[9819] in Perl-Users-Digest
Perl-Users Digest, Issue: 3412 Volume: 8
daemon@ATHENA.MIT.EDU (Perl-Users Digest)
Mon Aug 10 23:07:23 1998
Date: Mon, 10 Aug 98 20:01:27 -0700
From: Perl-Users Digest <Perl-Users-Request@ruby.OCE.ORST.EDU>
To: Perl-Users@ruby.OCE.ORST.EDU (Perl-Users Digest)
Perl-Users Digest Mon, 10 Aug 1998 Volume: 8 Number: 3412
Today's topics:
IMPORTANT: Rules for Posting to USENET (was: Re: variab <tchrist@mox.perl.com>
Re: Install error - activestate perl 5.005 (Heath Holcomb)
perl 5.005_01, Tk400.202 and AIX <cmihaly@fa.disney.com>
Re: PERL and HTACCESS <sme@planetpod.com>
Re: Perl Docs.. forget the original post <metcher@spider.herston.uq.edu.au>
Re: perlfaq - frequently asked questions about Perl (pa <uri@sysarch.com>
Re: perlfaq - frequently asked questions about Perl (pa (Nathan V. Patwardhan)
Re: socket: read on close filehandle ??? (Ronald J Kimball)
Re: strings (Mark-Jason Dominus)
taint checking and find.pm (Alex Krohn)
Re: Test Results - Asking a question (Ronald J Kimball)
Special: Digest Administrivia (Last modified: 12 Mar 98 (Perl-Users-Digest Admin)
----------------------------------------------------------------------
Date: 11 Aug 1998 02:44:52 GMT
From: Tom Christiansen <tchrist@mox.perl.com>
Subject: IMPORTANT: Rules for Posting to USENET (was: Re: variable inerpolation)
Message-Id: <6qob74$nc7$1@csnews.cs.colorado.edu>
Keywords: perl cgi
[courtesy cc of this posting sent to cited author via email]
In comp.lang.perl.modules, bday@cbr.eng.sun.com writes:
:Does anyone know if there is a way to have a variable name interpolated inside
:another variable name ? i.e.:
I already answered you in your clone posting. *PLEASE* do not clonepost
instead of crosspost. It wastes spaces, and tries the patience of those
whose help you need.
Please learn how to use your newsreader to make correct crosspostings and
to control followups. And while we're at it, I don't see anything about
modules in your question. Why did you clonepost it to the modules group?
I include the following bit of netiquette that no one seems to read anymore.
Please read all of it. Search for `cross-post' for the problem I'm
referring to. And check the headers of my article for good examples.
--tom, who just neoligismed "clonepost"
Path: csnews!coop.net!news.verinet.com!news.psd.k12.co.us!newsfeed.frii.net!news.internetMCI.com!newsfeed.internetmci.com!144.212.95.13!nntprelay.mathworks.com!panix!news.gw.com!do-not-use-path-to-reply
Date: Tue, 28 Jul 1998 08:00:18 GMT
Supersedes: <Evz28J.IB4@tac.nyc.ny.us>
Expires: Fri, 28 Aug 1998 08:00:18 GMT
Message-ID: <EwsowI.Kzq@tac.nyc.ny.us>
>From: netannounce@deshaw.com (Mark Moraes)
Subject: Rules for posting to Usenet
Newsgroups: news.announce.newusers,news.answers
Followup-To: news.newusers.questions
Approved: netannounce@deshaw.com (Mark Moraes)
Lines: 354
Xref: csnews news.announce.newusers:3010 news.answers:137189
Archive-name: usenet/posting-rules/part1
Original-author: mark@stargate.com (Mark Horton)
Comment: enhanced & edited until 5/93 by spaf@cs.purdue.edu (Gene Spafford)
Last-change: 16 Jan 1998 by netannounce@deshaw.com (Mark Moraes)
Changes-posted-to: news.misc,news.answers
This message describes some of the rules of conduct on Usenet. The rules
vary depending on the newsgroup.
Some newsgroups are intended for discussions and some for announcements
or queries. It is not usually a good idea to carry on discussions in
newsgroups that are designated otherwise. It is never a good idea to
carry on "meta-discussions" about whether a given discussion is
appropriate -- such traffic mushrooms until nobody can find articles
that belong. If you are unhappy with what some user said, send him/her
mail, don't post it.
