Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.misc > #6859 > unrolled thread

[CM] Mozilla reports Thunderbird usage growing

Back to article view | Back to comp.misc

Contents

  [CM] Mozilla reports Thunderbird usage growing RS Wood  <rsw@therandymon.com> - 2015-02-28 21:04 +0000
    Re: [CM] Mozilla reports Thunderbird usage growing VanguardLH <V@nguard.LH> - 2015-02-28 23:31 -0600
      Re: [CM] Mozilla reports Thunderbird usage growing Jim Jackson <jj@franjam.org.uk> - 2015-03-02 12:43 +0000
      Re: [CM] Mozilla reports Thunderbird usage growing per@hedeland.org (Per Hedeland) - 2015-03-04 08:15 +0000
        Re: [CM] Mozilla reports Thunderbird usage growing VanguardLH <V@nguard.LH> - 2015-03-04 16:35 -0600
          Re: [CM] Mozilla reports Thunderbird usage growing per@hedeland.org (Per Hedeland) - 2015-03-06 07:30 +0000

#6859 — [CM] Mozilla reports Thunderbird usage growing

From the «and not just because the competition sucks» department:
Title: Thunderbird continues to grow
Author:
Date: Sat, 28 Feb 2015 15:01:57 -0500
Link: https://blog.mozilla.org/thunderbird/2015/02/thunderbird-usage-continues-to-grow/

Comments[1]

Links:
[1]: https://news.ycombinator.com/item?id=9123702 (link)


--
Posting to comp.misc, sci.misc, and misc.news.internet.discuss

[toc] | [next] | [standalone]

#6865

RS Wood wrote:

> From the «and not just because the competition sucks» department:
> Title: Thunderbird continues to grow
> Author:
> Date: Sat, 28 Feb 2015 15:01:57 -0500
> Link: https://blog.mozilla.org/thunderbird/2015/02/thunderbird-usage-continues-to-grow/
> 
> Comments[1]
> 
> Links:
> [1]: https://news.ycombinator.com/item?id=9123702 (link)

Installation count does mandate or equate to usage count.  I've trialed
Thunderbird several times.  Once only lasted a week and another was
about a month.  Those hardly count as "usage growing".  In my latest
trial, I used Tbird for almost 6 months but eventually couldn't stand
all the bugs (some very old and still left open), workarounds, and
crappy add-ons for which there is no matrix chart showing which work (or
won't work) with which.  Does 6 months of using Tbird really count in a
"usage" count?  Software developers love to cite install counts for how
popular is there software.  Some even try to mislead further by citing
download counts.

http://emailclientmarketshare.com/
Tbird doesn't even make it to the top 10 in this list.

Just 2 months into 2015 is hardly cause to claim any e-mail client is
the leader or even growing.  The time span is too short to overcome
aberrant or "flyers" in the statistics, especially when citing
associated data, like installs or downloads, rather than real-use data
(what e-mail providers see as the most popular clients connecting to
them).

http://blogs.salesforce.com/company/2015/02/top-trends-from-2014s-email-client-market-share.html

With more and more users employing [inane, underwhelming, or crippled]
e-mail clients on mobile devices, the desktop marketshare for e-mail
clients is waning.  So even if Tbird did make a spurt in a couple months
during 2015, a blip is an attenuated venue is of minor importance.

[toc] | [prev] | [next] | [standalone]

#6872

On 2015-03-01, VanguardLH <V@nguard.LH> wrote:
>
> http://emailclientmarketshare.com/
> Tbird doesn't even make it to the top 10 in this list.
>

That is an interesting list. Add up the Apple clients and you get 46%
of all email clients are Apple!!!! Sorry but that does not look anywhere 
near sane.

Then down the page

"Since determining the client in which an email is opened requires images 
to be displayed, the data for some email clients and mobile devices might 
be over- or under-represented due to automatic image blocking."

Might account for a lot. It's like those "services" that people use to 
check "when" someone reads the email they send - they rely on HTML email 
and an embedded 3rd party "invisible" image. By these schemes I never read 
email :-) 

So back to the drawing board re. any sort of sane emaio client counts.

