[Fish-users] Non-orthogonal builtins?

[Philip Ganchev]

Hi all! Why were 'test' and 'echo' made builtins when we have the external
commands?
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] fish as a first programming language

[Philip Ganchev]

Patrick,

I think a shell has some advantages as an introduction to programming
and computing, and Fish is designed to be simple and consistent. Good
luck and keep us posted about how it goes!

On Tue, Jan 26, 2016 at 4:33 AM, Patrick  wrote:
> Hi Siteshwar
>
> Thanks for your post and I received another helpful one offlist too, lots to
> work with.
>
> I think these are all good suggestions but I think I will start with fish
> and then move towards these other options.
>
> Thanks again to all
>
>
>
>
>
>
> On 01/24/2016 01:16 PM, Siteshwar wrote:
>
> On Sun, Jan 24, 2016 at 9:08 PM, Patrick 
> wrote:
>>
>> Hi Diego, Hi Glenn
>>
>> He will turn 10 soon. He has a severe disability but is quite good with
>> computers and could type the whole alphabet, compensating for
>> uppercase/lowercase at 18 months of age.
>
>
>
> I would suggest to start with Squeak and Etoys. There is a talk by Alan Kay
> at Ted which has a brief demo of it. From what I understand, kids can do
> some real learning with it.
>
>
>>
>> I thought I could start with some custom commands to launch
>> videos/webpages he like. I was thinking we could explore the math command to
>> tinker with math and I could introduce him to some custom commands that take
>> an argument or two.
>>
>> Scratch might work, I considered this in the past and then forgot about
>> it. The thing is with fish I could write commands that would better suit his
>> interests.
>>
>> I have been writing an assisted speech device for him and his sister for
>> sometime. I have failed many times but right now i am looking at a
>> Javascript/HTML5 to emit words on clicks and a text user interface for a
>> more complicated application to emit full sentences with proper grammar.
>>
>> Thanks-Patrick
>>
>>
>>
>>
>> On 01/24/2016 10:11 AM, Diego Zamboni wrote:
>>
>> You don't say how old your son is, but I'd go for the tutorials at
>> code.org or Scratch (scratch.mit.edu), which are far more structured, visual
>> and geared for kids. My 7- and 9-year olds love them.
>>
>> --Diego
>>
>>
>> On Sun, Jan 24, 2016 at 2:03 PM, Patrick 
>> wrote:
>>
>> Hi !
>>
>> I thinking of introducing programming to my son. I am reading javascript
>> for kids but it seems to me that many of these concepts would be easier
>> in fish and a shell might be an even easier place to tinker than a
>> browser/text editor.
>>
>> I am looking for the lowest barrier to entry.
>>
>> What do you think?
>>
>> -Patrick
>>
>>
>> --
>> Site24x7 APM Insight: Get Deep Visibility into Application Performance
>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
>> Monitor end-to-end web transactions and take corrective actions now
>> Troubleshoot faster and improve end-user experience. Signup Now!
>> http://pubads.g.doubleclick.net/gampad/clk?id=267308311=/4140
>> ___
>> Fish-users mailing list
>> Fish-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/fish-users
>>
>>
>>
>>
>> --
>> Site24x7 APM Insight: Get Deep Visibility into Application Performance
>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
>> Monitor end-to-end web transactions and take corrective actions now
>> Troubleshoot faster and improve end-user experience. Signup Now!
>> http://pubads.g.doubleclick.net/gampad/clk?id=267308311=/4140
>> ___
>> Fish-users mailing list
>> Fish-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/fish-users
>>
>
>
>
> --
> Regards,
> Siteshwar
>
>
>
> --
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=267308311=/4140
> ___
> Fish-users mailing list
> Fish-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/fish-users
>

--
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=267308311=/4140
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] History search behaviour

[Philip Ganchev]

Is this correct? I get a different behavior with Fish 2.2.0:

echo test_string  # 
echo # ^P replaces the command line to:
set_color white; echo hello  # Something from my distant history that
does not appear in the output of the command 'history'.


On Fri, Dec 11, 2015 at 8:08 AM, Peter Flood  wrote:
> Not sure if this will help or not. If you type 'vim' the command you use
> most that start with vim is offered (greyed out), you can press ctrl+f to
> complete it and enter to run it, if instead you press up or ctrl+p the most
> recent command starting with 'vim' is completed, pressing up/ctrl+p again
> will complete the second most recent command etc etc.
>
> I think there's also a way to scroll through the most common commands
> (greyed out) but I don't know how to do that.
>
>
> On 11/12/2015 15:51, Michael Jones wrote:
>
> Thanks for the response. If I press shift + tab then I get a kind of search
> dialogue (which looks great!) for various possible commands but related to
> what's on the system rather than what is in my history. I am sorry if wasn't
> clear, I was hoping to learn how to navigate my command history in a manner
> which feels consistent whether I'm going for the last command I ran or one
> before that.
>
> My issue at the moment is that it feels like it is a different key to get to
> & run the last command versus getting to & running the second to last
> command.
>
> Thanks again,
> Michael
>
> On 8 December 2015 at 23:14, Jônatas Davi Paganini 
> wrote:
>>
>> Hi Michael! Have you tried shift + tab then tab again?
>>
>> On Tue, Dec 8, 2015 at 7:47 PM, Michael Jones 
>> wrote:
>>>
>>> Hi,
>>>
>>> I'm new to fish shell and generally loving it. A great experience so far.
>>>
>>> I have one behaviour I'd like to continue from my zsh setup which is
>>> searching the history. On zsh, if I type the start of a command like 'vim'
>>> then it doesn't suggest anything immediately but if I do a
>>> history-search-backward style key press then it completes to the last
>>> command I ran that started with 'vim' and I can press 'enter' to run it or
>>> repeat the history-search thingy to find the command before that, etc.
>>>
>>> So it is the same pattern for the last command or the one before it, just
>>> more key presses.
>>>
>>> With fish, if I type 'vim' then it ghosts in the last command I ran that
>>> starts with 'vim' and if I click history-search-backward thingy then I get
>>> the second to last command. So if I want to run the last command, I
>>> currently have to press 'right arrow' to complete the ghosting then enter
>>> and if I want the second to last then I have to press
>>> history-search-backward and then press enter. This small difference is
>>> throwing me off and I'd like to emulate the behaviour I had with zsh.
>>>
>>> Is that possible? I've had a little look at some of the fish scripts and
>>> the use of the 'commandline' command but I can't see how I would achieve the
>>> results I'm after. Any help would be much appreciated!
>>>
>>> Thanks,
>>> Michael
>>>
>>>
>>> --
>>>
>>> ___
>>> Fish-users mailing list
>>> Fish-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/fish-users
>>>
>>
>>
>>
>> --
>> atenciosamente,
>>
>> Jônatas Davi Paganini.
>> --
>> ideia.me | invent.to
>>
>
>
>
> --
>
>
>
> ___
> Fish-users mailing list
> Fish-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/fish-users
>
>
>
> --
>
> ___
> Fish-users mailing list
> Fish-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/fish-users
>

--
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] wishlist history and search

[Philip Ganchev]

Also, maybe up-arrow should also match directory history?

On Sun, Apr 20, 2014 at 4:14 AM, Philip Ganchev phil.ganc...@gmail.com wrote:
 So if the user enters text and presses up-arrow, Fish should match
 input words as substrings of commands in history. If it finds a
 matching command, Fish will suggest the command and the directory
 where it was executed. The user can:

 1. Press Enter to execute the command in the search directory, or
 2. Press Alt+right to cd to the historical directory, then:
 2.a. Press Alt+left to return to the search directory, or
 2.b. Press Enter to execute the command in the historical directory, or
 2.c. Press Ctrl-E Ctrl+U to clear the search and continue working in
 the histriocal directory.


 On Fri, Apr 18, 2014 at 5:40 AM, Martin Bähr
 mba...@email.archlab.tuwien.ac.at wrote:
 hi,

 quite often i want to search for multiple terms in the history when those 
 terms
 are not in a sequence.

 currently typing
 foo barup-arrow

 will search for exactly foo bar,
 but i'd also like to see any results that contain foo and bar with other
 stuff in between.

 maybe exact matches first, then matches with foo and bar in order 
 followed
 by matches with them in any order.

 btw, i saw someone suggesting to search based on the directory, that's a nice
 idea too, moreover, i am occasionally browsing the history to reconstruct 
 what
 i did in a certain directory.

 occasionally i am then coming across commands where i wonder in which 
 directory
 i was running them. so if the history could be augmented with the working
 directory of each entry, even if it doesn't get used in search, just having
 that information would be of great help.

 greetings, martin.

 --
 eKita   -   the online platform for your entire academic life
 hackerspace beijing -http://qike.info
 --
 chief engineer   eKita.co
 pike programmer  pike.lysator.liu.secaudium.net societyserver.org
 BLUG secretary beijinglug.org
 foresight developer  foresightlinux.orgrealss.com
 unix sysadmin
 Martin Bähr  working in chinahttp://societyserver.org/mbaehr/

 --
 Learn Graph Databases - Download FREE O'Reilly Book
 Graph Databases is the definitive new guide to graph databases and their
 applications. Written by three acclaimed leaders in the field,
 this first edition is now available. Download your free book today!
 http://p.sf.net/sfu/NeoTech
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users

--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] wishlist history and search

[Philip Ganchev]

So if the user enters text and presses up-arrow, Fish should match
input words as substrings of commands in history. If it finds a
matching command, Fish will suggest the command and the directory
where it was executed. The user can:

1. Press Enter to execute the command in the search directory, or
2. Press Alt+right to cd to the historical directory, then:
2.a. Press Alt+left to return to the search directory, or
2.b. Press Enter to execute the command in the historical directory, or
2.c. Press Ctrl-E Ctrl+U to clear the search and continue working in
the histriocal directory.


On Fri, Apr 18, 2014 at 5:40 AM, Martin Bähr
mba...@email.archlab.tuwien.ac.at wrote:
 hi,

 quite often i want to search for multiple terms in the history when those 
 terms
 are not in a sequence.

 currently typing
 foo barup-arrow

 will search for exactly foo bar,
 but i'd also like to see any results that contain foo and bar with other
 stuff in between.

 maybe exact matches first, then matches with foo and bar in order followed
 by matches with them in any order.

 btw, i saw someone suggesting to search based on the directory, that's a nice
 idea too, moreover, i am occasionally browsing the history to reconstruct what
 i did in a certain directory.

 occasionally i am then coming across commands where i wonder in which 
 directory
 i was running them. so if the history could be augmented with the working
 directory of each entry, even if it doesn't get used in search, just having
 that information would be of great help.

 greetings, martin.

 --
 eKita   -   the online platform for your entire academic life
 hackerspace beijing -http://qike.info
 --
 chief engineer   eKita.co
 pike programmer  pike.lysator.liu.secaudium.net societyserver.org
 BLUG secretary beijinglug.org
 foresight developer  foresightlinux.orgrealss.com
 unix sysadmin
 Martin Bähr  working in chinahttp://societyserver.org/mbaehr/

 --
 Learn Graph Databases - Download FREE O'Reilly Book
 Graph Databases is the definitive new guide to graph databases and their
 applications. Written by three acclaimed leaders in the field,
 this first edition is now available. Download your free book today!
 http://p.sf.net/sfu/NeoTech
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users

--
Learn Graph Databases - Download FREE O'Reilly Book
Graph Databases is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/NeoTech
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Refreshing ~/.config/fish/config.fish

[Philip Ganchev]

The '.' (dot) built-in is what you are looking for.

On Saturday, May 25, 2013, charlie charlie.f...@gmail.com wrote:
 Wish I had asked this earlier, I've been looking for a replacement for
'source' for a long time now.

 On Sat, May 25, 2013 at 8:32 AM, dag.odenh...@gmail.com 
dag.odenh...@gmail.com wrote:

 . ~/.config/fish/config.fish


 On Sat, May 25, 2013 at 3:23 PM, Sergiusz Kierat 
sergiusz.kie...@gmail.com wrote:

 Hi.
 I would like to know how to refresh fishshell after changing my
configuration.

 What's fish's equivalent to source .bashrc ?

 S.

--
 Try New Relic Now  We'll Send You this Cool Shirt
 New Relic is the only SaaS-based application performance monitoring
service
 that delivers powerful full stack analytics. Optimize and monitor your
 browser, app,  servers with just a few lines of code. Try New Relic
 and get this awesome Nerd Life shirt!
http://p.sf.net/sfu/newrelic_d2d_may
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users




--
 Try New Relic Now  We'll Send You this Cool Shirt
 New Relic is the only SaaS-based application performance monitoring
service
 that delivers powerful full stack analytics. Optimize and monitor your
 browser, app,  servers with just a few lines of code. Try New Relic
 and get this awesome Nerd Life shirt!
http://p.sf.net/sfu/newrelic_d2d_may
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users




 --
 Thanks!
 Charlie
 512-698-2009

--
Try New Relic Now  We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app,  servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Shell Completion

[Philip Ganchev]

Fish's highlighting of incorrect commands could be better if it highlighted
the characters where the command differs from the closest recognized
command, This could be from left to right, or some more sophisticated
matching algorithm like smallest string manipulation. Use a simple
algorithm if the typed command name is long.

As for suggesting a correction, Zsh's solution (with a Y/n prompt) works
for me. Yes, it is modal, but it's not obtrusive. Typing anything other
than 'y' has the same effect as the original command would, that is, an
error message that the command name does not exist.

An alternative solution is to put the suggested correction in the command
line so the user can press Enter to accept it, or Ctrl+C or Ctrl+U to
erase. If the user types any characters, they would be added to the end of
the command. This is modeless.


On Wed, Dec 5, 2012 at 11:19 PM, ridiculous_fish 
corydo...@ridiculousfish.com wrote:

 Hi Charlie,

 The autosuggestion tries to be smart.

 If your command prefixes a history item, that's used as the suggestion.
 Except that, when you run a command, it remembers which arguments are valid
 file paths (possibly relative), and only suggests that command if those
 paths are still valid.

