Progress, but not a complete fix
Here’s an update on my blog and what I now have come to call the caching problem.
I’ve been working on a fix—talking to various people at my host, Googling the situation and seeing what people on the Web have suggested—and at this point various solutions seem to have caused an improvement but not a complete solution.
Right now the most common form it seems to be taking is that recently-cached versions of the blog—or incompletely-updated versions–keep show, particularly with the comments section.
For example, a post might say it has 10 comments. But when the viewer clicks on the comments to read them, it then lists 14 and shows the contents of the full 14. But if the viewer clicks on the “refresh” arrow-icon, the counter gets up to speed with 14 and stays that way without reverting.
So the blog is essentially usually all here, there’s just a time lag with the counter and sometimes also with the display, although the latter is less common.
Things like this drive me nuts, because the blog is precious to me and a functioning comments section is precious to me. This has been going on for quite a few days now. Fortunately it’s not a serious problem—like, for instance, if the blog were down entirely (perish the thought!).
So I’m asking for your patience, and strongly suggest that you refresh every now and then. And don’t refrain from commenting—your comment will show up sooner or later, usually sooner. The counter might not acknowledge it right away, though.
One hint at what’s going on is that my host said they slipped in some cache plug-in about a month ago (without telling me), and the plug-in seems not to be compatible with the blog. So last night they removed it. I thought that would be the complete fix, but unfortunately there are still lingering issues.
My apologies for any inconvenience, and rest assured that I’m still working hard on it.
[ADDENDUM: For example, after I wrote and published this post, it didn’t show on the blog until I clicked on “refresh” and then it showed up immediately. But it should be showing up right away without the need to refresh.]
Oddly enough, during the past few years, I’ve had this problem 5-10 times with Neo’s blog, but have not had recent problems until today, January 10, 2017.
In each instance, I’ve been unable to view comments from the most recent day, but have had no problems with comments from previous days.
I’ve never had this problem with another blog. The problem isn’t specific to a particular operating system or browser. The problem isn’t susceptible to anything I can do to my own computers. Eventually, the problem disappears.
I’m not formally trained in IT, but simple logic suggests that the problem resides on the server that runs Neo’s blogging software. The operators of that server have installed the blogging software in a way that interferes with something else on their server. This is a longstanding, irregular problem that does not uniformly affect blog readers. It can’t be caused by a recent change to the server, unless that change has also been made in the past.
If I were Neo, I’d look for someone else to manage my blogging software on another server. If that’s too much trouble, or too expensive, then maybe the problem can be accepted as a small symptom of the cognitive breakdown that hits us all as we age. Thank you for this public service.
Please,
Avoid going to Facebook commenting.
Ira
So I’m asking for your patience, and strongly suggest that you refresh every now and then.
Sorry you’re having to deal with this kind of tech-trivia, Neo. Another thing that might help some readers is to flush the browser’s data (eg “Clear Browsing Data” under the “Chrome” menu item).
The problem is caused by faulty communications between a) the blog software, and b) the browsers connected to the site. The blog software is trying to be very clever and accomplish 2 competing goals at once:
1) Ensure that what a reader sees displayed on his browser exactly mirrors the state of the website (which changes rapidly as people comment and new posts are added) at the time when he connected or did the last refresh, and
2) Minimize the amount of communications sent between the website and the browser (mostly data downloaded from the website to the browser).
The simple, brute-force method would be for the website to just download everything each time a reader hits “Refresh.” But that would be very inefficient because probably over 90% of the data hasn’t changed since the last “Refresh” (eg a blog post from 3 days ago — and comments — has not changed at all, so there’s no need to send it again).
So I believe what it’s trying to do is to determine the differences between the current state of the website and the cached version held by my browser, and send only the differences, so my browser can synchronize its cache to mirror the current contents of the website.
As you might guess, this is a complex operation that can easily go wrong if the blog software is not extremely careful about how it does this (eg forgets to send the current number of comments). And I think it’s made even more complicated because there are many different browsers (and versions of each), some of which might manage its cached version slightly differently and/or have bugs.
I’ve tried viewing comments with several different kinds of computers and about every kind of browser that’s available. All have had Windows OS, so I can’t speak for Apple or Linux users. Clearing my browser cache, internet temp files, etc. has never had any effect on this issue. But that doesn’t mean that browser caching is unrelated to the source of the problem. I’m way out of my depth here, but I think that caching goes on at a few different levels between a commenter’s typing and my reading.
This problem may have worsened recently, but it’s been going on for a few(?) years.
P.S.
I forgot to mention that, at least for me, the problem with comments has now disappeared. I never had any problem viewing Neo’s posts.
Cornflour wrote:
Clearing my browser cache, internet temp files, etc. has never had any effect on this issue. But that doesn’t mean that browser caching is unrelated to the source of the problem.
My suggestion about clearing the browser data was just a guess, but evidently a wrong one.
I think my explanation above is plausible, but since then I’ve thought of others that are also plausible.
I’m way out of my depth here, but I think that caching goes on at a few different levels between a commenter’s typing and my reading.
I’m sure you’re right about that. These things are almost always much more complex than you’d imagine. I’m also “way out of my depth” here, but I’ve done a lot of software debugging, though nothing like this.
This problem may have worsened recently, but it’s been going on for a few(?) years.
I think I noticed problems occasionally, but nothing serious enough to make a big impression on me. I had been using Chrome on a Linux machine and now on a Mac for the last several months. That might be relevant.
Sporadic bugs can exist in software for many years. My guess is the source of the problem is in the blog software, the company whose servers host this blog (including their software) or some weird interaction between the two.