All threads show up as read.

[~graffiti]

Is the CSS on the site messed up? I'm not getting bolded text for unread threads plus "updated" isn't listed on threads that have been updated.

But some of the ones I've already visited are a little lighter blue than the others.

[Claudio_González1]

Jochem van Dieten wrote:

Claudio González wrote on 11/04/07 15:02:

Do you happen to have similar measurements of downolad times?

I wouldn't even know how to properly define download times. Time until

the first byte? Time until the last byte? But is that the last byte of

the first document or the last one? With or without caching? Do the

extra requests that are delayed to fill all the additional widgets count

or not?

BTW, the criterion for down that was used was "first byte takes longer

than 8 seconds". That 8 seconds was taken from the DW manual as the time

people are willing to wait before they move on and was about 4 times the

average time to the first byte.

I know this will sound terribly primitive to you, so I apologize beforehand.

You must be aware that a far from negligible fraction of the participants in these forums lack the technical knowledge to understand concepts that you manage so easily and well. I am, of course, one of them. So, for me, the "download time" of a web page is the time between the moment I click on its bookmark, and the moment my browser indicates that the page is ready, for example, by indicating "Done". Before the page is ready, it is simply not usable; for example, if the page for posting messages hasn't finished loading, I cannot write any text in the box.

This is the time that matters for me, and I must say that I find no noticeable difference between now and when Read Tracking was working. So, when asked if I want RT back at the expense of losing whatever speed increase we may have got, my answer is a very large YES.

I have no idea how the decrease in downtime that you mention affects the time my browser takes to announce that a page is ready; all I can say is that these forums are slow even with this lower downtime. I think this is why several participants who haven't understood the question we were asked have been saying that we must have both, RT and speed. And this is also why several of us have been asking to recover RT. Imperfect as it was, it did save a lot of time for all of us who navigate the forums through browsers.

And it worries me that, after being asked to vote whether we want RT back or not, we have not heard anything further about the possibility of getting it back,

[Phillip_M_Jones]

If Jive would quit trying fix this model -T of a system they have foisted upon us and  move up to the latest 4 or 5 x they currently at.

I've been to Jive main site to read at JVD request. I don't exibited all the strange problems we have. They are naturally using the very latest JIve version that out.

Another thing they are probably doing is using adaquate amounts of Server Farms to cover Traffic. Something Adobe is not doing.

So until we have the latest version of software for these Forums. And Have enough Server  Space with as much allocated for reserve, and they need for current traffic. Nothing is going to get fixed.

The odds of either of the above being done, is bewteen Slim and None. And, Slim Left a long time ago. Adobe is just pay lipservice and not trying to do anything. They are not interested in doing anything. They don't care. JVD can defend all he wants. They don't care and never will.

[Harbs.]

Claudio,

Jochem was talking about downtime, not access speed. Two different things...

Harbs

[Jacob Bugge]

In addition to what Claudio said, here is a copy of the two key posts by John and excerpt of the key post from Jochem specifying the difference in terms of downtime, thus showing the magnitude of (that part of) the performance issue:

Post #10 by John:

Jive turned off the "Read  Tracking" in response to some performance issues. Hoping to get it back  on after testing or a fix. Doesn't make me happy.

Lighter blue means that you have visited that link recently, not related to the new/unread messages.

Post #13 by John:

Read Tracking (identifying threads with new content since you last visit) does have an impact on performance.

Vote  time. Which is more important? Read Tracking or Speed? The initial hope  with read tracking is that it would do something similar to the old  system that kept a high-water-mark for each forum so you could just pick  up where you left off. But it never lived up to that.

Post #118 by Jochem:

Without read tracking, the average downtime has decreased from 2% to

0.5% on a daily basis.

Conclusion:

John asked us to vote between Read Tracking and Speed, and one key element of speed has now been quantified as a decrease in downtime of 2% > 0.5%.

We have not seen any quantification of download times.

Almost all voters have voted for Read Tracking, even with the knowledge that it  would  have an impact on performance to get it back, and I am convinced that all of us should prefer Read Tracking even with a much more significant difference in download times than in downtime, and therefore still cast the same vote.

For my part, no difference in download time would change my vote.

For most of us voting for Read Tracking I believe the lack of Read Tracking means far more delay than even a tenfold increase in download time, apart from the annoyance and needless effort, because:

1) We shall have to waste time considering/trying to remember whether each thread may be relevant (of interest and/or whether we have followed and maybe posted in it) because there is no certain marking of any kind, regardless of normal browser settings; at least in FF, the normal change of link appearance only occurs sometimes and is therefore of no help whatsoever;

2) After wasting time on 1) we shall have to waste time on opening each possibly relevant thread, read the last post(s), and consider/try to remember whether there is actually anything new, and if inconclusive we shall have to waste further time until certain.

By the way: In the context of the issues we are facing here, it is extremely hard to see that a difference in downtime between 2% and 0.5% could have the slightest significance.