 If there is no suggestion from history, it uses the same algorithm as
 tab-completions, except that it cannot expand subcommands, because it would
 be too slow.

 Also, 'cd' commands are handled specially, whether from history or
 completions.

 Regarding spell checking, at first blush, I think it might be OK as long
 as it's unobtrusive and doesn't do anything modal. It would be very bad to
 put up a prompt that you have to dismiss by typing Y or N. Also remember
 that fish has syntax highlighting and shows invalid commands in red, which
 is a form of spelling suggestions too.

 _fish

 On Dec 5, 2012, at 9:42 AM, charlie charlie.m.sand...@gmail.com wrote:

  How is the word completion determined, is it just the last command
 entered or is it weighted somehow by history usage ?
 
  And different topic, what would you think about adding spelling
 suggestions to mistyped commands, some sort of 'did you mean X ?  hit y to
 run this command'




 --
 LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
 Remotely access PCs and mobile devices and provide instant support
 Improve your efficiency, and focus on delivering more value-add services
 Discover what IT Professionals Know. Rescue delivers
 http://p.sf.net/sfu/logmein_12329d2d
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users

--
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] fish_fish

[Philip Ganchev]

On Thu, May 31, 2012 at 12:14 AM, ridiculous_fish
corydo...@ridiculousfish.com wrote:

 On May 30, 2012, at 1:17 PM, Maxim Gonchar wrote:

[...]
 4. I think fish is doing the right thing with false ; and sdlkfjsdklf.
 That semicolon means that the 'and' is the beginning of a new statement, and
 fish agrees with other shells (including fish trunk) that this is an error.
 Remove the semicolon (false and sdlkfjdsklf) and it does not error.

 This statement confronts all my fish experience. I use fish as default shell
 for the long time.
 What about echo 123 and echo 234 and echo 123; and echo 234?

If I might add to that: having 'and', 'or' and 'not' as separate
commands seems conceptually simpler to me. Also, it is more useful for
interactive use, because you can start executing a command like 'foo'
and then add 'and bar'.

--
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] (Was: Current state of fish in gitorius)

[Philip Ganchev]

On Wed, Mar 28, 2012 at 9:51 AM, Martin Baehr
mba...@email.archlab.tuwien.ac.at wrote:
 On Wed, Mar 28, 2012 at 08:44:18PM +0400, Maxim Gonchar wrote:
 It's definitely ineffective to choose one file with tab, when you can
 type several letters to choose it. When you have to choose several
 files, or several options from long lists. And when after pressing
 tab your line is completed by first completion from the list, you
 can not simply add one or two letters to get desired completion,
 because you need to delete the completion first.

 only if you hit one tab to many. if you are careful to stop hitting tabs
 as soon as the list appears, and then type the right letters to make the
 selection, then zsh does allow you to operate like fish.

 or if the zsh style selection can be done with arrow keys then those not
 wanting to use it would not be affected at all. (note that currently
 arrow keys are used to scroll through a long list of selections, so
 adding highlighting to a particular choice while doing so would be quite
 natural)

 There is also one point, at least for me. When typing fast enough,
 it's always annoying to type single key several times. It's more
 smooth and convenient when you need to press different keys.

 as said above, zsh style doesn't prevent doing that.

Yep, it would be great to have both ways of completing because each
one is the more efficient one in some circumstances. (No
configurability though!)

Spelling correction is also very useful, and can be done with a
completion menu instead of just after you hit enter.

A related blue-sky feature I've wanted since when Axel started on Fish
is a status line or lines below the prompt, to show context, messages
or automatic completion suggestions automatically, as you type. That
would allow to make shell features more discoverable, show spelling
corrections and other things.

Philip

--
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here 
http://p.sf.net/sfu/sfd2d-msazure
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Editing long function

[Philip Ganchev]

On Sun, Nov 6, 2011 at 1:03 AM, gmaxfl gma...@gmail.com wrote:
 Hi,

 when I want to edit a long function, for example 'funced help' part of
 the function, the function does not fit the screen and I can not find
 any way to scroll to the beginning of the function. Do anybody know how
 to do it?
 Or can it be fixed?

 regards,
 Maxim

That's something that was decided Axel, the legendary developer of
Fish, because he did not want to write ugly code that wraps lines. But
that would be better UI, so anyone who knows how can make it wrap
around.

Regards,
Philip

--
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] moving forward, what needs to be done

[Philip Ganchev]

OK, I'm not particularly attached to it.

I see why  has to be the final character. The real Fish prompt is
incorporated into the logo, though it's not the whole logo.


On Wed, Jul 13, 2011 at 10:47 AM, David Frascone d...@frascone.com wrote:

 I don't see a fish there.  I see a comic strip's attempt at profanity.


Heh. Don't we all find what we expect to? :-)


 Maybe reversed with the fish head pointing toward the prompt?:

 ~(*

 -Dave


 On Wed, Jul 13, 2011 at 8:36 AM, Myrddin Emrys myrd...@gmail.com wrote:

 The problem with that logo, nice as it is, is that the  should be on the
 right (as the final prompt character) rather than the left.

 And as far as pretty prompts... I'm of the opinion that they are a waste
 of space overall, but I think that it would be a nice little thing to have
 on, perhaps, the first prompt shown on a new terminal (not every prompt,
 just the first).

 On Tue, Jul 12, 2011 at 22:21, Philip Ganchev phil.ganc...@gmail.comwrote:

 One more try.


 *)~


--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] moving forward, what needs to be done

[Philip Ganchev]

OK good. Then what's wrong with the current Fish logo?


On Wed, Jul 13, 2011 at 7:13 PM, Terin Stock terinjo...@gmail.com wrote:

My beef with the ASCII art is that it doesn't seem all that friendly, in
 reality quiet the opposite. Hackers and programmers (arguably one in the
 same) like ASCII art, but not your average Ubuntu-er.
--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] moving forward, what needs to be done (was: Axel (Was: Fish Shell adoption)

[Philip Ganchev]

Agreed. And I would suggest that we register fishshell.org when its
old registration expires on September 18 (see
http://whois.domaintools.com/fishshell.org). We are not a .com, after
all.

On Sat, Jul 9, 2011 at 9:36 PM, Robert Shinka k...@unknownlayer.net wrote:
 https://gitorious.org/fish-shell indicates that the website for fish is
 fishshell.org, which should be changed to fishshell.com for the time being.

 If development for fish is going to remain on gitorious, I'll update the
 freshmeat project page to include that as well.

 --
 All of the data generated in your IT infrastructure is seriously valuable.
 Why? It contains a definitive record of application performance, security
 threats, fraudulent activity, and more. Splunk takes this data and makes
 sense of it. IT sense. And common sense.
 http://p.sf.net/sfu/splunk-d2d-c2
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users


--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] moving forward, what needs to be done

[Philip Ganchev]

I am not opposed to a new logo, though I don't consider it important.
There is no problem with someone having the trademark to the project's
logo. If you do design a new logo, it might be cool to incorporate the
Fish prompt, ~, into it. Or you can use other ASCII art:

*)

@

*~

Cheers,
Philip


On Fri, Jul 8, 2011 at 3:33 PM, Patrick Mc(avery
patr...@spellingbeewinnars.org wrote:
 Hi Terin, hi List

 I am afraid I don't know enough ruby to help.

 If Axel is done with us should we create another fish logo? I doubt he
 trademarked  fishshell but the fish shell logo is trademarked by default. I
 can create one and mail thumbnail photos to the list? It might take some
 time though. We should probably purge the project of liabilities, if not for
 anything else, then for Terin's sake as he is hosting it.

 -Patrick


 On 11-07-08 12:59 PM, Terin Stock wrote:

 It's a standard Linux stack. Right now the main site is static content. I'm
 currently working on migrating to Redmine

 --
 #Terin Stock
 Undergraduate, Computer Science (CISE), University of Florida

 On Friday, July 8, 2011 at 7:43 AM, Patrick Mc(avery wrote:

 So I don't actually know who is in control of the mailing list? Is it
 Terin? He is in control of fishshell.com right? What is fishshell.com
 running on in terms of a framework or is it all static content? Is there
 anyway to help Terin with the site? once we have a wiki again I would
 like to contribute but setting up the wiki may or may not be that simple.

 --
 All of the data generated in your IT infrastructure is seriously valuable.
 Why? It contains a definitive record of application performance, security
 threats, fraudulent activity, and more. Splunk takes this data and makes
 sense of it. IT sense. And common sense.
 http://p.sf.net/sfu/splunk-d2d-c2

 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users


 --
 All of the data generated in your IT infrastructure is seriously valuable.
 Why? It contains a definitive record of application performance, security
 threats, fraudulent activity, and more. Splunk takes this data and makes
 sense of it. IT sense. And common sense.
 http://p.sf.net/sfu/splunk-d2d-c2
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users



--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish Shell adoption / random comments

[Philip Ganchev]

On Tue, Jul 5, 2011 at 6:33 PM, Patrick Mc(avery
patr...@spellingbeewinnars.org wrote:

[...]
 How about something like this for a description:

 Fish allows for shell scripting with a modern, intuitive syntax. You can do
 nearly everything that you can do in a standard POSIX shell with Fish but
 due to the friendly syntax POSIX shell scripts must be ported to Fish and
 vice versa.
[...]

That's nice, though the description should probably mention the other
strengths of Fish. I think the design document
(http://fishshell.com/user_doc/design.html ) would help a lot. Also,
why do nearly everything, you can do anything? How about something
like this:

The friendly interactive shell is a command shell designed to be
simple, powerful, consistent and discoverable. It has syntax
highlighting, intelligent tab completion, helpful error messages and
easy history search. The syntax is cleaner than POSIX shell syntax and
lets do anything that you can do in POSIX shells.

Regards,
Philip

--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] question about the order of command substitutions and variable expansions

[Philip Ganchev]

On Sat, Feb 19, 2011 at 12:45 PM, weakish weak...@gmail.com wrote:
 In fish's user doc (section Combining different expansions):

 When combining multiple parameter expansions, expansions are performed
 in the following order:

   * Command substitutions
   * Variable expansions

 So command substitutions get performed before variable expansions?

 But give fish the following code:

    echo $(echo HOME)

 fish said:

    echo $(echo HOME)
         ^

 It seems variable expansion happens first here?

No, first Fish checks the syntax. The command you typed is
syntactically incorrect. Fish returns an error message, but not the
one you posted. It says:

fish: Did you mean (COMMAND)? In fish, the “$” character is only used
for accessing variables. To learn more about command substitution in
fish, type “help expand-command-substitution”.
echo $(echo $PATH)
 ^

--
The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE:
Pinpoint memory and threading errors before they happen.
Find and fix more than 250 security defects in the development cycle.
Locate bottlenecks in serial and parallel code that limit performance.
http://p.sf.net/sfu/intel-dev2devfeb
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish sucks (but your shell sucks more)

[Philip Ganchev]

On Wed, Dec 22, 2010 at 11:30 PM, Terin Stock terinjo...@gmail.com wrote:
 Just in cased you all missed it:
 http://markhansen.co.nz/fish-sucks/
 Discussion?

Thanks, Terin. The article itself is shallow and silly, but any
publicity is good publicity, right?

--
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and, 
should the need arise, upgrade to a full multi-node Oracle RAC database 
without downtime or disruption
http://p.sf.net/sfu/oracle-sfdevnl
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish sucks (but your shell sucks more)

[Philip Ganchev]

All 6 criticisms except the last one are design decisions that were
made in order to create a shell syntax that is simple and
orthogonal, and a shell that is interactive and easy to use. That's
the purpose of Fish. So I cannot see them being valid criticisms.
Certainly we should always be open to better design ideas, but
backwards compatibility is low on Fish's priorities because the goal
was to be free try something new.

It seems the author wants completions, syntax coloring and decent help
with Posix syntax.


On Thu, Dec 23, 2010 at 12:38 AM, Myrddin Emrys myrd...@gmail.com wrote:
 The points are valid, if presented in a combatative manner. I don't think
 it's a silly article.

 On Wed, Dec 22, 2010 at 23:00, Philip Ganchev phil.ganc...@gmail.com
 wrote:

 On Wed, Dec 22, 2010 at 11:30 PM, Terin Stock terinjo...@gmail.com
 wrote:
  Just in cased you all missed it:
  http://markhansen.co.nz/fish-sucks/
  Discussion?

 Thanks, Terin. The article itself is shallow and silly, but any
 publicity is good publicity, right?


 --
 Learn how Oracle Real Application Clusters (RAC) One Node allows customers
 to consolidate database storage, standardize their database environment,
 and,
 should the need arise, upgrade to a full multi-node Oracle RAC database
 without downtime or disruption
 http://p.sf.net/sfu/oracle-sfdevnl
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users



--
Learn how Oracle Real Application Clusters (RAC) One Node allows customers
to consolidate database storage, standardize their database environment, and, 
should the need arise, upgrade to a full multi-node Oracle RAC database 
without downtime or disruption
http://p.sf.net/sfu/oracle-sfdevnl
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] add to end of env variable array

[Philip Ganchev]

On Mon, Dec 13, 2010 at 2:10 AM, i...@whywouldwe.com
i...@whywouldwe.com wrote:
 Is there a way to add to the end of an
 environment variable that is an array
 without explicitly specifying the index?

set -Ux PATH $PATH /some/path

[...]
     set -Ux PATH[-1] /some/path

 will change the last element but

     set -Ux PATH[-0]

 gives `set: Array index out of bounds`

That would be a good feature.

--
Oracle to DB2 Conversion Guide: Learn learn about native support for PL/SQL,
new data types, scalar functions, improved concurrency, built-in packages, 
OCI, SQL*Plus, data movement tools, best practices and more.
http://p.sf.net/sfu/oracle-sfdev2dev 
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Running fish_indent on all fish functions.