Before posting, think about where your article is going. If it's posted
to a "comp", "humanities", "news", "misc", "soc", "sci", "rec" or "talk"
newsgroup, it will probably go to the sites on every continent with an
estimated audience of more than 3 million potential readers. Certain
articles are only of local interest (e.g. used car ads) and it is
inappropriate to post them to the whole world. Use the "Distribution"
feature to restrict distribution to your local area. If you don't know
how to use this feature, read the "Answers to Frequently Asked Questions
about Usenet" in another article in news.announce.newusers. (Note,
however, that some sites have broken software or improperly configured
news systems, so sometimes use of a "Distribution" header may not work.)
Be considerate with your use of network resources. Your individual
usage may not seem like much compared to the net as a whole, but in
aggregate, small savings in disk or CPU add up to a great deal. For
instance, messages offering thanks, jibes, or congratulations will
only need to be seen by the interested parties -- send these by mail
rather than posting them. The same goes for simple questions, and
especially for any form of "me too" posting.
To help minimize some transfer load and disk usage throughout the
Usenet, consider not only how many groups should carry your posting
over what distribution area, but also how long it will be useful. Many
kinds of postings -- such as those making announcements or offers --
have an obvious useful lifetime. Posted questions that aren't answered
within a decent interval probably won't be answered at all, and
announcements will have a limited lifetime. All such postings will be
using bandwidth to no purpose after a certain time. When making such
postings one should determine what that time interval is, based upon
the nature of the posting, the volume of articles on the newsgroup(s)
involved, and the habits of the audience, if known. Then include an
expiration date in the posting. This will mark the date after which
the article should not be retained at each site.
To include an expiration date in an article, when posting insert a
line in the header below the "Newsgroups:" line with the expiration.
For instance, type "Expires: 5 Feb 92" to have the article expire
after Feb 5, 1992. Most news software will also accept expiration
dates of the form "Expires: +5days". Please do NOT set expiration
dates far into the future simply to have the article stay around.
Many sites expire old articles no matter what the header indicates, so
you are unlikely to achieve much other than clutter the disk on a few
sites. Default expiration is normally in the range of 7 to 21 days,
depending on disk space at each site.
Don't post announcements regarding major news events (e.g. the space
shuttle has just exploded!) to news groups. By the time most people
receive such items, they will long since have been informed by
conventional media. If you wish to discuss such an event on the net,
use the "misc.headlines" newsgroup.
Announcement of professional products or services on Usenet is allowed,
provided suitable restraint is exercised. Since someone else is paying
the phone bills for this, it is important that it be of overall benefit
to Usenet. One of the few groups where such information is appropriate
is comp.newprod. comp.newprod is a moderated group; you can get the
submission guidelines from the article "Welcome to comp.newprod", posted
periodically to comp.newprod and news.answers. You can also get this
article by sending a mail message to mail-server@rtfm.mit.edu with the
single line:
send usenet-by-group/news.answers/newprod
If your posting is really relevant to some other newsgroup, particularly
one of the *.announce newsgroups, you may consider posting it there; some
moderators allow product announcements in the *.announce newsgroups. e.g.
an announcement about an Amiga product could go in
comp.sys.amiga.announce. Before you post any such announcements, make
sure that you carefully read all of the administrative documents for the
group. Also, read the regular messages in the group itself for at least
a week to make sure that your announcement is consistent with what other
people post. Of course, this is true for *any* post, but especially true
for commercial announcements.
General guidelines: Clearly mark your article as a product announcement
in the subject. Never repeat these -- one article per product at the
most; preferably group everything into one article. Advertising hype is
especially frowned upon -- stick to technical facts. Obnoxious or
inappropriate announcements or articles violating this policy will
generally be rejected. This policy is, of course, subject to change if
it becomes a problem.
There exists an alternative hierarchy called "biz" specifically for
commercial postings. See the articles "Alternative Newsgroup Hierarchies,
Part ...", posted periodically to several newsgroups, including
news.lists.misc. You can also get these articles by sending a mail message
to mail-server@rtfm.mit.edu with the lines:
send usenet-by-group/news.answers/alt-hierarchies/part1
send usenet-by-group/news.answers/alt-hierarchies/part2
Some newsgroups are moderated. In these groups, you cannot post
directly, either by convention or because the software prevents it. To
post to these newsgroups, send mail to the moderator. Examples:
Newsgroup Moderator Purpose
--------- --------- -------
news.announce.important announce@stargate.com Important announcements for everyone.
comp.std.unix std-unix@uunet.uu.net Unix standards discussion.
rec.food.recipes recipes@taronga.com Sharing favorite recipes.
Some newsgroups have special purpose rules:
Newsgroup Rules
--------- -----
news.announce.important Moderated, no direct postings, important things only.
misc.wanted Queries, "I want an x", "Anyone want my x?". No
discussions. Don't post to more than one xxx.wanted.
Use the smallest appropriate wanted (e.g. used car
ads to nj.wanted.)
