Hi All,

I have a somewhat radical suggestion in terms of browser support. I'm
actually not 100% convinced this is the way to go, but I think there are
enough pro's that it's worth considering and discussing. My suggestion
is that we cut down our browser support to two, cross platform browsers;
namely firefox and chrome.

I guess this suggestion was originally motivated by two major points: IE
is performing and behaving horribly, and IE9 will soon be out, meaning
we will have to support another environment, making the QA burden bigger.

*IE: *As we all know, there are several issues with IE:

* It's buggy: the most buggy of our supported environments,
  consequently meaning longer development time to avoid bugs and
  fixing those that we missed
* It's hard to style: IE8 doesn't support standards as well as other
  browsers (incl. css3 and html5)
* It's very slow compared to other browsers, giving users a bad
  experience with CollectionSpace which I believe is a very
  important and relevant point. Although slow, collectionspace is
  tolerable to run in all the supported browsers except IE. People
  using IE will blame the performance on cspace and not IE, which
  makes us look bad (or.. worse than we are)
* Forces OS: currently about 50% of windows users use XP, for which
  IE9 is not available. In the near future, we would likely need to
  support IE9 too (which will be default for most win7 users), which
  adds a lot of QA time. Furthermore, since such a big percentage of
  windows users are using XP, we will have to be backwards
  compatible with IE8 for quite a while.
* In all fairness, IE9 is rumored to be more standards compliant and
  faster (but as pointed out above, we will have to continue support
  for IE8 for quite a while)

*Safari: *There aren't many issues with Safari (the ones that are,
mainly have to due with minor stuff like keyboard accessibility) - yet,
according to all statistics, Safari has less support than FF and Chrome.
General statistics says ~4-6% uses Safari against 10-16% Chrome and
30-60% FF. Also, unlike Chrome and FF, Safari is only supported in a
single environment (Mac) and will furthermore have certain versions of
Safari compatible with the newest/older versions of OSX.

So the idea is to change to only support two browsers: Firefox and
Chrome. The reason for choosing these are that they are cross platform
browsers, popular, stable and relatively fast.

There are several pros for this:

* Signal value: we support free browsers that are available for all
  OS ( -- well ok; most popular OS)
* Only supporting two browsers significantly lessens QA burden
* Good debugging tools in both browsers
* Avoiding taking IE8 misbehavior/and standards deviance into
  account, styling and programming will be faster, and less bugs
  will surface and take up developer time.
* Although limiting the supported browsers, all OS' currently
  supported will have a choice of browser, and hence not cut off
  from using CSpace.

And of course there are some rather big cons too:

* Museums could potentially choose against CollectionSpace due to
  lack of support for IE (and/or Safari). This is a major issue, and
  something we do not want to happen.
* Potentially forces museums to switch browser - while the choice of
  supported browsers makes this easy for museums disregarding the
  environment they use, this might still cause some frustrations
  that will (rightly) be pointed towards CollectionSpace.

We currently have Tier 1 and Tier 2 support of browsers; Tier 1 meaning
that we include them in QA, and create blocking JIRAs (and generally
prioritize highly) and Tier 2 meaning we attempt to support them, but
that they are not included in our QA, and that we wont block on bugs in
these environment. If we decide that it is too drastic to completely cut
support for IE and Safari, we could downgrade these to Tier 2, meaning
that we would still fix bugs for these, etc., but that they would be
lower prioritized than our two main environments.

I talked to Megan earlier today about museums' willingness to have to
install certain browsers to be compliant with collectionspace and she
said it didn't sound unrealistic. Before making any decisions, we should
of course talk with the implementers as well as potential implementers -
large and small, to hear their take on this...

I realize that this would be quite a drastic change, and that there
might be very strong opinions on this, but I hope we can get a
constructive talk about this and figure out the best solution for
CollectionSpace. As said, I'm not 100% convinced that this would be the
way to go, but I do see a lot of gains and am quite optimistic on how
this would be received from the side of the implementers.

~Kasper

Hi,

The feeling from asking around at SMK is...

• dropping IE support doesn't create any serious problems for our IT department or end users
• as long as the markup is W3C valid then we'll use the browser which performs best
• the type of organisation which will be attracted to CollectionSpace are unlikely to have a strong preference for IE
• despite the above, dropping IE doesn't look so good from a marketing perspective

Cheers,
Chris

Developer, Corpus Project
Statens Museum for Kunst (National Gallery of Denmark)

Fra: talk-bounces@lists.collectionspace.org [mailto:talk-bounces@lists.collectionspace.org] På vegne af Kasper Galschiot Markus
Sendt: 15. december 2010 20:58
Til: talk@lists.collectionspace.org
Emne: [Talk] Reconsidering our browser support

Hi All,

I have a somewhat radical suggestion in terms of browser support. I'm actually not 100% convinced this is the way to go, but I think there are enough pro's that it's worth considering and discussing. My suggestion is that we cut down our browser support to two, cross platform browsers; namely firefox and chrome.

I guess this suggestion was originally motivated by two major points: IE is performing and behaving horribly, and IE9 will soon be out, meaning we will have to support another environment, making the QA burden bigger.

IE: As we all know, there are several issues with IE:

• It's buggy: the most buggy of our supported environments, consequently meaning longer development time to avoid bugs and fixing those that we missed
• It's hard to style: IE8 doesn't support standards as well as other browsers (incl. css3 and html5)
• It's very slow compared to other browsers, giving users a bad experience with CollectionSpace which I believe is a very important and relevant point. Although slow, collectionspace is tolerable to run in all the supported browsers except IE. People using IE will blame the performance on cspace and not IE, which makes us look bad (or.. worse than we are)
• Forces OS: currently about 50% of windows users use XP, for which IE9 is not available. In the near future, we would likely need to support IE9 too (which will be default for most win7 users), which adds a lot of QA time. Furthermore, since such a big percentage of windows users are using XP, we will have to be backwards compatible with IE8 for quite a while.
• In all fairness, IE9 is rumored to be more standards compliant and faster (but as pointed out above, we will have to continue support for IE8 for quite a while)
  Safari: There aren't many issues with Safari (the ones that are, mainly have to due with minor stuff like keyboard accessibility) - yet, according to all statistics, Safari has less support than FF and Chrome. General statistics says ~4-6% uses Safari against 10-16% Chrome and 30-60% FF. Also, unlike Chrome and FF, Safari is only supported in a single environment (Mac) and will furthermore have certain versions of Safari compatible with the newest/older versions of OSX.

So the idea is to change to only support two browsers: Firefox and Chrome. The reason for choosing these are that they are cross platform browsers, popular, stable and relatively fast.

There are several pros for this:

• Signal value: we support free browsers that are available for all OS ( -- well ok; most popular OS)
• Only supporting two browsers significantly lessens QA burden
• Good debugging tools in both browsers
• Avoiding taking IE8 misbehavior/and standards deviance into account, styling and programming will be faster, and less bugs will surface and take up developer time.
• Although limiting the supported browsers, all OS' currently supported will have a choice of browser, and hence not cut off from using CSpace.
  And of course there are some rather big cons too:
• Museums could potentially choose against CollectionSpace due to lack of support for IE (and/or Safari). This is a major issue, and something we do not want to happen.
• Potentially forces museums to switch browser - while the choice of supported browsers makes this easy for museums disregarding the environment they use, this might still cause some frustrations that will (rightly) be pointed towards CollectionSpace.
  We currently have Tier 1 and Tier 2 support of browsers; Tier 1 meaning that we include them in QA, and create blocking JIRAs (and generally prioritize highly) and Tier 2 meaning we attempt to support them, but that they are not included in our QA, and that we wont block on bugs in these environment. If we decide that it is too drastic to completely cut support for IE and Safari, we could downgrade these to Tier 2, meaning that we would still fix bugs for these, etc., but that they would be lower prioritized than our two main environments.

I talked to Megan earlier today about museums' willingness to have to install certain browsers to be compliant with collectionspace and she said it didn't sound unrealistic. Before making any decisions, we should of course talk with the implementers as well as potential implementers - large and small, to hear their take on this...

I realize that this would be quite a drastic change, and that there might be very strong opinions on this, but I hope we can get a constructive talk about this and figure out the best solution for CollectionSpace. As said, I'm not 100% convinced that this would be the way to go, but I do see a lot of gains and am quite optimistic on how this would be received from the side of the implementers.

~Kasper