Changes to the All search

Sling will like this

I don't look at search very often - almost never - so I don't know how long the changes have been in place but the search box is now resizeable, which it should always have been, but the biggest change is to each place's html page.

At the end of each listed item in the html page is a "GO" button that takes you straight to the item. It's a very good addition for most people but it means that I have some work to do.

Back when I was actually interested in the rankings, I kept changing my page as other people changed theirs - we were trying to outrank each other It became really ridiculous as the 'for sale' items were replaced more and more by keyword stuffing prims, and the informative aspect of the pages became less and less. So some of us now have "GO" TP links to stacks of keyword prims which may or may not be where the actual items are. Earlier today, I clicked on a GO link in my html page. It landed underneath the store and I fell to earth, coincidentally to the very spot that I'd started from.

Oh well. I'll have fix all those TP destinations or the castle will have a lot more people dropping in than usual

Surely this will fail epically for those who have set teleport routing to a fixed position.

However I like the idea.

I hadn't thought of that. It should get people to a proper place albeit not to the item itself.

It's all good, you still get the red beacon for those places.

Although, if what they are pointing at are search gaming ( oops! - 'optimisation') prims, then it should really be a hard Cheesy Beacon.

Karma!




As an aside - I sort of noticed Cheesy Beacons in the options but for maybe a year read it as Cheesy Bacon.

Our private frankenviewer is called Snotglob, so Cheesy Bacon would be a most appropriate addition. Thanks!

Haven't noticed that before but it is very handy ,How long has it been there?

I think this has been around for at least as long as I have, which is about five-six weeks now. But it hasn't been much use the few times I've tried it. I still arrive at a central TP point and have to look for the item.

As Viktoria mentioned, that beacon is still a huge help, even if there's a landing point set. It beats the hell out of what I used to do, which is to walk or fly, watching the coordinates at the top of the screen, to get to the coordinates displayed in the Search / All item listing. What we need now is a text search function within that Search/All browser window, in order to track down whatever on the page stupid GSA thinks is matching the search string. (I still end up copying the page URL, pasting it into a real browser, and text searching the page there. Pain in the ass.)

By the way, Phil: what do you think of Google's move to pageload times as a ranking factor for real web search? (I love it, personally, and wish we could do the same in SL Search, based on rez times. As Jeff Atwood says, if it's gonna suck, it's a win if it sucks quickly.)

Maybe because it was early days, I didn't realise what the beacon was. Trust me to miss the obvious.

When I joined SL 3 years ago I dropped out of seo and I haven't been back even though I still have a thriving seo forum and site, so it's news to me.

Off the top of my head, I don't like it as a ranking factor. Search engines are supposed to rank on relevance as best they can, and page load times have nothing to with relevance. If it makes a difference in the rankings, I can see it making things slower for users. I.e. website owners will start breaking perfectly good large pages up into many smalll ones, which the user will have to navigate to get all the information that was originally on one page. If it doesn't make much difference to the rankings, it doesn't really matter, but it's a move away from relvance, and, for that reason, it's bad imo. It's not as though pages are slow to load these days. It used to be that people would generall allow 10 seconds for a page to load, but bandwidth has increased enormously since then and the meat of pages is very quick to load now so it's not only bad, but also unnecessary.

If LL ranked on rez times, as you suggested, and nobody 'arranged' anything like Kitty suggested, it would be bad for SL users. It would mean that large stores that have lots of choice for people would be pushed down the rankings by tiny places that just have a very few things for people. Like the webpages, people would often have to visit many tiny places to find something they fancy, which would make the whole shopping process slower.

Now if we could just change the size of the tab.

I go for screen efficiency, I would resize it to 0.

Define 'relevance'.
All pages that mention "foo" are relevant to "foo".
Search engines use factors other than that very simple 'relevance' to try to rank matches.
I can see where page weight could be seen as being relevant to users' wants.


People can respond to Google punishing high-weight pages by redesigning pages to reduce the amount of bandwidth consuming bling, and paying greater attention to graphics content. They do not necessarily have to reduce the core information content of the pages. Should they reduce the information content of pages, they would run a very high risk of losing ranking as IBLs to the original page would get diluted over the new separated 'smaller' pages. They could also lose out on search hits for terms that used to be encompassed be the original page.
Keeping - or even extending - the original information content but engineering the presentation to be lower bandwith would be the best response.

That approach would work for SL as well as for Google.



I don't agree that a large store is necessarily better for SL users than a tiny store. If I'm looking for something specific, I would like to evaluate items as rapidly as possible. Large stores can work against that. Visiting a series of tiny stores could be faster than visiting a few large stores. It would also tend to present a greater variety of item design.

Large/tiny says nothing about the perceived quality.
Large/tiny says nothing absolute about the ease of finding the items, but a smaller place may be faster to search.





It was more that people generally *had* to allow 10 seconds, but increases in network bandwidth have reduced that to nearer the 3 to 4 seconds max that people find more acceptable.

A general reduction in page weight would be good for people who
1) are on a slow connection
and/or
2) have their download volumes capped by their ISP


Network bandwidth has to be paid for by somebody. Neither is it unlimited. Remember a while back when some ISPs wanted to bill Google for the bandwidth that their users consumed while accessing Google content? They saw Google benefiting from revenues generated by delivering ads over the ISPs' networks.


