Re: Free alternatives to Rietveld?

2017-01-10 Thread Federico Bruni
Il giorno lun 2 gen 2017 alle 12:10, Alexander Kobel 
 ha scritto:

Hi,

On 2017-01-02 11:50, Federico Bruni wrote:

Il giorno mer 28 dic 2016 alle 20:59, Graham Percival
 ha scritto:

On Wed, Dec 28, 2016 at 06:35:30PM +0100, Federico Bruni wrote:

 It would be possible to add a configuration option in git-cl so
 you can login in rietveld with a specific browser different from
 the default one?


Certainly!  In the source, I see:

-
If your browser is on a different machine then exit and re-run
upload.py with the command-line parameter

  --no_oauth2_webbrowser
-

IIRC that prints a URL string, which you can open with whatever
browser you want (on whichever computer you want).


Thanks! I'll try to remember when I need it.

There's a way to make it the default in my configuration? I see that
config values are saved in .git/config and there are some under
[rietveld]. I wonder if no_auth2_webbrowser=true may work...


Not sure about that; didn't spot anything from a (very) quick glance 
at the code. However, what you can do is to set up your own defaults:


alias git-cl-nooa="/path/to/git-cl --no_oauth2_webbrowser"
alias git-cl="BROWSER=/usr/bin/echo /path/to/git-cl"

Specifying the BROWSER environment variable instructs git-cl to use 
this application as browser - which I guess is what you actually 
wanted in the first place?


I've just had the chance to test these alias.

I've tested first the second alias. I've put this line in my bashrc:
alias git-cl="BROWSER=/usr/bin/midori /home/fede/src/git-cl/git-cl"

When I upload the patch, Midori is launched with the following link, 
which doesn't work (connection refused):

https://codereview.appspot.com/get-access-token?port=8001

If I remove ?port=8001 I see the access token, but I don't know how to 
use it.



So I tried the first alias:
alias git-cl="/home/fede/src/git-cl/git-cl --no_oauth2_webbrowser"

But it doesn't work either, because the option should follow the real 
command:


$ git-cl upload origin/master
unknown command: --no_oauth2_webbrowser
usage: /home/fede/src/git-cl/git-cl 


So I decided to use this:

alias git-cl-upload="/home/fede/src/git-cl/git-cl upload origin/master 
--no_oauth2_webbrowser"


And:

$ git-cl-upload
Usage: git cl upload [options] [args to "git diff"]

git-cl: error: no such option: --no_oauth2_webbrowser

WTF!
I guess that I'll temporarily change my default browser in GNOME 
settings.





___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2017-01-02 Thread Alexander Kobel

Hi,

On 2017-01-02 11:50, Federico Bruni wrote:

Il giorno mer 28 dic 2016 alle 20:59, Graham Percival
 ha scritto:

On Wed, Dec 28, 2016 at 06:35:30PM +0100, Federico Bruni wrote:

 It would be possible to add a configuration option in git-cl so
 you can login in rietveld with a specific browser different from
 the default one?


Certainly!  In the source, I see:

-
If your browser is on a different machine then exit and re-run
upload.py with the command-line parameter

  --no_oauth2_webbrowser
-

IIRC that prints a URL string, which you can open with whatever
browser you want (on whichever computer you want).


Thanks! I'll try to remember when I need it.

There's a way to make it the default in my configuration? I see that
config values are saved in .git/config and there are some under
[rietveld]. I wonder if no_auth2_webbrowser=true may work...


Not sure about that; didn't spot anything from a (very) quick glance at 
the code. However, what you can do is to set up your own defaults:


alias git-cl-nooa="/path/to/git-cl --no_oauth2_webbrowser"
alias git-cl="BROWSER=/usr/bin/echo /path/to/git-cl"

Specifying the BROWSER environment variable instructs git-cl to use this 
application as browser - which I guess is what you actually wanted in 
the first place?



Cheers,
Alexander

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2017-01-02 Thread Federico Bruni
Il giorno mer 28 dic 2016 alle 20:59, Graham Percival 
 ha scritto:

On Wed, Dec 28, 2016 at 06:35:30PM +0100, Federico Bruni wrote:

 It would be possible to add a configuration option in git-cl so
 you can login in rietveld with a specific browser different from
 the default one?


