Re: [NEW] editors/vis
On 12/20/2017 2:18 PM, Theo Buehler wrote: On Wed, Dec 20, 2017 at 07:53:31PM +0100, Klemens Nanni wrote: On Mon, Dec 11, 2017 at 01:18:24PM +0100, Klemens Nanni wrote: On Sun, Dec 10, 2017 at 08:11:56PM -0500, Brian Callahan wrote: This is not some famous piece of software like the GNU coreutils, where accidentally using the base utility vs. the GNU utility could really matter. Fair enough, I'm happy with a simpler port. They bleed into each other. When I load vis by itself, I get: 0% 1, 1 When I press i to enter insert mode, it becomes: 0% 1, 1 1 Then if I type, say, "OpenBSD" it shows: 100% 1, 1 71 Let's say I finish the sentence and now I wrote "OpenBSD is an operating system." it will display: 100% 1, 10311 Now let's say I press Enter to move to the next line, it says: 97% 2, 0311 I would say this is very broken. This is within tmux, which limited testing suggests makes a difference, but there are still bleeding issues when not in tmux. Specifically, outside of tmux, pressing i in a blank file still produces 0% 1, 1 1 I cannot reproduce this at all, whether in tmux or not. vis works fine for me in both xterm and st. Brian is still the only I heard of having this issue, can anyone else reproduce this? I've been using vis a bit over the past week and the port seems to work just fine. In particular, I was unable to see the problems bcallah described. Besides that, I renamed both binary and manual page to vise again to avoid conflicts; conflicts should be avoided even if it might be of such minor concern. Can anyone else please test this? If no further breakage comes up I'd really like to see this port go in. If bcallah no longer objects, I'm ok with importing this version of the port. I'm not planning on using this editor, so I do not object. ~Brian
Re: [NEW] editors/vis
On Wed, Dec 20, 2017 at 07:53:31PM +0100, Klemens Nanni wrote: > On Mon, Dec 11, 2017 at 01:18:24PM +0100, Klemens Nanni wrote: > > On Sun, Dec 10, 2017 at 08:11:56PM -0500, Brian Callahan wrote: > > > This is not some famous piece of software like the GNU coreutils, where > > > accidentally using the base utility vs. the GNU utility could really > > > matter. > > Fair enough, I'm happy with a simpler port. > > > > > They bleed into each other. > > > When I load vis by itself, I get: > > > 0% 1, 1 > > > When I press i to enter insert mode, it becomes: > > > 0% 1, 1 1 > > > Then if I type, say, "OpenBSD" it shows: > > > 100% 1, 1 71 > > > Let's say I finish the sentence and now I wrote "OpenBSD is an operating > > > system." it will display: > > > 100% 1, 10311 > > > Now let's say I press Enter to move to the next line, it says: > > > 97% 2, 0311 > > > > > > I would say this is very broken. This is within tmux, which limited > > > testing > > > suggests makes a difference, but there are still bleeding issues when not > > > in > > > tmux. Specifically, outside of tmux, pressing i in a blank file still > > > produces > > > 0% 1, 1 1 > > I cannot reproduce this at all, whether in tmux or not. vis works fine > > for me in both xterm and st. > Brian is still the only I heard of having this issue, can anyone else > reproduce this? I've been using vis a bit over the past week and the port seems to work just fine. In particular, I was unable to see the problems bcallah described. > Besides that, I renamed both binary and manual page to vise again to > avoid conflicts; conflicts should be avoided even if it might be of such > minor concern. > > Can anyone else please test this? If no further breakage comes up I'd > really like to see this port go in. If bcallah no longer objects, I'm ok with importing this version of the port. > > New tarball attached.
