Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
What do you want from an init system?
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4  Next  
Reply to topic    Gentoo Forums Forum Index Off the Wall
View previous topic :: View next topic  
Author Message
wildhorse
Tux's lil' helper
Tux's lil' helper


Joined: 16 Mar 2006
Posts: 148
Location: Estados Unidos De América

PostPosted: Tue Dec 31, 2013 1:16 am    Post subject: Reply with quote

do you use XML? no self-respecting programmer would
Back to top
View user's profile Send private message
Naib
Advocate
Advocate


Joined: 21 May 2004
Posts: 4476
Location: Removed by Neddy

PostPosted: Tue Dec 31, 2013 1:19 am    Post subject: Reply with quote

I'm not a programmer :) I do hardware :)
_________________
The best argument against democracy is a five-minute conversation with the average voter
Great Britain is a republic, with a hereditary president, while the United States is a monarchy with an elective king
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Tue Dec 31, 2013 3:48 am    Post subject: Reply with quote

Naib wrote:
does it use XML? no self-respecting program wouldn't

No, it uses Lua, Perl and PHP, with a side of Java. :^)
Back to top
View user's profile Send private message
erm67
Tux's lil' helper
Tux's lil' helper


Joined: 01 Nov 2005
Posts: 130
Location: somewhere in Berlusconia.

PostPosted: Tue Dec 31, 2013 8:45 am    Post subject: Reply with quote

Butts McCracken wrote:
anything with lua is dead by definition

hehe, it is part of LUCI the management system of Openwrt written in LUA, so it not strange that uses lua and with their requirement (32Mb RAM minimum configuration) probably LUA is the only modern scripting language that could be used.
Subsentient wrote:
erm67 wrote:

I think I will spend some time on
http://nbd.name/gitweb.cgi?p=luci2/procd.git;a=summary
and
http://nbd.name/gitweb.cgi?p=luci2/netifd.git;a=summary

They do a lot more, PID tracking mounting tasks network management and so on .....

PID tracking: Yes.
Autorestarted services: Yes.
Facilities for service control: Yes.
Networkmanagement: Best left to something else IMHO.
Mounting tasks: I cannot think of a good place in the boot order to put something like that,
so I left it out. If I figure it out, I might add it in Epoch 1.1.

well handle /etc/fstab is a minimum, user mounts probably useful, on-demand mounting optional ... a _netdev mount option is useful.
it should also take into account that it could be the origin or the destination of a pivot_root, being so small you could find customers in the initrd market :-)
But since now mounting /var and /usr on different partitions has become illegal probably thing are a lot easier.

BTW since I am going to pivot_root, epoch mounts /proc because it needs it, I do manually the mount job and .... umount /proc? will the pivot_root work and exec a new instance of epoch, wich will re-mount /proc in the new root?
_________________
Truck!!
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Tue Dec 31, 2013 9:57 am    Post subject: Reply with quote

erm67 wrote:
mdeininger wrote:
erm67 wrote:
Butts McCracken wrote:
erm67 wrote:
Butts McCracken wrote:
erm67, you should try this. I have some pull with the creator if there are any parts you are desperate to have in your router :D

:-) :-)
I guess I will go with the usual busybox and uClibc, the less I have to install the better.
Well here you go, it gives you something to play with over new year: https://github.com/jyujin/curie

I think I will spend some time on
http://nbd.name/gitweb.cgi?p=luci2/procd.git;a=summary
and
http://nbd.name/gitweb.cgi?p=luci2/netifd.git;a=summary

They do a lot more, PID tracking mounting tasks network management and so on .....
maybe finally something will replace NetworkManager
eINIT had that... source code's still around on the same site... just sayin', those're very basic features ;D.

