Pre vs HTC Hero

[gmanvbva]

I'm going to have to leave these subjects alone and concentrate on the Hero vs Pre topic.

I feel like we have gone completely off topic (in numerous different directions) and are bordering on completely derailing this thread.

[gmanvbva]

On topic.

Last night I tested tethering on the Hero and it worked perfectly. You have to follow a few very straight forward steps (to install the drivers and enable it) but it works nicely. From the boards it seems to be a mixed bag. It works for some and not for others. Worked for me right off the bat.

Tethering is one of the things that broke for me in 1.3.1 (using MyTether) and I rely on this. USB/BT tether appears to be fixed and possibly WiFi but the author is still working out the kinks and creating a new installer.

Also, over the past week my Hero is much better with both cell signal and EDVO signal.
Yesterday I was testing this at my home and while at a basketball game.
My Hero would access EDVO almost immediately and hold it. While my Pre would regularly cycle EDVO in/out trying to maintain a EDVO connection.

Edit: Multitasking on the Hero actually works very well. You simply hold the home key down for a couple seconds and it brings up a list of all applications open (by icon) and you can scroll (or touch) the application you would like to switch to. While at a basketball game I was surfing the net for bball scores and watching (and listening) NFL Live on the Hero (Miami vs Carolina). It worked very well. After a certain period of inactivity it would suspend streaming the NFL game in the background. To resume it all I had to do was switch to the application and press the play button and it started streaming again.

Battery life seems to be so much better with the Hero (compared to the Pre).

[grndslm]

Originally Posted by UF15

I think your main hurdle from understanding how computers work, and this is from a logical standpoint, is distinguishing what percentage of that 73% is solely due to graphic rendering.

Umm... Do *you* understand English. Engadget was already rendered. It dropped down to 20% I could switch from Engadget back to the Terminal and top showed 20% after the page was already rendered and Terminal had "rested". Then jumping to Engadget and pushing my finger around the screen like I would normally do got 73% once I switched back to Terminal. What do you think the 53% jump was due to? It was due to my frickin' finger. That's it. It's repeatable if you have even half a braincell.

Originally Posted by UF15

You see, you have to understand that just because the CPU is running at 73% (and I still believe something is wrong with yours), doesn't mean that all of that is from drawing and refreshing graphical frames. I think you just assumed that it is, and built all your other assumptions on top of that. Unfortunately, the foundation for your argument is wrong, and the CPU isn't being taxed because of graphical rendering as much as you had hoped.

If you can't figure this out on your own, you need to be fired.

Originally Posted by UF15

Saying 73% or 67% CPU usage means nothing unless you can prove it directly related to rendering what appears on screen. You have utterly failed to do this. Using that common sense you talk about, just think about all the other instructions, far more complex than graphics, a CPU is constantly tackling to keep every function of everything going.

You just don't understand a damm thing about a damn thing. How far back do we need to go? All the way back to teaching you English?? ... or should we jump into how the finger scrolling around Engadget is the cause of the 50% jump. I knew you were full of **it when you said "8%" was your entire test. Get off your high horse.

Originally Posted by gmanvbva

grndslm you did not make a single counter point based on anything factual or with any type of representative truths.

You're full of **it, too!! Why the frick don't you [QUOTE=grndslm]??!??

Originally Posted by gmanvbva

You obviously know nothing about CPU/GPU architectures or how operating systems employ them. Making statements (yet again) like that just show your true ignorance and arrogance.
At this point, your ignorance is only outdone by your arrogance.

Which statement?

Originally Posted by gmanvbva

PS: I think UF15 just caught you in a flat out lie.

Main Entry: lie
Function: verb
intransitive verb

1 : to make an untrue statement with intent to deceive
2 : to create a false or misleading impression

Thanks for the definition, yo. What lie?

Originally Posted by meyerweb

If the apps on the Pre, as it ships today, are really enough to meet all your needs, then your needs are extremely simple. I guarantee you that the currently available apps don't come close to what I get everyday with my P-OS Treo.

[QUOTE] me. Like I have said many times now... Shop Savvy is the only app I need that I can't get at the moment. All the rest of the issues I have solutions for myself.

Originally Posted by gmanvbva

Don't project your simplistic lifestyle onto all other smartphone users, please, as you obviously have no basis on which to know, nor even guess, the needs of millions of other people.

What do you need on another platform than a snappy OS that isn't wasting 20-40% CPU when you're just flicking things around the screen? I can't think of anything besides Shop Savvy... so what's your complex lifestyle in dire need of?

