On Wed, 21 Feb 2001, Geoff Hutchison wrote:
> Date: Wed, 21 Feb 2001 18:55:49 -0600
> From: Geoff Hutchison <[EMAIL PROTECTED]>
> To: "Joe R. Jah" <[EMAIL PROTECTED]>
> Cc: Gilles Detillieux <[EMAIL PROTECTED]>,
> [EMAIL PROTECTED]
> Subject: Re: 3.1.x v. 3.2.x (was [htdig] date range - PATCH for 3.1.5)
>
> I do not wish this thread to degenerate. Please do not take this as a flame.
>
> At 4:33 PM -0800 2/21/01, Joe R. Jah wrote:
> >I wouldn't interpret lack of volunteers to lack of interest. Overwhelming
> >majority of "interested" in HTdig, or any other public software for that
> >matter, are just users; they may not have adequate expertise, time, or
> >gumption to pull up their sleeves and delve into the code, but I assure
> >you, they are _interested_.
>
> Of course. No one doubts that. On the other hand, Gilles had a few points:
> a) It's quite a bit of work to organize a release. It's not just
> something you jump into for one or two patches that only a few people
> use. Keep in mind it wasn't until I stepped forward to put together
> 3.1.0b1 that there was such a beast. Until then it was a pile of
> patches to 3.0.8b2.
Yes, I remember your successful, tenacious, and single handed resolve to
jump a full notch from 3.0 to 3.1; in my previous message I meant to say that
the move from 3.1.4 to 3.1.5 was almost entirely done by Gilles;
therefore, I think he is most familiar with the ins and outs of it because
that's where 3.1 lingers, 3.1.5.
> b) Little development is happening right now because too few of us
> are spread too thin. "Many hands makes light work" is the saying that
> comes to mind. Coding expertise is not necessarily key, but testing
> and organization are.
That's exactly my point. The same workforce can organize a solved puzzle
a lot faster and easier than to solve a new one and organize it.
> >I know 3.2.0bx is moving forward, but it may not be the answer to all
> >situations. There may be sites that do not really require phrase
> >searching; they may not want to sacrifice speed and efficiency for that.
>
> There are many reasons beyond just phrase searching to use 3.2. When
I am not well versed about 3.2, but I can not enumerate any concrete,
widely demanded features/advantages other than phrase searching.
> I started laying out the tradeoffs for phrase searching, I was quite
> worried about the disk space requirements and whether it would be
> worth it. For most users, the compressed databases are *smaller* than
> 3.1 databases.
You don't get any argument from me here.
> >I think if you were to poll htdig users and ask them if they wanted a 3.2
> >release soon with a 50 to one indexing speed loss with phrase searching,
>
> Woah. I think you're jumping the gun here. I have quite a few reports
> of 3.2 indexing being *faster* than 3.1. Certainly I'd love to get to
> the bottom of the performance problems for those who see them. But
> let's not jump the gun and say "well my site takes a long time to
> index, so the code sucks."
I have spotted several messages in the list; all of which were tails of
slow indexing; I may have missed those that said it was faster than 3.1.
> Remember that many of the tradeoffs involved early in the development
> were for faster searching at the expense of slightly slower indexing.
> Poll the same users and tell them that 3.2 is 10-15x faster at
> searching and often more so when dealing with the "dread"
> backlink_factor or sort methods?
Unfortunately in my case I have both slower indexing (52 times,) and
slower search; however, I have not yet made a wide enough study to
quantify it. In any case, for a rapidly changing environment, like my
site, I require an indexing to be fast enough for me to do meaningful
experiments in one day. 3.1.5 indexes my site in 11 minutes now. I may
be willing to accept 20, or may be 40 minutes, but 9 and a half hours is
out of the question.
Please do not get me wrong; I do not mean to put down 3.2. I am all for
continuing the development of 3.2. I will continue to support it, test
it, and try to adopt it to my system. I'm just saying we should not
abandon 3.1 in limbo.
Regards,
Joe
--
_/ _/_/_/ _/ ____________ __o
_/ _/ _/ _/ ______________ _-\<,_
_/ _/ _/_/_/ _/ _/ ......(_)/ (_)
_/_/ oe _/ _/. _/_/ ah [EMAIL PROTECTED]
_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
To unsubscribe, send a message to <[EMAIL PROTECTED]> with a
subject of unsubscribe
FAQ: http://htdig.sourceforge.net/FAQ.html
