Google Revises Usenet Search

michaelmalak writes "Wednesday night, Google Groups announced in a thread the rollout of their revised 20-year Usenet archive search engine. Among the various 'improvements': ability to search by date has been eliminated, as has the ability to deep link to a single post. See the announcement thread for others' reaction." An anonymous reader writes "ZDNet has published some interesting insights into what makes Google tick. In this lengthy article, Google's vice-president of engineering, Urs Hölzle delves into the nuts and bolts behind Google's operations, what back-up mechanisms and hardware setup is in place and even some interesting homegrown technology like the Google File System (GFS)."

Comments

Hey Google: you're being evil...

[Anonymous Coward]

Try to search for a number using Beta and you'll see how broken it is.

Also, it creeped me out to no end discovering this morning that my Gmail cookie is really a "Google Accounts" cookie which will now be attached to my Usenet forays via Google as well. I personally don't want the line between public and private conversations to be muddied like that, and I definitely don't want a unified cookie straddling both domains.

Finally, the interface leaves a lot to be desired. The layout is cluttered and junky now whereas it was clean and simple before. I'm not enthralled by the Javascript hooks. Threading seems to be worse than ever (and still not done by message-ID or References - when I asked Google why this was via email, the response was "too difficult"... *boggle*) and the CLI-esque search ability is degenerating into a GUI mess; where one line of text and a CR would before get you to the page you wanted, it now can take that plus several additional mouse gestures and clicks.

This is a sad day, to see a useful tool become so f**ked up for no apparent good reason. I can only hope and pray for a reversion.

Re:Progress?

[forrestt]

Well, there might be a more practical reason than they simply don't care about standard HTML. It appears the main problem is they don't tell the doctype. That would take them an extra 118 bytes PER REQUEST to include the type. That means, according to the 1000 requests per second mentioned in the article, they are saving 115Kbps in transfer rates by not including the doctype. It doesn't seem like much, but it is the same thing that got airlines to stop serving food. And this is just the Doctype. I'm sure they cut bites out wherever they can.

Don't like how the Google Usenet archive evolves?

[martin-k]

If you don't like how Google's Usenet search engine and archive evolves (neither do I; Dejanews was tops for its time and things went downhill from there), help the competition... :-)

I already have an archive of around 600 million messages (nearly everything sans binaries from 2000 till today; just a couple of terabytes) and intend to create a public Usenet search engine. As I am using Usenet myself on a daily basis, I know what *I* want in a Usenet search engine, and that's quite different from what Google gives us.

Here's how you can help: Contact me at martin-k (at) softmaker.de if you have a private collection of Usenet postings that you want me to put in the database.

-mk