[toc] | [prev] | [next] | [standalone]

#6895

In article <mcu87e$nm7$1@news.albasani.net> VanguardLH <V@nguard.LH> writes:
>
>http://emailclientmarketshare.com/
>Tbird doesn't even make it to the top 10 in this list.

Not claiming the numbers are wrong, but I couldn't help noticing:

 "Calculated from 1.1 billion opens tracked by <link deleted> ..."

I.e. it seems it only counts clients that auto-confirm your e-mail
address for spammers - which, regardless of its other faults,
Thunderbird does not.

--Per Hedeland

[toc] | [prev] | [next] | [standalone]

#6900

Per Hedeland wrote:

> VanguardLH wrote:
>>
>> http://emailclientmarketshare.com/ Tbird doesn't even make it to the
>> top 10 in this list.
> 
> Not claiming the numbers are wrong, but I couldn't help noticing:
> "Calculated from 1.1 billion opens tracked by <link deleted> ..."
> I.e. it seems it only counts clients that auto-confirm your e-mail
> address for spammers - which, regardless of its other faults,
> Thunderbird does not.

Ah, you might be correct but for a different reason.  They say:

  Since determining the client in which an email is opened requires 
  images to be displayed, ...

That means they are doing tracking by inserting externally linked
images.  Most good e-mail client will, by default, block externally
linked content in HTML-formatted e-mails to prevent such tracking.
Spammers used it to determine who read their spewage.  MsgTag, and
others, employed the spammer trick to track e-mails (supposedly for the
benefit of the sender).  So there would be a bias in their statistics to
e-mail clients that were either misconfigured, deliberately configured
to allow externally linked content, or wimpy e-mail clients that have no
such protection.  

  ... the data for some email clients and mobile devices might be over- 
  or under-represented due to automatic image blocking.
  
So they recognize that the results are skewed to represent only
misconfigured, malconfigured, or limp e-mail clients that do block
externally linked content.

Yet I see Outlook is listed but by install-time default it has
externally linked content blocked.  I doubt the majority of Outlook
users are deliberately overriding the default of blocking externally
linked content by configuring Outlook to show it.  Or it could be there
is such on overwhelming number of Outlook users that those that do allow
externally linked content still flood their statistics.  That is,
Outlook would place much higher if all users were counted instead of
just those that allow externally linked content.

Litmus is a service that tries to let senders track if recipients have
opened the sender's e-mail.  They assume that opening an e-mail equates
to reading it.  Mostly but not completely true.  This appears to be the
same tracking service (using externally linked content) as MsgTag,
except Litmus adds other features, like previewing your e-mail in 30+
e-mail clients before sending it (so you can test how it will look to
recipients using various e-mail clients).  After experimenting with
MsgTag for awhile, it became very apparent that tracking opens of sent
e-mails using linked images (web beacons) was not accurate.  Almost
everyone to whom I sent e-mails either deliberately configured their
e-mail client to block external content or that was the install-time
default for their e-mail client.

I'm not sure what you mean by "auto-confirm".  That would require users
to click on a link in an e-mail to confirm that the e-mail got opened
(so the user could click on the link).  The string sent by the client
for the HELLO and EHLO commands does not identify the e-mail client,
plus it is the same commands use between SMTP servers (MTAs) to transfer
messages and those aren't e-mail clients per se.  The client adds an IP
address, hostname, or domain for the sender, not an e-mail program name.
Do *any* of the SMTP commands identify the e-mail program used as the
client?

If by "auto-confirm" you are referring to whether externally linked
content is blocked or not, that is not a Thunderbird exclusive feature.
Most "good" e-mail clients block external content by default.  I can
only speak for desktop e-mail clients.  The e-mail web apps on toy
computers (aka mobile devices) are often highly crippled.  For example,
although Outlook for the desktop includes the References header to allow
proper threading of replies in a discussion, the Outlook web app does
not.  So, for example, when using mailing lists or just corresponding
with someone and they use the Outlook web app, References is missing and
their reply starts a whole new thread (and threading based on the
Subject header is fraught with poor matching and leads to mixing up of
unassociated replies).  I cannot attest to how crappy web apps coded
small due to lack of space on their platform will perform regarding
blocking of external content. 