[Noel Carboni]

It would appear that good and helpful forum users are having a very reasonable conversation here without any further attention by the people running this forum.

You Adobe folks may feel you own this forum, but the users who help other users here (in total) have invested FAR MORE in it than you EVER WILL.

The forum seems to be spiraling into chaos. I think we deserve an update. Nay, you OWE us an update.  Or you can just ignore us. We'll all probably go away sooner or later.

Do we need to use a bigger font to make the request? If only that were possible!!!

You think I have an attitude? Just what have YOU done for ME lately? Meanwhile I've been helping your Photoshop users left and right, helping keep your paycheck coming so you can feed your kids.

-Noel

[John Hawkinson]

Noel, it is comments like that that turn reasonable people into enemies.

Even if you are right, how you say it can matter more than what you say.

It's easy to imagine thinking, "They don't pay me to put up with this abuse!" and just ignore you, regardless of the legitimacy.

Even I, a relatively new but active volunteer on these forums, am already disgusted by this thread.

Perhaps sending a private note (or a more formal or traditional business communication) might make more progress. It might also be a good forum to apologize for past behavior.

[Noel Carboni]

John Hawkinson wrote:

Noel, it is comments like that that turn reasonable people into enemies.

Even if you are right, how you say it can matter more than what you say.

It's easy to imagine thinking, "They don't pay me to put up with this abuse!" and just ignore you, regardless of the legitimacy.

Even I, a relatively new but active volunteer on these forums, am already disgusted by this thread.

I appreciate your advice on how to address people on a public forum.  However, when I intend to be abrasive - and it's not often - I'll be as abrasive as I intend to be, thank you.

We're reading here that by not maintaining a "thread has been read" flag for each user that the forum software is more stable.  Fair enough.  So fix the root problem and roll out a new version.  Or eliminate the feature.  But let everyone know what's going on, especially, if the users have been asked to share their opinions on whether the feature is worth saving.

This is not some entertainment portal that really doesn't matter to Adobe.  It's a public face of Adobe.  Adobe's own phone support people sends folks here.  What does it make Adobe look like when the forum fights their customers every step of the way?  You think that makes people WANT to go buy Adobe software in the store?

My own products depend in part on Adobe's products being successful.  It is more than an idle interest that drives me to care about how well Adobe is doing here.

-Noel

[Harm_Millaard]

Google has a great tool to measure the speed with which an internet page loads. It is called Page Speed and I use it every time to measure the performance of the pages I put on our website, PPBM5 Benchmark

Despite the fact that each of these results pages is currently about 1.8 MB in size, the page speed after optimization is around 85 and the other pages are around 92. Other sites, like DVInfo.net achieve figures in the 80's or 90's. Various other fora achieve figures like 92 for Creative Cow for instance. Adobe does not achieve anything over 71 without RT and with RT on it was around 69. Googles own pages achieve around 90 - 95.

This lackluster performance is with RT off, while DVInfo.net and Creative Cow achieve the much better performance with RT on.

It is appalling that Adobe can not give users better reliability, better performance and the much needed functionality of RT.

[Claudio_González1]

Thank you Harm for providing data that cannot be so easiliy dismissed as "anecdotal statistics that one person creates  by downloading occasionally from one location using one system either.". It is quite clear from your figures that these forums are very slow, and that the speed gain withour RT is marginal.

I find it most peculiar that we have lost so many features for reasons of solving different problems. I would think that the logical measure would be to attack the problems themselves, not to cripple the forums.

[Jochem van Dieten]

Harm Millaard wrote on 11/04/08 07:42:

Google has a great tool to measure the speed with which an internet page loads. It is called Page Speed

Page Speed does little to measure the speed with which a server returns

content. It measures how the structure of the HTML, JS and CSS in

combination with the caching configuration and number of the loaded

assets influence the speed with which a browser can render a site. Since

switching RT on or off does not influence HTML structure, JS or CSS,

number of assets or the configuration of the HTTP headers on the CDN

this tool is not supposed to notice any difference in page speed

measurements.

If you check the suggestions the tool offers you will notice that "use a

faster server" isn't one of them.

That isn't to say the tool doesn't give any suggestions for improvement.

Some of those are even improvements you can plug straight into AdBlock

or similar tools: stop loading the default avatar for people who didn't

specify one, don't load the web analytics stuff etc. My browser blocks

about 1/3 of all HTTP requests to these forums and I could optimize it

even more. In fact if you search the forums you may find some posts from

Chris Cox pointing out exactly the same issues as Page Speed reports

long before Google released the tool.

[Noel Carboni]

For what it's worth:

Others may not agree, but from where I sit the forums both feel a good bit faster and are more consistently available since the Read Tracking was removed.

-Noel

[the_wine_snob]

Noel,

I have not observed (seat-o-the-pants, but no real timing) an increase in speed. However, I am usually accessing via a wireless G connection, so perhaps the increase, that others see, are just not available to my laptop.

