Re: Default search changed by search and replace

On 20/07/10 04:08, Tim Chase wrote:
> On 07/19/10 15:50, Tony Mechelynck wrote:
>> On 19/07/10 19:56, Tim Chase wrote:
>>> Um, this has been vi behavior since the beginning,
>>> AFAICT...I've got an older version of nvi as well as vim6.2
>>> (yes, I can hear Tony complaining already ;-) and they both
>>> behave the same wa
>>
>> :-) I know you were joking, but I'll answer seriously:
>>
>> I'm not complaining. Keeping legacy versions side-by-side with
>> the latest and shiniest for comparison purposes is perfectly
>> all right. Where I'll "complain" is if someone barges in
>> saying "I've found a bug in Vim 6.2", or, as yesterday, "We
>> need a written and notarized assurance that Vim 6.3 supports
>> Windows 7".
>
> Fair enough -- I tend to just use whatever made it onto the system first
> or has the least resistance. But I can sympathize with expecting the
> latest release version (if not the latest-and-greatest build from
> source+patches) in order to address bugs. Usually if I provide an answer
> on the list that I know will trigger a bug in older versions, I'll at
> least mention it (the one I hit most frequently involves using "@:" to
> re-execute the command-line if it contains control-characters such as ^M
> in it; which I think was fixed in 7.0)
>
> In the case of of the v6.2, it's what came stock on my G4 iBook (OS X
> 10.4) and it suffices for just about everything I do in Vim. The only
> thing I occasionally miss is the string-text-objects (i"/i'/a"/a') that
> were introduced in 7.0 (which in 6.2 lists my name beside the request in
> todo.txt :)
>
> On my Win32 boxes at work, I usually just install whatever's current at
> the time I do the install and then don't bother upgrading; while my
> Linux boxes get whatever the repository upgrades me to when I "apt-get"
> the latest system updates.
>
> -tim
>
I think I could (almost) compile Vim in my sleep by now, that Mercurial
repository has done more to make it easy than I thought it would.
Staying with the wavefront also makes my life easier:
- :vimgrep -- I use that constantly, but I thought it appeared in some
6.x version. :help version7 says it didn't. Of course on Unix there is
external grep, but egrep's regexps aren't as well-documented (*nothing*
is as well-documented as something from Vim ;-) ).
- Scrolling back at the more-prompt. For very long lists (mappings,
highlight groups, autocommands, ..., even UTF-8 digraphs) it does come
handy.
- Separate mappings for Visual and Select mode, where in Visual an
alphabetic key is normally an action or a motion (which I might want to
remap) while in Select it replaces the selection (and I want it to go on
doing so).
- Lists? Dictionaries? Funcrefs? I don't use them much in the scripts I
write, but at least one third-party plugin that I love has a Funcref
that my colorscheme calls.
- Try blocks? I don't use them much myself in the scripts I write, but I
wouldn't bet that none of the plugins I use often requires them.
- Floating point? I could go without, but when using Vim as a glorified
calculator it spares me the job of remembering the orders of magnitude
right of a virtual ("scaled") decimal point -- which I could, I used a
slipstick in college and even in high school, but I make fewer errors
when I don't have to.
- Tab pages? I've only begun using them day-in-day-out (three minor
releases later ;-) though I'd done the preliminary work in 7.00aa as
soon as they appeared) but now I find them useful.
- CursorLine and CursorColumn highlighting: the latter is very useful to
keep indents where they belong, the former shows me how far a moderately
long wrapped line goes down the screen.
- Displaying hanzi in the U+20000 block the way they should be, and not
as question marks: that was bugfix 7.1.116.
- Maybe I'll start experimenting with the conceal and relativenumber
features now that they've been brought in.
- And of course a lot of others that don't come to my mind just now.

Best regards,
Tony.
--
Maier's Law:
If the facts don't conform to the theory, they must be disposed
of.

Corollaries:
(1) The bigger the theory, the better.
(2) The experiment may be considered a success if no more than
50% of the observed measurements must be discarded to
obtain a correspondence with the theory.

--
You received this message from the "vim_use" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php