[Philip Ganchev]

On Sat, Dec 4, 2010 at 10:33 PM, Dylan Smith dylan.ah.sm...@gmail.com wrote:
 I thought a good test case for fish_indent was to try it on all the fish
 functions (i.e. share/functions/*.fish). I used the following command:

 for f in *.fish; diff $f (../../fish_indent  $f | psub); end | vim -R -
 (Note: I used vim to show whitespace and for syntax highlighting)

 Here are my observations:

 In __fish_complete_ls.fish I found that and commands get split onto the
 next line:
 -       command ls --version /dev/null ^/dev/null; and set is_gnu --is-gnu
 +       command ls --version  /dev/null ^ /dev/null
 +       and set is_gnu --is-gnu
 A good heuristically for fish_indent would be to keep the and and or
 commands on the same line. Perhaps this will cause long lines in some
 cases, but if the and command gets split onto another line then perhaps
 it should be indented.

Absolutely not! and, or and not should go on separate lines. It
is easier to read code that has smaller chunks of code per line. The
traditional shell script mindset runs counter to that, but if you
change the mindset, it's all good. Fish is about doing the right thing
even if it means breaking with convention. Also, they are separate
commands, so there should be no exceptions for them compared to other
commands. That is the Law of Minimalism.

 Above you might also have noticed ^/dev/null get a space added between the
 redirection symbol and the filename. This frequently occurs, because it
 seems as if it is more common to omit the space, so should fish_indent be
 changed to omit this space?

No. It is easier to read code that has tokens separated by spaces. You
can certainly find shell script code that omits the spaces before or
after the redirection symbols, but I think that's just sloppiness and
poor style; there is no good reason for it.

[...]


Philip

--
What happens now with your Lotus Notes apps - do you make another costly 
upgrade, or settle for being marooned without product support? Time to move
off Lotus Notes and onto the cloud with Force.com, apps are easier to build,
use, and manage than apps on traditional platforms. Sign up for the Lotus 
Notes Migration Kit to learn more. http://p.sf.net/sfu/salesforce-d2d
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Login/startup problem

[Philip Ganchev]

It seems that an event fish_prompt is being emitted and it calls the
function __fish_on_interactive, which calls function
__fish_config_interactive, etc (as listed from bottom to top). It
does not make sense to me why a fish_prompt event would call that
function.

I would try to figure out where function my_keybindings is being
called and comment out the call, to see if it stops the error
messages.

Another strategy to try is to copy parts of ~/.config/fish/ until you
find the offending file.

Sorry, I don't have too much time to help right now.

Philip

On Mon, Nov 29, 2010 at 7:15 PM, Mike Miller fish-us...@mgmiller.net wrote:
 Hi all, was hoping you could help.

 I've copied my fish config to a second machine, changing the name of the
 fishd file.  Now when I log in there I get this series of errors, and
 everything is broken:

 Last login: Mon Nov 29 16:02:00 2010 from 10.0.2.2
 fish: Unknown command “my_keybindings”
 Standard input: begin; my_keybindings ;end
                        ^
 in . (source) call of file “-”,
        called on standard input,

 in function “__fish_reload_key_bindings”,
        called on line 213 of file
 “/usr/share/fish/functions/__fish_config_interactive.fish”,

 in function “__fish_config_interactive”,
        called on line 104 of file “/usr/share/fish/config.fish”,

 in function “__fish_on_interactive”,
        called on standard input,

 in event handler: handler for generic event “fish_prompt”

 The file does exist and works fine on my original computer:
[...]
 Anyone know how to fix it?
 Thanks,
[...]

--
Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
Tap into the largest installed PC base  get more eyes on your game by
optimizing for Intel(R) Graphics Technology. Get started today with the
Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
http://p.sf.net/sfu/intelisp-dev2dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Key binding to append --help to command line

[Philip Ganchev]

On Wed, Dec 1, 2010 at 12:39 PM, Grissiom chaos.pro...@gmail.com wrote:
 On Sun, Nov 28, 2010 at 2:12 AM, James Bowlin bow...@mindspring.com wrote:
 I think the alt-p key binding that appends |less to the command
 line is cool and useful.  I wrote a little function that mimics
 this behavior to insert --help in the command line.  I bound it to
 alt-h with:

 bind \eh __fish_help_arg

 Here is the simple function:

 function __fish_help_arg
        commandline -i -- --help
 end

 The reason I found this useful is because I'm lazy and I will always try
 -h first instead of --help.  Most of the time this works but when it
 fails it is a PITA.  Since alt-h is even easier for me to type I've been
 using that instead of -h.  This seems to streamline the discovery
 process.

 I've also added completion for funced (and my variations).  It is just
 one line:

 complete -c funced -xa (functions -na) --description Function


 That seems COOL! ;) As it is something new, should this go into fish-next?

 --
 Cheers,
 Grissiom

I think it should go into some kind of extras place, maybe the wiki;
but not in the main Fish code because I don't think it's a central
feature.

Philip

--
Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
Tap into the largest installed PC base  get more eyes on your game by
optimizing for Intel(R) Graphics Technology. Get started today with the
Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
http://p.sf.net/sfu/intelisp-dev2dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] A more robust help function

[Philip Ganchev]

Hi James,

Awesome! That's quite elaborate, and I think most of that is very
useful. Thanks for all the work!


On Sat, Nov 27, 2010 at 5:33 PM, James Bowlin bow...@mindspring.com wrote:
[...]
 NOTES:
[...]
 I used an ugly kludge to print out a list of all the browsers we
 looked for.  I create a temporary directory and touch a file with
 the name of each browser and then do an ls.  This allows me to
 take advantage of the formatting ls provides instead of just
 printing a long string that gets wrapped at the right edge of the
 terminal.  This code only runs when we can't find a browser which
 should be a rare occurrence.

Why not use the column command?


[...]
 The --text and --no-browser flags are to aid testing, although I
 find the --text flag to be very handy.  You can always display
 in text mode by setting $BROWSER to a known text browser.

 I realize I may have gone way overboard with the --verbose and
 --shutup flags.  These two flags erase and set the global
 $fish_quiet_help variable.  The default, of course, is verbose mode.
 I did not KISS.  One compromise would be to only allow the complete
 --shutup and --verbose and no abbreviations.  Then don't mention
 them in the simple help -h.

Personally, I think the --verbose, --quiet and --shutup options
and the fish_quiet_help environment variable add unnecessary
complexity. Fish aims to do the right thing without user intervention,
and aims for functionality through orthogonality and minimal
configurability. See http://fishshell.org/user_doc/design.html  Now, I
know that we want to lower the bar as much as possible, especially for
the help command. But adding these options does not do that.  I
think it's fine to show an info message every time the user asks for
help. Maybe I'm missing some important use cases?

The delay is an interesting idea. If we find it's annoying, maybe a
better solution is to have a static help page that has the message
could not find a topic at the top? And the help page would be
auto-generated with the usual index page.

--
Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
Tap into the largest installed PC base  get more eyes on your game by
optimizing for Intel(R) Graphics Technology. Get started today with the
Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
http://p.sf.net/sfu/intelisp-dev2dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Use of alt + arrow keys for dir history and forward/backward word

[Philip Ganchev]

I agree that Ctrl+Left and Ctrl+Right are more customary for moving
forward/backward by 1 word, and Fish should be usable on the console.
That should be the default key combo on Fish.

I don't see the argument against Alt+Up, Alt+Down for moving through
history words.

Philip


On Wed, Nov 24, 2010 at 8:49 PM, James Bowlin bow...@mindspring.com wrote:
 I use tabbed terminal emulators in X and I also often make use of
 the virtual consoles in Linux (ctrl-alt-F[1-6]).  AFAIK, the normal
 key combo for changing to the next/prev virtual console is alt +
 left/right arrow.  Because of this I make all my tabbed terminal
 emulators use this key combo as well for switching between
 terminal tabs.

 I've tried other solutions but one problem is that if I'm trying
 to fix someone else's Linux machine and I'm in the virtual
 consoles, then I get confused unless I'm used to using alt +
 left/right arrow for switching vconsoles on my machine.

 IMO, the binding in fish of alt + left/right arrow for directory
 history and moving forward and backward one word should be
 changed to ctrl + left/right arrow.  Inside of X-windows, it
 doesn't make much difference but I do think ctrl + arrow is more
 natural for moving by words.

 All the problems come in the vconsoles.  Normally alt + arrows are
 bound to switching vconsoles so not only do those combos fail to
 do what is expected in fish (dir history or move one word) they
 do something very unexpected: they change the vconsole.

 The problem in the vconsoles is that unless we run loadkeys (as
 root) alt + arrows changes vconsoles and all other $meta + arrow
 combinations do the same thing as the arrow keys alone.  So with a
 default setup ctrl + arrows will also be broken in the vconsoles
 but not nearly so broken as alt + arrows.  If we are willing and
 able to run loadkeys as root then we can map ctrl + left/right to
 alt + b and alt + f which are also mapped to the
 prevd-or-backward-word and nextd-or-forward-word commands which
 makes ctrl + left/right work in the vconsoles.  That is how my
 machine is configured now.

 Bottom line: I think the default alt + left/right bindings should
 be changed to ctrl + left/right and alt + left/right should be
 unbound.  While this would still be broken OOTB in the vconsoles,
 it is less broken.  Further, it is easily fixed by running
 loadkeys as root without disrupting the normal use of alt +
 left/right to change vconsoles.


 Peace, James



 --
 Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
 Tap into the largest installed PC base  get more eyes on your game by
 optimizing for Intel(R) Graphics Technology. Get started today with the
 Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
 http://p.sf.net/sfu/intelisp-dev2dev
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users


--
Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
Tap into the largest installed PC base  get more eyes on your game by
optimizing for Intel(R) Graphics Technology. Get started today with the
Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
http://p.sf.net/sfu/intelisp-dev2dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Current state of fish in gitorius

[Philip Ganchev]

On Thu, Nov 18, 2010 at 10:20 AM, i...@whywouldwe.com
i...@whywouldwe.com wrote:
 I use fish daily and it seems complete
 to me, I'm curious about Grissom's
 comment regarding devs, what new
 features would people like to see? Is
 there a list somewhere?

Yes, Axel's list from a long time ago:
http://fishshell.org/user_doc/index.html#todo

I also have other ideas for improvements, which are in the mailing
list archives but are not summarized anywhere. We should probably make
a wiki page with suggestions.

Philip

--
Increase Visibility of Your 3D Game App  Earn a Chance To Win $500!
Tap into the largest installed PC base  get more eyes on your game by
optimizing for Intel(R) Graphics Technology. Get started today with the
Intel(R) Software Partner Program. Five $500 cash prizes are up for grabs.
http://p.sf.net/sfu/intelisp-dev2dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish with no 'source' command?

[Philip Ganchev]

The command is .' (dot)

http://fishshell.org/user_doc/commands.html#source

It does not do both . and source because Fish is intended to have
a small set of features that do different things (the Law of
Orthogonality: http://fishshell.org/user_doc/design.html )

But maybe the documentation for the command should actually mention
the word source. Maybe it should be in the FAQ?


On Fri, Oct 8, 2010 at 5:47 PM, Italo Maia italo.m...@gmail.com wrote:
 Why the `source` command does not work with fish?

--
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2  L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] fish developing

[Philip Ganchev]

Is this the sort of thing you mean?

http://fishshell.org/doc2/pages.html

http://fishshell.org/doc2/annotated.html

http://fishshell.org/doc2/files.html


On Wed, Oct 6, 2010 at 11:37 AM, Nima Sahraneshin unix.n...@gmail.com wrote:
 Hi


 I am looking for a technical documen of fish for developing it . Can
 you help me?

 With thanks

 Nima Sahraneshin Samani
 Computer Science Student at University of  Shahrekord.

 --
 Beautiful is writing same markup. Internet Explorer 9 supports
 standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2  L3.
 Spend less time writing and  rewriting code and more time creating great
 experiences on the web. Be a part of the beta today.
 http://p.sf.net/sfu/beautyoftheweb
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users


--
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2  L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Globbage?

[Philip Ganchev]

I love the idea of interactive globbing!

--
ThinkGeek and WIRED's GeekDad team up for the Ultimate 
GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the 
lucky parental unit.  See the prize list and enter to win: 
http://p.sf.net/sfu/thinkgeek-promo
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Globbage?

[Philip Ganchev]

On Sat, Jun 5, 2010 at 1:06 AM, Martin Bähr
mba...@email.archlab.tuwien.ac.at wrote:
 On Fri, Jun 04, 2010 at 05:31:43PM +0300, Beni Cherniavsky-Paskin wrote:
 (B) there should be an easy one-press way to quote the current wildcard.
    Maybe pressing TAB on an unmatched wildcard should quote it?

 a slight problem with this is that if the user expects a match, but the
 match fails because of a typo, then it will be quoted, but since the
 user didn't want quotes he or she has to go back to remove the quote and
 correct the typo.

Fish needs an undo command in general.

 although, the idea could work if the tab toggles quoting.
 so hitting tab on a quoted argument removes the quotes again.

Toggles are confusing, in my opinion.

--
ThinkGeek and WIRED's GeekDad team up for the Ultimate 
GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the 
lucky parental unit.  See the prize list and enter to win: 
http://p.sf.net/sfu/thinkgeek-promo
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] help in OSX

[Philip Ganchev]

The problem is more general. For example, on Linux where help is
configured to use w3m:

~ help complete
w3m: Can't load file:///usr/share/doc/fish/builtins.html#complete.
~ ls /usr/share/doc/fish/builtins.html
ls: cannot access /usr/share/doc/fish/builtins.html: No such file or directory
~ w3m '/usr/share/doc/fish/commands.html#complete'
# Works
~ fish --version
fish, version 1.23.1

On Sun, Apr 4, 2010 at 3:00 AM, Michael Lachmann lachm...@eva.mpg.de wrote:
 On my machine, calling help only opens the main help page, and does
 not go to the specific page.
 So:
 help complete
 open the help page commands.html, instead of commands.html#complete
 (I have BROWSER set to open)
[...]
 Works, but maybe there is a more elegant solution?

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Uniquely add to path?