Google penalising unnecessarily high page-weight could be viewed as a Green move to moderate the ever increasing consumption of bandwidth.
It could also be viewed as Google saving themselves the bandwidth demands of crawling pages that are increasingly heavy.




Some years ago, I read a SciFi book in which communications bandwidth had become a very precious commodity. I forget the title, but the idea made sense.

Pish, you noobs should have been here before point to point teleporting

Without the beacon, you never found your destination when flying from the telehub...unless it was AT the telehub (and even then it was dicey since the telehubs were like little cities, 8 stories high, meant to trap you there forever.)

Which reminds me...get off my lawn!!

Search engines can record individual user actions and rank pages according to their apparent preferences, which is relevant to individual users. Page weight, however, is never relevant to either searchterms or users.

They could but, since loading time has nothing to do with relevance to a searchterm, they shouldn't have to do such things. It's not a search engine's business what 'bling', as you call it, is on a page, and using it as a ranking factor doesn't help their users in any way.

No response should be necessary, because no search engine should be doing it. But the most likely response, if it does make a difference in rankings, is to split larger pages into smaller ones and optimise each smaller page, but then you run into the problem that Google created years ago, where perfectly good pages can't even get into Google's index, regardless of how good they are in their field. That's another story though.

I didn't say that larger stores are necessarily better for shoppers than small ones although they often are. Suppose you want a ballgown for an event. If you could go to a store that sells a wide range of them, you are very likely to find one that you like, whereas if you go to stores that have only 2 or 3, you're more likely to have to visit a fair number of them before you come across one that you like enough to buy. Of course, you may find it in the first one, but on the whole I'd say that tiny stores means that shopping take longer.

Yes, I know.

Of course, but page weight has nothing whatsoever to do with relevance to a searchterm, and it's not a search engine's business to cater in favour of slow connections.

It could be, but it won't be. It will be seen for what it is - Google being typical Google by moving in wrong/bad directions.

Actually the fact that they were not here back then might be making this new feature irrelevant. If they don't know to fly to the beacon the beacon being there is useless.

I remember when I first got to SL someone had to explain that that big red line in the sky when I landed at the info hub was where I was trying to go to. Because I didn't know I spent about 4 days at infohubs and never found a darn thing I was looking for.

In my very early days I flew for ages in the direction of the red arrow, not knowing why it was on my screen. I didn't get to the end though. It was afterwards that someone explained what the arrow is. I must have had the map open and clicked on somewhere.

The GO TPs can be very good, but they can also be bad. For one thing, I don't get a beacon and, if I land where the item isn't, I'm stuck with searching for it. An example of the other thing is my fireplaces html page. The page is a list of fireplaces in the store that are set for sale, some of which are against a wall. On the other side of the wall are completely different items (Chaise Longues), and that's where the GO links for the fireplaces against the wall land people. Looking round from there, there are no fireplaces in sight. In this case, the wall seperates different parts of the store, but it could be an outside wall - in the sky. Not good.

Yeah this has long been an issue. I tested the Go links for two items, first item landed me outside the store, second item landed me perfectly. I say it has long been an issue because it happens when I TP from the TRU website, sometimes I end up outside the store.

OK, so they need a viewer change, a little sound clip of Zelda Rubenstein, "run into the light!"

It's probably to be systematic.
Perhaps an offset from the centre of the root prim of the object on its x oy y axis?

Alternatives:
For the keyword prim spammers: the solution is too obvious lol. If you were not unethical keyword spammers I would tell you.

For those worried about landings: Put your prims out in the middle of the store. I am doing this in my zindra store. Will eventually do my main store the same way.

If you had anything worth knowing, people might ask you, but you don't.

We can do that, but I don't see why anyone should need to rearrange the items in their stores to accomodate LL's lack of thought. Even then, only those who are aware of the problem can do anything about it, and I'm sure that the majority of people, to whom it would matter, aren't even aware of it.

My guess is that those links are auto-generated, to a specific offset from the product's root prim. Like "One Meter South of the center of the root prim". If that puts you in a wall, or outside the building, it's a mess. But in a lot of stores this can be a fantastic improvement.

I just saw this change myself in the last week, but I don't use search that often, so I don't know when it started.

It *would* be a fanstastic improvement IF it worked as intended. I just did another couple of GOs to fireplaces in my store, which is in the sky.

I did a search for "low prim fireplaces" (without quotes). I clicked on the #1 ranking, which is mine, and then I clicked on the first listing's GO link. I was on the ground below when I clicked it, and it worked fine - I landed in front of the fireplace. From that location I clicked the 2nd one in the list - to a firepplace that's only 2 meters to the right. I landed sort of inside the fireplace. So I TPed home (much higher in the sky than the store is) and tried the 2nd one again. This time I landed underneath the floor and fell to the ground. From there I tried it a third time and landed inside the fireplace again. The only way out of being inside the fireplace is to sit on something or TP out.

So it would be a really good improvement *if* it worked, but it doesn't work very well at all. Many places are on the ground so people won't fall but landing inside an object isn't very good.

Incidentally, I don't think it's a fixed offset because I landed in different places when GOing to the same item (under the floor and inside the item).

I can only think that whoever added the GO links did it because it sounds like a good idea, but without actually trying it out.

Now I need to consider fixing a landing point for the fireplaces section, so the people are guaranteed toy land somewhere sensible ... sigh.