Requests for sources, termcaps, etc. should go to the
"comp.sources.wanted" newsgroup.
rec.humor Clean humor only; anything offensive must be rotated;
no discussions -- humor only. Discussions go in
rec.humor.d
rec.arts.movies Don't post anything revealing part of a movie
without marking it (spoiler) in the subject.
rec.arts.* Same as movies -- mark spoilers in the subject line.
news.groups Discussions about new groups: whether to create
them and what to call them. Don't post yes/no
votes, mail them to the author
misc.test Use the smallest test group possible, e.g.
"test" or "ucb.test". Say in the body of the
message what you are testing.
If you're thinking of posting anything that was written by someone else
(eg. article, song, picture), make sure that you are familiar with the
copyright issues. If you're not sure about the copyright issues, then
find out before posting. For instance, you must not post anything that
you were allowed to see only because of a confidentiality agreement,
such as a UNIX source license.
It is generally considered rude to post private e-mail correspondence
without the permission of the author of that mail, and furthermore,
it's likely a copyright violation as well.
All opinions or statements made in messages posted to Usenet should
be taken as the opinions of the person who wrote the message. They do
not necessarily represent the opinions of the employer of that person,
the owner of the computer from which the message was posted, or anyone
involved with Usenet or the underlying networks of which Usenet is
made up. All responsibility for statements made in Usenet messages
rests with the individual posting the message.
Posting of information on Usenet is to be viewed as similar to
publication. Because of this, do not post instructions for how to do
some illegal act (such as jamming radar or obtaining cable TV service
illegally); also do not ask how to do illegal acts by posting to the net.
If you have a standard signature you like to append to your articles,
and you are running a form of news software that supports automatic
inclusion of a signature file, it is usually enabled by putting it in
a file called .signature in your home directory. The posting software
you use should automatically append it to your article. Please keep your
signatures concise, as people do not appreciate seeing lengthy signatures,
nor paying the phone bills to repeatedly transmit them. 2 or 3 lines are
usually plenty. Sometimes it is also appropriate to add another line or
two for addresses on other major networks where you can be reached (e.g.,
CompuServ, Bitnet). Long signatures are definitely frowned upon. DO NOT
include drawings, pictures, maps, or other graphics in your signature --
it is not the appropriate place for such material and is viewed as rude
by other readers.
If you post an article and remember something you've left out or realize
you've made a factual error, you can cancel the article and (if canceled
quickly enough) prevent its distribution. Then you can correct whatever
was wrong and post a new copy. In "rn", "trn", "nn" and "readnews",
(and probably most other newsreaders) an article that you posted can be
canceled with the "C" command. In "tin", use "D" (delete) to cancel
an article. Newer newsreaders typically offer "Cancel" from a menu.
Be aware, however, that some people may have already read the incorrect
version so the sooner you cancel something, the better.
Before posting a question to the net (especially one that you think
will be easy for experts to answer), consider carefully whether posting
is the most appropriate way to get the answer. There are many ways to
find answers without using up network resources and forcing thousands
of people to read your question (and several helpful volunteers to spend
time responding). Many newsgroups have a Frequently Asked Questions (FAQ)
list that is posted periodically (usually every few weeks), and they are
also usually cross-posted to news.answers. They usually have explicit
expiration dates set, so they shouldn't be expired until a new version
has been posted, so if you can't find the FAQ in either the newsgroup or
news.answers, there probably isn't one (thus, it's probably not useful to
post a question asking whether there is one). If you have local experts
(or simply more experienced users than yourself) at your site, try asking
them before posting. If you're trying to find where you can FTP software
or a newsgroup archive, try using the Archie service; see postings in
news.answers for details. Many newsgroups are also archived in Wide Area
Information Service (WAIS) databases; WAIS client software may be FTPed
from ftp.think.com, or you may use WAIS by telnetting to quake.think.com
and logging in as "wais". Finally, you should also check the manuals
for your system; if you don't, and you post a question that's answered
there, you'll likely receive a number of responses that scream "RTFM"
(Read the F*ing Manual).
Usually, it is sufficient to post any article to a single newsgroup; the
one that's most relevant to the subject of your article. If the article
is really relevant to multiple newsgroups, then "cross-post" to the
relevant newsgroups by posting the article only once with all newsgroups
named on the "Newsgroups" header line. For example:
Newsgroups: comp.fish,misc.sheep,talk.ketchup
would cause an article to be posted to comp.fish, misc.sheep, and
talk.ketchup simultaneously.
If you are using TIN, please do not use the "crosspost" function to
accomplish this. This is a misleadingly named command that really should
be called "repost."