[Philip Ganchev]

On Fri, Apr 2, 2010 at 7:28 PM, David Frascone d...@frascone.com wrote:
 I like the third function . . .. but I'd do a -d $x, since it should be a
 directory :)

Sure. Also, the test should be contains $x $PATH. And the test
should be done with two if-statements.

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Uniquely add to path?

[Philip Ganchev]

On Thu, Apr 1, 2010 at 9:46 AM, David Frascone d...@frascone.com wrote:

 There's a really cool feature in zsh that allows you to uniquely add a
 string to the path, if it did not exist before:

 typeset -U path

 This makes all items in path have to be unique, so then, if you do:

 path=(/foo/bin $path)
 path=(/foot/bin $path)

 You will NOT have two /foo/bin(s) in your path.

 Is there an equivalent way to do this in fish?

Not that I know of. Fish is designed to have a minimal set of builtins
that are as orthogonal as possible. But it might be useful to define a
function like:

function varuniq --description Return the unique arguments, merging
to the first occurrence of a word.
for x in $argv;
if not contains $x $ret
set ret $ret $y
end
end
end

Then use it like this:

set PATH (varuniq $PATH /home/joe/bin home/joe/src)


Or define:

function typeset
set rest argv[(seq 2 (count $argv))]
set $argv[1] (varuniq $argv[1] $rest)
end


Or even:

function path
for x in argv
if test -f $x and not contains $PATH $x
set PATH $PATH $x
end
end
end

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Does anyone use fish as their #1 shell (OS/X)?

[Philip Ganchev]

On Tue, Mar 30, 2010 at 3:48 PM, David Frascone d...@frascone.com wrote:
 Ok -- let's start slow -- my aliases need to be recreated -- which is fine.

It's better to write functions than use alias. alias is a sort of
a crutch; in Fish, it's a function that creates a simple function for
you, but functions in general are more powerful than aliases.

 I set the alias 'vi' to either be vim, or mvim depending on if I'm on my
 mac, or shelled into a linux box.  But, the type command does not seem to
 do what I think it should:

 codemon...@daves-mbp ~/.c/fish type vi
  -- type -Q -o tpPafh -- vi
 type: Could not find 'vi'
 codemon...@daves-mbp ~/.c/fish vi
 codemon...@daves-mbp ~/.c/fish type fg
  -- type -Q -o tpPafh -- fg
 fg is a builtin

Do these problems occur on Linux or only on MacOS? type, which and
alias work on my standard install of Fish 1.23.1 on Linux:

~ type type
type is a builtin
~ type vi
vi is /usr/bin/vi
~ type which
which is /usr/bin/which
~ type alias
alias is a function with definition
[...]
~ which vi
/usr/bin/vi
~ alias vi=vim
~ type vi
vi is a function with definition
function vi
vim $argv
end
~ alias hello vi
~ type hello
hello is a function with definition
function hello
vi $argv
end


 codemon...@daves-mbp ~/.c/fish alias
 alias: Expected one or two arguments, got 0
 codemon...@daves-mbp ~/.c/fish alias vi
 fish: Array index out of bounds
 /usr/local/share/fish/functions/alias.fish (line 19):             set body
 $tmp[2]

 ^
 in function 'alias',
     called on standard input,
     with parameter list 'vi'

Feel free to extend the alias function with these features:
1. Handling the case when there are no arguments
2. Querying for an alias. You could simply call the functions builtin here.

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish compatibility with other shells?

[Philip Ganchev]

The Fish wiki has a short page about this.[1] Feel free to expand it.
Yes, in Bash, variables are set using export, not set; see the
Bash manual page, section Shell Builtin Commands.

[1] http://fishshell.org/wiki/moin.cgi/BashToFish

On Thu, Mar 25, 2010 at 11:44 AM, chombee chom...@lavabit.com wrote:
 Is there a list somewhere of where fish's syntax is compatible and
 incompatible with other shells (sh, bash, zsh)?

 I'm trying to share as much of my shell configuration as possible between
 shells by putting it in files that are evaluated by each shell. For example
 I moved all my aliases into ~/.aliases and added . ~/.aliases to my fish
 config file. The alias command seems to be compatible, so other shells
 should be able to source the same aliases file.

 My fish config file also has a lot of variables set with the set or set -x
 command. From reading the man files of sh and bash and doing a quick test
 with bash, fish's set syntax does not seem to be supported by other shells.

 For example in fish:

    fish  set foo 'bar'
    fish  echo $foo
    bar
    fish 

 but in bash:

    bash:$ set foo 'bar'
    bash:$ echo $foo

    bash:$

 and also comparing my fish and bash config files fish seems to use set -x to
 set an environment variable, whereas bash uses export.

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] fish project homepage

[Philip Ganchev]

On Wed, Oct 7, 2009 at 12:56 PM, the jimmybot jimmy...@jimmybot.com wrote:
 ... is there anywhere one
 might find or share custom fish completions or nice little functions /
 scripts?

There was a wiki with completions and functions. I think it was on the
fishshell.org site.

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish doesn't expand ~ in environment variables

[Philip Ganchev]

On Mon, Oct 5, 2009 at 1:15 PM, Ken Kinder kkin...@gmail.com wrote:
 On Thu, Sep 17, 2009 at 2:20 PM, Isaac Dupree
 m...@isaac.cedarswampstudios.org wrote:

 Look, here is how it works in fish:

 .. set foo ~/hi
 .. echo $foo
 /home/me/hi
 .. set foo ~/hi
 .. echo $foo
 ~/hi

 Perhaps you were confused by quoting?

 Ah, I this was the command that was bugging me:

 set -x PYTHONPATH (pwd)
 echo $PYHONPATH
   - ~/whatnot

 The problem is that pwd seems to return ~ in the value. However, if you
 explicitly call /bin/pwd, it returns a proper path. Is fish using an
 internal pwd?

 type pwd
pwd is a function with definition
function pwd --description 'Print working directory'
echo $PWD | sed -e s|^$HOME|~|

end

So you can put this command inside
~/.config/fish/functions/config_interactive.fish :

functions --erase pwd

Or, redefine the function pwd, by creating and editing the file
~/.config/fish/functions/pwd.fish .

Cheers,
Philip

--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Fish doesn't expand ~ in environment variables

[Philip Ganchev]

Fish expands ~ when you access the variable, so why confusion and
why unnecessary typing?

On Thu, Sep 17, 2009 at 11:04 AM, Ken Kinder kkin...@gmail.com wrote:
 In Bash, when you set an environment variable to ~/foo, the shell
 automatically expands your home directory. However in fish, if you use the
 set command to set an environment variable, it does not expand ~. This leads
 to some confusion and unnecessary typing, IMHO.
 --
 Come build with us! The BlackBerryreg; Developer Conference in SF, CA
 is the only developer event you need to attend this year. Jumpstart your
 developing skills, take BlackBerry mobile applications to market and stay
 ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
 http://p.sf.net/sfu/devconf
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users

--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Possible Bug and question.

[Philip Ganchev]

On 2009-08-09, Daniel Baucom loudambia...@gmail.com wrote:
 Question:
 Is there an easier way to have fish reload the conf file than doing an:
 exec fish ?
[...]

. ~/.config/fish/fish.config

The dot builtin is the equivalent to source.

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] BUG: 100% CPU usage when used as login shell

[Philip Ganchev]

On Fri, Aug 7, 2009 at 5:29 PM, Ken Kinderkkin...@gmail.com wrote:
 Hi. I searched the mailing list and saw this post:

 http://www.mail-archive.com/fish-users@lists.sourceforge.net/msg02151.html

 I thought I would confirm the same bug, and the same solution. :(

What was a call to alias that didn't work and what was the modified
one that did work?

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] set with arrays, functions emitting arrays

[Philip Ganchev]

On Wed, Jun 17, 2009 at 2:44 PM, James Aguilaraguilar.ja...@gmail.com wrote:
 On Tue, Jun 16, 2009 at 19:55, Philip Ganchev phil.ganc...@gmail.com
 wrote:
[...]
 This is the closest to what I want.  Is there any way to get something like
 this:

 for var in [something]; echo $var[1] - $var[2]; end
 1 - a
 2 - b

 Where I can easily access each field separately (e.g. with array
 subscripts)?

 I know I can do
 for var in 1:a 2:b; echo $var; end

 But that doesn't give me easy access the fields using array
 indexing.  The best I can think of right now is to define a separate function get_fields (gf for short):

 function gf --description Get\ one\ field\ from\ a\ colon\ separated\ field\ 
 list.
 echo $argv[1] | awk -F: '{ print $'$argv[2]' }'
 end

 And then I can do this:
 [...] for var in 1:a 2:b; echo (gf $var 1) - (gf $var 2); end
 1 - a
 2 - b
 Which is what I want, but maybe a little more verbose than my true desire.

I don't understand why you would want to access the fields, but you can do:

for v in 1:a 2:b
# Create a new array variable
set w (echo $v | tr : \n)
# Use the new array variable
echo $w[2] - $w[1]
end

Interestingly, here you cannot use an array of arrays:

set x 1 a
set y 2 b
for v in x y
echo $$v[2] - $$v[1]
end
fish: Array index out of bounds
echo $$v[2] - $$v[1]

for v in x y
echo $$v[2] - $$v[1]
end
fish: Did you mean (COMMAND)? In fish, the “$” character is only used
for accessing variables. To learn more about command substitution in
fish, type “help expand-command-substitution”.
echo $$v[2] - $$v[1]

Personally, I think that $$v[2] ought to be made to work.

--
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] strange binding problem

[Philip Ganchev]

On Sun, Feb 1, 2009 at 10:39 AM, Axel Liljencrantz
liljencra...@gmail.com wrote:


 2009/1/14 Philip Ganchev phil.ganc...@gmail.com

...
 Actually, none of these commands work for me as you typed them.
 According to the help, they should be:

 bind '\C-p: history-search-backward'
 bind '\C-n: history-search-forward'
 bind '\M-.: history-token-search-backward'
 bind '\M-:: history-token-search-forward'

 Those work fine for me.

 Hi Philip,

 Seems like you're running an older fish version?

Hi Axel! Great to see you back in action. It's impressive how many
emails you are able to answer in one day :-)

I must have been using an older version or something. Now I have
1.23.0, and the help and init files are updated with the new notation
(\cq, etc). The old notation does not work. Unfortunately, the new one
doesn't work either:

 bind \cq history-search-backward

Now press Control+q, nothing happens.

 bind \cq 'echo swimming'

Control+q, nothing happens.

--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] BUG: Autocompletation files with '=' and '{'

[Philip Ganchev]

On Tue, Jan 27, 2009 at 8:51 AM,  jdinu...@uc.edu.ve wrote:
...
 When I press tab to autocomplete a filename with '=' or '{' in its name the
 fish console hangs. It happends in gnome gterminal and in the text conlose.
...
 I'm running fish 1.23.0-5 on Ubuntu-8.10 x64 installed from ubuntu's 
 repository.

Can't reproduce with Fish 1.22.3 on Ubuntu 8.04.

--
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] mimedb bug fix and fish on github

[Philip Ganchev]

James, Thank you for all your work!

Maybe you should ask Axel for control of the repository and the web
page. That way it might be a little less confusing to newcomers.

On Fri, Jan 16, 2009 at 2:08 PM, James Reeves jree...@weavejester.com wrote:
...
 Since Fish seems not to be actively maintained, I've also converted
 the Fish darcs repository to git, applied the mimedb patch, and
 uploaded it to github. If any programmers want to make any changes to
 Fish, it's now just that little bit easier to do so.

 The repository is at: http://github.com/weavejester/fish
...

--
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] strange binding problem

[Philip Ganchev]

On Mon, Jan 12, 2009 at 8:03 AM, Goran Josic jos...@lu.unisi.ch wrote:
 Hi guys,

 I have a strange problem.
 I used google but didn't found any useful information.

 In my .config/fish/config.fish I have:

 bind \cp history-search-backward
 bind \cn history-search-forward

 bind \e. history-token-search-backward
 bind \e: history-token-search-forward

 but the second two bindings are not working.

 What is strange is that they work if I type them as regular command.

Actually, none of these commands work for me as you typed them.
According to the help, they should be:

bind '\C-p: history-search-backward'
bind '\C-n: history-search-forward'
bind '\M-.: history-token-search-backward'
bind '\M-:: history-token-search-forward'

Those work fine for me.

--
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] wildcard ** makes fish shell freeze

[Philip Ganchev]

Hi, Goran

On Thu, Jan 1, 2009 at 2:48 PM, Josic Goran goran.jo...@lu.unisi.ch wrote:
...
 I don't know if this is a bug or not but when I type ** the fish shell
 freezes.

That's strange. It does not do that for me. It does this, and what I
would expect:

 **
fish: Illegal command name **
 fish --version
fish, version 1.23.0

--
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] is it possible to search the beginning of a a command like it matlab?

[Philip Ganchev]

Thanks, Abe.  Without having played with it, I also would think that
this algorithm would be confusing for users.  The relevance score may
be a nice idea, though it is arguably even more complicated.

I seem to recall that we once discussed using the position of the
cursor as an indicator; if the cursor is at char0, restrict the search
to strings that start with the query string. If we did not discuss it,
at least I was thinking about the idea.  Of course, I can't argue that
it's more obvious to users, but it may be.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK  win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] find exit code

[Philip Ganchev]

Couldn't the parsing be more sensitive to the possible errors, thus
avoiding a second pass?  For example, if '$' is followed by an illegal
character, check which character that is; if it's !, print an error
message like 

The '$' character begins a variable name. The character '!',
which directly followed a '$', is not allowed as a part of a variable
name.  To access words from the command history, press Alt+up.
For more information about command history, type help history.

If there's a space or no character after the $, print the message:

The '$' character begins a variable name. Variable names must be
longer than zero characters. To learn more about variable expansion
in fish, type help expand-variable.

And so on.

