Jeremy Huntwork wrote:
> On 7/28/10 11:00 PM, Thomas Pegg wrote:
>> I think describes what I think the CD should be to me. Very simple
>> doesn't need to be fancy.
>> Break it down to just a bootable system with a web browser (text, not
>> graphical simplifies maintenance) and all sources you need.
> 
> Alright, then we'll make that the main purpose, and we'll keep that 
> purpose clearly stated somewhere. Any future decisions about the CD 
> would have to fit in with that purpose. I was waiting to see if there 
> would be any more comments, but I guess we're in agreement.
> 
> So, given that, I think we can move to trying to define the contents of 
> the CD. Dan, I thought your idea of a survey was generally a good 
> suggestion, but I don't know if we need to do that just yet. Perhaps 
> after we finish discussing a general direction we can ask for 
> input/suggestions from outside this list.
> 
> So, a few thoughts about content:
> 
> 1. Obviously, the CD needs to fit the base requirements of LFS. The 
> easiest and traditional way of achieving that was to have the CD be 
> built using the same commands and packages as the LFS version it is 
> targeting.
> 
> 2. Support as much hardware as is reasonably possible. Basically, build 
> all stable drivers as kernel modules. Add some basic tools for advanced 
> disk configuration, e.g., mdadm, LVM2. And add as many userland disk 
> utilities as is reasonable, e.g., xfs, reiserfs, jfs...
> 
> 3. Provide tools needed to find answers online and get support. LFS's 
> main channels of support are the lfs-support mailing list and IRC. So we 
> need at least one IRC client. Many users will be able to read and send 
> mail through a browser. But should we include a mail client for those 
> that can't?
> 
> And that's essentially it, I believe. All of the above can be achieved 
> without X, and that would definitely be easier to maintain. However, I 
> must concede that X is much easier to work in (if you're not used to a 
> text browser or irssi, for example).
> 
> So then, if we provide X, I think we should keep that environment _very_ 
> simple. The question becomes how that is achieved, exactly.
> 
> Thoughts?
> 
> Jeremy
> 
> 
For the X environment the needed software over and above the chosen gui:
Seamonkey. it has the browser, email and irc tools in it.
a terminal.
anything extra isn't needed.

A note about Seamonkey though, the 2.x series has 2 issues that need to 
be considered if going with it.
1)It does randomly crash, even from just clicking a hyperlink. Some 
mozconfigs seem to really aggravate this, some reduce it dramatically. 
The debug mozconfig I've been using to try to get a core from the 
crashes dropped it to a very rare occurence.
2) the use of the mail/news client causes a Terminate and Stay actively 
Resident problem*. The current workaround is use the quit command. This 
completely shuts down all Seamonkey instances. [ quit from any component 
kills all active components ]

The second issue is a serious one for use to build LFS from a livecd. As 
everyone reading this can easily see.

These issues do not appear with the 1.9 series of Seamonkey, but 
unpatched security issues with the older, unsupported, version are also 
something to think about. A look the other day has the 2 unpatched 
vulnerabilities rated critical security flaws. Not truly a serious flaw 
for a livecd though. :D [ remote execution of local code, but no 
sensitive data to be exposed, nor permanent damage to os possible ]

Note: I have personal preference issues with one of the new features in 
the 2 series seamonkey. Namely I can't stand the message bars it went 
with. [ like Firefox, IE, and chrome have been using ].

* 250 MB+ of ram and cpu cycles on a dual core system of 6 to 15% being 
used even if you close all active components with clicking the "x" in 
the top corner.

Jaqui
-- 
http://linuxfromscratch.org/mailman/listinfo/livecd
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page

Reply via email to