Re: [Gnu-arch-users] Re: abrowse and rbrowse (what rbrowse has that abrowse doesn't)

[James Blackwell]

In lists.arch.users, you wrote:
> address@hidden (James Blackwell) writes:
>> > Well, to be fair, you don't actually _know_ that, do you?
>> 
>> Is there some significant point that you're trying to make, are you
>> trolling, or what? 
>
> No I'm not trolling.  Please don't be so sensitive; I know it's annoying
> to have to answer all sorts of dumb questions about your stuff, and be
> asked to justify what may seem perfectly obvious to you, but it's the
> way things work, and I think there are good reasons for it.  I am not
> posting to piss you off, I posting to clarify what are unsettled issues
> in my mind.
>
> [I get hugely exasperated by exactly the same thing, when people posit
> what seem to be to be absurdly misinformed objections to some feature
> I'm trying to add, but I grit my teeth and put up with it as best I can,
> because I know I'm sometimes going to be the one objecting.  Often I end
> up simply giving up, not because I believe in my position any less, but
> because the effort of defending it is simply too great -- and this is
> sad, but I know no real way around it.]


Ok! I can tell now that you know where I'm coming from. 


I figure that there's one of two things that rbrowse isn't doing quite
right from you. Either its the default options, or its the way that its
buffering input. 

First, the former. There's a patch waiting for Tom so that you can do
tla rbrowse [archivename/]REGEX, which should make it feel a lot closer
to abrowse, with one minor, intentional quirk. You can get sneaky, and
just type "tla rbrowse archivename/" and rbrowse a full remote tree.

The second issue is a bit more tricky. rbrowse and abrowse run in about
the same time: 

| Timing while just listing the tree 

   address@hidden:~$ time tla rbrowse -A address@hidden >
   /dev/null

   real    0m31.333s
   user    0m0.225s
   sys     0m0.100s
   
   address@hidden:~$ time tla abrowse -A address@hidden >
   /dev/null

   real    0m35.881s
   user    0m0.208s
   sys     0m0.130s

| Timing while getting patchlogs (-s) 
   address@hidden:~$ time tla abrowse -A address@hidden -s >
   /dev/null
   real    0m17.540s
   user    0m1.589s
   sys     0m0.545s

   address@hidden:~$ time tla rbrowse -A address@hidden -s
   real    0m17.487s
   user    0m1.581s
   sys     0m0.561s


It seems to me that both commands run in the same time. So, the question
is, "where do you think rbrowse is slower?". I'm aware of only one
difference between rbrowse and abrowse -- namely, rbrowse buffers the
patchlogs for a version and prints them out at once, abrowse prints them
as they find them.



-- 
James Blackwell          Please do not send me carbon copies of mailing
Smile more!              list posts. Such mail is unsolicited. Thank you!

GnuPG (ID 06357400) AAE4 8C76 58DA 5902 761D  247A 8A55 DA73 0635 7400