collection)
> +         (error "Function as a collection unsupported"))
> +        ((hash-table-p collection)
> +         (error "Hash table as a collection unsupported"))
> +        ((listp (car collection))
> +         (setq collection (mapcar #'car collection))))
[...]
> +      (setq completing-read-function 'ivy-completing-read)

Since completion-tables in the form of a function are pretty common
nowadays (and becoming every time more frequent), this is a real
problem:

   emacs -Q -l .../swiper/swiper-autoloads.el
   M-x ivy-mode RET
   C-x C-f
   C-x b

Notice how both C-x C-f and C-x b signal errors.


        Stefan

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Oleh Krehel]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>> +  (cond ((functionp collection)
>> +         (error "Function as a collection unsupported"))
>> +        ((hash-table-p collection)
>> +         (error "Hash table as a collection unsupported"))
>> +        ((listp (car collection))
>> +         (setq collection (mapcar #'car collection))))
> [...]
>> +      (setq completing-read-function 'ivy-completing-read)
>
> Since completion-tables in the form of a function are pretty common
> nowadays (and becoming every time more frequent), this is a real
> problem:
>
>    emacs -Q -l .../swiper/swiper-autoloads.el
>    M-x ivy-mode RET
>    C-x C-f
>    C-x b
>
> Notice how both C-x C-f and C-x b signal errors.

Yeah, that feature isn't completely ready yet.
I just have to read up on what to do if `collection' is a function.

Oleh

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

> Yeah, that feature isn't completely ready yet.
> I just have to read up on what to do if `collection' is a function.

BTW, the way I see ivy-mode working reminds me of icomplete-mode (but
with a "one-line-per-element" display, which is something that ido-mode
can already provide and which could be added to icomplete-mode fairly
easily).  Do you intend to move in the direction of integration with
icomplete-mode, or do you intend to make it provide a very
different interface?


        Stefan

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Oleh Krehel]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> Yeah, that feature isn't completely ready yet.
>> I just have to read up on what to do if `collection' is a function.
>
> BTW, the way I see ivy-mode working reminds me of icomplete-mode (but
> with a "one-line-per-element" display, which is something that ido-mode
> can already provide and which could be added to icomplete-mode fairly
> easily).  Do you intend to move in the direction of integration with
> icomplete-mode, or do you intend to make it provide a very
> different interface?

Initially, I wanted to just modify ido-mode to not scroll to the last
candidate from the first one on `previous-line'.  After I saw how
incomprehensible and weird the code was, I didn't want to modify it
anymore.

So I see ivy-mode as a replacement of ido-mode with vertical mod and
slightly different matching. It may not do as much as ido-mode, but the
code is very simple.

And actually, I did start with icomplete-mode as a template for
ivy-mode. Some similarities still remain.

Oleh

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Dmitry Gutov]

On 04/16/2015 04:37 PM, Stefan Monnier wrote:
>> +  (cond ((functionp collection)
>> +         (error "Function as a collection unsupported"))
>> +        ((hash-table-p collection)
>> +         (error "Hash table as a collection unsupported"))
>> +        ((listp (car collection))
>> +         (setq collection (mapcar #'car collection))))
> [...]
>> +      (setq completing-read-function 'ivy-completing-read)
>
> Since completion-tables in the form of a function are pretty common
> nowadays (and becoming every time more frequent), this is a real
> problem

It also needs to support obarrays, to be useful with e.g. 
`describe-function'.

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Oleh Krehel]

Dmitry Gutov <dgutov@yandex.ru> writes:

> On 04/16/2015 04:37 PM, Stefan Monnier wrote:
>>> +  (cond ((functionp collection)
>>> +         (error "Function as a collection unsupported"))
>>> +        ((hash-table-p collection)
>>> +         (error "Hash table as a collection unsupported"))
>>> +        ((listp (car collection))
>>> +         (setq collection (mapcar #'car collection))))
>> [...]
>>> +      (setq completing-read-function 'ivy-completing-read)
>>
>> Since completion-tables in the form of a function are pretty common
>> nowadays (and becoming every time more frequent), this is a real
>> problem
>
> It also needs to support obarrays, to be useful with
> e.g. `describe-function'.

