[prev in list] [next in list] [prev in thread] [next in thread] 

List:       sbcl-devel
Subject:    Re: [Sbcl-devel] What is 'high-security' feature expression?
From:       Roman Marynchak <roman.marynchak () gmail ! com>
Date:       2010-05-29 7:06:22
Message-ID: AANLkTikV1StVDYfmXiMXh79wEdW7HM2aQPEyWd4Xe7iv () mail ! gmail ! com
[Download RAW message or body]

[Attachment #2 (multipart/alternative)]


I also think that the performance impact will be small. So, I will rework
the macro as per Nathan's suggestion. But should maintainers  go one step
ahead and clean the entire SBCL sources from 'high-security' feature
expression at all? Or it is useful in some situations?

Regards,
Roman


2010/5/29 James Y Knight <foom@fuhm.net>

> On May 28, 2010, at 4:52 PM, Nathan Froyd wrote:
> On Fri, May 28, 2010 at 3:44 PM, Roman Marynchak
>
>> <roman.marynchak@gmail.com> wrote:
>>
>>>  I have seen this feature expression in several source files, for example
>>> in
>>> /code/sysmacs.lisp:
>>>
>>> It seems that it is some old CMUCL feature expression. Is it okay to get
>>> rid
>>> of it, at least in that macro? I need the stream type check to be made by
>>> default in order to fix this issue:
>>>
>>> https://bugs.launchpad.net/sbcl/+bug/586940
>>>
>>> The first step is to pass t as a second macro argument, and the second is
>>> to
>>> remove or enable 'high-security', as one may guess from PRINT definition:
>>>
>>
>> The second macro argument seems to be totally unused, so it's probably
>> best to just delete the argument from the macro and make the check the
>> default.
>>
>
> I suppose it was omitted because of the performance cost? Although,
> checking the type of an argument can't *really* have significant performance
> cost in an output function I wouldn't think.
>
> James
>
>

[Attachment #5 (text/html)]

I also think that the performance impact will be small. So, I will rework the macro \
as per Nathan&#39;s suggestion. But should maintainers  go one step ahead and clean \
the entire SBCL sources from &#39;high-security&#39; feature expression at all? Or it \
is useful in some situations? <br> <br>Regards,<br>Roman<br><br><br><div \
class="gmail_quote">2010/5/29 James Y Knight <span dir="ltr">&lt;<a \
href="mailto:foom@fuhm.net">foom@fuhm.net</a>&gt;</span><br><blockquote \
class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt \
0pt 0.8ex; padding-left: 1ex;"> <div><div></div><div class="h5">On May 28, 2010, at \
4:52 PM, Nathan Froyd wrote:<br> On Fri, May 28, 2010 at 3:44 PM, Roman Marynchak<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); \
margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> &lt;<a \
href="mailto:roman.marynchak@gmail.com" \
target="_blank">roman.marynchak@gmail.com</a>&gt; wrote:<br> <blockquote \
class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt \
0pt 0.8ex; padding-left: 1ex;">  I have seen this feature expression in several \
                source files, for example in<br>
/code/sysmacs.lisp:<br>
<br>
It seems that it is some old CMUCL feature expression. Is it okay to get rid<br>
of it, at least in that macro? I need the stream type check to be made by<br>
default in order to fix this issue:<br>
<br>
<a href="https://bugs.launchpad.net/sbcl/+bug/586940" \
target="_blank">https://bugs.launchpad.net/sbcl/+bug/586940</a><br> <br>
The first step is to pass t as a second macro argument, and the second is to<br>
remove or enable &#39;high-security&#39;, as one may guess from PRINT definition:<br>
</blockquote>
<br>
The second macro argument seems to be totally unused, so it&#39;s probably<br>
best to just delete the argument from the macro and make the check the<br>
default.<br>
</blockquote>
<br></div></div>
I suppose it was omitted because of the performance cost? Although, checking the type \
of an argument can&#39;t *really* have significant performance cost in an output \
function I wouldn&#39;t think.<br><font color="#888888"> <br>
James<br>
<br>
</font></blockquote></div><br>



------------------------------------------------------------------------------



_______________________________________________
Sbcl-devel mailing list
Sbcl-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-devel


[prev in list] [next in list] [prev in thread] [next in thread]