Running Linux, 3rd Edition by Matt Welsh, Matthias Kalle Dalheimer, and Lar Kaufman Unconfirmed error reports come from readers. They have not yet been approved or disproved by the author or editor and represent solely the opinion of the reader. This page was updated on January 30, 2002 Here's a key to the markup: [page-number]: serious technical mistake {page-number}: minor technical mistake : important language/formatting problem (page-number): language change or minor formatting problem ?page-number?: reader question or request for clarification UNCONFIRMED errors and comments from readers: (viii) The index lists "The K Development Enviroment" as being located on page 370, where one would actually find the "The K Desktop Enviroment". (42) Last line in Usenet Newsgroups section; related topis like comp.windows.x should be: related topic like comp.windows.x (46) serious omission; I am very surprised (I first wrote "shocked") to see that a chapter about installing Linux, in a book presented as a help to many relative newcomers, does not include any advice that you write down system configuration information prior to installing Linux. A newbie might quit for no other reason than Linux did not detect his video card, or sound card. When I installed RedHat 5.2 over Windows 95, I had forgotten to write down my SoundBlaster IRQ information. RedHat 5.2 wasn't able to detect the information. Thus, I had to reinstall Windows 95, get the information, then go back and reinstall RedHat 5.2 manually entering the information. This is not just an isolated example. Redhat 6.1 doesn't recognize my Riva TNT (it thinks it is a Riva128). I can upgrade to TNT drivers, but I have to *know* what my system is before I can even try to tell Linux to do this. I wouldn't put detailed Windows information into Running Linux, such as "click on start, click on start panel, click on..." blah, blah, blah. But *something* about getting system information should be said. If it was said, then wherever you said it needs to be moved to some other place, or at least referenced to where it was in chapter 2. I quite overlooked it. [102] The tcsh command (setenv VISUAL vi) probably shouldn't have a "$" prompt before it. tcsh uses ">" by default, I think; a lot of people use a csh-style "%" instead. (105) In the last line, the word "input" should be "output". [114] This error is present in the example in the middle of the page. It would clarify the chown example if distinct names for the owner and the group were used. As it is now I don't know whether chown owner.group file or chown group.owner file is the correct syntax. [139] In the last paragraph; The line reads "this means the kernel itself is using 1500K." I do not understand how the author arrived as this number. The closest calculation i could make was the following: 552K kernel code 384K reserved 464K data ----- 1400K total kernel memory use Is 1500K a typo? or miscalculation? If not, how does the author arrive at that number? {169} 1st complete paragraph; Last sentence of the first complete paragraph on p169 incorrectly refers to the type option of mount as -T. It should actually be -t. ?179? (1 and 2) Are you sure that is the error in the program? If you try access the last number far from the first row, what you get is the first number of row 2. It cannot fail the program. Remember in C the first index is a pointer to the rows. (Page 127 line 30 of "C Programming language" 2nd Edition (SPANISH: I do not know the exact page number in the English version) (187) At the very bottom of page 187, there is this line: "v To print verbose information when packing or unpacking archives" And then the option is repeated at the end of the same list with a more elaborate description, towards the top of page 188: "v To make tar show the files it is archiving or restoring--it is good practice to use this so that you can see what actually happens (unless, of course, you are writing shell scripts)" The top list "v" option should probably be combined with the bottom. (188) In the fourth paragraph; The option v of tar is explained twice. {192} In the gunzip example in the middle of the page; Giving the option -9 to gunzip does not make sense. It is not used and silently ignored. [192] In the middle of the page, the example command: gunzip -9c tarfile.tar.gz | tar xvf - I've never needed to use 9 with gunzip. I think it's a no-op here. The 9 is useful when *creating* the archive in the first place -- to tell gzip to "try harder". [194] At the top, it reads: tar cf - . | (cd ../to-stuff; tar xvf -) I'd recommend writing that as: tar cf - . | (cd ../to-stuff && tar xvf -) so that, if the "cd ../to-stuff" fails -- for instance, if the user makes a mistake in the pathname -- the "tar xvf" will not run (and try to overwrite the files in the "from-stuff" directory). (204) In the fifth paragraph: "(RPM package..." should be: "(RPM packages..." In the sixth paragraph 6: "choose whatever you like better:" should be: "choose whichever you like better" since there are only two possible options. (207) In the first paragraph of the "Building a New Kernel" section, the run-on sentence beginning with "This reduces the amount of memory used by...." should be two sentences. (222) In last sentence of the third paragraph: "parametes" should be: "parameters" (222) In the second paragraph: "a number if" should be: "a number of" (226) In the first, it reads "These tools allow you to take backups more or less 'by hand'". In the second paragraph, it reads "The simplest means of taking a backup is to use tar to archive...". Shouldn't these be "make" instead of "take"? I've