RE: tcsh hangs after updating to cygwin 1.5.7-1. Expires with Out of Memory
I have experienced exactly the same problem! On two different machines running XP. I'm running XP *and* I'm using tcsh on a regular basis and I'm not able to reproduce that effect. Well, I know of at least one way to cause this problem! My cygwin installation was running fine for over a year on an XP machine without any problems. Today, out of nowhere, it starts giving me the Out of memory error whenever I try to run tcsh. Looking in that directory revealed that the .history file was huge. This happened to me today also, on a 128 MB Windows 2000 SP4 machine. All was well until yesterday, when tcsh failed a couple of times, but it seemed to be inconsistent. Today tcsh crashed every time I invoked it without -f. I discovered that my .history file was 53 MB in size! This is with history 1024 savehist(1024 merge) in my .login. wc .history 0 3316299 53060768 .history Examination of the .history file shows it to be garbage. It contains nothing but a sea of repeated copies of $prompt (when in $HOME, post variable substitution). Could there be a problem with the history save or merge code? cygwin 1.5.12-1 tcsh 6.13.00-2 Cheers Michael -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
The tcsh manual vs tcsh initialization; fix for slow startup
1. The tcsh manual states: onintr is ignored if the shell is running detached and in system startup files (see FILES), where interrupts are disabled anyway. But: grep onintr /etc/csh.cshrc onintr - onintr Is the manual incorrect or is the initialization file conceptually flawed? 2. The tcsh manual also states: -f The shell ignores ~/.tcshrc, and thus starts faster. But -f also causes the shell to ignore /etc/csh.* and .login. Is the specification or the implementation at fault? 3. On a Pentium I-level computer, tcsh takes an aggravatingly long time to process the initialization files. I found that commenting out the lines that call source /etc/profile.d/complete.tcsh in /etc/csh.cshrc (or deleting the file /etc/profile.d/complete.tcsh) sped up this process dramatically. Cheers Michael -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
vi W10: Warning: Changing a readonly file on large read-write file
When attempting to edit a *writeable* file of zero lines but 53 MB in size (no newlines), vi appears to hang. Ctrl-C is followed by the response W10: Warning: Changing a readonly file following which it appears that one is editing a new file. Quitting without writing succeeds. cygwin 1.5.12-1 vim 6.3-1 Cheers Michael -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: ttmkfdir no longer needed
Due to this I am going to pull the link to ttmkfdir from our 'Ported Software' page. Perhaps someone would like to write a postinstall script that creates symlinks to the Fonts folder for Windows under /usr/X11R6/lib/X11/fonts/windows, One note: In the thread proposal for using windows truetype fonts I suggested to use the /usr/X11R6/lib/X11/fonts/truetype dir for truetype fonts, because this is the default for qt and kde. Are there any objectivities using this instead of /usr/X11R6/lib/X11/fonts/windows ? I object. :-) Shouldn't we be consistent with established Unix practice and put any local fonts in /usr/X11R6/lib/X11/fonts/local? Cheers Michael
Re: X/Cygwin icon proposal
Hi folks The new icon with alpha looks quite bad on Windows 2000 and earlier systems, with a thick white border -- see the attachment PNG. The tips of the X on some of the other versions of the icon also look slightly blunt (minor quibble), and the top and bottom rows are lost at 16x16. Presumably we shouldn't be setting the default to something that uses a feature unsupported by the majority of systems out there! Alpha is nice, but it is a new, optional feature; we still need to support low-colour desktops by default. Using the CVS icon as a starting point, I created a new icon using an outlined white square as the background. It is rendered at 16x16, 24x24 and 32x32 sizes, each for monochrome, 16 and 256 colours. It has the correct proportions of the thick and thin lines, properly anti-aliased and quantised. It's even rotationally invariant! :-) I have attached two files: a comparison of the icons in Overview.png, and the improved icon in Improved.ico. Cheers Michael attachment: Overview.pngattachment: Improved.ico
Re: X/Cygwin icon proposal