Re: [NEW] editors/vis
On Mon, Dec 11, 2017 at 01:18:24PM +0100, Klemens Nanni wrote: > On Sun, Dec 10, 2017 at 08:11:56PM -0500, Brian Callahan wrote: > > This is not some famous piece of software like the GNU coreutils, where > > accidentally using the base utility vs. the GNU utility could really matter. > Fair enough, I'm happy with a simpler port. > > > They bleed into each other. > > When I load vis by itself, I get: > > 0% 1, 1 > > When I press i to enter insert mode, it becomes: > > 0% 1, 1 1 > > Then if I type, say, "OpenBSD" it shows: > > 100% 1, 1 71 > > Let's say I finish the sentence and now I wrote "OpenBSD is an operating > > system." it will display: > > 100% 1, 10311 > > Now let's say I press Enter to move to the next line, it says: > > 97% 2, 0311 > > > > I would say this is very broken. This is within tmux, which limited testing > > suggests makes a difference, but there are still bleeding issues when not in > > tmux. Specifically, outside of tmux, pressing i in a blank file still > > produces > > 0% 1, 1 1 > I cannot reproduce this at all, whether in tmux or not. vis works fine > for me in both xterm and st. Brian is still the only I heard of having this issue, can anyone else reproduce this? Besides that, I renamed both binary and manual page to vise again to avoid conflicts; conflicts should be avoided even if it might be of such minor concern. Can anyone else please test this? If no further breakage comes up I'd really like to see this port go in. New tarball attached. vis.tgz Description: Binary data
Re: [NEW] editors/vis
On Sun, Dec 10, 2017 at 08:11:56PM -0500, Brian Callahan wrote: > This is not some famous piece of software like the GNU coreutils, where > accidentally using the base utility vs. the GNU utility could really matter. Fair enough, I'm happy with a simpler port. > They bleed into each other. > When I load vis by itself, I get: > 0% 1, 1 > When I press i to enter insert mode, it becomes: > 0% 1, 1 1 > Then if I type, say, "OpenBSD" it shows: > 100% 1, 1 71 > Let's say I finish the sentence and now I wrote "OpenBSD is an operating > system." it will display: > 100% 1, 10311 > Now let's say I press Enter to move to the next line, it says: > 97% 2, 0311 > > I would say this is very broken. This is within tmux, which limited testing > suggests makes a difference, but there are still bleeding issues when not in > tmux. Specifically, outside of tmux, pressing i in a blank file still > produces > 0% 1, 1 1 I cannot reproduce this at all, whether in tmux or not. vis works fine for me in both xterm and st.
Re: [NEW] editors/vis
Hi Klemens -- Please reply to everyone not just the list. On 12/10/17 19:58, Klemens Nanni wrote: On Sun, Dec 10, 2017 at 07:01:11PM -0500, Brian Callahan wrote: Attached is an updated port, combining both of your ports. I made both of you MAINTAINER, you can share. Looks good to me, thanks for fixing -Os. As for this port conflicting with vis(1), I don't really mind. /home/brian $ which vis /usr/bin/vis even after installing the port, which I think is correct. Set this port to an alias in your shell. I assume that's how a lot of people are going to run it anyway, something like: alias vi='/usr/local/bin/vis' Most probably, but renaming was done to prevent conflicts regardless of aliases and PATH. If this too much however, I won't object. This is not some famous piece of software like the GNU coreutils, where accidentally using the base utility vs. the GNU utility could really matter. However, the line and column display on the bottom right corner is very broken here. You two should fix that and resubmit. What does "very broken" mean? Both column and row index have always been correct for me. Please describe your problem in detail. They bleed into each other. When I load vis by itself, I get: 0% 1, 1 When I press i to enter insert mode, it becomes: 0% 1, 1 1 Then if I type, say, "OpenBSD" it shows: 100% 1, 1 71 Let's say I finish the sentence and now I wrote "OpenBSD is an operating system." it will display: 100% 1, 10311 Now let's say I press Enter to move to the next line, it says: 97% 2, 0311 I would say this is very broken. This is within tmux, which limited testing suggests makes a difference, but there are still bleeding issues when not in tmux. Specifically, outside of tmux, pressing i in a blank file still produces 0% 1, 1 1 ~Brian
Re: [NEW] editors/vis
On Sun, Dec 10, 2017 at 07:01:11PM -0500, Brian Callahan wrote: > Attached is an updated port, combining both of your ports. > I made both of you MAINTAINER, you can share. Looks good to me, thanks for fixing -Os. > As for this port conflicting with vis(1), I don't really mind. > > /home/brian $ which vis > /usr/bin/vis > > even after installing the port, which I think is correct. Set this port to > an alias in your shell. I assume that's how a lot of people are going to run > it anyway, something like: > alias vi='/usr/local/bin/vis' Most probably, but renaming was done to prevent conflicts regardless of aliases and PATH. If this too much however, I won't object. > However, the line and column display on the bottom right corner is very > broken here. You two should fix that and resubmit. What does "very broken" mean? Both column and row index have always been correct for me. Please describe your problem in detail.