never heard of "taking" a backup. (230) In the last sentence before section "Incremental Backups"; There is an extra "are" in that sentence. It should read: "Among those programs are the freely available taper and tob, as well..." [230] The first example shows \! in a command line, and the second paragraph after that explains why the \ is used. I don't think I've ever used a shell that treats a single ! by itself as a history reference or anything else... so a ! by itself shouldn't need quoting. I just tried this on the zsh, tcsh, and bash on my Red Hat 6.2 system, and it was true for all those shells. Same thing on page 232, second example. (249) In the fourth paragraph; The #9 video card is either using the S3 chipset or the XF86_S3 server, but not the XF86_S3 chipset. (249) In the last but one paragraph, explanation of mx; The last sentence should read: "...set a value slightly smaller than the expected maximum space available on a disk." Note the "maximum". {254} In the second if statement it should read: printf "$SENDEOF" instead of: printf "" Otherwise it doesn't make sense to set SENDEOF to a specific value. (258) In the fifth paragraph; The file's name is ".apsfilterrc". Note the "c" at the end. (280) 2nd paragraph from bottom.; "--after pressing Ecs--" should read "--after pressing Esc--" (281) There is an unwanted period after the word "screen". {293} In the last paragraph in "Searching and Replacing"; Entering "M-x string" will not replace a string. Probably the name of a replace function should be entered here, though I don't know which. The same thing with "M-x regexp". To replace all occurences of a string, I use the replace function and enter "!" at the first query. {301} In line 14 of LaTeX example; With LaTeX2e you should use "\emph{text}" instead of "{\em text}". Also, in the last but one paragraph on page 302, you should use "\textbf{text}" instead of "{\bf text}" and "\texttt{text}" instead of "{\tt text}". [304] Is the name of the program "eps" correct? Shouldn't it be "dvieps"? I am not aware of either program, so you maybe right. [315] In the end of section "Formatting Texinfo"; Maybe it would be good to mention the script "texi2dvi" which takes care of the necessary "tex" and "texindex" calls. (339) In the last but second paragraph; At least in SuSE, the text-based program is called "xf86config", not "ConfigXF86". (351) The second paragraph should start "In the Device section..". No "s". (364) The fourth paragraph should start "The pager, in the bottom-right corner..." (371) In the third paragraph, last sentence: "In addition, almost all KDE application are..." should read: "In addition, almost all KDE applications are..." "Application" needs to be plural. (385) In the third paragraph; There is an extra "g:" in the line before the "pattern ...unmatched"line. (393) The first sentence in the fourht paragraph should end: "...when you started xman." Not: "xterm". (398) In the last but third line: "in bug" should read: "works around a bug in..." (418) In the second paragraph, ABI should be API. (444) While trying to make a perl script run, I encountered some problems. For some odd reason, the version I have of last padded its output with spaces at the end of each line. This made the regexp /^(\S*)\s*.*\((.*):(.*)\)$/ fail, since the final $ sign didn't allow for this padding. My solution was just to remove the $ sign, so that the regexp looks like /^(\S*)\s*.*\((.*):(.*)\)/ instead. This fixed the problem as far as I can see. (494) The paragraph on the second-to-last line is almost a copy of the preceding one. [513] 2nd to last line; indent -troff importrtf.c | groff -mindent Is this example correct? The indent man page speaks nothing of the "troff" option. My version of indent (GNU indent 2.2.5) returns the following error: indent: unknown option "-troff" [515] The second paragraph reads: "Linux NET-2 also supports..." Is it right?, Why not NET-4? (588) The eighth paragraph currently reads: "The ISP accepts your mail from then Internet" It should read: "The ISP accepts your mail from the Internet" ?535? The first paragraph talks about how Winmodems use the host CPU to convert digital signals to analog signals. That description is generally what we term a SoftModem. If we take a modem and divide it up into Controller, DSP and Phone Interface we can then make the Winmodem definition a little more clear. A "standard" modem would include the Controller, DSP and Phone interface whereas a Winmodem would not include the Controller. For a Winmodem, the Controller based functions are moved to the host CPU. For a SoftModem, both the Controller and DSP functions have been offloaded to the host CPU. Using these definitions, a WinModem takes a very small portion of the available CPU bandwidth and then only when you are online. A SoftModem would require significant CPU bandwidth to handle the increased load of the Digital Signal Processing. {597} 5th paragraph; In this section, it says: "If smail is configured correctly on your system, there is a line in the /etc/services file that says:" blah...blah...blah... ... below that is the error: "There is also a line in the /etc/services file that says: SMTP stream tcp nowait root /usr/sbin/tcpd /usr/sbin/rSMTP -bs" This line is in the /etc/inetd.conf file and not the /etc/services file. (626) In paragraph 3, line 3, word 9: "manufacturer" should be: "manufacture" since it's used as a verb in present tense. (708) Second Column; The index entry for Devices is wrong. It gives p262 instead of 179 (at least 179 seems a better reference). I've had other problems with the index but this is the most recent.