View previous topic :: View next topic |
Author |
Message |
kubark42 n00b
Joined: 15 Feb 2004 Posts: 44
|
Posted: Sun Aug 08, 2004 11:58 pm Post subject: cAseLEss gEntOO |
|
|
wElcoME tO oUr efFoRt To mAkE a morE uSEaBle gEntOO.
We're embarking on ye ol' grand adventure, to go where no nerd has gone before. As inspired by this thread and Rinnan, we shall make a case insensitive gEntOO system. The point of this is to bring the gEntOO experience one step closer to what we think the average user expects from his computer. Agree with us, or don't, all input is appreciated. However, I do ask that if you feel philosophically inclined, that you contribute to the above-mentioned thread in order to keep this thread clear for technical discussions.
To wit:
rinnan wrote: | Elsewhere I've been warned that the "cvs" ebuild might break, so I'll start working on an ebuild patch for it when I encounter that . I will solve everything in the Gentoo way -- my approach will be thusly -- if a patch can be made which will cause it to work correctly on either a case insensitive or case sensitive file system (rather than just a case-sensitive one) I will submit it as a bug to be fixed. If it by some necessity changes the behavior in any way on a case-sensitive system, I will make it subject to a new USE flag (in my patches) called "case-insensitive" (rather long, but "disablekernelsupport" is longer, so hey, shouldn't break anything). This same switch will also change the default config files (for example baselayout or bash) such that case-insensitivity is activated for every tool (such as bash) which has it as an option. |
I couldn't say it any better.
So, any takers? |
|
Back to top |
|
|
Cinder6 l33t
Joined: 05 Aug 2004 Posts: 767 Location: California
|
Posted: Mon Aug 09, 2004 3:04 am Post subject: |
|
|
I have to agree with you, after reading the posts on the first topic.
Lol, when I first read the post, I thought you wanted to make Gentoo display random cases for letters (e.g. you type in "monkey" and it might give you "MOnKey"). _________________ Knowledge is power.
Power corrupts.
Study hard.
Be evil.
Ugly Overload |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 3:35 am Post subject: Thank you kubark42 |
|
|
Okay, good. I've done a little bit of coding to start with. I've created two new very small c programs that take standard input, uppercase it, or lowercase it. I know this can be done with other tools but I thought I'd start here. Using "uniq", "sort", and "slocate", and my new tools, "toupper" and "tolower", I've discovered the following things about my filesystem. My filesystem has:
* 610995 files in it, and
* 129 files which have at least one mate, in the same directory, which differs in filename only by case.
The vast majority of these are include files under
Quote: | /usr/include/fltk-1.1/fl/ |
At least one important other one is this one:
Quote: | linux-2.6.7-mm1/arch/i386/kernel/vmlinux.lds.s |
which has the mate: Quote: | linux-2.6.7-mm1/arch/i386/kernel/vmlinux.lds.S |
capital-S is an ld script used to make the kernel, small-s claims to be the same thing, but differs (seems simplified) and is generated during the build.
So it looks like it's possible to run into trouble compiling the kernel on a case-insensitive file system. I'm having trouble reinstalling my whole system on a new partition -- it comes down to the fact that reiserfs is not resizeable. Working on solving that now. In the meanwhile, I will try compiling the Linux kernel on jfs, created with the "-O" option (case-insensitive) and see where it breaks and work on a patch.
The code to "toupper.c" and "tolower.c" follows, I will just show "tolower.c", I bet you can figure out how "toupper.c" differs yourself.
Code: | #include <stdio.h>
#include <ctype.h>
int
main(int argc, char **argv)
{
char somecase, lowercase;
while(1)
{
somecase = getchar();
if (somecase == EOF) break;
lowercase = tolower(somecase);
putchar(lowercase);
}
exit(0);
}
|
I'll tackle installing Gentoo from scratch after I work on fixing the kernel problem, but if someone else wants to tackle that one (someone who has unpartitioned space on their harddrive for example) that would work out well. Otherwise I gotta do a huge backup/restore cycle.