Re: [NEW] editors/vis
Hi Carolyn and Klemens --
On 12/10/17 09:56, Klemens Nanni wrote:
On Thu, Dec 07, 2017 at 11:45:16PM -0800, Carolyn Saunders wrote:
I just made a rudimentary port of vis, "[a] vi-like editor based on Plan
9's structural regular expressions," which can be found at
https://github.com/martanne/vis. However, it exposes the binary "vis"
which conflicts with another program in OpenBSD's base. How should this
be rectified? This is my first port so I'm still unfamiliar with the
way this is usually handled.
I've been sitting on my unfinished vis port for way too long, thanks for
bringing it up again.
Some comments inline and my port attached as it's finally ready for
submission.
The version number in your port directory name should be stripped.
V = 0.4
GH_ACCOUNT =martanne
GH_PROJECT =vis
GH_TAGNAME =v${V}
DISTNAME = vis-${V}
GH_* is all you need here; zap V, DISTNAME is set automatically.
MAKE_FLAGS =BINDIR="${TRUPREFIX}/bin" \
CONFIGURE_STYLE=simple
CONFIGURE_ARGS= --mandir="${TRUEPREFIX}/man"
Why not just --prefix=${PREFIX} --mandif=${PREFIX}/man?
NO_TEST=Yes
vis has a test suite and it should be used. Running it through our ports
framework even revealed a libtermkey bug when TERM is unset.
My port works fine on amd64, I've been using it as my main editor
exclusively for over a year now. All tests pass.
Feedback?
Attached is an updated port, combining both of your ports.
I made both of you MAINTAINER, you can share.
As for this port conflicting with vis(1), I don't really mind.
/home/brian $ which vis
/usr/bin/vis
even after installing the port, which I think is correct. Set this port
to an alias in your shell. I assume that's how a lot of people are going
to run it anyway, something like:
alias vi='/usr/local/bin/vis'
However, the line and column display on the bottom right corner is very
broken here. You two should fix that and resubmit.
~Brian
vis.tgz
Description: Binary data
Re: [NEW] editors/vis
On Thu, Dec 07, 2017 at 11:45:16PM -0800, Carolyn Saunders wrote:
> I just made a rudimentary port of vis, "[a] vi-like editor based on Plan
> 9's structural regular expressions," which can be found at
> https://github.com/martanne/vis. However, it exposes the binary "vis"
> which conflicts with another program in OpenBSD's base. How should this
> be rectified? This is my first port so I'm still unfamiliar with the
> way this is usually handled.
I've been sitting on my unfinished vis port for way too long, thanks for
bringing it up again.
Some comments inline and my port attached as it's finally ready for
submission.
The version number in your port directory name should be stripped.
> V = 0.4
> GH_ACCOUNT = martanne
> GH_PROJECT = vis
> GH_TAGNAME = v${V}
>
> DISTNAME =vis-${V}
GH_* is all you need here; zap V, DISTNAME is set automatically.
> MAKE_FLAGS = BINDIR="${TRUPREFIX}/bin" \
>
> CONFIGURE_STYLE= simple
> CONFIGURE_ARGS= --mandir="${TRUEPREFIX}/man"
Why not just --prefix=${PREFIX} --mandif=${PREFIX}/man?
> NO_TEST= Yes
vis has a test suite and it should be used. Running it through our ports
framework even revealed a libtermkey bug when TERM is unset.
My port works fine on amd64, I've been using it as my main editor
exclusively for over a year now. All tests pass.
Feedback?
vis.tgz
Description: application/tar-gz
[NEW] editors/vis
I just made a rudimentary port of vis, "[a] vi-like editor based on Plan 9's structural regular expressions," which can be found at https://github.com/martanne/vis. However, it exposes the binary "vis" which conflicts with another program in OpenBSD's base. How should this be rectified? This is my first port so I'm still unfamiliar with the way this is usually handled. -- PGP Fingerprint: 5A32 BFC6 EAD3 3E07 2986 C35E 5635 4CE5 492F CAFB vis-0.4.tgz Description: application/tar-gz