You know the times were not ripe for einit, now systemd and upstart raised the bar about what someone wants from an init system ..... maybe it could be the right moment to resurrect the project :-)

What is really incredible is that it took so many years for linux people to realize that the whole initd/rc system needed a refresh. initd was doing the pid tracking, who doesn't remember 'X is spawning too fast disabled for 5 minutes' errors? Having everything simply started from scripts and tell the user that they can easily manually restart it with a simple command worked for some time, but eventually something better should appear.
Why don't you just re-enter the competition?
Openwrt guys are desperately seeking a (really small) tool for the job, just like the openwrt wiki for procd says there is no tool reasonably small (a few kb at most, it needs to be run on systems with 32Mb) around. Possibly that doesn't write anything (beside pids), and log to syslog or to the printk buffer.
The network daemon is probably too complex, it should be able to handle complex networks, several interfaces, not just bring eth0 up :-) probably the openwrt guys knows better :-)

My small router could potentially have 4 wi-fi SSID 802.11n, 4 SSID 802.11a, 4 eth, 1 wan, adsl, 4 atm, 4 ppp, 4 IPSEC vlan; handle that it is not a job for an init system anymore.
i'm pretty sure we had that networking there nailed... *shrug*. thing is i'm just not interested in writing this type of programme anymore, which is why I'd much rather see it left to people who are. like Subsentient here :).

Subsentient wrote:
erm67 wrote:

I think I will spend some time on
http://nbd.name/gitweb.cgi?p=luci2/procd.git;a=summary
and
http://nbd.name/gitweb.cgi?p=luci2/netifd.git;a=summary

They do a lot more, PID tracking mounting tasks network management and so on .....

PID tracking: Yes.
Autorestarted services: Yes.
Facilities for service control: Yes.
Networkmanagement: Best left to something else IMHO.
Mounting tasks: I cannot think of a good place in the boot order to put something like that,
so I left it out. If I figure it out, I might add it in Epoch 1.1.
seems solid to me. mounting is actually pretty neat to do in parallel, depending on the configuration. i remember my init had to automatically generate dependencies for that though, based on where the programmes were, which files they used and which file systems had to be mounted. Feel free to browse the repositories for ideas or code for later versions: http://git.becquerel.org/ ;)

Speaking of stuff from there, a custom bootchart daemon always seemed to come in handy :D
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
Butts McCokey
Advocate
Advocate


Joined: 23 Apr 2004
Posts: 3331

PostPosted: Tue Dec 31, 2013 12:30 pm    Post subject: Reply with quote

mdeininger wrote:
Speaking of stuff from there, a custom bootchart daemon always seemed to come in handy :D
you used to do cool graphical boot charts. Do you still have them? They might be useful to Subsentient
_________________
when you're sitting back, in your rose pink Cadillac
Making bets on Kentucky Derby Day
I'll be in my basement room, with a needle and a spoon
And another girl to take my pain away...
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Tue Dec 31, 2013 1:30 pm    Post subject: Reply with quote

erm67 wrote:

well handle /etc/fstab is a minimum, user mounts probably useful, on-demand mounting optional ... a _netdev mount option is useful.
it should also take into account that it could be the origin or the destination of a pivot_root, being so small you could find customers in the initrd market :-)
But since now mounting /var and /usr on different partitions has become illegal probably thing are a lot easier.

BTW since I am going to pivot_root, epoch mounts /proc because it needs it, I do manually the mount job and .... umount /proc? will the pivot_root work and exec a new instance of epoch, wich will re-mount /proc in the new root?

You have a good point with the pivot_root. I forgot about that completely, but yes, that's a REALLY important feature to me. Damn, I have broken the general rules of release candidates so many times, next time I'll call it "1.1 Not Done Yet 1".

As far as fstab handling, I really think it's better to use 'mount' for mounting actual disks, but if I think I can do it inexpensively, I'll add it, perhaps.
Might as well, since we aren't thinking like an RC anymore...

I also really don't want to hear about being an unresponsive upstream provider after all this...
Back to top
View user's profile Send private message
erm67
Tux's lil' helper
Tux's lil' helper


Joined: 01 Nov 2005
Posts: 130
Location: somewhere in Berlusconia.