The kernel problem is interesting because it's slightly more difficult to fix within an ebuild -- although I suppose in the end it's going to have to be something like an extra patch to the Makefiles.
rinnan |
|
Back to top |
|
|
PowerFactor Veteran
Joined: 30 Jan 2003 Posts: 1693 Location: out of it
|
Posted: Mon Aug 09, 2004 3:48 am Post subject: Re: Thank you kubark42 |
|
|
rinnan wrote: |
At least one important other one is this one:
Quote: | linux-2.6.7-mm1/arch/i386/kernel/vmlinux.lds.s |
which has the mate: Quote: | linux-2.6.7-mm1/arch/i386/kernel/vmlinux.lds.S |
|
There's a post about that here, unfotrunately the link to the patch itself is dead. But I think the explanation of the fix is pretty clear and simple. |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 4:04 am Post subject: Luckily... |
|
|
Luckily the link is not dead, merely malformed. The patch is bigger than what I would have expected. Here is the correct link:
http://mirror.vtx.ch/lfs/patches/downloads/linux/linux-2.6.7-build_on_case_insensitive_fs-1.patch
EDIT building now... untarred Linux kernel and applied patch to kernel while it was on a case-insensitive file system.
EDIT2: builds without error on jfs -O
rinnan
Last edited by rinnan on Mon Aug 09, 2004 5:12 am; edited 3 times in total |
|
Back to top |
|
|
PowerFactor Veteran
Joined: 30 Jan 2003 Posts: 1693 Location: out of it
|
Posted: Mon Aug 09, 2004 4:11 am Post subject: Re: Luckily... |
|
|
rinnan wrote: | Luckily the link is not dead, merely malformed. |
Heh, my bad, didn't notice the space.
That's bigger than I would have thought too, but then I don't know much about the kernel build system. |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 5:19 am Post subject: Re: Luckily... |
|
|
PowerFactor wrote: | rinnan wrote: | Luckily the link is not dead, merely malformed. |
Heh, my bad, didn't notice the space.
That's bigger than I would have thought too, but then I don't know much about the kernel build system. |
I've examined the patch more carefully. The key is that it patches the filenames in every architecture, and there are quite a few.
EDIT: My next step, to do tomorrow, is to modify the existing patch to work against gentoo-dev-sources, stuck to the end. This is my first big patch job for the kernel, wish me luck.
rinnan |
|
Back to top |
|
|
blaster999 l33t
Joined: 09 May 2004 Posts: 902 Location: Between keyboard and chair
|
Posted: Mon Aug 09, 2004 6:08 am Post subject: |
|
|
In some unixes (unices?) filename.c means it is written in 'plain' C, while file.C means it is in C++. Frankly speaking, I don't think that making commands like "cAT /Usr/srC/linuX/MakefILE" valid is what a casual user needs. Make a layer that ignores the case (e.g. a specially formed bashrc or a patch) if you need. In point'n'click environment (offered by most WM and DE) you don't have to care about the case.
If you really hate case-sensitivness, access files by their inodes - inodes are case-insensitive _________________ 60s: sex, drugs, rock'n'roll
90s: sux, bugs, drag'n'drop
---
Some multimedia keys refuse to work? See my mini-howto:
https://forums.gentoo.org/viewtopic.php?p=1896734#1896734 |
|
Back to top |
|
|
kallamej Administrator
Joined: 27 Jun 2003 Posts: 4975 Location: Gothenburg, Sweden
|
Posted: Mon Aug 09, 2004 6:21 am Post subject: |
|
|
Moved from Other Things Gentoo as it is not a support request. _________________ Please read our FAQ Forum, it answers many of your questions.
irc: #gentoo-forums on irc.libera.chat |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 6:37 am Post subject: |
|
|
blaster999 wrote: | In some unixes (unices?) filename.c means it is written in 'plain' C, while file.C means it is in C++. Frankly speaking, I don't think that making commands like "cAT /Usr/srC/linuX/MakefILE" valid is what a casual user needs. Make a layer that ignores the case (e.g. a specially formed bashrc or a patch) if you need. In point'n'click environment (offered by most WM and DE) you don't have to care about the case.
If you really hate case-sensitivness, access files by their inodes - inodes are case-insensitive |
While I agree that the user probably doesn't need the command "cAT /Usr..." to work, I do believe that what the user does need is consistency and not to have to worry if they are in "case sensitive mode" or "case insensitive mode" or if a certain process needs case sensitivity and another doesn't -- therefore I must disagree with any layering approach. I believe that that would make the final product even more complicated and indecipherable rather than less, which is the goal of this project. Though this all borders on the philisophical -- there's another thread for that.