As you and I both use these forums in about the same way - answering others' questions - the Read Tracking is very, very useful to me, and I would suppose to you, as well. As an example, and looking only at one particular forum, in the Premiere Elements forum, the top 6 most-prolific posters spend the vast majority of their time, helping others. To not have Read Tracking makes that task very difficult. One has to search the Last Post data, to see if they were the last poster, or if someone (the OP maybe?) had posted to a thread more recently. Then, one needs to look at the Last Post, once the thread is opened to make sure that it has been seen, and responded to. One could turn on e-mail notification, but in my case, and in yours too, that would mean a ton of e-mail.

Before the elimination of Read Tracking, but WHEN the forums were working OK, I did not find the "normal" speed all that bad. It could have been faster, but it was not a big issue with me. Now, when things went wrong, the speed seemed to be a secondary issue to aspects, such as the servers being down, or the constant log-outs.

Just my thoughts,

Hunt

[Noel Carboni]

Oh, I agree with you, as I have stated in the thread above somewhere.  Read Tracking is a great feature to have.

About a month ago I moved up to a super fast computer, IE9 (also a month+ old) is faster than ever, and I have a fiber internet connection.  This was all in place and I'd gotten used to it well before the Read Tracking was disabled.

I didn't do any solid measurements before Read Tracking was turned off, using the F12 Developer Tools feature of IE9, but this feature can show the detailed timings of retrieved data.

Just refreshing the top page of the Photoshop Windows forum right now, after having cleared the cache, I see I have had to receive well over a megabyte of data (!!), and the whole thing took about 3 seconds total to load up.

However, with the cache intact the whole thing takes about 1 to 1.5 seconds to load.  I can honestly say I virtually never saw load times under several seconds with Read Tracking enabled.  It was usually 5 to 10 seconds.  Painfully slower than it is now!

The other thing I'm not seeing lately at all is an attempt to connect to the forum met with a "We're sorry, we're offline; try later" response.  I suspect this is the reduction in "Downtime" mentioned above.

It's interesting that the 69 different requests for data are made to no fewer than 3 different servers...  forums.adobe.com, wwwimages.adobe.com, stats.adobe.com.  There are even two requests to nothing at all!  (http:///).

-Noel

[the_wine_snob]

Noel,

When things were working (in between major slowdowns), I never felt that the speed of the Adobe forums was that big an issue, and I have the bottleneck of Wireless G to the laptop. Again, and just based on my "seat-o-the-pants" observations, I'd say that page refreshes were in the 1 - 2 sec. range for me.

Though it's been said above, please bring Read Tracking back. [I know that work is progressing in that direction, so I will just sit patiently, and spend minutes, instead of seconds, figuring out if I have read a certain thread already.]

Hunt

[Noel Carboni]

You have knowledge that work is progressing in the direction of restoring the unread indication?  That sounds comforting, though I haven't heard a thing here to support it.

If not, time is passing.  Soon we will all only vaguely remember having the forum help us help new posters.  Then it will be all but forgotten, passed down in beer song only as a great legend of the past, a time when technology actually worked.

-Noel

[Claudio_González1]

Noel Carboni wrote:

...

time is passing.  Soon we will all only vaguely remember having the forum help us help new posters.  Then it will be all but forgotten, passed down in beer song only as a great legend of the past, a time when technology actually worked.

-Noel

You mean, like the previous version of these forums? There are some of us left who still remember...

And I haven't read anything here hinting at work in progress to restore Read Tracking.

[Noel Carboni]

Ooh, ooh, it's back!  At 8:36 pm EDT on 4/13 I'm seeing bold topics again!

Ooh, ooh, let's see if everything's slowed down again...

-Noel

[Noel Carboni]

Speed's not too bad, but then there doesn't appear to be a lot of folks on right now...

Might be a hair slower than it was a few hours ago, but the jury is still out...

Pages are all coming up for me in a second or two.

On trying to post this response, I saw one of these for the first time in a few weeks just now, though...

-Noel

[Guest]

since we're voting and nothing seems to be getting done, i say we go back to webcrossing's webx and be done with it.

[Claudio_González1]

dave milbut wrote:

since we're voting and nothing seems to be getting done, i say we go back to webcrossing's webx and be done with it.

And what makes you think that something will be done about this proposition? (which by the way is far from new; as far as I can remember, there were many persons asking asking the same thing during the testing stages of this "improved" form of the forums, and for some months after their implementation).

[pwillener]

Thank you for restoring the original functionality!

[Jim_Simon]

Halle-freakin'-lujah!

[Jacob Bugge]

Thank you very much, John.

What a relief.

And the forums seem a good deal faster, too, at least now.

[Jacob Bugge]

Over here, all forums have definitely been faster today, with the Read Trackníng, whatever is measured or experienced elsewhere.

In the past we have seen quite different forum behaviours from different places, and usually I am in one of the lucky corners.

Also, there might be something else fixed, along with Read Tracking.

[Noel Carboni]

Yeah, it's not bad, and it's not consistently slower, just intermittently so for me.

Out of curiosity, Jacob, what's your latency, as measured per my note above?

-Noel