This content has been marked as final. Show 194 replies
Excuse me for being extremely skeptical of this. If it doesn't work for people, how do you know the rate?...Because you can see if someone who goes to Classic has the right version of Flash installed - even if they opt-out of going. That is how that image on the right of the signin page works. It detects if you have flash, have an older version of flash, have no flash and gives you feedback to do what is needed. And yes it is a very small percentage of folks who don't have flash, but vocal and frankly important. And your problems with it are real, and even for me inside of Oracle, annoying. But I also know they are being fixed, and there is some value to some of those annimations. I KNOW for a fact why every annimation was added, and there was specific work to avoid gratuitous annimations. They should be there to add "context". For example, clicking an article and going to the article browser from the knowledge page. The annimation is supposed to communiate that, "hey the same list of articles I was just viewing is there on the right, so I don't need to go back and forth to click the next article". Stuff like that.
But having said that, maybe there should be a switch to turn off all of the annimations? I think there is one in Windows as well. Have you tried that one? What do you think?
And that is the same screen shot I took a year ago and filed a bug for!
But it is fixed in the next release (and the current support.oracle.com).
I have to say this was one of my favorite bugs that I saw (and no I didn't write this code).
It made me laugh, but not in the "hah hah" sort of way.
I think if you clear your cache it might go away. ;->
wbfergus wrote:Funny huh? But you didn't go to the URL for the MOS home page. I am guessing you went to a previous URL?
Okay, my position on the the new MOS Flash system is well documented, but most of my opinions expressed were base on earlier experiences, so I figured it worth a shot to see what has improved since my original impressions.
I am connecting from my Windows XP Home system (broadband cable) through an encrypted VPN to one of my office computers, and then connecting to MOS from the office computer across T3 lines. Just entering the URL for the login page took 45 seconds to load. After it did, directly next to the "User Name" field for login, it says "LOGIN_TEXT_NEW_1" with "LOGIN_TEXT_NEW_2" directly below it. This alone is simple Jr. Developer screwup 101. This kind of stuff should not be showing up on a production system, this is development stuff.
The reason why I say this is that screen with the silly properties files is a flash UI screen, and you can only get there from a deep link URL.
But that "embarrising" link is fixed in the new release. Your right is should never ever ever show up in production. Ever. ;->
Okay, I'm ready to login, start entering my id, and I fat-finger an entry. I look at the field, and WOW!, that text is really small! I re-adjust my glasses, squint, and have to move closer to the screen. My eyes aren't what they used to be, even 6 years ago before I needed glasses. I correct my mistake and finally press "Enter".Me too. I have been legally blind without glasses since 5th grade. Have you tried making the font bigger on the Settings > Personalization screen? Large works ok when you also have a nice wide monitor.
WOW! (again, but in a good way). I guess the delay to simply display the login screen was to pre-load a bunch of stuff, as the login itself no longer went through the marketing sales pitch 'splash screen', and to actually show me the first usable screen only took about 10 seconds, instead of minutes. This is a major improvement!- Right, the login screen you saw already had loaded the application, maybe from Cache. But I think it is faster with the new support.oracle.com, and will be faster yet with when you are migrated.
Next, I select one of my databases and look at what is presented in "System Health" (upper left corner). It shows 0 (zero) critical, 3 Warnings, and 3 Informational, but no details on what those are. I click on 'Patch advive" and everything is blank, so no help there and I go back to "Health checks". I move the mouse around some more, and still no details appear as to what the 3 critical and 3 informational warnings are for. Okay, this is pretty useless, so I create another remote desktop connection to my server and login to Database Control.-- When you say "No details" are you saying you clicked on the bar and drilled down into the details screen and there were no details? Or you never clicked on the bar?
I think all of the details should be on the next screeen. Try there and see what you think.
Okay, I now have absolutely no confidence in MOS. Database control show that I have 15 Critical, 21 Warnings, and doesn't even display a count for Informational. I am pretty darn sure that the numbers shown by Database Control directly on the server are going to be a heck of a lot more accurate than what MOS shows, especially since there is such a big discrepency between the numbers. So, this brings up the big question, "Why even bother trying to show this, if what MOS is going to show is wrong?". Maybe I might like this part, if one of the security geeks goes poking around to see which systems are how far out of compliance (so this one machine looks better than it really is), but formy own use, it is totally worthless because it is so inaccurate. At least Database Control agrees with MOS in that this server doesn't need any patches.DB Control as a) a whole lot more checks, b) because there is an agent a whole lot more it can check.
But I would be interested in knowing IF the ones checked by MOS are a subset of the ones DB Control found?
I don't know why you say the MOS ones are wrong, if we didn't look at the detail yet? Again, maybe just a subset issue?
I next go over to the "Service requests" section and try to find SR's I've created and also have the "Include Hard Closed" box checked. I first search for my last name and wait. Nothing gets returned, which I know is wrong, so I connect to another machine at the office and connect to "Classic". Quick and fast. I go to "Service Requests", and right away, even without conducting a search, it show my last one from August 5, 2009. I copy the SR number and go back to the first machine and search MOS for that SR number. I wait, and again, nothing is found.typing your name or a piece of your name returns no results?
