From: ibuprofin@painkiller.example.tld.invalid
On Mon, 22 Oct 2012, in the Usenet newsgroup alt.os.linux.mandriva, in article
, Adam wrote:
>Moe Trin wrote:
>> So, you're taking over the security and compatibility
>> _responsibility_ rather than the distro?
>As you've noticed, this started its own very helpful subthread, with
>reasons for and against both approaches. I finally had to make a
>sysadmin-type decision, and will revert to the Mandriva repositories'
>versions of Firefox/Thunderbird/Seamonkey RSN.
Distributions evolved over the years and have been growing much more
than Topsy ever considered. When this racket started, a full
distribution (such as Soft Landing Systems version 1.03 from August
1993) was small enough that the whole deal was downloadable in one go
using a typical 33.6 modem without tying the phone up all evening.
The main reason was that they didn't include that much stuff. (You may
not recall, but in 1993, the web was the weak sister compared to
"gopher" and a distro may not have included a browser - lynx from
ukans.edu or the original Mosaic from uiuc.edu, and I don't recall
either doing mail or news.) It was relatively easy for a home user to
maintain. But there's an enormous difference between a full
distribution of 35-40 Megabytes total, and the soup-to-nuts bloated
distributions of today, which might have as many as 30 thousand
packages in-toto. Representative: the ChangeLog file for the 3.0
and 3.1 kernels is larger than the entire 2.0.33 kernel source:
6010694 Oct 18 1996 linux-2.0.23.tar.gz
6347251 Jul 22 2011 ChangeLog-3.0
96675825 Jul 22 2011 linux-3.0.tar.gz
6022933 Oct 24 2011 ChangeLog-3.1
97333832 Oct 24 2011 linux-3.1.tar.gz
and those aren't the whoppers - the ChangeLogs for 2.6.25 through
2.6.39 were all larger (2.6.30 was 7.9 Megs).
>> Those older versions on the distro repositories - are they old, and
>> unpatched for recent security updates? If that's the case, you may
>> want to be looking at a different distro
>That's what I have a test system for. :-)
Not many people think to look at the updates issue. Yes, the packages
included and the versions thereof, but how does the distribution
respond to bugs and security problems?
>> is your kernel up to date? (3.0.46, 3.2.31, 3.4.14, 3.5.7 and 3.6.2
>> were _all_ released by kernel.org last week, 3.2.32 on Sunday.)
>Actually, before I sent "stolid" off for repair, I was using the
>latest kernel (3.2.x) in the Mandriva 2011.0 repositories, which is
>essentially a generic version. Since it's come back, I've stayed with
>the latest Mandriva-customized kernel which ATM is 2.6.39.4-5.1. I'm
>not sure what features the newer kernels offer.
I don't believe 2.6.39 is being maintained by kernel.org. The last
2.6.39 was 2.6.39.4 on 8/03/11. (3.0-rc1 was a branch from 2.6.39 back
on 5/30/11, and 3.0 came out on 7/22/11.) Version 2.6.35.14 was
released on 3/14/12, 2.6.27.62 on 3/18/12, 2.6.34.13 on 8/20/12 and
2.6.32.60 about 2 weeks ago. These later 2.6.x releases are all
back-ports as I understand it.
As for what features, see ftp://ftp.kernel.org in /pub/linux/kernel/
in v2.6/longterm/ and v3.0/ - the files you're looking for are the
ChangeLog-$VERSION. While we're up to 3.6.3, Linus stopped issuing
those files for 3.2, 3.3, 3.4 and 3.5. The ChangeLogs for the
incrementals (3.5.1, 3.5.2 and so on) are still being released. Is it
worth the effort? That's a harder question to answer, but that's what
the people at your distribution are supposed to be handling for you.
[HP computer mouse]
>> I rarely move the mouse beyond about 8" from the front edge, and
>> that still leaves the cable essentially flat. The cable is getting
>> more side-to-side motion than anything else.
>That probably explains why your mice last longer. My drawer came
>with a wrist rest and I've added another, higher one for the mouse,
>so mine is often close to the far edge.
I'm used to having lots of space for the mouse - partially because the
mouse was scaled less sensitive (1:5 was more normal) when I was doing
most of my early work with one (engineering schematic drafting).
[slow dialup connections]
>> I don't know if it would show up in 'lsusb -v' or not, but you might
>> compare what eris and stolid report.
>No difference in what they report for the USB-RS232 adapter, except
>for what's due to the two different versions of 'lsusb'.
No idea then - I haven't seen a 9600 BPS connection in so long
>> One obvious point would be to look to see if the driver does v.90
>> or v.92
>I tried some of the "AT+" and "AT-" settings, and actually got one
>v.92 connection at 52000/21600. That showed it's possible.
Without the manual, it's hard to say but that does sound as if the
"modem" isn't being initialized correctly - that old "ATZ" problem.
The "NVRAM" settings (yeah, I know) have something bizarre in them.
Try resetting to factory defaults, and writing that to NVRAM.
>I haven't been able to get that driver to compile under 3.x kernels
>yet. It /is/ working under CentOS with kernel 2.x. A next step
>would be to try the precompiled packages for earlier releases and
>kernels, both 64- and 32-bit versions (all OSs there are 64-bit).
What kind of error messages? Do you have all of the assorted -devel
packages installed?
>Okay, I tried it. It's a USR 5686-05 with v.92. I had it working
>(through a RS232-USB adapter) on stolid for a little while, under
>circumstances I wasn't able to duplicate.
I had a 5686-03 for a while, and sold it for some reason. It was a
fine modem.
>Connection speeds were 45333/ARQ to 50666/ARQ, so I don't think it's
>any improvement over the external ModemBlaster, the new USB modem,
>or eris's Compaq winmodem with Linuxant drivers.
Were those speeds here, or at the previous place?
>So about the only thing left is to get eris's winmodem using v.92
>more consistently, or move the USB modem from stolid to eris when
>necessary. Looks like both external modems go back to my box of
>computer pieces. There's another PCI winmodem in there that I might
>try on eris.
I'd probably keep only one plugged in at a time - both noise/loading
issues and avoiding exposing them to the same lightning surge at the
same time.
>What with having to do new installations of the NetZero and Juno
>software because of the new distros on both boxes, I found out a few
>more things, so I revised netzero-juno.txt to include those and
>posted the new version on my home page.
Lessee...
Re: internal PCI modems - my understanding is the the USR 5610C is
[continued in next message]
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
