[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Warning on unreleased versions.

From: Alan Mead
Subject: Re: Warning on unreleased versions.
Date: Wed, 2 Mar 2016 16:34:03 -0600
User-agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.6.0


Maybe you understand this very well, but I think a better way to think about this is that the PSPP team releases a release of PSPP (which is source code) and then platform-specific "packagers" release binaries for their platform, possibly making decisions along the way.  For example, Friedrich is the packager for Debian and he might make changes if PSPP fails Debian tests.  A packager might possibly be very active, porting bug fixes and releasing new packages.  Maybe this is bad behavior, but packagers have sometimes made choices that were even upsetting to the "upstream" project (I'm thinking about Red Hat's controversial decisions about gcc) because the packages are primarily concerned with their platform.

By this logic, there is no "official" binary release of PSPP for any platform, including therefore that there will never be an official Windows release of PSPP.

I see the logic of "recommending" 0.8.5 but I'm not sure I agree. I believe that all of the 2015 snapshots available for Windows on have an extremely serious bug triggered by having a non-ASCII character in Windows paths.  So, many non-English speakers will hit this bug which will be a show-stopper for them.  Isn't it better to recommend to them that they use the latest 0.9.0-g745ee3 snapshot?  I think Harry's practice of supplying a latest snapshot package and also some older packages is an extremely practical approach. I also think it's a decision that properly belongs to the packager.

Finally, the opposing view is that 0.9.0-g745ee3 contains all kinds of other changes, including really big changes like a switch to GTK+ 3 (I think?).  I think the remedy for this would be to have a process for testing Windows packages so that they don't have to have "untested" labeling.


On 3/2/2016 2:00 PM, Mark Hancock wrote:
I've attached a screenshot of what I think is a much better solution (to change the website). I think this simple change would alleviate the need for any messages in the application (though still wouldn't object to "(unstable)" or "(development)" in the title bar, just not a long instructional message).

Inline image 1

This can be done by adding the following HTML:

<ul style="margin-top: 0; margin-bottom: 0">
  <li>The latest stable release is: 0.8.5 (<strong>recommended</strong>)</li>
  <li>The latest development release is: 0.9.0</li>

You could obviously make these numbers be dynamically loaded from somewhere, if that makes updating less cumbersome.

This suggestion is assuming you have no control over the page. If you do, it would be much better to redesign it so that it has a salient link to the most stable recent release, and a harder-to-find-and-accidentally-download link to the dev version.


On Wed, Mar 2, 2016 at 2:35 PM, Friedrich Beckmann <address@hidden> wrote:

> Am 02.03.2016 um 20:26 schrieb John Darrington <address@hidden>:
> On Wed, Mar 02, 2016 at 07:52:34PM +0100, Friedrich Beckmann wrote:
> 2 Then, as a first post-release task, let's replace the existing splash screen with an startup window
>   which must be acknowledged.  Then we don't need to have any warning in the titlebar, or in the output window.

Please, do not let us sacrifice functionality for this purpose. Nobody wants to acknowledge this. Nobody. Yes, nobody.
Please remember software that does want you to acknowledge something like this. You hate it. For good reason.

Please remember the reason for the splash screen as we discussed it already: Inform the user that something
has happened when he started the software, if the startup time is long. Thats it. The logo is perfect and the warning
message does not harm this purpose.


Pspp-users mailing list

Pspp-users mailing list


Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.

science + technology = better workers

+815.588.3846 (Office)
+267.334.4143 (Mobile)

I've... seen things you people wouldn't believe...
functions on fire in a copy of Orion.
I watched C-Sharp glitter in the dark near a programmable gate.
All those moments will be lost in time, like Ruby... on... Rails... Time for Pi.

          --"The Register" user Alister, applying the famous 
            "Blade Runner" speech to software development

reply via email to

[Prev in Thread] Current Thread [Next in Thread]