Is this only when Hard Closed is checked?
That sounds like a bug.
I tried this on the internal version and it works (I even typed just a piece of a name so it would be doing a contains search), so I think this is working for the next release.
Okay, after this, conducting two simple, yet very common tasks, I have decided that MOS is much faster than it used to be, but it is far less reliable than the "old way" of using Database Control and "Classic Metalink" combined (which is still faster than MOS).I know I am not the only one reading these blogs. So, yes we are trying to address the issues. Maybe we should set up a "complaint" fest. I could host a web conference. You can (in tern) play around with the new release, you can all complain in person and we can either try to address the issue or file bugs. But I would just ask that you play nice, we might have a lot of folks who want to join the call!
I hope someone from Support management reads this and takes these points to heed, as MOS in it's current form are useless, except to show pretty pictures to high level management with no technical background or training.
I see that now you can tell when we are running one of the Firefox flash blocking plugins. In the last week or 2 I have seen a message in that upper right window, "Why you need to unblock Flash Player to use [MOS]". Whereas with SeaMonkey (I have not figured out if there are Flash blockers for that - since I don't use it that much, and rarely for sites with Flash - I make sure to avoid them if they hit the SM browser session with Flash "bombs" too much), it just states that Flash Player 9 is needed (which I do have installed on SM from back in the days when Flash was not such a "problem" with the few sites I used SM for, mostly internal), so that is not very indicative of whether my SM setup will run MOS (I think it does, but I only tried that with a CSI that I have very limited use/access for, so not much impact or usage).
The fact that we have Flash blockers should tell you something, i.e., we consider Flash a "problem" more than a "solution".
Interestingly, I have recently gone through a Forrester analysis of Oracle's new, 11g Fusion Middleware development platform release, and it seems that it would be well-suited for Oracle to use to develop MOS as a showcase for its features and usefulness - a high profile "eat your own dogfood" opportunity. Or is that what is being done by plugging Flash coding into the development framework? I am curious ...
Edited by: ROC on Oct 27, 2009 9:16 AM
I'm sorry I have to pick you up here.
--- And I think the support.oracle.com experience is much faster, no one wants to wait. Right ----
Umm. No. In fact Not at all. It's actually brutally slow on every system I've tried, Classic ML beats it hands down.
Things get even worse when you're VPN'd to a remote Desktop somewhere on a client site, then RDP'd inside the firewall, then open a VNC desktop to a Linux server, then have to open a browser at their end to access a silly Flash interface.
My Oracle Support is Oracle's Windows Vista, which was basically a complete obstacle to getting anything done, especially Oracle related, which is the primary reason business adoption of Vista is a single digit percentage. Can you image the outrage if Microsoft decommissioned Windows XP on everyone's PC and forced an upgrade to Vista that nobody wanted? This is exactly what is happening with My Oracle Support. My workstation can happily run the latest versions of e-Business Suite, Hyperion, SOA Suite and several virtual machines. The moment I log into My Oracle Support, my CPU goes to 100%, the fan kicks into to turbojet hairdryer mode and the temps rise 30 degrees Celsius. As long as I want to use my computer to heat the room, it is ok, but will be lame come summer...
RichMill wrote:Not from my chair, it isn't faster. 21 Seconds from click of Login screen to first dashboard using the Flash stuff vs 14 seconds for Classic. (And Flash still does not display on my Palm Treo.)
And I think the support.oracle.com experience is much faster, no one wants to wait. Right?
I wonder where the MOS acceptance testing was performed?
As I've said a few times, the concept is OK. But it seems that the usability testing was isolated to a set of cases that did not represent my environment or needs.
(And the lack of ack for Oracle's own UI dev tools is a serious kick in the dev tool communty's behind.)
I am completely baffled by the reasoning that is going into forcing the new support website onto my. My experience is like most others: "Classic MetaLink" is fast(er) than "My Oracle Support". For us, the old screens are a LOT faster. What is with that? Are they blind? I would think that Uncle Larry would notice this and do something (or not).
I just got the notice, again, about the retirement of Classic ML, and wondered if anything had changed from the one a few weeks ago wherein I noticed that an HTML version would remain available with the "core" functions for us non-Flash'ers. That is the good news. However, something that had not registered with me before is that the HTML functions will be within the new MOS - they really are retiring the underlying Classic Metalink site! Is that "bad news"???
There seem to be a lot of erroneous behaviors noted with MOS in this thread, and I have to wonder if those are in the core functions that we can access via HTML interface, or if they are artifacts of the Flash code, or some combination. Can we "preview" the HTML-on-MOS interface to test this out??
I completely understand the reasons to refurbish Classic MetaLink. IMO, it never really was that good of a website. The new one: holy cow! What a disaster. Larry? Larry? Do you know, or care, about the differences between "skinny guy in a suit" who never looks past the dashboard and the ordinary not-so-pretty-guy/gal who has to actually get some work done --- with YOUR software? If so, please, please, do something to make this necessary change a little bit more practical.