Certainly!  In the source, I see:

-
If your browser is on a different machine then exit and re-run
upload.py with the command-line parameter

  --no_oauth2_webbrowser
-

IIRC that prints a URL string, which you can open with whatever
browser you want (on whichever computer you want).



Thanks! I'll try to remember when I need it.

There's a way to make it the default in my configuration? I see that 
config values are saved in .git/config and there are some under 
[rietveld]. I wonder if no_auth2_webbrowser=true may work...





___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-31 Thread Han-Wen Nienhuys
You could switch to Gerrit for code review. It is a native git tool,
so it needs less awkward scripts.

There are some hosted providers (such as gerrithub), or you can run
your own version of Gerrit if you have inclination to set it up.
Setting up your own account authentication infrastructure is pretty
annoying, and running it yourself is unlikely to save you time.

On Tue, Dec 27, 2016 at 1:33 PM, Simon Albrecht  wrote:
> Hello everybody,
>
> just now I tried to login with Google in order to close my two recent
> Rietveld reviews. However, Google decided that despite entering a
> verification code from e-mail it couldn’t confirm me being authorised to
> login. So I’m currently out. Maybe they want to force me to give a mobile
> phone number?
>
> Whatever the reason for this weirdness, I think it would really be better if
> we had a code review tool which didn’t rely on external login providers. Is
> there really no free alternative?
>
> Best, Simon
>
>
> ___
> lilypond-devel mailing list
> lilypond-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/lilypond-devel



-- 
Han-Wen Nienhuys - hanw...@gmail.com - http://www.xs4all.nl/~hanwen

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-28 Thread Graham Percival
On Wed, Dec 28, 2016 at 06:35:30PM +0100, Federico Bruni wrote:
> It would be possible to add a configuration option in git-cl so
> you can login in rietveld with a specific browser different from
> the default one?

Certainly!  In the source, I see:

-
If your browser is on a different machine then exit and re-run
upload.py with the command-line parameter

  --no_oauth2_webbrowser
-

IIRC that prints a URL string, which you can open with whatever
browser you want (on whichever computer you want).

Cheers,
- Graham

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-28 Thread Federico Bruni
I've been trying to definitively stop using my gmail account recently and 
lilypond review is still forcing me to log in to a browser with my google 
account. Not a big deal, since my contributions are about 3 per year.. But a 
free alternative would be great.

Small workaround suggestion (not useful for Simon's problem, just for the main 
privacy issue):

It would be possible to add a configuration option in git-cl so you can login 
in rietveld with a specific browser different from the default one?
___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-28 Thread David Kastrup
Simon Albrecht  writes:

> Thanks for the heads-up on previous discussions, now I know more about
> the pitfalls.
>
> Also I see now that the ultimate free alternative would likely be the
> one Jan N. brought up in that old discussion, namely exchanging
> patches via e-mail only. Which makes sense with git’s design, however
> the side-by-side diff features of Rietveld are excellent and I
> wouldn’t miss them.
>
> I hope my problems with the Google login will soon cease (I wrote a
> complaint and didn’t hold back my annoyment too much).

Mr. Google is enough of a split personality that spanking his behind
will not cause a reaction at his head.  You are just making life less
pleasant for the person dealing with your mail.

Support personnel are trained to deal with annoyed customers, but those
dealings do not really involve passing on the message.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread Simon Albrecht
Thanks for the heads-up on previous discussions, now I know more about 
the pitfalls.


Also I see now that the ultimate free alternative would likely be the 
one Jan N. brought up in that old discussion, namely exchanging patches 
via e-mail only. Which makes sense with git’s design, however the 
side-by-side diff features of Rietveld are excellent and I wouldn’t miss 
them.


I hope my problems with the Google login will soon cease (I wrote a 
complaint and didn’t hold back my annoyment too much).


Best,
Simon


___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread Graham Percival
On Tue, Dec 27, 2016 at 02:39:09PM +, James wrote:
> On Tue, 27 Dec 2016 13:33:11 +0100
> Simon Albrecht  wrote:
> 
> > Whatever the reason for this weirdness, I think it would really be 
> > better if we had a code review tool which didn’t rely on external
> > login providers. Is there really no free alternative?
> 
> We had a 'similar' discussion back in 2013
> 
> http://lists.gnu.org/archive/html/lilypond-devel/2013-09/msg00351.html

Heh, yes!