The top 10 list given shows the mobile e-mail apps are leading.  Well,
yes, more users are using smartphones for e-mail now then for more
robust and RFC-complient desktop e-mail apps.  They mention "Outlook"
but no differentiation between Outlook for desktop or web app.  So you
could say that this is a top 10 list for the crappiest e-mail clients;
i.e., those that don't block or cannot be configured to block external
content.

I'm sure even you know, though, that the number of Thunderbird users is
a small cadre when compared to the desktop Outlook app.  Mobile devices
are more used now for e-mail than desktop apps so those apps would show
more usage than Thunderbird.  So there are many reasons why Tbird would
get shoved before their top-10 threshold.

Got any better sites reporting on recent e-mail client marketshare?

[toc] | [prev] | [next] | [standalone]

#6911

In article <md81ba$sop$1@news.albasani.net> VanguardLH <V@nguard.LH> writes:
>Per Hedeland wrote:
>
>> VanguardLH wrote:
>>>
>>> http://emailclientmarketshare.com/ Tbird doesn't even make it to the
>>> top 10 in this list.
>> 
>> Not claiming the numbers are wrong, but I couldn't help noticing:
>> "Calculated from 1.1 billion opens tracked by <link deleted> ..."
>> I.e. it seems it only counts clients that auto-confirm your e-mail
>> address for spammers - which, regardless of its other faults,
>> Thunderbird does not.
>
>Ah, you might be correct but for a different reason.

No, the reason you describe at some length is exactly the one I'm
referring to.

> So there would be a bias in their statistics to
>e-mail clients that were either misconfigured, deliberately configured
>to allow externally linked content, or wimpy e-mail clients that have no
>such protection.  

Yes.

>Yet I see Outlook is listed but by install-time default it has
>externally linked content blocked.  I doubt the majority of Outlook
>users are deliberately overriding the default of blocking externally
>linked content by configuring Outlook to show it.

I've been fortunate enough to never use Outlook, but my limited
experience with Windows software is that it has a tendency to keep
popping up dialogues where you have to click "Yes"/"OK"/"Go ahead" if
you are to get anything useful done. I can well imagine that users
eventually get into the habit of doing this without actually paying
attention to what the popup is saying, so they may have inadvertently
enabled "always fetch remote content without asking" (assuming Outlook
has such a setting). Or even done it intentionally just to get rid of
the popups.

>I'm not sure what you mean by "auto-confirm".  That would require users
>to click on a link in an e-mail to confirm that the e-mail got opened
>(so the user could click on the link).

I obviously mean just the opposite, that the e-mail client fetches
remote content without any action from the user beyond "opening" the
message - since the latter is what the "opens" above refers to IIRC.
I.e. if they are really supposed to count *message opens*, they *have*
to be counting only clients that automatically fetch remote contents.

> The string sent by the client
>for the HELLO and EHLO commands does not identify the e-mail client,
>plus it is the same commands use between SMTP servers (MTAs) to transfer
>messages and those aren't e-mail clients per se.  The client adds an IP
>address, hostname, or domain for the sender, not an e-mail program name.
>Do *any* of the SMTP commands identify the e-mail program used as the
>client?

I don't know why you think SMTP is in any way involved when *reading*
(or "opening") an e-mail message, but anyway what *I* wrote above was
"auto-confirm your e-mail address for spammers", nothing about IP
addresses or client program names. The e-mail address confirmation in
the spam case is achieved by having "personalized" messages, where the
link includes the address of the recipient in one way or another. I
don't know how they figure out the e-mail client name, and don't
particularly care.

--Per Hedeland
per@hedeland.org

[toc] | [prev] | [standalone]

Back to top | Article view | comp.misc

csiph-web