By posting a single article to all the newsgroups you wish to reach, the
news software is able to transfer a single copy. Furthermore, users with
"smart" newsreaders will see the article only once. Making separate
postings of your article for each newsgroup you wish to reach tends to
annoy readers rather than emphasize the message content as well as waste
computational resources.
All newsreaders should have two ways to post a news article. First,
there is an original posting; this is used whenever you are starting a
new topic. Second, there is a "followup"; this is used when you are
posting a response to another news article. In several newsreaders,
including "rn", the "f" command usually generates an original posting
if your current position is at the end of the newsgroup, but a followup
when you have a current article; you can also use the "Pnews" command
outside of rn to make an original posting.
The news posting software does special things in the second case that
indicates to the news system that this article is "related" to the
article to which you are following up. First, the newsreader adds
"Re: " before the existing subject line to tell people that this is
"regarding" a previous article. Second, the software adds a "References"
line that contains the Message-ID of the article you are following up.
This header is used by threaded news readers such as "trn" to follow
"threads" of discussion.
It is important that these two posting methods not be confused. Don't
follow up to articles without using the newsreader's "followup" mechanism.
Conversely, don't use the followup mechanism to post an article that
is an unrelated thread. Violating this convention sometimes leads to
confusion and annoyance of users with threaded newsreaders.
When posting a followup, be careful about newsgroups. The article that
you're responding to might have been cross-posted to several newsgroups,
and by default your followup will go to ALL of those newsgroups. Or the
article might have a Followup-To line in its header, and in that case,
by default your followup will go where the Followup-To line says -- which
might not be the newsgroup where you're reading the article. You should
ensure that your article is posted only to newsgroups where its actual
content is appropriate. Sometimes it's better to leave the newsgroups
on your own article the same as they were, but put a Followup-To line in
its header to confine followups to an appropriate group. In any case,
it's best for articles that have a Followup-To line to be posted to
whatever groups are mentioned in that line, and to mention in the text of
the article that followups are redirected. The idea is for the threads
of articles to make sense in each newsgroup where the articles appear,
for people who don't read the others.
If you don't see your posting immediately, don't assume it failed and
try to repost it at once. Some sites have set up the local software
to process news periodically. Thus, your article will not appear
immediately. If you post again, you will have multiple copies of the
article in circulation.
If the news system rejects a followup due to "more quoted lines than
new text," please do not use "filler" lines to make up for this.
Instead, if after careful editing, you have more to quote than to
write, change the citation character. For example, in the display
editor vi, you could use the incantation:
:%s/^>/</
Be careful not to do the very similar:
:%s/>/</
which will affect >'s that are not being used as the citation
character. (In particular, it will damage the "References" line in the
article header.)
In preparing an article, be aware that other people's machines are
not the same as yours. The following is a list of things to keep
in mind:
* Keep your lines under 80 characters, and under 72 if possible (so that
the lines won't get longer than 80 when people include them when
responding to your postings). Most editors have a fill or format mode
that will do this for you automatically. Make sure that it
actually puts ("hard") newline characters into the file, rather
than just wrapping the displayed lines on your screen.
* Right justified text may look "prettier" in some sense, but it
is almost always harder to read than leaving ragged right
margins; don't justify your articles.
* Most special control characters will not work for most readers.
In fact, the space character is about the only one
you can be sure will work consistently. Even tabs aren't always
the same from machine to machine, and should be avoided. Many mail
agents will strip or remap control characters.
* Pictures and diagrams should not use embedded tabs.
* Refer to articles by Message-ID, and never by article number.
Article numbers vary on every news system, Message-IDs are always
preserved throughout the network.
* What you think is the previous article is unlikely to be so elsewhere.
* Submissions in a single case (all upper or all lower) are
difficult to read.
In general, when a mailing to somebody fails, DON'T post a message about
it! Think for a moment: you are trying to send something to someone
on ONE system. Your message might go through (at most) TEN systems on
the way there. Posting a message in the news sends it to many tens of
thousands of systems throughout the world! There is no way to justify
adding to the news load of all those machines simply because you cannot
determine how to get your mail through.
If your message is important, contact someone who knows more about the
mail system and who might be able to help you get your message through.
Your local system administrator, for instance, or the admin of the
next site "upstream," might be able to help. You can also send mail to
"postmaster" at one of the major Usenet sites. Almost all of these
people would rather see an occasional plea for help in their mailbox than
read another broadcast in the news system. If your message is *really*
important, pick up the phone and try to call the other person.
