On Feb 8, 2008 10:40 AM, Martin Bochnig <mb1x at gmx.com> wrote: > First, let me point out, that there seems to be a deep language barrier > between myself and 99% of you. I write in a foreign language and it doesn't > only take a while, but also limits me in the way of how I can try to express > or "bring over" my message.
One thing that may help is that if we get other members of the Conary evaluation project involved in the conversation. Let's update the project page with the work done to date, and bring the community into the conversation.. Alot of work has gone into IPS and without clear and calm presentation of data, you risk upsetting the people who have been working on the project. (One of the big things holding back adoption of Conary was that the port would require unknown resources to adopt, now that that this costs have been mostly paid, an evaluation is fairly straightforward). > It wasn't planned, nor intended this way. > I wanted to announce it much later, after having written a 1:1 comparison. > But Shawn Walker's CG proposal forced me to act, as I have been quiet for > two months, yet did do something I think would be good for the project. I'm > not so much into writing fancy blogs, which seems to be a drawback. Martin let's hold off from making blanket statements, I think a feature comparison of the various packaging systems on the market is in order now. Let's calmly present the data, and let the data sell itself. > You're welcome to like Conary, > > work on it, propose a project for sponsorship, whatever, but you'll be > > much more likely to get support from me if you actually sell me on what > > Conary can do, not just make assertions of comparison to something else > > without facts to support the assertions. > > I thought somebody would use google?? Just to be clear there is a project already approved to evaluate conary as an approved opensolaris.org packaging system: http://www.opensolaris.org/os/project/conary-eval/ I do need to update this page, with a status, but what I can say now is that it looks like the key question of whether or not Conary could adapted to Solaris is answered. The answer is "yes, with a good bit of work". Unfortunately a good bit of that work had to be done to come to this answer. Luckily, that means that most of this work is behind us, and will be able to move fairly quickly from a Conary evaluation project to a conary integration project. A couple of strengths that Conary has that may peak interest: 1) Full rollback support. (This exists today and is not ZFS dependent.) 2) Ability to install directly from source packages in a repository. (Conary also fully supports binary packages as well, and like IPS can be adapted to ingest system V packages.) 3) Currently running on Sparc and x86, and has built in support for other architectures. We can not say at this time a full comparison is difficult because I am having difficulty tracking down the design criteria specification and requirements documentation for IPS. (Any help here is appreciated, especially if this is an internal Sun document.). Cheers, Brian -- - Brian Gupta http://opensolaris.org/os/project/nycosug/ -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/ogb-discuss/attachments/20080208/40d9f844/attachment.html>