Hi folks This is a little long, but I have combined several points rather than bombard the list with multiple messages. Thanks for your patience. :-) ___ Nahor wrote: What is New Alpha? I sent a few on the mailing list. Was it icon_test9 (attached again here)? This one has 24bit icons, hopefully the prefered format on systems not supporting the alpha channel (crossing fingers). The file you attached has issues under Windows 2000 (does not show icon picture in Explorer). But it was the latest version of your icon at the time. Original was from X.exe. I don't care about the majority of the systems out there. I care about the majority of the system using Cygwin/XFree. And that can be very different. The majority of Cygwin users are not typical gamers. They are more likely to be similar in profile to hackers such as the Linux or BSD folks -- and those are well known to be frequently using older generations of hardware (and hence software). Industry is still receiving PC's preloaded with Windows 2000 -- which will be supported until 2007! Remember, 2 years after Windows 2000 came on the scene, IT organisations were still DEPLOYING Windonts NT! Two years after the debut of Windows 2000, the number of *new* Windows NT server licenses matched the number of Windows 2000 licenses. And that's just the new liceneses -- just think of the huge installed base. And as for desktops, by 2002 75% of desktops in industry were Windows 9x! Uh? I don't get your point. I personally don't buy a machine just to run unix. I use it to do other stuff (mostly compilation) that do make use of CPU power. So I have a recent machine, so I have XP. I assume that quit a dew (most?) geeks using Cygwin/XFree would be in the same case. But it's just a guess. I'm sure that many Cygwin users have brand new machines. But I am equally sure that many more have older systems. The baseline for support today must clearly be pre-XP systems. The other thing, IMHO, is that the alpha icon on non-alpha system, while not the best icon that can be on such system, is not completely ugly either. Frankly, I disagree. I wouldn't have put in the effort of designed a new icon set if I thought it were OK! :-) So between an icon that looks best on recent machines but not as good on older ones and one that looks best on older machines but not as good as it can be on recent ones, I prefer to think future/progress/whatever and take the first. The problem is that the rest of the software world disagrees. It is standard software practice to support as many platforms as possible with the *default* install, even if it is not as flashy as the others. Sure, you can have an option to enable alpha -- but don't make it the default. Do you really want someone installing X/Cygwin for the first time to be confronted with an amateurish-looking icon? That was my first impression. From a technical perspective, aesthetics are secondary -- but in the real world, first impressions last. Between the CVS and your improved, I prefer the one in CVS. The thin lines is acutally too thin in 16x16, the line is too blury on yours, the white background seems to wash over the black line. You originally said that my original monochrome X was ugly due to blocky edges, but that is exactly the problem with your icon on Windows 2000 systems! :-) The lines in Improved.ico (why the quotes?) are actually in exactly the correct anti-aliased proportion to represent the X logo within the limits of the bitmap. The CVS icon is incorrectly proportioned. I do not argue that you personally prefer your version. That is of course a subjective choice! However, Improved.ico has the proportions of the original X vector logo; you may prefer something that looks different, but that then is something different, not a faithful rendering of the X logo. ___ Ago wrote: If you can build ico files with both alpha and non-alpha icons why not include your version with alpha channel and for non-alpha either the boxed (which I liked) or a plain two-color variant. cygwin is unix. unix is simple (shell and stuff) and this is the opposite of the bubble-gum os WinXP with alpha channel. Hear hear! :-) ___ Earle wrote: - Default to a safe setting for anything that's not critical. - You'll save TONS of user grief, and by extension, your own. Agreed. Looking really nasty under OSs earlier than XP is a bug I'd say. Plus it's probably rechnically an invalid icon resource under those OSes so you may wnd up causing a boom (hey, under 95 or 98 it doesn't take much to crash the system!) Strongly agreed. You've not very familiar with how a shortcut is made, are you? Make the 1st icon in the file the clean X-in-a-white-box that's been there for some
X/Cygwin icon proposal
Hi folks I've attached my suggestion for a X/Cygwin program icon. It does not feature the speckled white border problem exhibited in the Cygwin/X screenshots at http://x.cygwin.com/. It's also efficient, at 518 bytes. :-) There is another new X icon in X/Cygwin CVS that features a white background in a black square, but do we want an icon with a background? It seems more normal to me to have a transparent background. (The CVS icon has other issues like a blue 24x24 bitmap and is asymmetrical (not rotationally invariant at 180 degrees, for the OCD.) :-) What do you think? Cheers Michael attachment: X.ico
xcalc glitches
Hi xcalc occasionally doesn't see the X server: xcalc Error: Can't open display: 127.0.0.1:0.0 It gives a warning when it opens correctly: xcalc Warning: No type converter registered for 'String' to 'Bitmap' conversion. Why is this? This is a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4. Thanks Michael
XFree86 Start Menu shortcuts broken -- run.exe at falt?