--
I hate programs that chdir --Boyd Roberts
------------------------------
Date: Tue, 11 Aug 1998 02:08:30 GMT
From: holcojh5@REMOVEMEwfu.edu (Heath Holcomb)
Subject: Re: Install error - activestate perl 5.005
Message-Id: <35d0a654.6338328@enews.newsguy.com>
On Mon, 10 Aug 1998 15:47:39 -0700, "Mary Gough" <gough@cruznet.net>
wrote:
>I am using the win 98 operating system and have no trouble installing perl.
>However, certain features don't work for example when executing the program
>perlse during the install I get:
>
>C:\Perl\5.005\bin\MSWin32-x86-object>perl perlse.pl
>Can't locate Win32/OLE.pm in @INC (@INC contains:
>C:\PERL\5.005\lib/MSWin32-x86-
>object C:\PERL\5.005\lib C:\PERL\site\5.005\lib/MSWin32-x86-object
>C:\PERL\site\
>5.005\lib C:\PERL\site\lib .) at perlse.pl line 1.
>BEGIN failed--compilation aborted at perlse.pl line 1.
>
>No example programs are 'unpacked' etc.
>
>Can anybody help?
>
>Thanks,
>
[posted and emailed]
According to what I read at Activestate's website, build 500 of
ActivePerl (aka Perl 5.005) was missing some files, which evidently
included the Win32::OLE module. Build 501 is available as well as
another download which will install the missing files.
Hope this helps...
Heath
--
Heath Holcomb
holcojh5@REMOVEMEwfu.edu
http://www.wfu.edu/~holcojh5/
delete REMOVEME to email
------------------------------
Date: Mon, 10 Aug 1998 18:36:21 -0700
From: Chris Mihaly <cmihaly@fa.disney.com>
To: perlbug@perl.com
Subject: perl 5.005_01, Tk400.202 and AIX
Message-Id: <35CFA015.8AF797A@fa.disney.com>
I am trying to build Perl 5.005_01 and a few Perl packages on AIX
4.1.3. I have been able to build perl itself and several other modules,
but when I try to build Tk400.202, I get into problems.
First, it appears that there is a bogus file in the distribution,
pTk/Tix which is a sym link to an invalid address,it's an absolute path
to somebody's machine. This causes perl Makefile.PL, on AIX to hang
forever trying to get to this sym link. Strangely, on IRIX and Solaris,
it just ignores the sym link. I resolved this by removing the bogus sym
link and creating an empty file, this seems to make Configure happy for
that part.
Second problem is a bit later on, except that now it core dumps. Here
is the output of configure:
(hermes:Tk400.202) > /rel/extern/rel1.6/bin/perl Makefile.PL
/rel/extern/rel1.6/bin/perl is installed in /rel/extern/rel1.6/lib/perl5
okay
Generic gettimeofday()
/usr/bin/X11/xmkmf suggests /usr
Using -L/usr/lib to find /usr/lib/libX11.a
Using -I/usr/include to find /usr/include/X11/Xlib.h
Checking if your kit is complete...
Looks good
Note (probably harmless): No library found for -lpt
Note (probably harmless): No library found for -lsocket
Note (probably harmless): No library found for -lnsl
Writing Makefile for Tk::Bitmap
Writing Makefile for Tk::Canvas
Writing Makefile for Tk::Contrib
Writing Makefile for Tk::DragDrop
Writing Makefile for Tk::Entry
Writing Makefile for Tk::Ghostview
Writing Makefile for Tk::HList
Following modules are required but not installed:
HTML::Parse
Writing Makefile for Tk::HTML
Writing Makefile for Tk::IO
Writing Makefile for Tk::InputO
Writing Makefile for Tk::Listbox
Writing Makefile for Tk::Menu
Writing Makefile for Tk::Menubutton
Writing Makefile for Tk::Mwm
Writing Makefile for Tk::NBFrame
Writing Makefile for Tk::Photo
Writing Makefile for Tk::xpm
Writing Makefile for Tk::Pixmap
Checking if your kit is complete...
Looks good
Segmentation fault (core dumped)
The core dump is in the pod directory. I tried to trace it down via
perl -d, but couldn't follow the extensive eval nestings, although
strangely, with the -d it just hangs, it never core dumps.
Anyone seen this problem and have an idea of how to resolve it?
Thanks
Chris
--
Christopher Mihaly Email: cmihaly@fa.disney.com
Walt Disney Feature Animation Phone: (818) 526-3231
500 S. Buena Vista St. Fax: (818) 560-9388
Burbank, CA 91521-4806
------------------------------
Date: Mon, 10 Aug 1998 16:54:53 -0700
From: Stewart Eastham <sme@planetpod.com>
Subject: Re: PERL and HTACCESS
Message-Id: <35CF884C.752F8f36@planetpof.com>
> >I am trying to get a user through a .htaccess protected directory
> >through PERL. (i.e. I want to have the user enter their name and
> >password into a HTML form instead of the htaccess pop-up box.). Is this
> >possible?
