Any Way to DESTROY My HTTP Version?

[Sally Stitts]

According to the GSC, it is taking me forever to get to 100% HTTPS.
I finally got down to a HTTPS/HTTP indexed page ratio of 450/8.
Now, it is GOING bad again - 450/10.

The Screaming Frog says I am good to go. No HTTP references anywhere.

Is there a way to tell Google "Kill the d@mned http version - ALL the remnants1"? Make them GO AWAY FOREVER.
There SHOULD be a way, there MUST be a way, to eliminate the parasitic old junk.

With HTTP, I had 200K incoming links.
With HTTPS, it is now down to 22K, and continuing to drop rapidly.
My traffic is dropping massively each month.
Was switching to HTTPS a total mistake? Looks like it. EVERY TIME I take G's advice, it never fails. WORSE.
.

[keyplyr]

Any Way to DESTROY My HTTP Version?

You should not have an HTTP version. HTTP or HTTPS are not physical things. They are protocols, a language. This is how the browser (or other agents) talks to the server.

If you have a properly functioning redirect for all pages, you have successfully done your end of it. Now leave it alone.

Plus... as far as traffic, it doesn't matter if the index is showing the old HTTP page. Your redirect takes them to HTTPS.

You have no control over any Search Engine indexing process. There are a lot of things going on with Google, Bing, etc. Indexing is likely taking longer right now.

You could make sure your sitemap.xml is corrected to show only HTTPS link paths and resubmit to each SE, but it is still up to them to index those pages.

With HTTP, I had 200K incoming links.
With HTTPS, it is now down to 22K, and continuing to drop rapidly.
My traffic is dropping massively each month.

You still have the same amount of incoming links, they are now tallied by protocol at GSC, but they still exist at the source and they will get redirected to HTTPS. This is not responsible for any drop in traffic.

Some older browsers (IE on XP for example) will not be able to load your HTTPS pages, but other than that, there is no drop in traffic inherent to the HTTPS protocol.

Was switching to HTTPS a total mistake? Looks like it. EVERY TIME I take G's advice, it never fails. WORSE.

Note: you really need to do more reading. There's no reason for this level of misconception about HTTPS at this point in time.

Related links:

What Will Happen if I Don't Switch to HTTPS? [webmasterworld.com]

Downsides of not using HTTPS [webmasterworld.com]

Why HTTPS Matters [developers.google.com]

[tangor]

What he said. If your ducks are in a row you are gold. BTW, you don't really believe every thing GSC tells you, right?

Most of this garbage comes from external links you have nothing to do with and can't control, and is LARGELY TRANSITORY in the first place.

HTTP redirected to HTTPS erases the HTTP as far as your visitor is concerned. G, on the other hand, has its own problems and you can't fix those either.

[not2easy]

You should have the old http domain listed in GSC and the new https domain also. When you view the old version there should be fewer and fewer indexed URLs as they increase on the new https version.

Be sure that the sitemaps have https URLs and if they are referenced in robots.txt they are referenced as https. As long as everything is 301'd to the new URLs they should not be accessible as http and incoming links end up at the new URLs.

[Sally Stitts]

You should have the old http domain listed in GSC and the new https domain also.

I do.That's where the data is coming from. And by "version", this is what I meant (GSC protocol).

As long as everything is 301'd to the new URLs they should not be accessible as http and incoming links end up at the new URLs.

I will be talking to my host today, Sounds like the problem lies there. It is a "plesk" system.
I wish I understood it. More importantly, I wish my HOST understood it.

I was told that I didn't need to 301 in my htaccess file, because the redirect is done in the plesk panel. "All taken care of". I think not.
My lack of programming skills is my special form of SELF-EXECUTION. D@MN IT. This is not what I want.
How many hundreds of dollars must I pay someone here to get me set up on a new host that works?

Two days ago -
HTTP version = 2 incoming links (for 2 months)
HTTPS version = 22,000

Today -
HTTP version = 4,433 (unbelievable) Sound like a host problem to you?
HTTP version = 20,000

I'm just about done with this game. Between host issues and Google issues, it is no fun any more.
The more I work, the faster I go backward.
Thank you folks for the confirmation. Can anyone suggest a host THAT WORKS? Been with them since 2003.
From $500/ day in 2013 to my current $5/ day. The switch to HTTPS killed me, and I don't know how to fix it.

And just to add insult to injury, I paid my SSL certificate (GeoTrust/ RapidSSL SHA256 CA) for 3 years in June 2017, and now Google tells me that my cert is NO GOOD, and must be replaced ASAP, due to some Chrome BS. Screw Chrome.
.