I'm not worried about ".C" and ".c" since that's a rarely used convention, and of course doesn't actually matter anyway as long as there isn't a ".c" and a ".C" file that are otherwise exactly the same (or differ only by filename case). If we run into it, we'll have to patch it.
I will add this, looking carefully over the files which would create problems moving over to case insensitivity, I notice that almost all are related to building software, such as post-process files for builds, source files (particularly include files) and so on. There are some "man" pages which have this problem too, but they are named after API calls which exhibit the need for case-sensitivity. (note that I am not calling for an end to case sensitivity for everything, specifically not for API's or passwords). The man pages might be tricky, if they depend on the filenames as a sort of database, then we'll have to patch "man" and "makewhatis" as well as any other tool that sorts through and displays man pages. (luckily most such tools are wrappers for other tools). There are also a few "problem children" in "/usr/share/terminfo" so I'll have to explore and sort that out as well.
I have confirmed that running executables by calling them with a case that differs from what their natural name is (for example, "GREP what file1 file2 file3") works just fine if the binary is on a case-insensitive file system. It even refers to itself as "GREP" for example:
Quote: | [amigatoo:erik:/mnt/jfs] grep
Usage: grep [OPTION]... PATTERN [FILE]...
Try `grep --help' for more information.
[amigatoo:erik:/mnt/jfs] GREP
Usage: GREP [OPTION]... PATTERN [FILE]...
Try `GREP --help' for more information.
[amigatoo:erik:/mnt/jfs] |
This is without any case-insensitive options turned on in bash. This is what I expected, but I wanted to check.
rinnan |
|
Back to top |
|
|
placeholder Advocate
Joined: 07 Feb 2004 Posts: 2500
|
Posted: Mon Aug 09, 2004 6:53 am Post subject: |
|
|
Case-sensitivity when used in conjunction with tab-completion can come quiet in handy. Taking it out all seems stupid to me, but I'm not going to be a dick about it so if you really want to make a work-around then as long as I don't _have_ to use it then it's fine with me because it's not my time you're using. |
|
Back to top |
|
|
Rafje Tux's lil' helper
Joined: 19 May 2004 Posts: 109 Location: .be
|
Posted: Mon Aug 09, 2004 7:26 am Post subject: |
|
|
If you log into a console using your login and password in all uppercase, then your whole console session will be in uppercase. This comes standard out of the box, no need to patch anything.
If I remember correctly, this is a historic feature left over from medieval times when some terminals simply didn't have lowercase.
(Just so that you know how I feel about case-insensitivity: I do not want to go back to medieval times, thank you. You can pry my case-sensitivity from my cold dead fingers )
Cheers |
|
Back to top |
|
|
blaster999 l33t
Joined: 09 May 2004 Posts: 902 Location: Between keyboard and chair
|
Posted: Mon Aug 09, 2004 7:41 am Post subject: |
|
|
Well, I can't login in all uppercase When I type my login all upper, the prompt for password is also uppercase (like this: "PASSWORD:"), but it rejects password either upper- or lowercase.
I am sure most people who want cASe-inSensITIve gENtoO want simply case-insensitive tab-completion. _________________ 60s: sex, drugs, rock'n'roll
90s: sux, bugs, drag'n'drop
---
Some multimedia keys refuse to work? See my mini-howto:
https://forums.gentoo.org/viewtopic.php?p=1896734#1896734 |
|
Back to top |
|
|
kubark42 n00b
Joined: 15 Feb 2004 Posts: 44
|
Posted: Mon Aug 09, 2004 9:18 am Post subject: |
|
|
Well, it looks as if portage were broken straight off the bat. I get this error when installing gentoo from 2004.2 on a new case insensitive jfs partition, made using jfs -O:
Code: |
#>emerge system
/usr/lib/python2.3/fcntl.py:7: DepricationWarning: the FCNTL module is deprecated; please use fcntl
DeprecationWarning)
Traceback (most recent call last):
File /usr/bin/emerge, line 2517, in ?