>> yes. did you have another question about it?
The follow-up question would be, then, HOW does one do it?
> >The other thing I am wondering, along the same lines, is it possible for
> >PERL to retrieve the name and password that a user has entered into a
> >.htaccess pop-up box? Then, based on their login, I could output a
> >page, through a PERL script, with conditional outputs.
>
> with mod_perl you can get both.
I downloaded the mod_perl module, but couldn't quite figure out how to use it
to serve my purposes. I also downloaded the HTTPD module, which seemed to
have something related to authenticating .htaccess user through PERL, as
well. Again, though, I couldn't figure out which code is directly applicable
to what I need to do.
Any sample code would be greatly appreciated. Thanks
stewart eastham
sme@planetpod.com
------------------------------
Date: Tue, 11 Aug 1998 12:30:06 +1000
From: Jaime Metcher <metcher@spider.herston.uq.edu.au>
Subject: Re: Perl Docs.. forget the original post
Message-Id: <35CFACAE.86EC7B4E@spider.herston.uq.edu.au>
Dave Till wrote:
>
> I don't think you have to read all the Perl man pages before using Perl.
> To quote from the Preface to the Camel:
>
> Most important, you don't have to know everything there is to know
> about Perl before you can write useful programs. You can learn Perl
> "small end first". [...] Many of the ideas in Perl are borrowed from
> natural language, and one of the best ideas is that it's okay to use
> a subset of the language as long as you get your point across. Any
> level of language proficiency is acceptable in Perl culture.
>
> What the Perl gurus *have* been saying all along is that the answers
> to many questions can be found in the man pages or in the FAQs. (From
> my own experience, I've found this to be true.)
>
<snip>
You're right. I read that preface when I was starting to learn Perl,
was impressed by it, and put it into immediate practice.
However, I've seen people on this ng get annoyed when people ask a
question about something that is already explained in (for example):
1. The camel book (fair enough).
2. Not in the camel book, but in perlre (still fair enough).
3. Not in either the camel or perlre (where they thought it would be)
but in perlop (yeah, OK, they should have grepped).
4. In neither the camel, perlre or perlop, but somewhere in one the the
many parts of the FAQ list (ditto, but I'd be giving points for trying).
5. In none of the above, but answered on this ng and therefore in
DejaNews (fatigue allowance here, especially for newcomers).
Several people have kindly posted their search strategies when trying to
find things in the documentation, under subject lines like "Things I do
before posting to the newsgroup). When we're getting search strategies
with five, six and more stages we're in a different culture to one where
the procedure is:
1. Consult *the* manual.
2. Call support.
All this really means is that the latter culture does not have as many
resources at its disposal. Give me the freeware system anyday. But
lets all keep an eye out for those suffering from culture shock, hey?
--
Jaime Metcher
------------------------------
Date: 10 Aug 1998 22:18:03 -0400
From: Uri Guttman <uri@sysarch.com>
Subject: Re: perlfaq - frequently asked questions about Perl (part 0 of 9)
Message-Id: <x77m0g5kis.fsf@sysarch.com>
>>>>> "IZ" == Ilya Zakharevich <ilya@math.ohio-state.edu> writes:
IZ> [A complimentary Cc of this posting was sent to Nathan
IZ> V. Patwardhan <nvp@shore.net>], who wrote in article
IZ> <6qnrdp$49j@fridge.shore.net>:
>> : and check out tkman which uses polyglotman internally. i think tk
>> will : run on winblows so you could run this. it is an excellent
>> gui viewer of : man pages and texinfo pages.
>>
>> You refer to Windows as "winblows" and yet you refer people to a
>> GUI-based tool? Something is fishy in Denmark.
IZ> For some tasks (reading docs is one of them) GUI-based tool pretty
IZ> often blow command-line tools away (*). And I do not see how this
IZ> relates to M$ bashing (is it correct that Win* has no format for
IZ> online books?!).
thanx ilya for that backing. i don't dislike gui's (i prefer command
lines to do real work) but i hate micro$hit in general. i have many
windows on my sun (including tkman, netscrape, emacs and may xterms) and
i do mouse around. but xwindows is a proper system which i can customize
and control. winblows is just what i call it. sneeze and reboot!
uri
--
Uri Guttman ----------------- SYStems ARCHitecture and Software Engineering
Perl Hacker for Hire ---------------------- Perl, Internet, UNIX Consulting
uri@sysarch.com ------------------------------------ http://www.sysarch.com
The Best Search Engine on the Net ------------- http://www.northernlight.com
------------------------------
Date: 11 Aug 1998 02:41:53 GMT
From: nvp@shore.net (Nathan V. Patwardhan)
Subject: Re: perlfaq - frequently asked questions about Perl (part 0 of 9)
Message-Id: <6qob1h$ask@fridge.shore.net>
Uri Guttman (uri@sysarch.com) wrote:
: and control. winblows is just what i call it. sneeze and reboot!
