Services & ResourcesWolfram ForumsMathGroup Archive

[Date Index] [Thread Index] [Author Index]

Re: Re: Strange behavior with loading Experimental package

• To: mathgroup at smc.vnet.net
• Subject: [mg105038] Re: [mg105004] Re: Strange behavior with loading Experimental package
• From: Daniel Lichtblau <danl at wolfram.com>
• Date: Wed, 18 Nov 2009 07:01:29 -0500 (EST)
• References: <200911171016.FAA17366@smc.vnet.net>

Nasser M. Abbasi wrote:
> ref (me)
> 
> I found out that if I use CleanSlate[] then this problem does not show 
> up any more. So I am using this package now to remove packages. So, case 
> closed. One should always use CleanSlate!
> 
> << "Utilities`CleanSlate`"
> SetOptions[CleanSlate, Verbose -> False];
> << "Experimental`";
> Names["Experimental`*"]
> 
> CleanSlate["Experimental`"];
> 
> << "Experimental`";
> Names["Experimental`*"] ----> OK now, symbols are reloaded ok.
> 
> --Nasser

If you try
Names["Experimental`*"]
after running that CleanSlate, I suspect you will find that the names 
are still there. That is, CleanSlate did not remove them.

Let me respond to the earlier question, about inability to regain access 
to Remove'd Experimental` functions by a second application of Get. A 
few packages, e.g Experimental.m and Developer.m, only serve to place 
their corresponding context onto $Contextpath. They do not actually load 
any functions, because their functions are already built into the 
Mathematica kernel (in the corresponding contexts). When you remove 
their symbols, you effectively disable those functions. At that point, 
rRepeating Get can only change the $ContextPath, it cannot restore the 
removed symbols.

Daniel Lichtblau
Wolfram Research


>   "Nasser M. Abbasi" <nma at 12000.org> wrote in message news:...
>   Version 7.0, on windows.
> 
>   I load this package using Get (not Needs), then remove all its 
> symbols, then do Get again on it, but the symbols in this package do not 
> reappear as expected
> 
>   In[1]:= << "Experimental`"
>   Names["Experimental`*"]
> 
>   Out[2]= { ...}
> 
>   In[3]:= Unprotect["Experimental`*"]
>   Remove["Experimental`*"]
>   Out[3]= {...}
> 
>   In[6]:= << "Experimental`"
>   Names["Experimental`*"]
>   Out[7]= {}   <--- EMPTY !
> 
> 
>   But look at another package picked by random, it behaves as expected
> 
> 
>   In[9]:= << "ANOVA`"
>   Names["ANOVA`*"]
>   Out[10]= {...}
> 
>   In[11]:= Unprotect["ANOVA`*"]
>   Remove["ANOVA`*"]
>   Out[11]= {....}
> 
>   In[13]:= << "ANOVA`"
>   Names["ANOVA`*"]
>   Out[14]= {....}   <--- OK, not empty
> 
> 
>   I know, I know, it is called "Expriemental" so should not complain, 
> but thought to point it out just the same as it was affecting something 
> I am doing...
> 
>   --Nasser

• References:
  • Re: Strange behavior with loading Experimental package
    • From: "Nasser M. Abbasi" <nma@12000.org>