[svlug] Redhat 6.2 formatting HD took over 3 hours!!!

[Bill Jonas]

On Thu, 8 Jun 2000, Rick Moen wrote:

>Irrelevant to my objection.  It went like this:
<snip of quotations>
>J.C. was purporting to respond to my quoted statement to Hart.  But,
>as such, it was a non-sequitur.  Which is what I said.

You said that it was "Irrelevant to the preceding discussion."  In a
thread, I take "the preceding discussion" to mean the thread, the
subject of which is the topic of conversation.  

  Vincent> It took over *3 hours* to format and partition and
  Vincent> stuff, and not a single "progress" meter or anything. Hell,
  Vincent> it didn't even tell me it was working on different partitions
  Vincent> -- it just said, "preparing / filesystem" or some stupid message the
  Vincent> whole time.

  Vincent> Now, I love RedHat's package over all the other's I've seen/tried, but
  Vincent> this is just rediculous that it should take that long to format and
  Vincent> doesn't even give an indication of time remaining or elapsed or anything.

  Zawodny> I wish the installer had said,
  Zawodny> "this could take half the day on big drives..."

Then you said:

  Moen> Nonetheless, you pretty much always need to do it on IDE.  On
  Moen> SCSI, there's already a "hotfix" provision at the hardware level.

Continuing:

  Zawodny> Yeah, I know. I just wished the installed had some simple logic in it
  Zawodny> that says:

  Zawodny>   if (drive_is_big)
  Zawodny>   {
  Zawodny>       tell_user_to_come_back_tomorrow();
  Zawodny>   }

  Hart>    but I have raised the issue of giving users
  Hart>    feedback (a warning when they enable the option and some feedback
  Hart>    during the check) with our install engineering team.

You responded, quoting the part that I omitted above:

  Moen> begin  hartr at redhat.com quotation:
 
  Moen> > 2) You enabled bad block checking when you installed. This most
  Moen> >    definitely does take a very long time. There is nothing we can do
  Moen> >    about the time it takes....

  Moen> That would _not_ be Red Hat Software, Inc. but rather Ted T'so, who
  Moen> writes the ext2 utilities, yes?  No, there's not a lot he could do, but
  Moen> there's something Daevid could do:  Use SCSI.

Then JC Lawrence replied, quoting the first 1.5 lines of your above
response, and adding a "Yes, but..." response, getting back to what had
been the topic originally:

  Lawrence> There is little preventing RH's installer from observing
  Lawrence> just before they fork mke3fs, "large partition size plus block
  Lawrence> checking," and thus put up an alert before forking mke2fs.

  Lawrence>   Moen> > No, there's not a lot he could do...

  Lawrence> Actually mke2fs does do screen IO during its block checking pass (it 
  Lawrence> lists and updates a displayed value of how many blocks it has
  Lawrence> checked).  Whether this value is visible during RH's install given
  Lawrence> any possible piping/redirection/virtual_console use is another
  Lawrence> matter, but is again, under RH's control.

He wasn't arguing that it would decrease mke2fs format and bad-block
check time on IDE a bit at all.  He was merely implying that it would be
a nicety for the user.  And then the thing that baked my noodle
slightly:

  Moen> Irrelevant to the preceding discussion.  That would not affect 
  Moen> _the time it takes_ by iota.  Please read more carefully.

>You then went on a long flight of fancy, tendentiously misrepresenting 
>the preceding exchange, and rather ironically claiming _I_ needed to 
>pay closer attention.

I fail to see how I have misrepresented 20 messages in the thread prior
to you post on Thu, 8 Jun 2000 11:21:46 -0700.  If you feel that way,
then you are certainly entitled to your own opinion and your own view
about that, just don't expect me to be convinced.  Just as I remain
unconvinced by your arguments.

>Thus:  You are wasting my time.  Please don't do that.

I fail to see how I or anyone else has forced you to reply.  I refuse to
accept responsibility for your actions; specifically, the time you spent
of your own free will composing 2 replies of your own to two replies of
mine.

I can see where you're coming from.  You were obviously referring to 4
of the immediately previous messages, while I was looking at 20.  Thank
you for pointing out the messages to which you were referring,
specifically.  However, I still assert that in the context of the thread
(which is the context that, in general, every remark on a mailing list
is made -- the context of a thread), the remarks were not "irrelevant",
though it would certainly appear that they are, based on the 4 messages
you quoted.

Let's drop this.  It's becoming increasingly obvious that either one of
us can find quotes to support our individual contentions.  If nothing
else, that should be a point on which we can both agree.

Bill
-- 
>Ever heard of .cshrc?             | "Linux means never having to delete
That's a city in Bosnia. Right?    |  your love mail." -- Don Marti
(Discussion in comp.os.linux.misc  |  http://www.netaxs.com/~bj/
on the intuitiveness of commands.) |  http://www.harrybrowne.org/