| 1 |
On Mi, 2016-02-17 at 10:53 +0000, Duncan wrote: |
| 2 |
> Michał Górny posted on Wed, 17 Feb 2016 07:47:06 +0100 as excerpted: |
| 3 |
> |
| 4 |
> > On Tue, 16 Feb 2016 22:48:08 -0600 Ryan Hill <rhill@g.o> |
| 5 |
> > wrote: |
| 6 |
> > |
| 7 |
> > > On Mon, 15 Feb 2016 15:35:12 +0100 Michał Górny <mgorny@××××××.or |
| 8 |
> > > g> |
| 9 |
> > > wrote: |
| 10 |
> > > > On Mon, 15 Feb 2016 14:37:41 +0100 "Justin Lecher (jlec)" |
| 11 |
> > > > <jlec@g.o> wrote: |
| 12 |
> > > > > On 15/02/16 13:59, Michał Górny wrote: |
| 13 |
> |
| 14 |
> > > > > > Don't mix echo with ewarn. |
| 15 |
> > > > > Why? |
| 16 |
> > > > Because they won't go through the same output channels. |
| 17 |
> > > |
| 18 |
> > > That's kinda the point. You want a blank (unstarred) space to |
| 19 |
> > > separate |
| 20 |
> > > out the "important" messages from the generic spew put out by the |
| 21 |
> > > package manager/eclasses/build system that you have no control |
| 22 |
> > > over. |
| 23 |
> > |
| 24 |
> > This is not just that. Different output channels mean that: |
| 25 |
> |
| 26 |
> > - There is no guarantee of correct output order! The empty lines |
| 27 |
> > may |
| 28 |
> > move randomly over the text. |
| 29 |
> |
| 30 |
> Good point! (Of course the others are too, but this one could be |
| 31 |
> particularly damaging to the intended communication.) |
| 32 |
> |
| 33 |
> > > If you have several different messages you want a blank space in |
| 34 |
> > > between them so you can use e* to create whitespace within the |
| 35 |
> > > message |
| 36 |
> > > to avoid the wall of text syndrome while still making it clear |
| 37 |
> > > where it |
| 38 |
> > > begins and ends. |
| 39 |
> |
| 40 |
> > > You're right that using echo means the whitespace doesn't get |
| 41 |
> > > saved by |
| 42 |
> > > the elog system. A while back someone proposed we add espace for |
| 43 |
> > > exactly this reason but IIRC they were laughed down, which is a |
| 44 |
> > > shame. |
| 45 |
> > |
| 46 |
> > So... to summarize your point. You shouldn't use the correct |
| 47 |
> > function |
| 48 |
> > that is saved in elog which is primary way of getting info because |
| 49 |
> > you |
| 50 |
> > find it more convenient to have empty non-'starred' lines that |
| 51 |
> > don't |
| 52 |
> > actually get to elog and make elog a mess? |
| 53 |
> > |
| 54 |
> > If you really don't like empty 'starred' lines (and I actually like |
| 55 |
> > them |
| 56 |
> > since they make separation between packages cleaner), why not |
| 57 |
> > submit a |
| 58 |
> > patch for Portage and make 'elog' with no arguments output log line |
| 59 |
> > without a star? That's a trivial solution that doesn't require |
| 60 |
> > extra |
| 61 |
> > functions for the sake of inventing elogspace, ewarnspace, ... |
| 62 |
> |
| 63 |
> It is at least possible to use say blank ewarn between elog lines, or |
| 64 |
> the |
| 65 |
> reverse, so while there's no totally blank separator, there's at |
| 66 |
> least a |
| 67 |
> different color to the star on the starred-blank-line separator. |
| 68 |
> |
| 69 |
> Similarly, if there's more than one "topic" to the messages, and |
| 70 |
> they're |
| 71 |
> of different severity, the severities can be interspersed to get |
| 72 |
> color |
| 73 |
> separation. |
| 74 |
> |
| 75 |
> I believe I've seen both techniques used to good effect in a few |
| 76 |
> packages |
| 77 |
> in the past, but I can't name any off the top of my head. |
| 78 |
> |
| 79 |
|
| 80 |
For all those who care, I've updated the eclass under: |
| 81 |
|
| 82 |
https://github.com/gentoo-science/sci/pull/588 |
Archives