[edited by: Sally_Stitts at 2:18 pm (utc) on Jun 5, 2018]

[not2easy]

It sounds like you are not handling your .htaccess file yourself. Since the Apache default for rewrites is a "302" temporary redirect unless the 301 flag is added, automated redirects often are not permanent. If you have not checked your own headers, you can't see how it is set up. Wherever you host, if you rely on the host for managing these things you may have the same problems. It might pay you to learn to manage your own .htaccess file. The Apache forum can help with specifics: [webmasterworld.com...] See the Library in that forum - under the Forum Options Button at the top of any thread.

There is a current discussion of affordable hosting here: [webmasterworld.com...]

Generally, you get what you pay for so unless you're paying for a managed service package, you should be learning to manage your own settings.

[MayankParmar]

Not very related, but you can use HSTS, to ensure that the HTTP pages are never indexed even if if something goes wrong in your setup.

[lucy24]

Never mind about who issues the redirects; that's a separate question. Are the redirects being sent out? For this you don't need to consult your htaccess or control panels; you need to look at your access logs. Most sites maintain separate logs for http and https, making it very easy to see. If it's all one file, you can also test it yourself by requesting http://whatever-it-is for a few randomly selected URLs--both page and non-page files--and ensuring that you end up on https.

Note, too, that just because something doesn’t need to be done in your own htaccess, doesn’t mean it can’t be done in your own htaccess. A common example is with/without www. A host may let you pick a preferred form, but this just means that they have their own redirect in addition to--not instead of--anything you might do at your end.

[Sally Stitts]

you need to look at your access logs

My host does not provide these.
Also, MOST OF THE DATA is missing from my AWStats, such as
Visitors/ Domains,
Robots/ Spiders,
Operating Systems,
Browsers,
Connect to Site From, etc.

So, I don't have a lot to work with, as you can see.
Today, I will find the necessary code to add a 301 redirect to my htaccess.
All it has right now is - ErrorDocument 404 /custom404page.htm, which I was forced to add.
WAY TOO LATE, but I might be able to stop the bleeding. But since I have already "bled out", I don't see much hope for recovery.
I "screwed the pooch", and I think the pooch may be gone forever. All my hard work = a withering death. Expletive.
.

[keyplyr]

In all respect Sally, in today's environment, you must use a server or hosting service that supplies access to raw server logs. There are just too many things you absolutely need this data for.

You cannot rely on software like AWStats to give you enough information. These programs miss too much.

[Sally Stitts]

keyplyr - Yes, indeed. I am finding this out, painfully. I have been too trustful, for far too long.

Is this the proper code to add to the top of my htaccess file?

## redirect to https:// ##
RewriteEngine On
RewriteCond %{HTTPS} off
RewriteRule (.*) https://%{HTTP_HOST}%{REQUEST_URI} [R=301,L]

Thank you.

[edited by: not2easy at 9:10 pm (utc) on Jun 5, 2018]
[edit reason] fixed readability [/edit]

[not2easy]

Depending on whether you want to use "www" or not - this is without:

RewriteCond %{HTTPS} !on [OR]
RewriteCond %{HTTP_HOST} !^(example\.com)?$
RewriteRule (.*) https://example.com/$1 [R=301,L]

is the most commonly suggested form.

And it should be the last thing at the end of your .htaccess file (for non-wordpress pages).

[keyplyr]

@Sally Stitts - Depending on how your host has things set up, you really need to ask them about it because if they are already redirecting to HTTPS adding another redirect may cause extra server hits.

The simple way for most is:

RewriteEngine On
RewriteCond %{HTTPS} !=on
RewriteRule ^ https://%{HTTP_HOST}%{REQUEST_URI} [L,R=301]

[added]
Looks like not2easy & I were posting at the same time.

[Sally Stitts]

Depending on how your host has things set up, you really need to ask them about it because if they are already redirecting to HTTPS adding another redirect may cause extra server hits.

I have sent a link to this thread to my host, which saves a lot of time and re-explaining.

To be consistent, I always use www. (15 years now)
And THE LAST THING, thank you, I didn't know that (non-wordpress, home spun).

As soon as I see (find) the www version, I will jam it right in there.
I may not (don't) understand it, but I am very vigilant about copying characters exactly.

I may not ever recover, but it would be nice to stop the carnage.
Thanks again.
.