On Thu, Oct 9, 2008 at 7:55 PM, Beni Cherniavsky [EMAIL PROTECTED] wrote:
 On Thu, Oct 9, 2008 at 5:39 AM, Suraj N. Kurapati [EMAIL PROTECTED] wrote:
 As was already discussed at least once, if we want to automatically
 offer hints to users from other shells, we want a system orthogonal to
 the normal parser.  For starters, triggering hints by simple string
 search against the command should be enough (if needed we can extend
 it to
[...]
 * If the command parsing/execution fails, any matching hints should be
 shown of course.
 * Normally, hints should be displayed transiently while typing and then go 
 away.
 * An easy way to say don't show this hint again might be useful.
 And perhaps otherwise inspect/manage hint status.

Those are great ideas, and I have submitted to the list a detailed
design for the transient messages. But it requires lots of
programming.  I think it's better to avoid allowing the user to
control the hints, though.

 If properly done, this could become a built-in learn-as-you-go tutorial.

 BTW, do we want an interactive tutorial for fish?  Perhaps encouraging
 people to spend several minutes early on to learn the basics of fish
 is the most useful thing to do.

 I'm ready to write one, it there is agreement that it's a good idea.
 I think I'd start with targeting it at complete shell newbies.  Later,
 I might add a which shells are you familiar with question at the
 start which would add the relevant transition notes.

You are free to write one, but I don't think it's a good idea.  I
having a tutorial for something makes potential users feel overwhelmed
by the effort required to learn it.  If possible, we should make the
use of Fish in itself be an interactive tutorial, as you said.  This
is possible even without the complicated modifications you suggested
above.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK  win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] something identical to bash's !$

[Philip Ganchev]

Press Alt+UpArrow repeatedly until you get the word you need in the buffer.

On Sat, Aug 30, 2008 at 3:43 AM, Alex Greif [EMAIL PROTECTED] wrote:
 I am completely new to fish and missing the good old history
 abbreviations lile !$
 Is there a way to simulate this behavior?

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK  win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100url=/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] making fish scripts exit on unexpected errors?

[Philip Ganchev]

On Thu, Jun 26, 2008 at 1:45 AM, Dwayne C. Litzenberger [EMAIL PROTECTED] 
wrote:
 In POSIX shells (including bash), you can do something like this:

  #!/bin/sh
  set -e
  false # The shell will exit with a non-zero status here
  echo 'This will never get executed'

 Is anything similar to the set -e at the top of the script possible in a
 fish script?

Not as far as I know.  You can use the and built-in inside the script:

#!env fish
echo blah
and false
and echo 'This will never get executed'

Or maybe you can write a function that does that.  Something like:

function and_all
for x in $argv
if not $x ^/dev/null
return $status
end
end
end

Then the script could use it like:

#!env fish
and_all \
false \
echo 'This will never get executed'

It's a bit ugly and slow, but it's more explicit than in Posix.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] argh

[Philip Ganchev]

On Thu, Jun 26, 2008 at 4:37 PM, Beni Cherniavsky
[...]
 [EMAIL PROTECTED] ~ if true
   echo long long long long long
 long long long long long long long long long lon
 g long long long long long long
[...]

I think this would be better until we are able to lobby for terminals
to not copy indentation spaces, or to display new lines (or lack of)
visually.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] argh

[Philip Ganchev]

It would be nice if the terminal could display newlines visibly so
that ellipses would not be necessary.  But even without that, I prefer
lines to be broken without ellipses.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] environment variables and arrays

[Philip Ganchev]

On Tue, Jun 17, 2008 at 2:50 AM, David Bremner [EMAIL PROTECTED] wrote:
 Philip == Philip Ganchev [EMAIL PROTECTED] writes:
Philip Perhaps it should only convert
Philip variables it knows should be converted, like PATH,
[...]
 This really seems like a question of the operating environment or
 individual applications more than something the shell can necessarily
 know about.  I know it is unfishy, but it might require configuration.

Fish already remembers idiosyncracies of different programs to do
completion.  Variable syntax is similar to that. Requiring the shell
remember such info is to due a lack of design of early Unix systems, I
think.

 I'm just thinking out loud here, but two options that
 come to mind are
 1) a builtin to declare variable as an array
 split ':' LS_COLOURS
 2) automagically deal with : delimited strings as arrays.
  set foo 'a:b:c'
  foo[1]='d'
  echo $foo
would yield a:d:c

Option (2) is just to have all Fish arrays be colon-delimited, right?
That may be the simples solution, but seems a bit unsatisfying
conceptually.  We may as well use Posix for everything.


Philip I can't make it work in Bash or Tcsh. How is : inside a
Philip file name distinguished from a delimiter?

I meant putting a:b as one entry inside the PATH environment variable.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] environment variables and arrays

[Philip Ganchev]

On Mon, Jun 16, 2008 at 9:01 AM, David Bremner [EMAIL PROTECTED] wrote:
 The automatic conversion of environment variables containing ':' into
 arrays seems to be not so great to me. In particular it leads to mystifying
 experiences like the following.
[...]
From a practical point of view this means that any URLs inherited from
 the environment are mangled. [...]
[...]
The only alternative that comes to mind is to special case PATH.

There is no way around special-casing which variables must be
converted from colon-separated to array and which should not.
Currently Fish converts all variables except those it knows it should
not.   For example, WWW_HOME and DISPLAY are not converted.   Perhaps
it should only convert variables it knows should be converted, like
PATH, LD_LIBRARY_PATH, MANPATH, CLASSPATH, PYTHONPATH, LS_COLORS,
XDG_DATA_DIRS, GTK_RC_FILES, SESSION_MANAGER(?).

 Not to mention that ':' is AFAIK a valid
 character in a path (my home directory used to be on SFS, which uses ':').

I can't make it work in Bash or Tcsh. How is : inside a file name
distinguished from a delimiter?

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Bug with help command

[Philip Ganchev]

On Sun, Jun 8, 2008 at 8:55 AM, Stefan Majewsky [EMAIL PROTECTED] wrote:
 Fish is very cool. However, `help` opens the help pages in Firefox, but I
 normally use Konqueror. You should not call the firefox command to display the
 help, but rather `xdg-open` which selects the application associated with the
 given file's mime type (Konqueror, in my case; or Firefox for Gnome users, or
 whatever one chooses in the DE's config dialogs).

There have been many discussions on the list about how to display
help.  Maybe xdg-open would be a good solution.  For now there is the
BROWSER environment variable, and users can always redefine the help
script.

Regards,
Philip

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] defining key bindings

[Philip Ganchev]

I'm trying to define a key binding in my configuration. The only thing
mentioned in the documentation is to use the bind command.  But if
you do that in my user config.fish or config_interactive, it does
not take effect.

It only works if I define a new function that calls function
fish_default_key_bindings and sets any additional bindings:

function my_key_bindings
eval fish_default_key_bindings
bind \e\[3~ delete-char
end

Then the function has to be invoked by setting the universal variable
fish_key_bindings, say in your config_interactive:

set -U fish_key_bindings my_key_bindings

If you do not eval fish_default_key_bindings and do not define the
right bindings, the shell becomes unusable. You have to log in as
another user whose default shell is not Fish, to fix the problem, if
you know how.  Is there a simpler and safer way?  You should be able
to just call bind in config_interactive to modify the default
bindings.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] exec $VALUE doesn't work

[Philip Ganchev]

On Fri, Jun 6, 2008 at 12:50 AM, Martin Bähr
[EMAIL PROTECTED] wrote:
 On Thu, Jun 05, 2008 at 05:54:05PM -0400, Philip Ganchev wrote:
 On 6/2/08, Piotr Husiatyński [EMAIL PROTECTED] wrote:
  var=$(dmenu_path | dmenu); exec $var
 function temp_func; dmenu_path | dmenu; end
 temp_func
 functions -e temp_func

 that function will run dmenu_path | dmenu every time, but in the above
 case $var can be reused multiple dtimes while dmenu_path | dmenu is only
 run once.

True.  Since var is not local, the equivalent in fish is just
omitting the last line (functions -e temp_func).  I agree, though,
that it may be useful to be able to declare functions as local.

 in this example it may not matter, but in other cases it can make a
 difference if running the command has sideeffects or changes the result.

Not sure what you mean.
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] exec $VALUE doesn't work

[Philip Ganchev]

On Fri, Jun 6, 2008 at 7:51 PM, Martin Bähr
[EMAIL PROTECTED] wrote:
 On Fri, Jun 06, 2008 at 07:08:26PM -0400, Philip Ganchev wrote:
 On Fri, Jun 6, 2008 at 12:50 AM, Martin Bähr
 [EMAIL PROTECTED] wrote:
  On Thu, Jun 05, 2008 at 05:54:05PM -0400, Philip Ganchev wrote:
  On 6/2/08, Piotr Husiatyński [EMAIL PROTECTED] wrote:
   var=$(dmenu_path | dmenu); exec $var
  function temp_func; dmenu_path | dmenu; end
  temp_func
  in this example it may not matter, but in other cases it can make a
  difference if running the command has sideeffects or changes the result.
 Not sure what you mean.

 var=$(dmenu_path | dmenu); first executes dmenu_path, and if that fails
 it runs dmenu. the output is saved and stored in $var. $var may be
 reused many times.

 temp_func would execute dmenu_path every time it is called.

 (and now that i look at it, it doesn't even do what is expected here,
 namely take the string that dmenu_path dmenu print, and run that as a
 command.)

You are right, my bad.  The content of var is created when you run
dmenu_path | dmenu, but can be executed at another time.  Maybe
something like this would work:

dmenu_path | dmenu  dmenu_file
. dmenu_file

And what is wrong with using eval $var here??

set var (dmenu_path | dmenu); eval $var
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] exec $VALUE doesn't work

[Philip Ganchev]

On 6/2/08, Piotr Husiatyński [EMAIL PROTECTED] wrote:
[...]
 var=$(dmenu_path | dmenu); exec $var
[...]

That can be translated as:

function temp_func; dmenu_path | dmenu; end
temp_func
functions -e temp_func

More verbose, but perhaps clearer.  If I recall correctly, the reason
for disallowing execution of variables is that it makes syntax
checking impossible.
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] weird things happen in output redirects

[Philip Ganchev]

I guess what's confusing is the talk about token types, which is are
apparently something specific to Fish and not explained in the manual.
 Why doesn't the message just say

fish: Expected redirection specification, got \n (newline)
echo  (echo)
   ^
or

fish: Invalid file name   for output redirection
echo  (echo -n)
   ^

I think the error message repeats the problematic token so that the
user can see it. If it said 'Can't redirect to ', the user would
wonder which part of the command she typed produced the .  But the
message needs to say that the token expands to  or to newline, which
are invalid file names.  And it does not need to mention tokens.

What do you suggest the messages should be?

On 6/5/08, Isaac Dupree [EMAIL PROTECTED] wrote:
   set filename abcde
   set --erase not_a_filename

   echo out  $filename
   rm $filename
   #given that, this is weird:
   echo out  $not_a_filename
 fish: Expected redirection specification, got token of type
 $not_a_filename
 echo out  $not_a_filename
 ^

   echo out  (echo $filename)
   rm $filename
   echo out  (echo $not_a_filename)
 fish: Invalid IO redirection
 echo out  (echo $not_a_filename)
 ^
   echo out  (echo)
 fish: Invalid IO redirection
 echo out  (echo)
 ^
   echo out  (echo -n)
 fish: Expected redirection specification, got token of type (echo -n)
 echo out  (echo -n)
 ^

 Hmm. I'm a little confused (although luckily it seems to be only the
 messages that are confusing me -- it seems quite sensible so far in
 *whether* it accepts a redirection)

 -Isaac

 -
 Check out the new SourceForge.net Marketplace.
 It's the best place to buy or sell services for
 just about anything Open Source.
 http://sourceforge.net/services/buy/index.php
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users


-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] default browser?

[Philip Ganchev]

On 6/5/08, Frederik 'Freso' S. Olesen [EMAIL PROTECTED] wrote:
 Isaac Dupree skrev:
[...] Now I tried help function like
it told me to, since function --help doesn't work, and not only did it
shuffle me off to a graphical application -- but to Konqueror, even
though I'm running GNOME where my default browser is Firefox!  (of
course, since I do also have parts of KDE installed - though not running
- that might be nontrivial to fix? But it is generally well-defined
whether gnome or kde or neither is running at any point in time of the
current X session)

 fish, as most CLI applications, use the BROWSER environment variable to
 determine which browser to launch. Apparently, you have BROWSER set to
 konqueror.
 See http://fishshell.org/user_doc/index.html#variables-special for
 information on this and other environment variables affecting how fish
 operates.

I have always said that Fish should try to use a terminal-based HTML
browser by default.  The BROWSER variable specifies the default
browser for serious web browsing, Java script, flash, etc.  Instead,
Fish help needs to start up instantly and avoid window switching.
Fish help should just look for any browser known to exist on Unix
systems, starting with text-based ones.  If the user wants to change
that, she can redefine the function help in file
~/.config/fish/functions/help.fish.  Or it can be specified using a
separate variable TERMINAL_BROWSER, which would presumably be used
also for other terminal applications.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] weird things happen in output redirects

[Philip Ganchev]

On 6/5/08, Isaac Dupree [EMAIL PROTECTED] wrote:
 Philip Ganchev wrote:
 I guess what's confusing is the talk about token types, which is are
 apparently something specific to Fish and not explained in the manual.
  Why doesn't the message just say

 fish: Expected redirection specification, got \n (newline)
 echo  (echo)
^
 or

 fish: Invalid file name   for output redirection
 echo  (echo -n)
^

 I think the error message repeats the problematic token so that the
 user can see it. If it said 'Can't redirect to ', the user would
 wonder which part of the command she typed produced the .  But the
 message needs to say that the token expands to  or to newline, which
 are invalid file names.  And it does not need to mention tokens.

 What do you suggest the messages should be?

 in that case I'd rather mention  (expanded from (echo -n))
 or anyway mention both the value and the expression, for best
 understanding, somehow.

 BTW. newline is technically a valid filename (at least on my
 filesystem/OS, a normal linux) -- you just don't want it to be!