Inferior "OS" or not, blanket trashing of said product also insults
the work that a number of people have done to (1) create a complete
Perl distribution for the win32 crowd, (2) merge the AS code with the
core distribution, and (3) build fully-functional, Unix-like tools for
said environment (like Cygwin, mingw32, U/WIN, etc).
--
Nate Patwardhan|root@localhost
"Fortunately, I prefer to believe that we're all really just trapped in a
P.K. Dick book laced with Lovecraft, and this awful Terror Out of Cambridge
shall by the light of day evaporate, leaving nothing but good intentions in
its stead." Tom Christiansen in <6k02ha$hq6$3@csnews.cs.colorado.edu>
------------------------------
Date: Mon, 10 Aug 1998 21:41:23 -0400
From: rjk@coos.dartmouth.edu (Ronald J Kimball)
Subject: Re: socket: read on close filehandle ???
Message-Id: <1ddkdtd.1w5f7u2scvb4N@bay1-160.quincy.ziplink.net>
Michael Agbaglo <byteshifter@shifted-bytes.de> wrote:
> This is a multi-part message in MIME format.
> --------------6212C57626F51740D94A619D
> Content-Type: text/plain; charset=us-ascii
> Content-Transfer-Encoding: 7bit
Please don't do that.
> Read on closed filehandle <S> at ./x line 8.
>
> [...]
>
> $S = IO::Socket::INET->new('localhost:25') or die "can't connect to socket";
>
> print while(<S>);
Since the connection to your Socket is stored in $S, you might want to
read from $S instead of the unused filehandle S...
print while(<$S>);
--
_ / ' _ / - aka - rjk@coos.dartmouth.edu
( /)//)//)(//)/( Ronald J Kimball chipmunk@m-net.arbornet.org
/ http://www.ziplink.net/~rjk/
"It's funny 'cause it's true ... and vice versa."
------------------------------
Date: 11 Aug 1998 01:34:41 GMT
From: mjd@plover.com (Mark-Jason Dominus)
Subject: Re: strings
Message-Id: <6qo73h$vms$1@netnews.upenn.edu>
In article <6qnvui$hfj$1@marina.cinenet.net>,
Craig Berry <cberry@cinenet.net> wrote:
>Well, one could argue that the zip code should be treated as a string
>value, not a numeric value.
Well, it *is* being treated as a string value, not a numeric value,
but that's exactly what you said you didn't want. You just said you
wanted it to parse as octal:
>: >I've often wished that 077 and 0xCA in input data would parse as
>: >octal and hex respectively.
Since you seem to be professing two contradictory opinions, I'm going
to stick with my assessment that you haven't thought it through.
>I was referring (a bit elliptically, true) more to Lisps use of list
>representation for both data and code, and the resulting easy blurring of
>the two.
There are two parts to that. One part, the wonderful and important
part, has to do with closures and their use in object-oriented
programming, that you use to write things like this:
;; Make a counter object that supports `increment' and `get' methods
;; Please pardon my syntax errors if there are any
(defun make-counter ()
( let ((x 0))
(lambda (operation)
(cond ((eq operation 'increment)
(setq x (+ x 1)))
((eq operation 'get)
x)
(t
(error "Unknown operation: " (make-string operation)))
)
)))
;; Convenience methods for counters
;; Is x data? Or is it code? Who knows? Who cares?
(defun increment (x) (x 'increment))
(defun get (x) (x 'get))
;; Try them out
(setq c1 (make-counter))
(setq c2 (make-counter))
(increment c1)
(increment c1) ; Is c1 a data object? Or a function? Whatever!
(increment c1)
(get c2) ; yields 0
(get c1) ; yields 3
But it seemed to me that your complaint didn't have anything to do
with this; instead it was about treatment of input data and of literal
constants in the source, and had a lot more to do with eval and
backquote and string->symbol and the like. These are important
features of Lisp, but not nearly as central or as important to the
`data as code' notion as the use of closures is.
I sent followups to comp.programming since this no longer appears to
have anything to do with Perl.