PostPosted: Tue Dec 31, 2013 2:03 pm    Post subject: Reply with quote

Subsentient wrote:
erm67 wrote:

well handle /etc/fstab is a minimum, user mounts probably useful, on-demand mounting optional ... a _netdev mount option is useful.
it should also take into account that it could be the origin or the destination of a pivot_root, being so small you could find customers in the initrd market :-)
But since now mounting /var and /usr on different partitions has become illegal probably thing are a lot easier.

BTW since I am going to pivot_root, epoch mounts /proc because it needs it, I do manually the mount job and .... umount /proc? will the pivot_root work and exec a new instance of epoch, wich will re-mount /proc in the new root?

You have a good point with the pivot_root. I forgot about that completely, but yes, that's a REALLY important feature to me. Damn, I have broken the general rules of release candidates so many times, next time I'll call it "1.1 Not Done Yet 1".

pivot_root mounts and initrd go hand in hand :-) everybody would be happier without busybox :-)
Well handling mounts should be easy, I'd suggest to use the (now ubiquitous) _netdev mount option, it is ignored by the kernel and the user is required to identify network mounts with that option in fstab, could as well become a 'standard' feature.
It is really easy to implement just a few lines of code, if the mount point has a _netdev option it will be mounted only after UserSpecifiedTarget has completed. Previous attempts to automatically identify network moint point failed :-) (systemd adds _netdev to nfs filesystems but now everybody uses nfs4 ......).
_________________
Truck!!
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
jonnevers
Veteran
Veteran


Joined: 02 Jan 2003
Posts: 1594
Location: Gentoo64 land

PostPosted: Tue Dec 31, 2013 7:49 pm    Post subject: Reply with quote

Butts McCracken wrote:
anything with lua is dead by definition

lol, what?
Back to top
View user's profile Send private message
Butts McCokey
Advocate
Advocate


Joined: 23 Apr 2004
Posts: 3331

PostPosted: Wed Jan 01, 2014 4:47 am    Post subject: Reply with quote

jonnevers wrote:
Butts McCracken wrote:
anything with lua is dead by definition

lol, what?
lua = dead
_________________
when you're sitting back, in your rose pink Cadillac
Making bets on Kentucky Derby Day
I'll be in my basement room, with a needle and a spoon
And another girl to take my pain away...
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Wed Jan 01, 2014 11:28 am    Post subject: Reply with quote

Butts McCracken wrote:
mdeininger wrote:
Speaking of stuff from there, a custom bootchart daemon always seemed to come in handy :D
you used to do cool graphical boot charts. Do you still have them? They might be useful to Subsentient
aye, they really might... it's a pretty simple programme, too:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/jobs/bootchart.c;h=4522e844ff2fdff7967ec85fdc5a90c196317d03;hb=testing

... i think I might've ported that around a couple times too. Certainly better than the original shell script from the bootchart devs, heh. Meh, getting all nostalgic. I just noticed how the mounting logic got a bit long though:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/modules/mount.c;h=7e57d0d532fde24423800587b7f77f168dbc81af;hb=testing

lots of weird stuff with dependencies and parsing fstab. But mounting things or doing a pivot_root are actually quite easy, they're just plain ol' system calls that look pretty much the same as the command line programmes :).

Oh, also, @Subsentient: this programme in particular comes in really handy if you do mounting tasks yourself:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/linux/last-rites.c;h=e052bd7890541b7099ce64f45d3c3ff4939dc207;hb=testing

you can just exec() it from your init and it'll get EVERYTHING unmounted. Even the root filesystem.
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Thu Jan 02, 2014 12:02 am    Post subject: Reply with quote

mdeininger wrote:


Oh, also, @Subsentient: this programme in particular comes in really handy if you do mounting tasks yourself:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/linux/last-rites.c;h=e052bd7890541b7099ce64f45d3c3ff4939dc207;hb=testing

you can just exec() it from your init and it'll get EVERYTHING unmounted. Even the root filesystem.