 Anyway, it shouldn't be tokenized, basically: it's a single entity, just
 like a function argument is.  And even if the variable expands to '1',
 it writes to the file named '1'.

 how about

 fish: Invalid file name  from (echo -n) for output redirection

 Actually, fish more typically repeats the whole expression and points at
 the troublesome part with a ^, instead of naming the subexpression
 inline in the error. (although I'd prefer it to also color the text
 range it's mentioning, which would be easier for the eye to pick out and
 notice, generally)


 On 6/5/08, Isaac Dupree [EMAIL PROTECTED] wrote:
   set filename abcde
   set --erase not_a_filename

   echo out  $filename
   rm $filename
   #given that, this is weird:
   echo out  $not_a_filename
 fish: Expected redirection specification, got token of type
 $not_a_filename
 echo out  $not_a_filename
 ^

   echo out  (echo $filename)
   rm $filename
   echo out  (echo $not_a_filename)
 fish: Invalid IO redirection
 echo out  (echo $not_a_filename)
 ^
   echo out  (echo)
 fish: Invalid IO redirection
 echo out  (echo)
 ^
   echo out  (echo -n)
 fish: Expected redirection specification, got token of type (echo -n)
 echo out  (echo -n)
 ^

 Hmm. I'm a little confused (although luckily it seems to be only the
 messages that are confusing me -- it seems quite sensible so far in
 *whether* it accepts a redirection)

 -Isaac

 -
 Check out the new SourceForge.net Marketplace.
 It's the best place to buy or sell services for
 just about anything Open Source.
 http://sourceforge.net/services/buy/index.php
 ___
 Fish-users mailing list
 Fish-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/fish-users





-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] unfortunate command-line effects

[Philip Ganchev]

On Fri, May 30, 2008 at 9:52 AM, Isaac Dupree [EMAIL PROTECTED] wrote:
 I hate it!  Frequently when I'm using fish and editing something e.g.
 from my history, it will disappear and I can't get it back!  Sometimes
 it will mysteriously turn purple and then I have no choice but to keep

What disappears?  What turns purple?  The history?  The Fish prompt?
The fish process?

 hitting random keys until it's all deleted and normal again, because
 using the arrow keys for history fails to work how it's normally
 supposed to anymore!
[...]
Have you tried hitting ^U and typing reset?

What version of fish?  What kind of terminal? What operating system?

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Environmental Variables

[Philip Ganchev]

On Mon, May 26, 2008 at 11:16 AM, Sorin Ionescu [EMAIL PROTECTED] wrote:
 Hello,

 Should not something like this be in the global config.fish?
[...]

I think those scripts are useful for converting your init scripts to
Fish, but why do every time you log in? Just do it when you install
Fish.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Bazaar completion

[Philip Ganchev]

I found a Zsh completion script - would that help to create a Fish one?


_baz
Description: Binary data
-
This SF.net email is sponsored by: Microsoft 
Defy all challenges. Microsoft(R) Visual Studio 2008. 
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Ubuntu 7.10 package of 1.23.0

[Philip Ganchev]

On Jan 23, 2008 5:54 PM, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 On Jan 23, 2008 9:17 PM, James Vega [EMAIL PROTECTED] wrote:
  Unfortunately, it looks like fish won't make it into Hardy since Ubuntu
  is already past the point where they automatically import packages from
  Debian.

 That's a shame. Oh well.

 Just six more months until the next Ubuntu release... :-)

With all this flurry of discussion, Fish is set to evolve so fast that
no matter when Ubuntu freezes their import, Fish will be obsolete in a
week.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Non host-local universal variables?

[Philip Ganchev]

I have thought for a while how clean it would be to store variables as
files, similar to how functions are stored as files.  Then you can
read and change them outside Fish, even when Fish is not running.
That's conceptually simpler, a form of init file.  It also seems to me
technically simpler to implement robustly than making sure the daemon
does not die, etc.

But for the rest of the proposal, is there enough reason to do it?
First, it may be more humane to have the user set variables
explicitly; as you said this happens rarely. It may be that setting
them automatically is not the right thing to do in many cases.  If
they are updated automatically, each Fish instance has to know where
to get or send its updates, and configuring that may be too much
overhead for most use cases.  Also, if the network connection breaks,
when it is re-established updates have to be merged and there may be
conflicts.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] [bugs] Tinkering with Fish 1.23.0 on Mac OS X

[Philip Ganchev]

On Jan 22, 2008 1:16 PM, Nick Pilon [EMAIL PROTECTED] wrote:
[...]
 1) Man pages for commands that are in subsections. This is easy enough
 to fix. For example:
 lastwords(1m)- print syscalls before exit. Uses DTrace
[...]
 2) Man pages with alternative names with just plain weird formatting.

Would it help to read command names from the path for binaries first?
There may still be man pages for commands that are not installed in
the system when the completion files are generated.  Maybe then when
the user tries to complete this command, Fish should show the
generated info and show mark the command as non-existent?

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Suggestion: extend case insensitivity of completions

[Philip Ganchev]

I think it would be useful to be able to type echo $fish-Tab and
see completions for fish_color_command etc. if there is no match for
$fish-.  I often forget whether a given command, file or variable
name has underscore or hyphen.  For example I type echo
$fish-color-Tab, then have to delete and try with underscores.

For ergonomic bonus, completions could insert a missing hyphen,
underscore or escaped space (\ ) if it's the only match; for example
file\ name for filenaTab.

On a related note - shouldn't wild cards like ? work for
completions?  They don't in Fish 1.22.1.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Newline escaping problem

[Philip Ganchev]

On Jan 18, 2008 5:01 AM, Stefano Sabatini
[EMAIL PROTECTED] wrote:
[...]
 [EMAIL PROTECTED] ~/./f/completions set -l list (echo foo bar foobar barfoo)

By the way, you can achieve the same using

set -l list foo bar foobar barfoo

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] sudo !! doesn't work.

[Philip Ganchev]

On Jan 13, 2008 5:53 AM, Martin Bähr [EMAIL PROTECTED] wrote:
 On Sun, Jan 13, 2008 at 02:18:19AM -0500, Philip Ganchev wrote:
  I was not suggesting that the command should execute with root
  permissions all the time.  It should execute with the least
  permissions it can to do the job, but ask for a password only if it
  needs more permissions.  For example if the user executes rm myfile
  but myfile is owned by root, only then ask for a password and
  execute as root.

 but this is not something that the shell has any chance of figuring out.
 since the suggestion was that fish should know when to call sudo. fish
 can't help here.

I know. It's an unfortunate consequence of the poor design of the paradigm.

 and in your example, this is exactly a case where it should not ask for
 the password. the program can not know why you are trying to remove a
 root-owned file without root permissions. maybe you misstyped the
 filename and you don't actually want to remove the file.

I don't understand.  If you execute the command, you want to remove
the file.  This is no different with sudo.

 also sudo has a mode where it remembers that you typed the password for
 say 15 minutes, and won't ask again. if sudo were called automaticly
 (as has been suggested) that would mean that for those whole 15 minutes
 you are essentially running with root permissions, because they would be
 invoked automaticly every time they are needed.

 this should make clear that calling sudo automaticly is just a very bad
 idea. and every time asking for a password is just as bad because we'll
 get many password prompts where we'd have to abort. which means the user
 is forced to decide which action to take (type password or abort).
 eventually it will happen that the wrong decision is made. as it
 is now, the user has to decide to either do nothing, or rerun the
 command with appropriate permissions. as doing nothing is always an easy
 and safe choice, this is much preferable.

 changing access needs to be a concious decision which you get by
 prefixing sudo to the command manually. it should not be an automatic
 option.

I think you are saying that the user becomes habituated to entering a
password, and whenever he is asked, he gives it without thinking.  But
if he has to run a separate command to ask for privleges, that makes
him think before doing something with them.  I agree, but this
inconvenience is not the only way to alert the user that he is getting
privileges.  For example the user can be prompted like this: *** You
are about to get privileges that will allow you to greatly damage your
system!!! ***.  That is different than other prompts when he is
entering a password, so I think he would be alert.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] sudo !! doesn't work.

[Philip Ganchev]

On Jan 11, 2008 7:38 PM, Martin Bähr [EMAIL PROTECTED] wrote:
 On Fri, Jan 11, 2008 at 04:56:57PM -0500, Philip Ganchev wrote:
  Ideally, you should not have to type sudo.  If a command requires
  privileges, the user should be able to type the command and be asked
  for the password.  If she does not give the right password, the
  command is not executed.

 not a good idea.
 most programs do something useful both with and without root privileges.

 eg: rm.
 in most cases you do not need root access to remove a file,
 in most other cases where removing fails you do not want to switch to
 root to remove the file, but you just want to leave the file alone.
 only in a very few cases you really want to remove the file with root
 permissions.

I was not suggesting that the command should execute with root
permissions all the time.  It should execute with the least
permissions it can to do the job, but ask for a password only if it
needs more permissions.  For example if the user executes rm myfile
but myfile is owned by root, only then ask for a password and
execute as root.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] sudo !! doesn't work.

[Philip Ganchev]

On Jan 11, 2008 6:34 PM, Isaac Dupree [EMAIL PROTECTED] wrote:

  But just doing nothing when I type sudo !! is not very friendly.  At the
  least, it should pop up a message telling me the fish way to do it.
 
  This is about as complicated as fully implementing !!.

 code complexity, user simplicity: it's a tradeoff, right?

Not exactly.  This is one example of a user habit that is actually
inefficient.  It's fine for the special cases of !!, !$ and !^,
but in general editing using Bash's substitution mechanism is a
needless mental burden and a risk.

 [...] However, if
 '!!' was a command name then sudo !! would make sense in fish without
 special character interpretation, so just giving a message for !! would
 complain about some correct cases ( ` the backtick is also treated as an
 ordinary character in this silent way, I've noticed, which would have no
 troubles whatsoever if there weren't these habits from other shells).

How about:

function !!
 eval (echo $history[2])
end

By the way, I get this behavior with Fish 1.22.1, is it expected?

 function h1; tail -n 1 ~/.config/fish/fish_history; end
 echo -n
 h1
function h1; tail -n 1 ~/.config/fish/fish_history; end

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] sudo !! doesn't work.

[Philip Ganchev]

On Jan 11, 2008 4:13 PM, Endolith [EMAIL PROTECTED] wrote:
 Beni Cherniavsky [EMAIL PROTECTED] writes:
  Alt+Up appends the last word of the last command, so it won't do
  what you want in general.
  But what's wrong with UpHomesudo ?

 That's fine I guess.  Both this and !! are kind of awkward to type.

What would be less awkward than those two?  You can bind a key combo
to prepend sudo to your line, I guess.

 But just doing nothing when I type sudo !! is not very friendly.  At the
 least, it should pop up a message telling me the fish way to do it.

This is about as complicated as fully implementing !!.

 At best, it should know which programs require sudo and tack it onto the
 beginning without me needing to specify it with a message explaining the
 implications.

Ideally, you should not have to type sudo.  If a command requires
privileges, the user should be able to type the command and be asked
for the password.  If she does not give the right password, the
command is not executed.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] configuration lost and request for noclobber in set

[Philip Ganchev]

On 10/17/07, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[...]
 On 10/17/07, Fernando Canizo [EMAIL PROTECTED] wrote:
[...]
  The only thing left is that sometimes one will really want to truncate a
  file. In bash this is achieved with this syntax (when noclobber is set):
 
  echo nothing | file_to_be_truncated
 
  Now the syntax could be anything, it's not important, but it's important
  to have a way to overcome the default.

 That works very well, but there is one problem. Fish allows you to
 redirect any file descriptor to a pipe using something like:

 make 2| less

 The above will use the less pager to scroll through the output of
 make. But that clashes with the suggested syntax for clobbering files.
 We could chose to use something other than bash does, but when it
 comes to picking arbitrary squiggles for operators, I see a pretty
 strong benefit in making the same choices as the shell that must not
 be named does in order to make migration between shells easier.

The trouble is the 'noclobber' option in the shell that shall not be
named. It makes the '' operator modal, so when people write '' they
expect something different if they are used to having the option set
than if they are not.  Since we don't want to replicate this bad
effect in Fish, whatever '' means in Fish, it will be incompatible
with other shells for some users.

So another solution is to habituate to using tab completion to avoid
overwriting the file.  If the file name completes, then the file
exists and you are about to overwrite it.

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now  http://get.splunk.com/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Unknown command “_”, after installing Fish from darcs

[Philip Ganchev]

Hi,

I installed Fish on a new machine from the darcs repositories, and
built it without errors.  When I start it, I get error messages like:

fishfish: Unknown command _
/usr/local/share/fish/config.fish (line 1): _ '%s: Warning: The
directory %s has been removed from your PATH because it does not
exist\n'
^
in command substitution
called on line 56 of file /usr/local/share/fish/config.fish,

in . (source) call of file /usr/local/share/fish/config.fish,
called on standard input,
...

fish: Unknown command __fish_print_help
...

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Bug: eval function does not work for browsers that run in the terminal, as used in help function

[Philip Ganchev]

Function help invokes function eval to start a web browser:

eval $fish_browser file://$__fish_help_dir/$fish_help_page

Unfortunately, if the browser is a terminal browser, such as links or
w3m, this does not work.  For example, w3m outputs an error Error
occured while reset 800b: errno=5, while links opens the page but
does not respond to input.Less does not work either.  I guess the
. builtin does not work as expected?

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] how to do cls=clear

[Philip Ganchev]

On 5/21/07, shirish [EMAIL PROTECTED] wrote:
[...]
how can I do cls=clear (an alias) permanently. Thank you.