I'm aware of problems with the existing setup -- in fact, my first
few patches were delayed by half a week because I had problems
setting it up.  Quite apart from the loss of time, that was a huge
drain on my energy and motivation.

A week ago, I started investigating alternatives.  As always, my
primary motivation is *not* adding any additional burden to our
experienced developers.  I have not yet reached the point at which
it would be useful to discuss any proposals on -devel, though.

(As we can see from the 2013 emails, a very wide-ranging
discussion quickly becomes bogged down and fails to produce
results.  If anybody is very interested in the "alternatives"
question and is willing to do 2-5 hours of work that will probably
end up being discarded, feel free to contact me off-list.)

Cheers,
- Graham

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread James
On Tue, 27 Dec 2016 13:33:11 +0100
Simon Albrecht  wrote:

> Hello everybody,
> 
> just now I tried to login with Google in order to close my two recent 
> Rietveld reviews. However, Google decided that despite entering a 
> verification code from e-mail it couldn’t confirm me being authorised
> to login. So I’m currently out. Maybe they want to force me to give a 
> mobile phone number?
> 
> Whatever the reason for this weirdness, I think it would really be 
> better if we had a code review tool which didn’t rely on external
> login providers. Is there really no free alternative?
> 

We had a 'similar' discussion back in 2013

http://lists.gnu.org/archive/html/lilypond-devel/2013-09/msg00351.html

(just in case anyone is interested).

James

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread Simon Albrecht

On 27.12.2016 14:09, Urs Liska wrote:

While we're at it: is there really no tool that integrates with Git directly?

I find it pretty inconvenient and partly unreliable that the author is 
responsible himself that what is pushed actually is what has been reviewed.


Well, I had a brief look at some of the free tools on 
, and IIUC 
most of them are intended mainly for hosting the source code and also 
integrate a tool for code review. So sensibly they would do just that 
which you request.
So one /conceivable/ solution would be developing a code review tool for 
Savannah.


Best, Simon

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread Urs Liska


Am 27. Dezember 2016 13:33:11 MEZ, schrieb Simon Albrecht 
:
>Hello everybody,
>
>just now I tried to login with Google in order to close my two recent 
>Rietveld reviews. However, Google decided that despite entering a 
>verification code from e-mail it couldn’t confirm me being authorised
>to 
>login. So I’m currently out. Maybe they want to force me to give a 
>mobile phone number?
>
>Whatever the reason for this weirdness, I think it would really be 
>better if we had a code review tool which didn’t rely on external login
>
>providers. Is there really no free alternative?

While we're at it: is there really no tool that integrates with Git directly?

I find it pretty inconvenient and partly unreliable that the author is 
responsible himself that what is pushed actually is what has been reviewed.

Urs

>
>Best, Simon
>
>
>___
>lilypond-devel mailing list
>lilypond-devel@gnu.org
>https://lists.gnu.org/mailman/listinfo/lilypond-devel

-- 
Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail gesendet.

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Free alternatives to Rietveld?

2016-12-27 Thread Thomas Morley
Hi Simon,

2016-12-27 13:33 GMT+01:00 Simon Albrecht :
> Hello everybody,
>
> just now I tried to login with Google in order to close my two recent
> Rietveld reviews. However, Google decided that despite entering a
> verification code from e-mail it couldn’t confirm me being authorised to
> login. So I’m currently out.

just checked with on of my Rietveld-issues which I forgot to close ...
No problems here.
So I can't help.

> Maybe they want to force me to give a mobile
> phone number?

They always want my phone number. I never gave it to them and I'll never do so.
Every time they ask I simply do "Go back one page" in my
FireFox-browser and continue from there.
So far it always worked.

Cheers,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Free alternatives to Rietveld?

2016-12-27 Thread Simon Albrecht

Hello everybody,

just now I tried to login with Google in order to close my two recent 
Rietveld reviews. However, Google decided that despite entering a 
verification code from e-mail it couldn’t confirm me being authorised to 
login. So I’m currently out. Maybe they want to force me to give a 
mobile phone number?


Whatever the reason for this weirdness, I think it would really be 
better if we had a code review tool which didn’t rely on external login 
providers. Is there really no free alternative?


Best, Simon


___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel