WINDOWS users of Tidy: roadmap feedback requested

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

WINDOWS users of Tidy: roadmap feedback requested

Jim Derry
Dear users of HTML Tidy and LibTidy on the Windows platform,

The developers of HTML Tidy would like your feedback on this one question:

Q: Do you categorize support on Windows XP and earlier as a WANT or as a NEED?
A:

Thank you for your support of HTML Tidy over the years. As the developers of HTML Tidy make progress on localization efforts for HTML Tidy we find ourselves having to make some hard decisions regarding Windows support.

To be a "good citizen" in the localization world HTML Tidy should support the use of POSIX positional parameters in format strings (e.g., "%1$s %2$s", "silly", "clown" would be translated as  "%2$s %1$s", "tanto", "payaso" in Spanish in order to render the correct word order).

Support for this standard was added to Windows Vista, meaning that HTML Tidy would no longer function on Windows XP systems and earlier without sacrificing positional parameters.

Although [Windows XP has been EOL'd](http://windows.microsoft.com/en-us/windows/end-support-help), there are organizations with continuing service-level agreements, as well as [nearly 11% of current market share](https://www.netmarketshare.com/operating-system-market-share.aspx?qprid=10&qpcustomd=0).

Alternatively we could choose _not_ to support positional parameters while maintaining pre-Vista compatibility on Windows. This could lead to awkward translations in some languages. Supporting positional parameters could be deferred to the indeterminate future when support for Windows XP is deemed less critical.

Please note that Unix, Mac OS X, and Linux systems are not affected by this decision as they are POSIX compliant.

Therefore I invite you to discuss this matter. Thank you.

--
---
Jim Derry
Clinton Township, MI, USA
Nanjing, Jiangsu, China PRC
Reply | Threaded
Open this post in threaded view
|

Re: WINDOWS users of Tidy: roadmap feedback requested

Geoff McLane

Hi Jim,

I will preface this with this is just a reply from a mainly
/WINDOWS/ person... it is just "in my opinion"...

Re: POSIX compliant

First I was surprised about the netmarketshare link. I do not
know how they gathered the information but if I read it correctly
some nearly 90% of the world is Windows. Full stop.

And Unix, Linux, Mac OS, others account for just 10% or so!!! Very
interesting indeed, if true...

So why are we talking POSIX compliant? Just for that 10% or so?

Why does the other 90% not count as a quasi standard of some sort?

Re: Windows XP

Next I would categorize support for Windows XP as WANTED and
NEEDED.

I still run an XP machine, mainly because its hardware may not
support later Win versions. But it runs perfectly, and runs some
utilities and tools that I did not want to buy, install again
elsewhere. It does its job...

And while I know XP has been EOL'd, I note there are still
occasional MS updates for it... Apparently they also feel they
can not totally abandon this base...

I would however agree that maybe we should **not** need to support
earlier than this like say ME, or heaven forbid Windows 95.
But OTOH, as indicated earlier up-to-the-present tidy has had no
special requirements, so would probably compile and run...

Tidy certainly still compiles fine in my XP, using MSVC8 (2005).

It is a shame I do not still have MSVC6 (1998?), but think Tidy
would still compile and run...

Re: POSIX positional parameters (PP)

Well, as suggested above, I would **not** see chasing a standard
supported only by 10% as a "good citizen" act, despite understanding
that that 10% can be very **noisy**, vociferous even, argumentative over
the so called "standards", which I will **not** get into...

And I try hard to take all the MS knocks very quietly... Not because
I think them right or wrong... just that potentially 90% of my user
base might use that system...

Now while I understand these PP's may be needed if we needed say to reverse
any adjective+noun in other languages, and perhaps other phrase ordering,
but does tidy's warnings and error messages have any where this would be
used, needed?

Some rough stats usually help. Of the 444 msgs ids, only some 90 have PPs,
and most of them with just 1, which can not be swapped. And I see none
of the paired form "%1$s %2$s", but I did not do a complete analysis, but
will in time, if needed...

And even if there are, could not another phrasing be just as good?

So at this point in time, I would suggest tidy simply does **not** need
positional parameters!

Now, I will **not** argue the point on this. If someone feels it just
has to be that way, then ok, go for it... I do not want to stand in the
way of perceived **progress**, perhaps despite some lost windows support,
or at the very least some special WIN32 only code...

But will always be left wondering WHY?

Remember again, this is just my personal opinion!

Regards,
Geoff.


Reply | Threaded
Open this post in threaded view
|

Re: WINDOWS users of Tidy: roadmap feedback requested

Richard A. O'Keefe
In reply to this post by Jim Derry
Can I suggest a very simple solution?
Use an open source implementation of printf(),
rather than relying on whatever your C library gives you.

My use of Windows has been mainly under Cygwin and
Microsoft "Services for Unix Applications".



Reply | Threaded
Open this post in threaded view
|

Re: WINDOWS users of Tidy: roadmap feedback requested

Jim Derry
I’ve not managed to find a free C library the implements printf without too many dependencies on the rest of a runtime library. For example BSD seems like an obvious choice, but has dependencies on too much BSD-specific imports. GNU lib, aside from the licensing issue, has similar dependency problems. None of the “portable” implementations I’ve looked at support the POSIX extensions.


> On Jan 26, 2016, at 8:15 AM, Richard A. O'Keefe <[hidden email]> wrote:
>
> Can I suggest a very simple solution?
> Use an open source implementation of printf(),
> rather than relying on whatever your C library gives you.
>
> My use of Windows has been mainly under Cygwin and
> Microsoft "Services for Unix Applications".
>
>


Reply | Threaded
Open this post in threaded view
|

Re: WINDOWS users of Tidy: roadmap feedback requested

Ralf Junker
On 26.01.2016 02:11, Jim Derry wrote:

> I’ve not managed to find a free C library the implements printf
> without too many dependencies on the rest of a runtime library.

"Trio is a fully matured and stable set of printf and string functions
designed be used by applications with focus on portability or with the
need for additional features that are not supported by standard stdio
implementation."

   http://daniel.haxx.se/projects/trio/

According to its man page, Trio supports positional modifiers:

   http://daniel.haxx.se/projects/trio/trio_printf.shtml

Sources:

   http://sourceforge.net/p/ctrio/git/ci/master/tree/

Ralf


Reply | Threaded
Open this post in threaded view
|

Re: WINDOWS users of Tidy: roadmap feedback requested

Jim Derry
Thanks for that. I'll definitely check it out.

On Tue, Jan 26, 2016 at 5:38 PM, Ralf Junker <[hidden email]> wrote:
On 26.01.2016 02:11, Jim Derry wrote:

I’ve not managed to find a free C library the implements printf
without too many dependencies on the rest of a runtime library.

"Trio is a fully matured and stable set of printf and string functions
designed be used by applications with focus on portability or with the
need for additional features that are not supported by standard stdio
implementation."

  http://daniel.haxx.se/projects/trio/

According to its man page, Trio supports positional modifiers:

  http://daniel.haxx.se/projects/trio/trio_printf.shtml

Sources:

  http://sourceforge.net/p/ctrio/git/ci/master/tree/

Ralf





--
---
Jim Derry
Clinton Township, MI, USA
Nanjing, Jiangsu, China PRC