I would rather recommend that you write a cls function.  The alias
function above creates one, but it is only useful for short functions,
and anyway it is a workaround.   I think it will be more useful for
you to learn how to write functions in general.

function cls --description Alias for the 'clear' command
clear
end

Put this in a file called ~/.config/fish/functions/cls.fish

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Bug: $ interpreted as a variable reference

[Philip Ganchev]

The Fish parser treats any $ character that is unquoted or quoted
inside double quotes as a variable reference. Unfortunately, the user
may try to use $ in a regexp inside double quotes.  For example:

 grep -E -v ^$ myfile
fish: Did you mean (COMMAND)? In fish, the '$' character is only used
for accessing variables. To learn more about command substitution in
fish, type help expand-command-substitution.
grep -E -v ^$ myfile
   ^

This can be achieved by using single quotes instead:
 grep -E -v '^$' myfile

But the user may not think of that.  Unfortunately, I don't see a way
to make the parser smarter in general.  So either the error checking
should be less strict, or the error message should be more general:

fish: Incorrect variable reference.  The $ character is used for
variable reference when it is unquoted or inside double quotes ().
You can use single quotes (') instead.  To learn about variables, type
help variables; about variable substitution type help
expand-command-substitution.

or

fish: Incorrect use of $.  When the $ character is unquoted or
inside double quotes (), it is interpreted as a variable reference.
To avoid this, put it inside single quotes (').  To learn about
variables, type help variables; about variable substitution, type
help expand-command-substitution.

I'm not sure why the command substitution part is useful.

Incidentally, a more radical idea used in YubNub solves this and
another problem: use a builtin called var for all variable
operations, including setting, getting and removing:
 var A --set-global hello bye
 var A
hello bye
 echo say: (var A) now!
say: hello bye now!
 var A --erase

This is more verbose but perhaps more intuitive, as it uses the same
syntax for setting and reading.  Also, the builtin may be easier to
remember and discover because its name is mnemonic.  It also solves
the above problem, because $ would never be special.

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] [fish-users] How to show apropos output in fish_pager?

[Philip Ganchev]

Hi,

I have been thinking how to make it easier to find the right commands
to use.  In programs like Deskbar, MacOS Spotlight and Quicksilver,
the user types some words and sees a list of commands whose
descriptions match all the words as strings.

On the Unix shell, the command apropos (or man -k) is used in the
reverse order.  First you think: I want a command that does X, then
realize that you should ask the apropos database, and type apropos
and the search terms.  It may be easier if you can type the search
terms and then invoke the search, because you think of them before you
think of apropos.  This is a small step towards an object-verb order
of commands; the traditional order of Unix commands is verb-object.

I bound Alt+s to the a function that does this.

cat  ~/.config/fish/fish_inputrc
\M-s: apropos_words (commandline -po)
fish: Job 1, cat  ~/.config/fish/fish_inputrc  terminated by
signal SIGINT (Quit request from job control (^C))

function apropos_words -d Find commands whose descriptions match all
the arguments
if test $argv
set s
for w in $argv
set w (echo $w | tr -d '[:space:]')
if test $w
set s $w|$s
end
end
apropos -r $s
end
end

But it would be nice to paginate the output it is too big.  Any ideas
how this can be done (elegantly)? I can use the more command but
this requires the user to press 'q' before continuing to type the
command.  I tried to use fish_pager as
apropos -r $s | sed -r 's/ - /\\4/' | fish_pager 0 ''
but sed treats \\4 as a reference, and anyway it is computationally
intensive and unnecessary to transform the output just so that it can
be shown the fish_pager.

Ideas about the general problem of how to use object-verb order are
also welcome.

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] [fish-users] Documentation bug: commandline -b

[Philip Ganchev]

Hi

The documentation and completion for commandline -b say -b or
--current-buffer (select the entire buffer (default)), but executing
commandline -b gives an error.

 commandline -b
commandline: Unknown option -b

   commandline - set or get the current commandline buffer


   Synopsis
   commandline [OPTIONS] [CMD]

commandline: Type 'help commandline' for related documentation

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] fish and iTerm

[Philip Ganchev]

On 4/4/07, Claes Nästén [EMAIL PROTECTED] wrote:
[...]
 Please try to run with the standard Terminal application for a while and see
 if that helps?

Also, run Fish inside Bash or Sh.  Are any messages printed in iTerm
when Fish dies?

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Manual pages for

[Philip Ganchev]

Hi,

With the default darcs Fish installation, I could not get manual pages
for builtins like set.  After doing the following, it now works.

cd /usr/local/share/fish/man
sudo mkdir man1
cd man1
sudo ln -s *.1 *.1.gz man1/
set -gx MANPATH (manpath | tr : '\n') /usr/local/share/fish/man
echo set -gx MANPATH (manpath | tr : '\n') /usr/local/share/fish/man
 ~/.cofig/fish/config.fish

As I understand the manual of the man command, man expects manual
pages to be stored in a directory hierarchy like man/man1,
man/man2, etc. in a directory that is in listed in the MANPATH
variable.

I do not know if the original problem, the hierarchy standard or the
manpath command are Debian-specific.  And should MANPATH be set in
config.fish or config_interactive.fish?

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Bug in history token search for parenthesis

[Philip Ganchev]

Hi, sorry to report another bug without a patch :-/

ls ..
ls (Alt+Up

results in

..ls (

Then pressing Alt+down results in

 (


This only happens with parens; in fact for me, only with ( but that
may be due to my specific search history.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Bug while completing which parameter

[Philip Ganchev]

Hi, I'm not sure if this bug has been reported already.  I typed
which ls and hit the Tab key.

~ which ls*** glibc detected *** fish: realloc(): invalid pointer:
0x0808d0d0 ***
/lib/tls/i686/cmov/libc.so.6(__libc_realloc+0x31c)[0xb7ec7efc]
fish[0x8079ad7]
fish(al_push+0x11)[0x8079cb1]
fish(halloc+0xac)[0x8079f1c]
fish(completion_allocate+0x25)[0x8056b95]
fish[0x806cca8]
fish[0x806cd98]
fish[0x806cd98]
fish(wildcard_complete+0x36)[0x806ce46]
fish[0x806d22a]
fish[0x806de5f]
fish[0x806da88]
fish[0x806da88]
fish(wildcard_expand+0x2c)[0x806df0c]
fish(expand_string+0x5bc)[0x805e83c]
fish(complete+0xeb2)[0x8058122]
fish[0x80535d9]
fish(builtin_run+0x72)[0x8050d22]
fish(exec+0x8f3)[0x805c023]
fish(eval+0x8ff)[0x806625f]
fish(exec_subshell+0x58)[0x805a9d8]
fish[0x805c6bc]
fish[0x805c6fb]
fish(expand_string+0x65)[0x805e2e5]
fish(eval_args+0x158)[0x8063fb8]
fish[0x8056aa9]
fish[0x8056d46]
fish(complete+0xaa8)[0x8057d18]
fish(reader_readline+0xac0)[0x806a440]
fish(reader_read+0x290)[0x806b690]
fish(main+0x638)[0x807e278]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xdc)[0xb7e748cc]
fish[0x804ea81]
=== Memory map: 
08048000-08096000 r-xp  03:03 620479 /usr/local/bin/fish
08096000-08098000 rw-p 0004d000 03:03 620479 /usr/local/bin/fish
08098000-080ba000 rw-p 08098000 00:00 0  [heap]
b7cc6000-b7cd r-xp  03:03 2023798/lib/libgcc_s.so.1
b7cd-b7cd1000 rw-p 9000 03:03 2023798/lib/libgcc_s.so.1
b7cd1000-b7d0b000 r--p  03:03 1110233
/usr/local/share/locale/en/LC_MESSAGES/fish.mo
b7d0b000-b7d14000 r-xp  03:03 2024355
/lib/tls/i686/cmov/libnss_files-2.4.so
b7d14000-b7d16000 rw-p 8000 03:03 2024355
/lib/tls/i686/cmov/libnss_files-2.4.so
b7d16000-b7d1e000 r-xp  03:03 2024357
/lib/tls/i686/cmov/libnss_nis-2.4.so
b7d1e000-b7d2 rw-p 7000 03:03 2024357
/lib/tls/i686/cmov/libnss_nis-2.4.so
b7d2-b7d32000 r-xp  03:03 2024352
/lib/tls/i686/cmov/libnsl-2.4.so
b7d32000-b7d34000 rw-p 00011000 03:03 2024352
/lib/tls/i686/cmov/libnsl-2.4.so
b7d34000-b7d36000 rw-p b7d34000 00:00 0
b7d36000-b7d3d000 r-xp  03:03 2024353
/lib/tls/i686/cmov/libnss_compat-2.4.so
b7d3d000-b7d3f000 rw-p 6000 03:03 2024353
/lib/tls/i686/cmov/libnss_compat-2.4.so
b7d4f000-b7d82000 r--p  03:03 835224
/usr/lib/locale/en_US.utf8/LC_CTYPE
b7d82000-b7d83000 r--p  03:03 835225
/usr/lib/locale/en_US.utf8/LC_NUMERIC
b7d83000-b7e5a000 r--p  03:03 835227
/usr/lib/locale/en_US.utf8/LC_COLLATE
b7e5a000-b7e5b000 rw-p b7e5a000 00:00 0
b7e5b000-b7e5d000 r-xp  03:03 2024349/lib/tls/i686/cmov/libdl-2.4.so
b7e5d000-b7e5f000 rw-p 1000 03:03 2024349/lib/tls/i686/cmov/libdl-2.4.so
b7e5f000-b7f8c000 r-xp  03:03 2024346/lib/tls/i686/cmov/libc-2.4.so
b7f8c000-b7f8e000 r--p 0012c000 03:03 2024346/lib/tls/i686/cmov/libc-2.4.so
b7f8e000-b7f9 rw-p 0012e000 03:03 2024346/lib/tls/i686/cmov/libc-2.4.so
b7f9-b7f93000 rw-p b7f9 00:00 0
b7f93000-b7fca000 r-xp  03:03 359553 /lib/libncurses.so.5.5
b7fca000-b7fd2000 rw-p 00037000 03:03 359553 /lib/libncurses.so.5.5
b7fd2000-b7fd4000 rw-p b7fd2000 00:00 0
b7fd4000-b7fd5000 r--p  03:03 835316
/usr/lib/locale/en_US.utf8/LC_TIME
b7fd5000-b7fd6000 r--p  03:03 835317
/usr/lib/locale/en_US.utf8/LC_MONETARY
b7fd6000-b7fd7000 r--p  03:03 835230
/usr/lib/locale/en_US.utf8/LC_MESSAGES/SYS_LC_MESSAGES
b7fd7000-b7fd8000 r--p  03:03 835251
/usr/lib/locale/en_US.utf8/LC_PAPER
b7fd8000-b7fd9000 r--p  03:03 835319
/usr/lib/locale/en_US.utf8/LC_NAME
b7fd9000-b7fda000 r--p  03:03 835320
/usr/lib/locale/en_US.utf8/LC_ADDRESS
b7fda000-b7fdb000 r--p  03:03 835321
/usr/lib/locale/en_US.utf8/LC_TELEPHONE
b7fdb000-b7fdc000 r--p  03:03 835322
/usr/lib/locale/en_US.utf8/LC_MEASUREMENT
b7fdc000-b7fe3000 r--s  03:03 771985
/usr/lib/gconv/gconv-modules.cache
b7fe3000-b7fe4000 r--p  03:03 835323
/usr/lib/locale/en_US.utf8/LC_IDENTIFICATION
b7fe4000-bfish: Job 1, fish terminated by signal SIGABRT (Abort)

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] bugs and a few suggestions (1.22.3)

[Philip Ganchev]

On 3/12/07, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[...]
 On 3/3/07, Guillaume [EMAIL PROTECTED] wrote:
[...]
  Alt-P on an empty line may instead print pager [cursor] to be more useful.

 Good point. I'll add that to the todo list. Patches are welcome!

Is this better than cursor | pager?  It looks simpler, but that
difference is one more thing for the user to understand unconsciously.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] The trouble with casing

[Philip Ganchev]

On 3/5/07, Myrddin Emrys [EMAIL PROTECTED] wrote:
 I dunno. The fact that it displayed the text, and re-displayed the command
 line, is a pretty blunt stick indicating that there was no match. In
 particular, you'll see the list of matching files for the new match
 listed... if that doesn't send up warning signals (assuming they're not what
 you wanted to delete), it's hard to imagine what would.

That's my feeling too.  If completion behavior is consistent, then it
is predictable and so users will easily habituate to reacting to a
non-completion correctly.


 On 3/5/07, Martin Baehr [EMAIL PROTECTED]
 wrote:
  On Mon, Mar 05, 2007 at 11:59:08AM -0600, Myrddin Emrys wrote:
   linux rm ab*tab
   Ab* Absolute.txt, Abstract.txt, Abrasive.txt...
   linux rm Ab*
 
  this is exactly what i am afraid of, because if i do not notice that the
  case chsnged and my intention is to remove only ab* but NOT A* then this
  is dangerous. same with any situation where i'd overwrite a file.

Maybe it would be useful to somehow indicate that the command has been
changed by a completion.  That way you do not have to watch the whole
string, looking for the a one-character change.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Things that keep my from switching to fish.

[Philip Ganchev]

On 2/22/07, Brendan Rankin [EMAIL PROTECTED] wrote:
 Overall, I like the principles that fish employs, however, I can't seem to
 figure out how to easily access past commands and/or command's arguments using
 !$, !!, etc.

You can browse previous commands by pressing UpArrow and DownArrow,
and separate words from previous commands with Alt+UpArrow and
Alt+DownArrow. Once you press Space, the word browsing is reset to the
beginning.

You can read that and more about fish by executing help editor or
visiting file:///usr/local/share/doc/fish/index.html#editor or
http://www.fishshell.org/user_doc/index.html#editor

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] [fish-users] Search highlight extends to end of line when returning to query line, in Xterm

[Philip Ganchev]

I suspect this is not a bug with Fish, but still it has been bugging
me.  In Xterm, when you do a history search then return to the
original query command, the query remains highlighted, as well as the
whole line.  The rest of the line (at least) should not be
highlighted.  This does not happen in gnome-terminal.

And perhaps the query should not be highlighed either, to indicate
that you are back where you started before the search.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] less wishlist items

[Philip Ganchev]

On 11/3/06, Beni Cherniavsky [EMAIL PROTECTED] wrote:
[...]
 I see a simple solution: when the cursor was in the first line and you
 press Up, you should land you at the *first* line of the previous
 command, not the last.  Similarly, Down should land you in the last
 line of the next command. This way, consecutive Up or Down presses
 will  navigate the history one command at a time, while retaining the
 ability to one line at a time within a command. [...]

