[horde] horde replacement
Gunnar Wrobel
wrobel at horde.org
Sun Mar 4 22:27:46 UTC 2012
Zitat von Reindl Harald <h.reindl at thelounge.net>:
> Am 04.03.2012 19:36, schrieb Thomas Spuhler:
>> It's actually not that bad if you can install it from a Distro and all the
>> deps are pulled
>
> the distros are mostly not up-to-date
>
> so since we are PCI scanned weekly and security is one of our
> most important things beside performance / deployment i built
> horde, imp, mimp, dimp, kronolith... over years at my own as
> RPM but there is no way to do this any longer for Horde since
> the last major release
Of course there is. I admit that I didn't do a fantastic job with the
"components" scripts. But if there would have been some additional
interest I'm pretty certain we would have been able to get the
repackaging working pretty reliably. So far only Mathieu Parent from
Debian invested some time in that and as far as I know it helped him
getting the initial drafts of the Debian packages pretty quickly.
>
> a webmail is a must-have feature, but it must have not all
> the options of horde so i am only accepting a complex and
> feature rich software if it is deployable on many hosts
> and no "pear upgrade" is no acceptable deployment
As mentioned in my other mail we have no intention to force you to use
PEAR. On the developer side we need to focus on one packaging format.
To me PEAR is a solution for now. I don't like it that much, I would
look forward to better options but right now I don't see them.
If your comment is geared at the fact that you dislike the modular
nature of Horde but rather go for a monolithic block of software: This
is not where I see the future of web application development. Free
software distributions are extremely successful in powering the cloud
- because they are very modular and allow for many usage scenarios.
Typical web application nowadays are not - and that despite the fact
that the cloud is getting more complex by the minute. If free software
wants to be successful in that area as well I don't see the monolithic
blobs of software as the future. One of the main reasons I stick with
Horde: because they went early on for this modular structure. To me
this is the perfect choice in the long run.
>
>> but the horde folks are not very helpful in fixing the bugs
>> in the packages that would make it easier for packagers.
>
> horde folks are also not helpful in fixing any bugs
> since years randomly includes are failing and replace the
> relative include do dirname(__FILE__) is a perfect fix
>
> but not for the horde-developers, abnswer in a bugreport
> years ago "fix your environemnt" - sorry i am webdevleoper
> since years and i never developed anything where the
> envirnoment has to be fixed for paths
>
> a webapp has to know it's root-folder and use absolute paths
> first for clean running and second fro performance
>
>> I guess we all have choices and if it get's too difficult, they
>> will not be included in the Distros anymore.
>
> i fear they are not interested
I think all of the Horde developers disagree.
>
> but who cares, form now on i have to maintain only a single RPM
> instead a bundle of them and no questions what all of this
> thousands of options in the inerface are doing from users
>
> i understand that the horde-developers are thinking partly
> the univserse is around their life-work, but this is not
> the case and sooner or later they will be the only users
> if updates are randomly breaking as in the last months
> since H4 often happened - with a package-manager you can
> hanlde such problems much easier as with pear, but upstream
> is no longer interested in their userbase
Of course we do. We offer a free software and such a software is
pointless without users that like to use it. But of course there is no
way to make everyone happy. We try our best at finding some middle
ground. We have these mailing lists to take different opinions into
account and your opinion counts as well. But I must say that you also
received plenty of counter arguments to many of your postings. If the
bottom line on our side is that we have the impression that most users
support the current path we took with PEAR then I hope you are able to
accept that we continue on that road.
Cheers,
Gunnar
--
Core Developer
The Horde Project
e: wrobel at horde.org
t: +49 700 6245 0000
w: http://www.horde.org
pgp: 9703 43BE
tweets: http://twitter.com/pardus_de
blog: http://log.pardus.de
More information about the horde
mailing list