incubator-ooo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dennis E. Hamilton" <dennis.hamil...@acm.org>
Subject RE: [patch] Removal of Windows build requirement on unicows.dll - issue 88652
Date Thu, 29 Sep 2011 00:55:11 GMT
The only reason for testing the installer on Win32 platforms older than Windows 2000 is to
find out how they fail.  If they don't fail that is interesting too, but I understand it is
not part of any agreed support.  If they fail, they won't be fixed.

I guess Oliver is our source for fresh Windows builds to try testing.

 - Dennis

Funny, I just assumed you were a developer.  That's probably because I am a long way from
having been a professional developer.  Umm, well, 3-4 years but there are giant gaps between
developer gigs (e.g., 15 years before, then 10 years before that, etc.)

-----Original Message-----
From: Marcus (OOo) [mailto:marcus.mail@wtnet.de] 
Sent: Wednesday, September 28, 2011 15:42
To: ooo-dev@incubator.apache.org
Subject: Re: [patch] Removal of Windows build requirement on unicows.dll - issue 88652

Am 09/28/2011 11:56 PM, schrieb Dennis E. Hamilton:
> Marcus,
>
> I share your thinking about this.
>
> If you or Oliver or someone can put a recent successful trial build where I can get my
hands on the installer version, I will be happy to perform platform-confirmation installs.
 This would help me set up a routine for doing that kind of QA in the future. (For different
internationalization variations, I can't do that alone and there needs to be help from the
NL community.)

I'm not a developer nor have I tried to build OOo from source for 
myself. (yes, I know, shame on me. ;-( ) However, I can offer some help 
to get some basic tests done for Win2000. A running version in a VM 
should be enough.

> MY THINKING
>
> I agree a smoke test against Windows 2000 and even Windows 98 would be good.  It would
be useful if the colleague here who still has a Windows 95 installation could confirm some
things too.

I remember that some code parts were deleted for supporting Win95/98/ME 
in the past. So, I'm pretty sure that it will not work on these Win 
versions.

So, please correct me if I'm wrong but when we speak about "supporting 
older versions than WinXP" it's only Win2000 we speak about.

Marcus



> If there is a dependency on a newer API entry, that typically shows up at load time or
shortly thereafter.
>
> I suspect Unicode retrofitting will be the deal breaker, but it is useful to find out
and to let users know.  The other prospect is if there is dependency on a JVM or even VC++
RTL that is not supported that far back.
>
> If an artificial cut-off can be avoided, that is a good thing.  It is necessary to do
some sort of minimal testing to see if the current install works or not, and if it doesn't,
what users who try it should expect.
>
> I am happy to test for that.  It is within my competence and, I believe, the capabilities
of the Windows 7 Virtual PC.
>
>   - Dennis
>
> -----Original Message-----
> From: Marcus (OOo) [mailto:marcus.mail@wtnet.de]
> Sent: Wednesday, September 28, 2011 12:04
> To: ooo-dev@incubator.apache.org
> Subject: Re: [patch] Removal of Windows build requirement on unicows.dll - issue 88652
>
> Am 09/28/2011 01:39 PM, schrieb Rob Weir:
>> On Tue, Sep 27, 2011 at 8:46 PM, Dennis E. Hamilton
>> <dennis.hamilton@acm.org>   wrote:
>>> I don't think the vendor support lifetime for a consumer OS has bring the
>>> end of application support on that OS.  What is known is that there will
>>> be further service packs, maybe not even OS security patches, but it isn't
>>> as if they decay and die.  Many machines run much longer than the support
>>> life of the OS, and upgrades may not be feasible.
>
> +1
>
> I don't see a direct need to drop any OS support only because it is to
> old or it seems to be. To point to Microsoft and tell the users "they
> don't support it anymore, so we drop the support too" isn't a good argument.
>
> When we leave the baseline at Windows 2000 (or whereever it is at the
> moment) and tell the user we can give a "guarantee" (don't take this
> word to seriously ;-) ) for WinXP and newer, it should be OK. Then there
> is still a possibility to get it installed and started on Win2000.
>
>> The nice thing is a user of Windows 98 or 2000 can still download old
>> versions of OOo and run them.  And they can do that for free.  And
>> they always will be able to do this.
>>
>> The question is not whether we retroactively support for older
>> versions of Windows.  They question is whether we maintain that
>> support going forward, in new releases of the product.
>
> Yes, and as long as there are no real technical problems I don't see a
> need to drop the support.
>
> If there *is already* or *will be* a technical limitation (e.g., API
> things or system integration) that is a hurdle for going on in
> supporting newer Win versions, then we have a good reason to drop the
> support for older versions.
>
> Otherwise IMHO not.
>
> Marcus
>
>
>
>>> Outgrowing the size of machine that an older OS runs on (and might be
>>> limited to) is a different matter, as is relying on API functions that are
>>> not supported that far back.
>>>
>>> I don't have an opinion about the Win2k versus Windows XP SP2+ choice for
>>> OOo.  I am just curious to know what the current platform boundaries are
>>> and might become for purposes of QA.
>>>
>>>    - Dennis
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: Michael Stahl [mailto:mst@openoffice.org]
>>> Sent: Tuesday, September 27, 2011 15:50
>>> To: ooo-dev@incubator.apache.org
>>> Subject: Re: [patch] Removal of Windows build requirement on unicows.dll - issue
88652
>>>
>>> On 27.09.2011 22:22, Rob Weir wrote:
>>>> On Tue, Sep 27, 2011 at 4:08 PM, Dennis E. Hamilton
>>>> <dennis.hamilton@acm.org>   wrote:
>>>>> What is the oldest Windows OS version that Apache OOo 3.4(-dev) will
>>>>> be supported on?  How does that compare with the oldest Windows OS
>>>>> version that the last stable release (3.3.0?) of OpenOffice.org is
>>>>> supported on?  (If there is a JRE dependency, that is another variant
>>>>> to consider.)
>>>
>>> AFAIK OOo 3.x Windows baseline is NT 5.0 (Windows 2000);
>>> AFAIK this OS version is no longer supported by the vendor.
>>>
>>>> I'd recommend supporting Windows XP and beyond.   XP is officially
>>>> supported by Microsoft until April 2014.   I'm certainly not making any
>>>> effort to maintain or test support for earlier versions.  Of course,
>>>> that doesn't prevent anyone else from testing and patching to support
>>>> earlier versions.
>>>
>>> no objection from me to raising the baseline to WindowsXP; IMHO trying to
>>> support an OS that the vendor doesn't support any more doesn't make sense.


Mime
View raw message