So if the cursor is on the first line, pressing Up would search for
the whole command up to the cursor (as currently), but if the cursor
is on any other line, it would search for single lines containing the
string from the start-of-line till the cursor.  And likewise with
Down on the last line.  Right?

 [...] It gives you
 kind-of-modal functionality without really adding modes - the only
 state is the position of the cursor, which should be obvious enough.

I think this is almost as modal as my suggestion for moving among
lines of the same commad -- in both cases the behavoir depends on the
cursor line.  And for both a separate keyboard shortcut is probably
still a better solution, despite the great number of shortcuts.
Something like:

Up = search backwards for command containing the substring from
the start of the current command till the cursor
Shift+Up = search backwards for line containing the substring from
the start of current line till the cursor
Shift+PageUp = scroll the output up one screen (in common
terminals and terminal emulators)
Control+Up = move cursor to the previous line (similar to
Control+Left = move to prevoius word)

or

Up = move cursor to the previous line (as in an editor)
PageUp = search backwards for command containing the substring
from the start of the current command till the cursor
Control+PageUp = search backwards for line containing the
substring from the start of current line till the cursor
Shift+PageUp = scroll the output (in common terminals and terminal
emulators)


 The only price you pay is that switching directions while browsing the
 history still requires going through all the lines of a command (but
 Alt- / Alt- should solve it).

Switching direction is a pretty common operation, right?  And having
to move the cursor to the beginning is a bit complicated, I think.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

[Fish-users] Bug: prompt goes to default after pager exit

[Philip Ganchev]

Hi

Fish seems to be reverting to the default prompt after the pager
exits.  This only happens when the comand line is empty or contains
only spaces.


Welcome to fish, the friendly interactive shell
Type help for instructions on how to use fish
~ functions fish_prompt
function fish_prompt --description Write\ the\ prompt
set_color $fish_color_cwd
prompt_pwd
set_color normal
set_color green
echo 
set_color normal
echo  
end

~ Tab
[pager exits]
~ Enter
[EMAIL PROTECTED] ~ functions fish_prompt
function fish_prompt --description Write\ out\ the\ prompt

# Just calculate these once, to save a few cycles when displaying the pr
ompt
if not set -q __fish_prompt_hostname
set -g __fish_prompt_hostname (hostname|cut -d . -f 1)
end

if not set -q __fish_prompt_normal
set -g __fish_prompt_normal (set_color normal)
end

set -l prompt_color (set_color $fish_color_cwd)

printf '[EMAIL PROTECTED] %s%s%s \n' $USER $__fish_prompt_hostname 
$prompt_color
(prompt_pwd) $__fish_prompt_normal
end

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Bug: prompt goes to default after pager exit

[Philip Ganchev]

On 10/23/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 On 10/23/06, Philip Ganchev [EMAIL PROTECTED] wrote:
[...]
  Fish seems to be reverting to the default prompt after the pager
  exits.  This only happens when the comand line is empty or contains
  only spaces.
[...]
[...]
 Is your fish_prompt function defined in ~/.config/fish/functions?
 What is the value of $fish_function_path on startup, and what is its
 value after the prompt reverts?
 Does this always happen, or only occasionally?
 Is it always the pager that causes this or have you found some other
 way to trigger it?

 Thanks for the report.

Ah!  My function was defined in a file called
~/.config/fish/functions/interactive.fish, sourced at startup.  This
is how fish used to work a long time ago, and I have somehow missed
the change.  I never re-read the documentation, because everything
worked as expected.  Should teach me not to stick around one project
for so long :-)  Putting the function in
~/.config/fish/functions/foo.fish fixes this.

That was the same cause for the other problem I reported earlier, and
I was equally surprised.

From the documentation I understant that there are now some functions
(called autoloaded functions) that are not kept in memory, but instead
treated similar to external commands.  When you issue a command, fish
looks for a file by that name in the directories listed in variable
PATH; if it does not find it, it looks for a file with the right
name in the directories listed in variable fish_function_path; if it
finds that, it sources it and invokes the function.  Something like
that?  That could be clearer in the documentation as well as the error
messages.  I will write something as soon as I've got it.

But the function was defiend before the pager was invoked.  Why isn't
it defined after it exits?  And this only happens when there is no
command on the comman line, not for example when the pager completes
lsTab.

That also relates to my question from another post about differences
between scripts and functions.  In a sense, you were thinking ahead
and came close to unifying them.  But I think there is more to be said
about the relative advantages.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Changed expansion rule for '**'

[Philip Ganchev]

On 10/21/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 On 10/19/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 [...]
  ?: Matches any character except '/'
  *: Matches any substring not including the character '/'
  **: Matches any substring

 Seems like most people like this change. Patch is in the darcs tree.

I see that your implementation works just as I was suggesting.  That
is, fish* matches a directory called fish.d, not just a file.  So
I was needlessly worried about what you meant.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Please test the darcs tree if you can

[Philip Ganchev]

Since a few versions ago, I have been getting an error message when I
start Fish:

fish: Could not autoload item functions, it is already being
autoloaded. This is a circular dependency in the autoloading scripts,
please remove it.

Apparently it's because my file
~/.config/fish/functions/functions.fish invokes ``functions foo -d
Bar''.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Directory transition patch in darcs

[Philip Ganchev]

On 10/19/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[...]
 Fish tries to move existing config files to the new location. This
 move will only be performed once, on the first time an interactive
 shell is launced after the upgrade. The updarde works for me, but it's
 not yet widely tested. It may crash, kill your config or in rare cases
 eat babies. Please try it out and let me know if there are any issues.
 But back up your fish config files before doing so.

Hmm... The installation does not quite work for me:

fish/darcs sudo make install
Password:
...
/usr/bin/install -c -m 755 -d /usr/local/etc/fish
/usr/bin/install: `/usr/local/etc/fish' exists but is not a directory
make: *** [install-force] Error 1

bash$ killall fish
bash$ fish
fish: Warning: The directory ./bin has been removed from your PATH
because it does not exist
fish: Warning: The directory /tools has been removed from your PATH
because it does not exist
Welcome to fish, the friendly interactive shell
Type help for instructions on how to use fish
fish
fish ls -d .fish*
.fish*   .fish.d/.fish_history   .fish_inputrc*
.fish~*  .fishd.asparuh   .fish_inputrc~*
fish
fishecho $XDG_CONFIG_HOME
fish
fish ls .config/fish
fishd.asparuh


I found this patch intriguing: Make any character except '/' legal in
function names.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Changed expansion rule for '**'

[Philip Ganchev]

On 10/19/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[...]
 ?: Matches any character except '/'
 *: Matches any substring not including the character '/'
 **: Matches any substring

 To me, this seems much more intuitive.
[...]

So fish** would match fish, fish.d and fish.d/ggg.

Is there a way to, say, copy all files and directories whose names
that start with fish?  Currently you would write ``cp -r fish*''.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] bug: ctrl-w does not use whitespace as a word boundary

[Philip Ganchev]

On 10/12/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[rearranged]
 I have once again edited the relevant function, when moving/deleting
 to the left, the boundary character should never be deleted/moved
 past. But I am beginning to see a very clear trend that whenever you
 unbreak one aspect, another one breaks. The current implementation is
 doubtlessly not an exception. To save me some time, I have decided
 that I will not consider future suggestions for this function unless
 they contain a decent analysis of common usecases as well as a patch
 to the move_word function to implement the suggested behaviour. Sorry.

Thank you for the consideration, and sorry for the trouble.  I did not
intend to make so much work for you, and I would have been able to
create a patch for this.  I think it is a reasonable condition.  This
kind of patch is simple enough.

[...]
 On 10/6/06, Philip Ganchev [EMAIL PROTECTED] wrote:
[...]
  On Aug 27, 2006, Axel Liljencranz wrote:
   A simpler rule, that will most of the time do what you propose would be
   to only lump together a series of _identical_ boundary token. E.g. '
' is a single boundary, but ' ~. ./' is 6 boundaries.
 
  I find this a much more intuitive and useful behavior for delete-word.
 
  The other one is delete consecutive characters that belong to the
  same set, where each character is in one of three sets: alphanumeric,
  whitespace, or boundary.  This is how Vi works.
 
  Emacs and Bash work somewhat opposite of Fish.  All chars up to an
  alphanumeric char, then up to a non-alphanumeric char, are deleted.
 

 They do, but they don't allow you to delete one directory level in a
 path, for example, which I often find quite useful. And because I got
 into the hornets nest of adding this third level, we have a much
 harder time of defining a sane criteria. I am still hoping that we
 will get a rule that is more useful than the one in bash, but it is
 far harder than I initially thought.

You do not want to focus on this any more, and that is fair enough.  I
am just curious to understand what problem you found with deleting a
directory level.  That's exactly what I find easy in Bash's
delete-word: ^w deletes to the next directory level; unless the file
name contains boundary characters, in which case you will have to
press Backspace and/or ^w again.  None of the methods we discussed
avoid that.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] bug: ctrl-w does not use whitespace as a word boundary

[Philip Ganchev]

On 10/19/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 My bash deletes to whitespace on ^W, so all directory levels are
 munched in a single go. Fish does stop at '/', which I would have
 thought was what you wanted. I'm sure I'm missing something here, but
 I have no clue what it is.

Hm, strange.  The Bash 3.1 manual says:
   backward-kill-word (M-Rubout)
  Kill the word behind point.  Word boundaries  are  the  same  as
  those used by backward-word.
   backward-word (M-b)
  Move  back  to the start of the current or previous word.  Words
  are composed of alphanumeric characters (letters and digits).

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Changed expansion rule for '**'

[Philip Ganchev]

On 10/19/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 On 10/20/06, Philip Ganchev [EMAIL PROTECTED] wrote:
[...]
  Is there a way to, say, copy all files and directories whose names
  that start with fish?  Currently you would write ``cp -r fish*''.

 Unfortunatly no, not that I know of. 'cp' omits directories and
 flattens the file list. It would of course be possible to write a
 simple script to do such things, though.

Just to clarify:initially, you were talking about changing what **
matches, not *.  Now with the new rules, 'fish*' would not match a
directory called fish.d??  Currently it does.

If not, this seems as inconevenient as having to write fish{,**/}*
to match all files starting with fish and files in directories whose
name starts with fish, no?  Wouldn't it be quite surprising to
someone who is used to using ``cp fish* dest''?  They would have to
learn to use --parents.  You can hide that, using a script, but you
would have to do that for other commands that use directories too, or
there will be even more surprise.  In the end, it's probably best if
fish* matches directory names as well as file names.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] Directory transition patch in darcs

[Philip Ganchev]

On 10/19/06, James Vega [EMAIL PROTECTED] wrote:
 On Fri, Oct 20, 2006 at 01:29:04AM +0200, Axel Liljencrantz wrote:
[...]
  Previously, function names could only consist of alphanumerics and
  underscores, just like variables. This isn't very logical, though,
  since regular commands can have any legal file name. This patch
  changes the naming criteria. Functions may now have any legal file
  name, and '/' is the only character which is forbidden in file names.

Hmm, it's good to remove differences in the ways functions versus
other commands are invoked.  That way the user does not have to worry
whether what he is invoking is a function.  But actually this is not
related to allowed characters in function names, because at invocation
time the command (function or other) is already defined.  I guess it
does allow for any  command to be wrapped inside a function by the
same name.

That brings up a question I had wanted to raise.  Functions are simiar
to scripts in files.  For example, instead of defining a cp
function, you can write an executable script in ~/bin/cp and put ~/bin
before /bin on the path.  I guess they each have relative advantages:
- functions can be defined interactively on the command line
- you know where a script is defined, in case you want to modify it
Have I missed any, and would it be better to not have functions at all?

 That depends on the filesystem.  This is the case for most (if not all)
 *nix filesystems, but I think OS X's filesystem has more restrictions.
 If I recall correctly, POSIX only specifies which characters *must* be
 allowed in filenames, and that's a fairly small list.  Anything beyond
 that is just icing on the cake.

If so, some functions will not work on some systems :-o  And there is
a discrepancy between variable names and function names.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users

Re: [Fish-users] bug: ctrl-w does not use whitespace as a word boundary

[Philip Ganchev]

On 10/16/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
 On 10/17/06, Philip Ganchev [EMAIL PROTECTED] wrote:
  On 10/16/06, Axel Liljencrantz [EMAIL PROTECTED] wrote:
[...]
  j[4AܚAA�'
  mbrtowc: Invalid or incomplete multibyte or wide character
 

 This log is very helpful. See all the 'mbrtowc' lines? That is a
 function to convert from multibyte narrow character sets to wide
 character sets. If fishd is running into errors with this, it probably
 means that there is an issue with the locale. I'm going to rewrite a
 bit of code to make fish more stringent here, probably by making sure
 that the on-the-wire character set is always UTF-8, even if the locale
 is e.g. C.

 I probably wasn't seeing this problem because none of my universal
 variables contain non-ascii data. I'll be darned if I know why this
 problem would be surfacing right now for both you and Martin,
 though...

Maybe somehow due to installation.  I installed the darcs version as
my default shell for the first time soon before the bug appeared.
Before that I had tested it by starting it from an installed released
version.  Did anyone else try installing the darcs version yet?


 Anyway, thanks for the help. I hope to have a patch for this today or
 tomorrow, hope you'll have time to try it out.

 
 
  Then the last line seems to repeat for the next 11106 lines.
 
 
  I don't know where the /afs/cs.pitt.edu/usr0/philip/... comes from;
  maybe in a log or something.  That's my directory on AFS, but no AFS
  client is running now and rarely ever.

 That doesn't happen to be ~, or an alias for it, then?

/afs/cs.pitt.edu/usr0/philip is ~ on a remote system, but there is no
reference to it on my local home directory, either as a (broken) link
or as text in a file.  And the references are to the lib/*
subdirectories.
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Fish-users mailing list
Fish-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fish-users