Originally Posted by gmanvbva

Grndslam clearly has no idea what he's talking about when it comes to processors, nor what a GPU will and won't speed up.

You're full of it, too?!? What don't I know? I gave numbers that were very much in line with my estimations.

Originally Posted by gmanvbva

He says:

but it appears he knows very little, as opposed to thinking he knows almost everything. He even know what each and everyone of the millions of smartphone users really need. Ignore list time. He hasn't had anything useful to say in ages.

I said that because Palm doesn't know sheet about making a big splash. If they knew what I know about selling, they'd have sold 5 times as many phones. Very simple. If they knew what I knew and what everybody else knew... they never would have even allowed the creepy lady to go past one commercial. That's just unexplainable. Sell the phone, not some lesbian's metaphorical thoughts!

Originally Posted by gmanvbva

To help you a little since you apparently can not grasp this.

The GPU is not rendering the web page for display simply because it shows it on a screen. The CPU is responsible for rendering the web based code (html, javascript, php, etc.) by the use of a rendering engine employed by the browser in use.

Umm... the page rendered before I started flicking my finger around to make it jump the 50+%. The CPU should render the javascript... but the GPU should have no problem whatsoever with flicking the page all around, zooming in, flicking between cards, etc. Changing between cards is something the GPU could "render". Drop-down menus are something the GPU could help "render". The GPU would also help with "render" web pages in a graphical [output] sense, nut an instructional [input] sense. Even the camera would "render" pictures quicker. Rendering is not solely limited to pulling up an html/javascript/css page, ok?

I know the difference... and the page was COMPLETELY DONE RENDERING in an instruction sense before I start clicking the Engadget page around. I had two cards open. You can all try this out for yourself. Try as hard as you can... you'd probly be able to get it up to 100%. I am too busy considering I just shot down a "constant 8%" claim with 4 minutes of testing... yet you guys missed that.

Originally Posted by meyerweb

This is hilarious! grndslm provides no, nada, none, zed, zero research of his own, then demands others provide it. This man has an ego the size of Montana.

Umm... I provided it above. I gave an estimation of 40% just as a [what I thought was reasonable] guesstimation. And what do you know... I get 20% right off the bat just from switching between cards and 50% right off the bat with scrolling in the web app. Did you miss that part or what?

I performed the test just like any logical person would have done.

Originally Posted by meyerweb

Can we please get back to talking about the Hero. There's a small chance it could be my next phone.

Dear God... I hope you don't care to continue discussions with me if you won't take my word.

[gmanvbva]

Please fix your quotes.

[grndslm]

Is that right? Think it looks perfect to me. Pretty sure I responded to every statement.

After having some more time to play with just one browser and one page of PreCentral loaded in landscape, zoomed into column width.... 80-90% CPU no problem when scrolling at a decent rate!! That is definitely my most preferred way to use the browser. And like I said from the get-go... the Multi-Mod patch would be worth your while. I wasn't joking when I said scrolling in the browser is the biggest CPU tax you have on such a device. This is common sense if you'd just pop a top and look for yourself.

A 60+% jump from scrolling in a browser. Wow. I thought 40% was pushin' it.

[gmanvbva]

You have quotes attributed to the wrong individuals

[grndslm]

Help me out, buddy. Point it out!

[UF15]

Originally Posted by grndslm

Umm... Do *you* understand English. Engadget was already rendered. It dropped down to 20% I could switch from Engadget back to the Terminal and top showed 20% after the page was already rendered and Terminal had "rested". Then jumping to Engadget and pushing my finger around the screen like I would normally do got 73% once I switched back to Terminal. What do you think the 53% jump was due to? It was due to my frickin' finger. That's it. It's repeatable if you have even half a braincell.

I find your insults to be desperate and distasteful, especially since I am trying to help you understand how computer architecture works. What I see from you is a lack of understanding, or perhaps an overly simple and naive understanding, of just what the GPU is responsible for.

For starters, you still have been unable to actually prove that you are seeing sustained usage of 73% from the CPU from just scrolling. What you are likely seeing are single bursts and peaks, and even then it concerns me that you are either reading the data wrong or your Pre is faulty (probably both). You then work under the premise that the only thing involved with scrolling a screen of rendered HTML in a running app is redrawing pixels on the screen. That is an incredibly ignorant and naive understanding of what is going on and even the entire GDI system.