I'm currently writing mounting and fstab support. pivot_root support will be easy too. I have mounting support working partially, but I'm working on loopback support an unmounting. Fstab parsing is complete. I had to create a new file, mount.c, to add this. None of this is in git yet.
Back to top
View user's profile Send private message
Mardok45
n00b
n00b


Joined: 21 Jun 2008
Posts: 64
Location: Right behind you

PostPosted: Thu Jan 02, 2014 12:14 am    Post subject: Reply with quote

Butts McCracken wrote:
jonnevers wrote:
Butts McCracken wrote:
anything with lua is dead by definition

lol, what?
lua = dead

Ruby, by that definition, is also dead, but it's still around and kicking butt. So maybe Lua and Ruby are more like Zombies or Jesus depending on how you want to look at it.
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16138
Location: Colorado

PostPosted: Thu Jan 02, 2014 1:46 am    Post subject: Reply with quote

/. recently had an entry titled Is Ruby Dying? (queue appropriate references to questions in titles).
_________________
lolgov. 'cause where we're going, you don't have civil liberties.

In Loving Memory
1787 - 2008
Back to top
View user's profile Send private message
Butts McCokey
Advocate
Advocate


Joined: 23 Apr 2004
Posts: 3331

PostPosted: Thu Jan 02, 2014 2:16 am    Post subject: Reply with quote

mdeininger wrote:
Butts McCracken wrote:
mdeininger wrote:
Speaking of stuff from there, a custom bootchart daemon always seemed to come in handy :D
you used to do cool graphical boot charts. Do you still have them? They might be useful to Subsentient
aye, they really might... it's a pretty simple programme, too:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/jobs/bootchart.c;h=4522e844ff2fdff7967ec85fdc5a90c196317d03;hb=testing

... i think I might've ported that around a couple times too. Certainly better than the original shell script from the bootchart devs, heh. Meh, getting all nostalgic. I just noticed how the mounting logic got a bit long though:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/modules/mount.c;h=7e57d0d532fde24423800587b7f77f168dbc81af;hb=testing

lots of weird stuff with dependencies and parsing fstab. But mounting things or doing a pivot_root are actually quite easy, they're just plain ol' system calls that look pretty much the same as the command line programmes :).

Oh, also, @Subsentient: this programme in particular comes in really handy if you do mounting tasks yourself:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/linux/last-rites.c;h=e052bd7890541b7099ce64f45d3c3ff4939dc207;hb=testing

you can just exec() it from your init and it'll get EVERYTHING unmounted. Even the root filesystem.
I didn't realise you wrote it all yourself!
_________________
when you're sitting back, in your rose pink Cadillac
Making bets on Kentucky Derby Day
I'll be in my basement room, with a needle and a spoon
And another girl to take my pain away...
Back to top
View user's profile Send private message
energyman76b
Advocate
Advocate


Joined: 26 Mar 2003
Posts: 2041
Location: Germany

PostPosted: Thu Jan 02, 2014 2:49 am    Post subject: Reply with quote

Butts McCracken wrote:
jonnevers wrote:
Butts McCracken wrote:
anything with lua is dead by definition

lol, what?
lua = dead


and you are an expert in scripting languages... since when?
_________________
Study finds stunning lack of racial, gender, and economic diversity among middle-class white males
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Thu Jan 02, 2014 5:32 am    Post subject: Reply with quote

Subsentient wrote:
mdeininger wrote:


Oh, also, @Subsentient: this programme in particular comes in really handy if you do mounting tasks yourself:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/linux/last-rites.c;h=e052bd7890541b7099ce64f45d3c3ff4939dc207;hb=testing

you can just exec() it from your init and it'll get EVERYTHING unmounted. Even the root filesystem.


