Am 04/03/2014 01:09 PM, schrieb Rob Weir:
On Wed, Apr 2, 2014 at 5:19 PM, Marcus (OOo)<marcus.m...@wtnet.de>  wrote:
Am 04/02/2014 09:22 PM, schrieb Roberto Galoppini:

2014-04-02 21:15 GMT+02:00 Marcus (OOo)<marcus.m...@wtnet.de>:

Am 04/02/2014 06:20 PM, schrieb Roberto Galoppini:

   2014-04-01 21:30 GMT+02:00 Marcus (OOo)<marcus.m...@wtnet.de>:


   Am 03/31/2014 11:56 PM, schrieb Kay Schenk:


    On Mon, Mar 31, 2014 at 1:48 PM, Rob Weir<robw...@apache.org>
wrote:


    On Mon, Mar 31, 2014 at 4:43 PM, Marcus (OOo)<marcus.m...@wtnet.de>

wrote:

   Am 03/29/2014 09:36 PM, schrieb Roberto Galoppini:


    2014-03-28 21:24 GMT+01:00 Marcus (OOo)<marcus.m...@wtnet.de>:


    Am 03/13/2014 10:01 PM, schrieb Marcus (OOo):


    Am 03/09/2014 06:08 PM, schrieb Marcus (OOo):


    Am 03/08/2014 12:09 AM, schrieb Andrea Pescetti:


    Rob Weir wrote:


    http://linux.softpedia.com/get/Office/Office-Suites/



   Apache-OpenOffice-253.shtml




      Or maybe a disclaimer in the voting thread email?




   Andrew's comments show clearly that these editors do not
care

to be
careful or factual, or even read those disclaimers,
unfortunately.

We can be successful only if we manage to block their
downloads.

   They



   link to our binaries hosted on SourceForge (which is fine). Just


thinking loud, but if it was possible (on the Sourceforge side) to

deny
all download requests that do not come from the
openoffice.orgor

   the



   sourceforge.net domains, then the project would effectively be in


control. The embargo could be lifted just after the release.



   For me this sounds like a great idea.


Maybe we should start with denying all download requests that
some

   from



   these bad websites.



@Roberto:
Can you tell us if this possible? If yes, is it much effort for
you?


   Do you see a chance to get this implemented? I think it could

help to
stop some bad websites to do bad things with our software.


   @Roberto:

Maybe you haven't seen this up to now.


   Thanks for heads up Marcus, sorry for not having noticed this

thread
before.




   It would be great if you can tell us if it's possible to exclude

some
domains / IP addresses from downloading our software?


   I need the domain list and I'll check out with our SiteOps if

that's
doable. Feel free to send me a list with a direct message.



- chip.de
- computerbase.de
- softpedia.com

This would be the domains from this thread that could be blocked
from
downloading from Sourceforge. Obviously needs to be extended in the

   future.


   Remember, the next will happen with the AOO 4.1.0 RC. ;-)


*Of course*, this is just for the time frame as long as the new
version

   is


   not officially announced. As soon as the release is public, the
block


   will


   be removed.


@all:
I think this could help to limit the downloadability like we want to
see
until the official release. What you think?


   I don't know.  Won't this just cause confusion?  They point to the

files, go to test them, see the links don't work, and then get weird
errors and spend an hour trying to debug it.  We don't want to
needlessly annoy them, since their only fault is enthusiasm.   Is
there a way we can give a useful error message in this case like,
"This version of Apache OpenOffice has not yet been officially
released.  Links to these files are disallowed until the release is
officially approved"  or something like that?


   To be honest, I don't care about a few days were a special set of

domains
were not able to access for a few days. For me they are a bit too
enthusiastic. And as you said in a previous post it's to protect us as
best
as possible.


    +1 This seems sufficient to me.



@Roberto:
Do you see a technical way to return a predefined error message when
the
release builds are already on Sourceforge but not yet officially
released
and published?


Our SiteOps team looked into this, here our findings:


Great :-)


   One provider (chip.de) serves via Akamai CDN, one provider (

computerbase.de)
serves via their own FTP server, and softpedia.com lists SourceForge as
an
external mirror and passes traffic through our download redirector flow
(not direct to a mirror).

The first two cases are things we can't control.

In the third case, we can indeed redirect this traffic by referrer to a
different landing page if one is provided. Maybe we want to have a
openoffice.org page explaining that's a release-candidate and it's
served
only for testing purposes and its use on a daily basis it is not
recommended.

How does that sound?


I'm pretty sure that these kind of downloaders do not care about
disclaimers - less then ever when located somewhere else. ;-)

So, either we disable the entire download for the specific timeframe or
at
least a text as substitute (like "This release is not yet public. Please
stay tuned and come back when it is announced."). But this text has then
to
be on Sourceforge in the same location.


Yes, that's doable in the way Kay described. And yes, we would add the
text
and disable downloads.


Just to be sure, is this limited to a special subdir like
".../files/milestones/"? Or also, additionally for ".../files/"?


I'm wondering if the "staging" bit can help as best solution. I would
expect that the new location is not public *and* not known *and* not
useable/functional for the normal non-admin user *until* we remove the
bit.
Am I right?



In past we extended the 'staging' period of time for weeks, this could be
done again if necessary and it's definitely a more effective way to share


Good to know. I thought you had extended the timeframe permanentelly. ;-)


files only with a selected audience (admins). Would that work, or you want
to be able to share those files with a larger audience?


I don't think it's relevant to a wider audience.

We still speak about the timeframe between starting uploading to Sourceforge
and the official announcement. Within this timeframe it should be possible
for admins to test the download webpage with scripting - to see if it will
work - but there must not be no way to download the builds from the public.

So, I would prefer to go with the "staging"-bit as:
- it is already available
- there is no confusion for the public
- it's easy to delete the bit to make the release public
- and (please don't get me wrong ;-) ) we can do it alone without the help
of you or someone else of Sourceforge.

What do others think about this?


My original recommendation was to put a more prominent warning in the
[VOTE] emails.  The problem is this:  if we are to have a public vote
then the files we're voting on must also be public.  Whether this is
on SourceForge or people.apache.org, it is publicly known that this is
the 4.1 RC and some websites will download and copy onto their
websites.   I don't think there is any technological means to prevent
this.  We only have our ability to educate about the risks.

Sure, to put a note into the VOTE mail is nice but wouldn't help to avoid early downloads. I'm sure they will just smile and download anyway. ;-)

OK, maybe I've misinterpreted your recommendation.

And yes, there is no 100% protection against these kind of downloaders. However, we can give them a hard time and to use a staged folder can be a real help.

So, this is my opinion. :-)

Marcus



     Then we can exclude requester that we don't want (e.g., malware


"distributors").


Also in time frames with Beta or RC releases it can help us to
steer

   who



   is able and when it is possible to download OpenOffice like we want
to


see

until the real release date is reached.




    Thanks


Marcus



      Sure, sites could still copy all binaries being voted upon
and
offer


   them locally, but this would require a more significant effort.
on

their
side.


   And more HDD space and more own bandwith - which is also not

what

   they



   want.



Marcus

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
For additional commands, e-mail: dev-h...@openoffice.apache.org

Reply via email to