[Sally Stitts]

@nottooeasy and keyplyr
And yet ANOTHER issue, Every time I look for code, it is always different.
The examples you two provide differ in EVERY line.
So, perhaps you can understand how a non-programmer like me can have big issues with trying to "get things right".

I have been told to use code that results in every page not being available. That is gone in about 1 minute, as I toss it.
This has happened several times.
I don't think it is possible for two entities to present the same code. Just too many variables.
I must cut and try every one, and test the heck out of it.

Looks like I am going to go to every place I know, and get THEIR take on it. And see if ANY are the same.
Then the testing begins.
.

[keyplyr]

If you're using WP then I would use the code not2easy suggested and make sure to install it at the very bottom of you htaccess file.

Note: always make a copy of your htaccess file as a back-up before updating, in case you need it.

[not2easy]

For WP, the WP snippet comes after the canonical rewrites, but this is not for a wordpress site (from previous posts).
The only difference for a www. site is to use www. in the rules:

RewriteCond %{HTTPS} !on [OR]
RewriteCond %{HTTP_HOST} !^(www\.example\.com)?$
RewriteRule (.*) https://www.example.com/$1 [R=301,L]

If you are working with your host and you cannot check your headers, you might want to hold off editing .htaccess until they contact you regarding this topic.

[Sally Stitts]

I guess you missed my statement above -

And THE LAST THING, thank you, I didn't know that (non-wordpress, home spun)

To reiterate, NOT WordPress, home spun.
But thanks for trying.
.

[lucy24]

I have been told to use code that results in every page not being available.

It's got nothing to do with available or not available. The essence of a redirect is that it doesn't matter whether the originally requested URL (including protocol and host) exists or not; all that matters is that the requests get correctly redirected.

Yes, I realize that “find a new host” seems like an awfully annoying and unhelpful response. But, honestly, find a new host.

The two suggested redirect codes are actually identical except that one version also covers canonicalization (with/without www).* Once you've moved to a host that lets you control your own RewriteRules, you can apply to the Apache subforum for help on the details. Or possibly the IIS subforum, if that's where you end up ;) (There exist at least three lesser-known servers aside from those two--but they're the server equivalent of moving to a Linux operating system. If you don't already have the confidence and know-how, don't do it.)


* One of the two also involves stuff in the target that I personally don't approve of, but that is neither here nor there.

[No5needinput]

@SallyStitts Are you using hosting from your ISP? Just looking at your domain name and it looks like the long funny addresses ISP's would give their customers. (I know your site because of your past Google probs years ago)

[not2easy]

but this is not for a wordpress site

You might have missed that part. I know that your site is not a wordpress site.

I should have ignored the WP comment but didn't want some WP user stopping by to think that the posted code belongs at the end of their file.

[Sally Stitts]

@No5needinput - I am not hosting with my ISP. THAT would be a nightmare.
Most days, I must reset my ISP-supplied modem about 10 times. It simply loses its mind, and won't communicate.
The speed is good, but the connectivity is a nightmare.
I hate my ISP because of this, and because the TV sucks big time. Reading online, I have lots of company.
A few days ago, my new bill went up to $210 (up $10), for a MINIMUM tv-Internet package.

Who is my only choice? Think cable. Think big. Think least responsive. Think most arrogant. Think most expensive. You got it!

I guess there is another so-called choice - ATT. But I have read the reviews. The worst I have ever read. Average is 1 star out of 5, with many comments saying they wish they could give less than 1.
.

[not2easy]

Not to expand on the topic drift, but afaik ATT delivers via wifi. I looked into it as an option because my cell is ATT but I'm lucky to find 1 or 2 bars and often it says LTE so it did not seem like a good idea. I call it the middle-of-nowhere tax.

[Sally Stitts]

A little bit more.

In my GSC, I had site maps in both my HTTP and HTTPS versions. However, both were identical HTTPS site maps.
Maybe this was not cool?
Then, I noticed that the HTTP site map Google coughed up was ANCIENT (2 years old), and a totally HTTP site map,
even though I had updated it 20 plus times over the last 8 months.

Is it necessary to totally destroy ANY SITE MAP that may still exist in the HTTP URL?
I immediately blew that sucker's brains out (removed the site map from the HTTP version ENTIRELY, never to return.)

I love ALL data points. Do you like this one?
Still waiting for recovery.
.

[keyplyr]

The URL of the old property sitemap may still *appear* with the HTTP protocol, that would be normal since that's what was being used with that property.

But in reality, there should now only be one sitemap in existence, the one with all HTTPS paths.