I'm currently writing mounting and fstab support. pivot_root support will be easy too. I have mounting support working partially, but I'm working on loopback support an unmounting. Fstab parsing is complete. I had to create a new file, mount.c, to add this. None of this is in git yet.
nice, that's great to hear :).

just saying though, go nuts on the code in git if you have any use for it. no need to reimplement everything if you find something useful to ya ;).

Butts McCracken wrote:
I didn't realise you wrote it all yourself!
meh, what can I say :P. might just as well revive the testing branch and put out an "einit-1" for old times' sake, heh. still don't intend to do any significant improvements though. i guess at this point it just is what it is...
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
erm67
Tux's lil' helper
Tux's lil' helper


Joined: 01 Nov 2005
Posts: 130
Location: somewhere in Berlusconia.

PostPosted: Thu Jan 02, 2014 9:14 am    Post subject: Reply with quote

mdeininger wrote:
Subsentient wrote:
mdeininger wrote:


Oh, also, @Subsentient: this programme in particular comes in really handy if you do mounting tasks yourself:

http://git.becquerel.org/?p=einit/core.git;a=blob;f=src/linux/last-rites.c;h=e052bd7890541b7099ce64f45d3c3ff4939dc207;hb=testing

you can just exec() it from your init and it'll get EVERYTHING unmounted. Even the root filesystem.


I'm currently writing mounting and fstab support. pivot_root support will be easy too. I have mounting support working partially, but I'm working on loopback support an unmounting. Fstab parsing is complete. I had to create a new file, mount.c, to add this. None of this is in git yet.
nice, that's great to hear :).


Great, if it's done before I finish setting up my automatic brouting rules I am going to flash my modem with epoch :-)
Not that this will bring you many users since I am updating the firmware just for myself .........
My plan is to install the rootfs on squashfs (ro) and if an nvram key is set pivot_root to a aufs union between the ro nvram and a rw partition possibly to an external usb stick to avoid as much as possible to stress the nand flash.
Right after boot I have to setup epoch to check the nvram key, than insmod the modules, wait for the usb key to magically appear, mount it, create the union and than pivot_root (or switch_root like busybox calls it) :-)
_________________
Truck!!
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Fri Jan 03, 2014 1:08 am    Post subject: Reply with quote

erm67 wrote:

Great, if it's done before I finish setting up my automatic brouting rules I am going to flash my modem with epoch :-)
Not that this will bring you many users since I am updating the firmware just for myself .........
My plan is to install the rootfs on squashfs (ro) and if an nvram key is set pivot_root to a aufs union between the ro nvram and a rw partition possibly to an external usb stick to avoid as much as possible to stress the nand flash.
Right after boot I have to setup epoch to check the nvram key, than insmod the modules, wait for the usb key to magically appear, mount it, create the union and than pivot_root (or switch_root like busybox calls it) :-)


I actually had decided not to write the mounting support, even though I got most of it done already, but your story persuades me to proceed. pivot_root support is almost ready. I'll commit that and then resume work on the mounting support. This will take a few days, but it should be worth the wait if all goes well.
Back to top
View user's profile Send private message
erm67
Tux's lil' helper
Tux's lil' helper


Joined: 01 Nov 2005
Posts: 130
Location: somewhere in Berlusconia.

PostPosted: Fri Jan 03, 2014 12:59 pm    Post subject: Reply with quote

Subsentient wrote:
erm67 wrote:

Great, if it's done before I finish setting up my automatic brouting rules I am going to flash my modem with epoch :-)
Not that this will bring you many users since I am updating the firmware just for myself .........
My plan is to install the rootfs on squashfs (ro) and if an nvram key is set pivot_root to a aufs union between the ro nvram and a rw partition possibly to an external usb stick to avoid as much as possible to stress the nand flash.
Right after boot I have to setup epoch to check the nvram key, than insmod the modules, wait for the usb key to magically appear, mount it, create the union and than pivot_root (or switch_root like busybox calls it) :-)


I actually had decided not to write the mounting support, even though I got most of it done already, but your story persuades me to proceed. pivot_root support is almost ready. I'll commit that and then resume work on the mounting support. This will take a few days, but it should be worth the wait if all goes well.