If not mydepgraph.xcreate(myaction):
File /usr/bin/emerge, line 1109, in xcreate
myeb=portage.portdb.xmatch(bestmatch-visible:, mydep)
File /usr/lib/portage/pym/portage.py, line 4699, in xmatch
myval=best(self.xmatch(match-visible,None,mydep,mykey))
File /usr/lib/portage/pym/portage.py, line 4710, in xmatch
myval=match_from_list(mydep,self.xmatch(list-visible,None,mydep,mykey))
File /usr/lib/portage/pym/portage.py, line 4696, in xmatch
myval=self.gvisible(self.visible(self.cp_list(mykey)))
File /usr/lib/portage/pym/portage.py, line 4791, in gvisible
myaux=db[/][porttree].dbapi.aux_get(mycpv, [KEYWORDS])
File /usr/lib/portage/pym/portage.py, line 4543, in aux_get
mylock = lockfile(mydbkey,unlinkfile=1)
File /usr/lib/portage/pym/portage.py, line 89, in lockfile
fcntl.flock(myfd,fcntl.LOCK_EX)
AttributeError: 'module' object has no attribute 'flock'
#>
|
Wiser minds can divine the meanings of this archaic text. The obvious guess is that it's an overlapping of two functions, FCNTL and fcntl. Should I file a bug report? |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 10:35 am Post subject: |
|
|
kubark42 wrote: | Well, it looks as if portage were broken straight off the bat. I get this error when installing gentoo from 2004.2 on a new case insensitive jfs partition, made using jfs -O:
Code: |
#>emerge system
/usr/lib/python2.3/fcntl.py:7: DepricationWarning: the FCNTL module is deprecated; please use fcntl
DeprecationWarning)
Traceback (most recent call last):
File "/usr/bin/emerge", line 2517, in ?
If not mydepgraph.xcreate(myaction):
File "/usr/bin/emerge", line 1109, in xcreate
myeb=portage.portdb.xmatch("bestmatch-visible:, mydep)
File "/usr/lib/portage/pym/portage.py", line 4699, in xmatch
myval=best(self.xmatch("match-visible",None,mydep,mykey))
File "/usr/lib/portage/pym/portage.py", line 4710, in xmatch
myval=match_from_list(mydep,self.xmatch("list-visible",None,mydep,mykey))
File "/usr/lib/portage/pym/portage.py", line 4696, in xmatch
myval=self.gvisible(self.visible(self.cp_list(mykey)))
File "/usr/lib/portage/pym/portage.py", line 4791, in gvisible
myaux=db["/"]["porttree"].dbapi.aux_get(mycpv, ["KEYWORDS"])
File "/usr/lib/portage/pym/portage.py", line 4543, in aux_get
mylock = lockfile(mydbkey,unlinkfile=1)
File "/usr/lib/portage/pym/portage.py", line 89, in lockfile
fcntl.flock(myfd,fcntl.LOCK_EX)
AttributeError: 'module' object has no attribute 'flock'
#>
|
Wiser minds can divine the meanings of this archaic text. The obvious guess is that it's an overlapping of two functions, FCNTL and fcntl. Should I file a bug report? |
Probably but we have to track it first. I can't really look at it without reinstalling...
Can you try an experiment? Build it on a case-sensitive file system, such as reiser, and then just try copying it to a case insensitive one and see if it can just run for starters.
After that we can chase the bugs -- I want to be able to file bug reports with patches. Remember, although I believe that many will benefit from case-insensitivity in the file system, there's enough voices against it that we're (whoever "we" end up being, as a group) will have to fix all the bugs and do all the work -- which means we will have to submit patches, not just bugfixes.
I've converted my swap partition to a jfs -O system. I will try to install a new Gentoo onto it and try to duplicate your error.
I think your idea about FCNTL and fcntl is right on track, although as functions they should still work case-sensitively, it's just as filenames, somewhere in there, it's breaking.
rinnan |
|
Back to top |
|
|
Titeuf l33t
Joined: 19 Jun 2004 Posts: 759 Location: Middelkerke, Belgium
|
Posted: Mon Aug 09, 2004 11:03 am Post subject: |
|
|
blaster999 wrote: | Well, I can't login in all uppercase When I type my login all upper, the prompt for password is also uppercase (like this: "PASSWORD:"), but it rejects password either upper- or lowercase.
I am sure most people who want cASe-inSensITIve gENtoO want simply case-insensitive tab-completion. |
Once it's in uppercase you have to fill in again your username and password in normal case.