------------------------------
Date: Tue, 11 Aug 1998 01:25:39 GMT
From: alex@gossamer-threads.com (Alex Krohn)
Subject: taint checking and find.pm
Message-Id: <35cf9c91.53273808@coburg>
I'm having a problem with taint checking and the find module. I have:
sub is_tainted {
not eval {
join ("", @_), kill 0;
1;
};
}
taken from the Camel book to see if something is tainted or not. Then
in my code:
sub checkspace {
# -----------------------------------------------------
# Check for allowed disk space to determine whether we can allow
# editing or uploads.
#
use File::Find;
my ($directory) = shift;
my ($size, $used_space, $free_space) = 0;
if (&is_tainted ($directory)) { &cgierr ("$directory is
tainted!"); }
&find ( sub { $size += -s }, $directory );
$used_space = int ($size / 1024);
$free_space = ($config{'allowed_space'} - $used_space);
return ($free_space, $config{'allowed_space'}, $used_space);
}
The script dies with:
'Insecure dependency in chdir while running with -T switch at
/usr/local/lib/perl5/5.005/File/Find.pm line 133.'
which is the &find call. If I comment the line out, the script works
as expected.
Now $directory shouldn't be tainted as I run it through a is_valid
function which matches a reg exp on it and returns an $okversion.
Something like:
($dir =~ m,^([A-Za-z0-9\-_/]*)$,) ?
($okdir = $1) :
(return ($dir, "Illegal Characters in Directory. Please use
letters, numbers, - and _ only."));
and $okdir is returned. Any ideas what's going on?
Thanks for the help!
Alex
------------------------------
Date: Mon, 10 Aug 1998 21:41:26 -0400
From: rjk@coos.dartmouth.edu (Ronald J Kimball)
Subject: Re: Test Results - Asking a question
Message-Id: <1ddkeaw.gn5wc1f002hoN@bay1-160.quincy.ziplink.net>
John Porter <jdporter@min.net> wrote:
> I *would* gripe about giving away this secret to the unwashed
> masses; after all, you only *claimed* to have read the docs;
> you didn't give any evidence that you actually *had*.
Oo, I didn't think of that. :-(
So, how would someone give evidence that they had actually read the
docs?
> In fact, in this example at least, you showed that you
> actually had *not* read the docs. A FAQ is still a FAQ.
The entire post was based on John Moreno's suggestion in the cited post.
That was the question he proposed to ask.
And I proved him right. :-)
--
_ / ' _ / - aka - rjk@coos.dartmouth.edu
( /)//)//)(//)/( Ronald J Kimball chipmunk@m-net.arbornet.org
/ http://www.ziplink.net/~rjk/
"It's funny 'cause it's true ... and vice versa."
------------------------------
Date: 12 Jul 98 21:33:47 GMT (Last modified)
From: Perl-Request@ruby.oce.orst.edu (Perl-Users-Digest Admin)
Subject: Special: Digest Administrivia (Last modified: 12 Mar 98)
Message-Id: <null>
Administrivia:
Special notice: in a few days, the new group comp.lang.perl.moderated
should be formed. I would rather not support two different groups, and I
know of no other plans to create a digested moderated group. This leaves
me with two options: 1) keep on with this group 2) change to the
moderated one.
If you have opinions on this, send them to
perl-users-request@ruby.oce.orst.edu.
The Perl-Users Digest is a retransmission of the USENET newsgroup
comp.lang.perl.misc. For subscription or unsubscription requests, send
the single line:
subscribe perl-users
or:
unsubscribe perl-users
to almanac@ruby.oce.orst.edu.
To submit articles to comp.lang.perl.misc (and this Digest), send your
article to perl-users@ruby.oce.orst.edu.
To submit articles to comp.lang.perl.announce, send your article to
clpa@perl.com.
To request back copies (available for a week or so), send your request
to almanac@ruby.oce.orst.edu with the command "send perl-users x.y",
where x is the volume number and y is the issue number.
The Meta-FAQ, an article containing information about the FAQ, is
available by requesting "send perl-users meta-faq". The real FAQ, as it
appeared last in the newsgroup, can be retrieved with the request "send
perl-users FAQ". Due to their sizes, neither the Meta-FAQ nor the FAQ
are included in the digest.
The "mini-FAQ", which is an updated version of the Meta-FAQ, is
available by requesting "send perl-users mini-faq". It appears twice
weekly in the group, but is not distributed in the digest.
For other requests pertaining to the digest, send mail to
perl-users-request@ruby.oce.orst.edu. Do not waste your time or mine
sending perl questions to the -request address, I don't have time to
answer them even if I did know the answer.
------------------------------
End of Perl-Users Digest V8 Issue 3412
**************************************