Don't worry, I am still in the learning/documentation phase of layer 2 firewalling/routing :-) I will probably use epoch also without pivot_root using an initrd with busybox .....
All the examples and I found use a static network setup, what I am trying to do is automatically configure the bridge after the dhcp lease is obtained, ATM I am looking at how openwrt does that :-)
_________________
Truck!!
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Mon Jan 06, 2014 6:13 am    Post subject: Reply with quote

https://github.com/Subsentient/epoch/commit/b0471a2befa1791b03dcbca1648f49b4689f7171

pivot_root support is functional.

I am probably going to wait for 1.1 before I implement the built-in mounting support, because that's going to be buggy at first, and I don't want to release a half-baked 1.0, and I'm not willing to put off 1.0 for another month. It should be easy to use busybox mount to make pivot_root work fine. It's been tested and works as advertised.
If I can get a couple guaranteed and serious testers (not the kind who are never there when you need them) then I'll implement the mounting support.
Back to top
View user's profile Send private message
erm67
Tux's lil' helper
Tux's lil' helper


Joined: 01 Nov 2005
Posts: 130
Location: somewhere in Berlusconia.

PostPosted: Mon Jan 06, 2014 6:20 pm    Post subject: Reply with quote

Good pivot_root was the most important bit :-)

I can do a few tests in the next days :-)
_________________
Truck!!
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
dE_logics
Advocate
Advocate


Joined: 02 Jan 2009
Posts: 2220
Location: $TERM

PostPosted: Mon Jan 06, 2014 6:30 pm    Post subject: Reply with quote

Quote:
What do you want from an init system?


systemd
_________________
Buy from companies supporting opensource -- IBM, Dell, HP, Hitachi, Google etc...
Disfavor companies supporting only Win -- Logitech, Epson, Pioneer, Kingston, WD, Yahoo, MSI, XFX, Huawei
My blog
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Mon Jan 06, 2014 6:58 pm    Post subject: Reply with quote

erm67 wrote:
Good pivot_root was the most important bit :-)

I can do a few tests in the next days :-)


Alright then, I'll begin work on it again later today. :^)
Join #epoch on Freenode if you're an IRC guy. I will pop in when I'm ready for testers, but that could be a few days.
I'm easily exhausted and write only a few lines of code a day, because any more and my concentration disintegrates.
EDIT: see post below.


Last edited by Subsentient on Mon Jan 06, 2014 9:00 pm; edited 2 times in total
Back to top
View user's profile Send private message
Subsentient
n00b
n00b


Joined: 09 Dec 2013
Posts: 2
Location: The blue void

PostPosted: Mon Jan 06, 2014 8:59 pm    Post subject: Reply with quote

@erm67:
The more I think and look and read, the more I think that adding mount support to Epoch would be a mistake. Let me explain.

* Epoch is intended to provide you with everything that an init system *must* have to do it's job well, and an integrated mount command
does not fall under that category. Stuff like service management with commands, e.g. 'epoch restart sshd' needs to be there because
there are not too many clean ways of doing this outside of the init system. A mount command is easily done outside of Epoch with
minimal space taken up.

* The extra memory usage will ALWAYS be there if built into Epoch, with a mount command like busybox, it's freed once the program exits.

* It is such a large feature that it requires a dedicated source file. Nothing else has it's own source file. Nothing.

* It will uglify the syntax of epoch.conf, with an ever-expanding list of CAPITAL_SPECIAL_OPTIONS for ObjectStartCommand.

* Certain things that should be simple but are not, e.g. loopback mounting, will require too much code and too much work.

I hope this makes things clearer as to why I am reluctant.

-Sub
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Off the Wall All times are GMT
Goto page Previous  1, 2, 3, 4  Next
Page 2 of 4

 
Jump to:  
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