But lowercase letters are shown as uppercase |
|
Back to top |
|
|
Frodg l33t
Joined: 11 Feb 2004 Posts: 761
|
|
Back to top |
|
|
Titeuf l33t
Joined: 19 Jun 2004 Posts: 759 Location: Middelkerke, Belgium
|
Posted: Mon Aug 09, 2004 12:22 pm Post subject: |
|
|
Frodg wrote: | https://forums.gentoo.org/viewtopic.php?t=208052&highlight=hands
|
This can be a fun way to waste your spare time |
|
Back to top |
|
|
castorilo Apprentice
Joined: 25 Dec 2002 Posts: 157
|
Posted: Mon Aug 09, 2004 2:57 pm Post subject: Re: Thank you kubark42 |
|
|
rinnan wrote: | Okay, good. I've done a little bit of coding to start with. I've created two new very small c programs that take standard input, uppercase it, or lowercase it. |
Here is an alternative:
To convert to uppercase:
Code: |
echo "TeStInG" | tr "[:lower:]" "[:upper:]"
|
To lowercase:
Code: |
echo "TeStInG" | tr "[:upper:]" "[:lower:]"
|
|
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Mon Aug 09, 2004 11:54 pm Post subject: |
|
|
Pwnz3r wrote: | Case-sensitivity when used in conjunction with tab-completion can come quiet in handy. Taking it out all seems stupid to me, but I'm not going to be a dick about it so if you really want to make a work-around then as long as I don't _have_ to use it then it's fine with me because it's not my time you're using. |
Pwnz3r -- first, don't worry! It's not like a case-insensitive Gentoo is going to sneak up on you and bite you. You _really_ have to make an appointment, so to speak. The purpose of this project is to make it possible, and, hopefully, easy, which at this point it is not. Ideally we want to make it as simple as:
#1: Choosing a case-insensitive file system during installation.
#2: Setting the "case-insensitive" USE flag. (it would be off by default)
#3: And building your Gentoo system. (profit!)
This will require quite a few changes to make it possible, even to the basic installation, probably, because some processes (including the Gentoo portage scripts) break on case-insensitive file systems. It's worth noting that this involves choosing alternative filenames for some things, mostly. For example, take the kernel patch we talked about above. You could patch the kernel with that patch on your current case-sensitive file system and it would build exactly the same, no difference. The same will be true for all of the changes, ideally. If we run into a change which would make the behavior in any way differnet on a case-sensitive system, it will be ruled by the "case-insensitive" USE flag and will not be applied by default.
I just wanted to clarify all this as there seems to still be some confusion.
A couple more thoughts --
#1: This is not an "official" Gentoo project but I do hope that the changes will make it in eventually. I would like to see an "out of the box" Gentoo distribution which makes case-insensitivity as easy as described above.
#2: There's another thread, linked above, about the philosophical aspects of this. This thread is for technical discussion to make it actually work.
rinnan |
|
Back to top |
|
|
fernandotcl Veteran
Joined: 20 Nov 2003 Posts: 1396 Location: Sao Paulo, Brazil
|
Posted: Tue Aug 10, 2004 1:25 am Post subject: |
|
|
rinnan wrote: | blaster999 wrote: | In some unixes (unices?) filename.c means it is written in 'plain' C, while file.C means it is in C++. |
I'm not worried about ".C" and ".c" since that's a rarely used convention, and of course doesn't actually matter anyway as long as there isn't a ".c" and a ".C" file that are otherwise exactly the same (or differ only by filename case). If we run into it, we'll have to patch it. |
You can get errors when trying to compile C code with g++. You'll probably want to patch the makefiles. Don't forget assembly's ".S" files.
I wish good luck to you guys, although I think case-insensitivity is useless (and not well accepted by many, many people). Just be careful, because this projects looks a lot like the last Gnome fork.
Also, don't forget that you're going to go trough useless work, since everything could be accomplished with simple shell aliases, file system overlays, ZSH, and a lot of other solutions described in the other thread.
Oh, and keep it away from me, OK? |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Tue Aug 10, 2004 3:10 am Post subject: More info about python |
|
|
I've started to build a new Gentoo system with case-insensitivity. Python, as we know, is giving problems, and I've done more research on it.
Okay, good news and bad news about python on FCNTL/fcntl.
Check out this link:
http://python.fyxm.net/peps/pep-0235.html
This is a known situation python. It works by "architecture" -- for example, "mac" or whatever. So case sensitivity is munged with architecture, which is odd, because some Unixes are case-insensitive (and are threated as a seperate architecture even though they really aren't). I suppose that case sensitivity ought to be a completely seperate variable. In Windows, it's an option (you can force python to be either case-sensitive or case-insensitive). We'll have to patch the python libraries to be able to work the same way under Linux.