It already does, you're looking at an outdated version.  The only thing
that doesn't work so far is the completion in the directory of Info
(regular Info nodes work fine). Also, I have yet to encounter a
collection that is a hash table.

Oleh

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Dmitry Gutov]

On 04/18/2015 12:25 AM, Oleh Krehel wrote:

> It already does, you're looking at an outdated version.

Right, sorry. I had an older stand-alone version of ivy installed as a 
dependency from swiper, and it came first in the load-path.

I probably won't be the only one with that problem, though.

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

> It already does, you're looking at an outdated version.  The only thing
> that doesn't work so far is the completion in the directory of Info
> (regular Info nodes work fine). Also, I have yet to encounter a

For other interesting completion cases, you can try the completion of
revision names in Bzr-controlled files (e.g. C-u C-x v = s: TAB
or C-u C-x v = ancestor: TAB).


        Stefan

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

>> It already does, you're looking at an outdated version.  The only thing
>> that doesn't work so far is the completion in the directory of Info
>> (regular Info nodes work fine). Also, I have yet to encounter a

> For other interesting completion cases, you can try the completion of
> revision names in Bzr-controlled files (e.g. C-u C-x v = s: TAB
> or C-u C-x v = ancestor: TAB).

Or in case you don't have any Bzr tree at hand, try "M-x man RET 3 SPC c TAB"


        Stefan

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Oleh Krehel]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>>> It already does, you're looking at an outdated version.  The only thing
>>> that doesn't work so far is the completion in the directory of Info
>>> (regular Info nodes work fine). Also, I have yet to encounter a
>
>> For other interesting completion cases, you can try the completion of
>> revision names in Bzr-controlled files (e.g. C-u C-x v = s: TAB
>> or C-u C-x v = ancestor: TAB).
>
> Or in case you don't have any Bzr tree at hand, try "M-x man RET 3 SPC c TAB"

Can you explain what "3 c" means?  I'm getting:

- 31 entries for "3 c" with ivy-mode
- 2868 entries for "c 3" with ivy-mode
- 180 entries for "3 c" with plain completion

If I understand correctly, there are 11410 manual entries on my system,
and I'm just filtering through them with ivy-mode. I don't see an error
here.

Oleh

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

> Can you explain what "3 c" means?  I'm getting:

You can specify the manual as "<section-number> <command-name>",
so "3 cos" gives you the "cos.3" manpage.


        Stefan

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Oleh Krehel]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>> Can you explain what "3 c" means?  I'm getting:
>
> You can specify the manual as "<section-number> <command-name>",
> so "3 cos" gives you the "cos.3" manpage.

OK, then I'm fine with "3 cos" not working, as long as "^cos 3" works.
It's a design decision not to flip the regex parts while fuzzy matching.
So there's no error here.

Oleh

Re: [elpa] master 6a098c6 15/19: Add ivy-mode

[Stefan Monnier]

>>> Can you explain what "3 c" means?  I'm getting:
>> You can specify the manual as "<section-number> <command-name>",
>> so "3 cos" gives you the "cos.3" manpage.

> OK, then I'm fine with "3 cos" not working, as long as "^cos 3" works.
> It's a design decision not to flip the regex parts while fuzzy matching.
> So there's no error here.

Not sure what you mean by flip the regex, since the completion doesn't
do any flipping.  It's simply that you can do

   M-x man RET 3 SPC cos RET
or
   M-x man RET cos(3) RET

and both work (no completion involved).  Admittedly, part of the reason
why they both work, is the desire to be able to see the completion
candidates within a particular section without having to move point
backward within the minibuffer.


        Stefan