Hi My Cygwin root is C:\Program Files\Cygwin. The space appears to cause the automatically-created Cygwin-XFree86 shortcuts to fail. Run.exe pops up a dialog saying that it can't find Files\Cygwin\usr\X11R6\bin\run.exe xclock -display 127.0.0.1:0.0 for example. It occurs even if the shortcut is invoked as C:\Progra~1\Cygwin\usr\bin\ This is on a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4, with XFree86-bin-4.3.0-6. Cheers Michael [Previously posted in error to wrong list, doh.] cygcheck (abbreviated).out Description: Binary data
Cygwin 1.5.5-1: whois misbehaviour
Hi By my reading of the man page, the Cygwin whois appears to ignore the --version and --verbose options that are listed in the man page and/or built-in help. It also appears to me that it ignores the WHOIS_SERVER environment variable: setenv WHOIS_SERVER whois.stanford.edu whois mbax No whois server is known for this kind of object. whois [EMAIL PROTECTED] NO MATCH This is Cygwin DLL 1.5.5-1 running on Windows 2000 SP4. Cheers Michael cygcheck (abbreviated).out Description: Binary data -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
rxvt-2.7.10-3 man page errors
Hi There's a curious glitch in the rxvt man page: igl-freehand:~ man rxvt /tmp/tmp /usr/bin/tbl:standard input:310: `.' not last character on lin /usr/bin/tbl:standard input:310: giving up on this table /usr/bin/tbl:standard input:776: `.' not last character on lin /usr/bin/tbl:standard input:776: giving up on this table This is on a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4. Cheers Michael -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Cygwin 1.5.5-1: cp to network 100% slower than cmd's copy
Hi I am aware than Cygwin network operations are slower than native Windows operations, but is a factor of 2-3 reasonable? time cp 15MBfile /cygdrive/i/1 0.18u 2.04s 0:19.96 11.1% time cp 15MBfile //igl/home/2 0.21u 1.77s 0:20.20 9.8% time cmd /c copy 15MBfile i:\3 1 file(s) copied. 0.02u 0.02s 0:09.89 0.4% On longer files the operation has been closer to 3 times slower. Why is the CPU usage so much higher (~20x?) This is on a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4. Cheers Michael cygcheck (abbreviated).out Description: Binary data -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
XFree86 Start Menu shortcuts broken -- run.exe at falt?
Hi My Cygwin root is C:\Program Files\Cygwin. The space appears to cause the automatically-created Cygwin-XFree86 shortcuts to fail. Run.exe pops up a dialog saying that it can't find Files\Cygwin\usr\X11R6\bin\run.exe xclock -display 127.0.0.1:0.0 for example. It occurs even if the shortcut is invoked as C:\Progra~1\Cygwin\usr\bin\ This is on a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4, with XFree86-bin-4.3.0-6. Cheers Michael cygcheck (abbreviated).out Description: Binary data -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
XFree86 fails to see/render TrueType fonts
Hi XFree86 fails to render TTF fonts on my computer. This is true for the bundled fonts/TTF/* as well as the Windows fonts. There are two problems: 1. XFree86 does not appear to include the X11 fonts/TTF directory in the font path by default. This can be shown by removing the Luxi fonts in the fonts/Type1 directory and rehashing. The server should use the TTF versions instead, but it does not see these fonts at all. This can be fixed using xset fp+, but surely the TTF font directory should be included in the path by default? 2. After xset fp+ and rehash, the TrueType fonts are listed in e.g. xfontsel. However, they are not rendered (or are rendered blank)! The same is true for my custom-installed TrueType fonts. They worked in XFree86 3.x, however, using the same installation procedure. Other fonts are fine, but all TrueType fonts are effectively invisible. This is on a fresh installation of Cygwin DLL 1.5.5-1 running on Windows 2000 SP4, with XFree86-bin-4.3.0-6. Cheers Michael cygcheck (abbreviated).out Description: Binary data -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: fvwm-2.4.7-2 install missing /usr/local/etc/fvwm/system.fvwm2rc?
In followup to my earlier message:
INSTALL.fvwm from the fvwm source tarfile (fvwm.org) suggests
copying system.fvwm2rc to /usr/local/etc/fvwm/system.fvwm2rc
to provide a system default. This file does not appear to be
present in the Cygwin package, but it seems that installing
this file to the correct place would be the most transparent
way to streamline fvwm setup.
The Cygwin version of fvwm2 appears to be configured slightly
differently from the default. The appropriate directories for
system.fvwm2rc are
${datadir}/fvwm -- /usr/X11R6/share/fvwm
and
${sysconfdir} -- /etc (why not /usr/local/etc or even
/etc/X11?)
As a possible alternative to supplying this default configuration file,
what about installing fwwm-icons along with a configuration
pre-generated by FvwmScript-Setup95?
Cheers
Michael
Error reports and queries
Hi 1. I installed OpenSSH (and therefore OpenSSL) under Cygwin. My shell is tcsh. When my shell starts up, it gives an error because the last line of /etc/profile.d/openssl.csh has an error in the last line. This should be endif not fi. 2. If tcsh is invoked from the command line (for example), the MANPATH variable gains a duplicate openssl entry. This is duplicated again each time tcsh is nested. 3. The default configuration for tcsh is quite different to most systems' default tcsh configuration: noglob is set, command-line typo-correction is enabled, etc. Any chance that this could be changed to a setup that is more typical on Unix systems? 4. I am a power user, not an administrator. But inside my Cygwin shell I can read, edit and write arbitrary files in /, /etc, /home/Administrator etc. This is true both with and without ntsec. Is this proper behaviour, to follow the NTFS permissions and disregard the Posix permissions? 5. It appears that ntsec allows *listing* permissions, but the real permissions are unchanged with or without ntsec. Is this right? 6. What does ntea actually do? The guide isn't very clear whether this is a good thing, or exactly what it does: if it simply allows tracking chmod permissions, how do these interact with the ntsec permissions? Thank you Michael -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