So save your insults. I am trying to have an informative discussion here about something we both want; GPU access. I want it simply because I think it will make animations and scrolling SMOOTHER. However, I am not so myopic that I think it will relieve the CPU so much that it would as much as double the speed of the Pre from its current state.

Originally Posted by grndslm

If you can't figure this out on your own, you need to be fired.

I have figured it out. I monitor the CPU usage, and I also see how much of it is because of drawing pixels. You haven't. You just see CPU usage, and even then I (and everybody else here) know you are just fabricating those numbers.

Originally Posted by grndslm

You just don't understand a damm thing about a damn thing. How far back do we need to go? All the way back to teaching you English?? ... or should we jump into how the finger scrolling around Engadget is the cause of the 50% jump. I knew you were full of **it when you said "8%" was your entire test. Get off your high horse.

Frankly, I think I have displayed a better understanding of the English language than yourself in this thread . . . but I digress.

I also have mentioned above my qualms with your simplistic "Engadget Scrolling" test above, which you have ignored for multiple posts. Look at other animations on the device, ones that require less user interaction, and then monitor the CPU spikes to see if what you have claimed is true or not. I have not been able to replicate your claims on either of the devices I am working with, and have found my data to be almost opposite of yours.

Again, we both want the same thing: GPU access. We both agree that it will help the overall snappiness of the device and rid us of annoying choppiness that plagues WebOS. My points of contention over the issue have been your ignorant and unrealistic ideas of how much of a difference it will make, how easy it is to implement and how soon it will be implemented. I've explained to you what we can realistically expect GPU access to do to the device, and you have made outlandish and mightily contested claims to the contrary. I, and others that are smarter than both of us, have also explained how it isn't just a matter of switching the GPU on, that there needs to be javascript access to the GPU and then a lot of things will need to be rewritten so the GPU can understand and take over those tasks. Lastly, Palm is a tiny company that does not have the resources to come up with a new technology that will allow javascript to do these things right now. There are only a few hopes for us that it will even be feasible within 12 months right now. One is the single GPU guy they just brought in (that is a lot of work for one guy) and the other is WebGL, which we aren't even sure if it is capable of doing the things we talked about right now, or doing it with decent performance.

You have approached this topic with great hubris and a closed mind. You are setting yourself up for disappointment.

[awrnsmn]

I have the Hero and Pre, the Hero at this time runs circles around the Pre. I still want my Pre to become the device we all are hoping for. (hoping to get back on subject)

[swagner]

Day 2 of my Hero after a few months with my Pre. No hacks, patches or rooting. it runs smoothly, battery runs circles around the Pre. he bigger screen is nice and the build quality is better than the Pre.

Yep-- the OS is a bit tougher to navigate-- its a pain to get to settings and such. More diverse apps are nice to have access to. If the Pre were built this well and the app catalogue had more it would be no contest, but right now- I'm keeping the Hero.

Now I just have to figure out how to root/tether (unless I can figure out how to tether w/o root).

[gmanvbva]

You can tether without root. At least I was able to. If you need help, shoot me a quick pm and I can give you a link and let you know what worked for me.

[Brain_ReCall]

Originally Posted by grndslm

Umm... Do *you* understand English. Engadget was already rendered. It dropped down to 20% I could switch from Engadget back to the Terminal and top showed 20% after the page was already rendered and Terminal had "rested". Then jumping to Engadget and pushing my finger around the screen like I would normally do got 73% once I switched back to Terminal. What do you think the 53% jump was due to? It was due to my frickin' finger. That's it. It's repeatable if you have even half a braincell.

Wait wait wait, you're using the Terminal app to do your benchmarking? Seriously?

First off, top itself does take CPU time. Second, not to knock the guys at webos-internals, I don't think the Terminal app was really designed for efficiency or speed. (Incase you're unaware, the Terminal app works as a browser plugin, taking command outputs and rendering them pixel-by-pixel. I'm trying to look but I'm not sure what their refresh rate is.)

If you're going to perform any sort of meaningful benchmark, then you need to SSH into your Pre and run top through that. There is just no way in hell your Pre is "idling" at 20% unless you have some sort of active app, like, say, THE TERMINAL APP! On top of that (bad pun), switching apps brings in some other possible effects, such as webOS swapping memory.

Originally Posted by grndslm

Somebody's a pretty crappy computer engineer.

Please leave the computer engineering to the actual computer engineers.

[berdinkerdickle]

<<closed>>

Some of you guys need to learn how to converse without having to violate Forum Guidelines

It shouldn't be that difficult.