The good news is that, in a way, case sensitivity is a "solved" problem in python, the bad news is that the current solution isn't flexible enough to encompass our case, so we'll have to modify it.
EDIT: let me clarify that, when I refer to case-sensitivity with python, I am referring only to case-sensitivity as it deals with the filesystem. As far as I can tell, internally, python, like most programming languages, is case-sensitive as part of it's core definition, and this is not affected by any options or settings.
EDIT2: Any python guru's out there wish to comment? Especially those with porting experience? Religious commitment to case-insensitivity not required to offer help!
EDIT3: It turns out that Quote: | export PYTHONCASEOK=true | does all sorts of wonders for getting past things like the FCNTL problem. Let's see how far it gets. This variable was originally designed for Windows, it tells python that it needs to expect the scenario that the file system may return a case-insensitive match.
rinnan |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Tue Aug 10, 2004 4:20 am Post subject: emerge system begins |
|
|
works, which is to say it begins. My computer is very slow, I'll keep you updated.
rinnan |
|
Back to top |
|
|
kubark42 n00b
Joined: 15 Feb 2004 Posts: 44
|
Posted: Tue Aug 10, 2004 8:00 pm Post subject: |
|
|
Nice work. You're doing better than I. I'm stuck because I don't have an internet connection, and thus have to download things over a wireless link onto my laptop while I squat in front of a wirelessly equipped building. I'll see if I can't follow your instructions. I did do a stage3 install, because in theory that should work right out of the box, but I didn't get a chance to try it because I'm in the middle of a "heat" test. As soon as I figure out if 5 diskless computers and monitors closed into a lockbox melt themselves down or not I'll get right to the jfs.
Kenn
P.S. It'd be interesting to try emerge with Cygwin once we've ironed a few things out. Windows would be another good case-preserving filesystem to experiment with. |
|
Back to top |
|
|
rinnan Tux's lil' helper
Joined: 24 May 2003 Posts: 138 Location: Honolulu, Hawai`i
|
Posted: Wed Aug 11, 2004 2:07 am Post subject: |
|
|
kubark42 wrote: | Nice work. You're doing better than I. I'm stuck because I don't have an internet connection, and thus have to download things over a wireless link onto my laptop while I squat in front of a wirelessly equipped building. I'll see if I can't follow your instructions. I did do a stage3 install, because in theory that should work right out of the box, but I didn't get a chance to try it because I'm in the middle of a "heat" test. As soon as I figure out if 5 diskless computers and monitors closed into a lockbox melt themselves down or not I'll get right to the jfs. |
That sucks (your internet situation)-- I feel for you. I'd go crazy! :p
Good thinking with stage3. I'm going to continue working with stage1, because I think I'll catch more bugs that way, but we do need to check both paths seperately.
kubark42 wrote: | P.S. It'd be interesting to try emerge with Cygwin once we've ironed a few things out. Windows would be another good case-preserving filesystem to experiment with. |
Oh man, that would be interesting. For now though I want to attack that later.
Right now I'm working on two things -- first, I need to modify the kernel patch so that it works with more kernels (the current one only works with the vanilla kernel 2.6.7) -- maybe a script or something. It really looks just like the ".S" was changed to ".asm" in about 30 or 40 places. And second, I want to chase down this problem while emerging "sys-devel/gettext-0.12.1" (part of "system" in stage1):
Code: | checking for stdlib.h... yes
checking for string.h... yes
checking for memory.h... yes
checking for strings.h... yes
checking for inttypes.h... yes
checking for stdint.h... yes
checking for unistd.h... yes
checking dlfcn.h usability... yes
checking dlfcn.h presence... yes
checking for dlfcn.h... yes
checking for i686-pc-linux-gnu-g++... gcc
checking whether we are using the GNU C++ compiler... no
checking whether gcc accepts -g... no
checking dependency style of gcc... none
checking how to run the C++ preprocessor... /lib/cpp
configure: error: C++ preprocessor "/lib/cpp" fails sanity check
See `config.log' for more details.
configure: error: /bin/sh './configure' failed for autoconf-lib-link
!!! ERROR: sys-devel/gettext-0.12.1 failed.
!!! Function econf, Line 362, Exitcode 1
!!! econf failed
[amigatoo:root:/] |
|
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|