Pager in core should be enabled by default

Gregory Szorc gregory.szorc at gmail.com
Wed Nov 25 13:23:46 CST 2015


On Wed, Nov 25, 2015 at 11:00 AM, Bryan O'Sullivan <bos at serpentine.com>
wrote:

>
> On Wed, Nov 25, 2015 at 1:10 PM, Augie Fackler <raf at durin42.com> wrote:
>
>> “because git does it” doesn’t really convince me much UI-wise. I get that
>> git is popular, but that doesn’t make them right. Subversion certainly has
>> more users and has never supported automatic pagering of output.
>>
>
> Well, Subversion is wrong about almost everything, so probably not a good
> example :-)
>
> When I get back to doing work next week, I’ll turn on pager and force
>> myself to use it for a couple of weeks. Last time I tried this experiment I
>> really hated it and found it to be an awful experience, so I’ve been pretty
>> consistently against pager-by-default.
>>
>
> That's fair. And if you run into problems, writing them up would be
> helpful so we can either address them or declare them fatal to the idea.
>
> We have a large amount of accumulated experience with the pager extension
> enabled by default at Facebook, and only know of one usability issue: if
> less is the pager and it isn't invoked with -K (quit on interrupt), it's
> often necessary to reset the terminal to get it out of a wonky state.
>
> We've fixed that locally via configuration as follows:
>
> [pager]
> pager = LESS=FRKX less
>

So I don't forget, when I set up a new Ubuntu VM the other day, I'm pretty
sure the $LESS isn't defined in the default shell, meaning pager doesn't
kick in by default. I'm not sure how people feel about searching for and
using `less` by default, but not having a pager actually running feels just
as bad as having to enable the extension.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://selenic.com/pipermail/mercurial-devel/attachments/20151125/a05b4fd9/attachment.html>


More information about the Mercurial-devel mailing list