[Bioc-devel] name for new BioC package
mtmorg@n@b|oc @end|ng |rom gm@||@com
Tue Feb 7 13:54:14 CET 2023
If you choose to introduce a second package, I would suggest immediately deprecating MoonlightR. The two versions co-exist for a development and release cycle for comparison / opportunity for users to adjust, and then there is only MoonlightR2.
This avoids confusion for new users. You can focus development & support efforts on the new version. Users wishing to compare results can use the version of R / Bioconductor where both packages were available. Users wishing to rigorously reproduce previous results are in the unenviable position of re-creating an R environment from the original analysis; they do not benefit from having MoonlightR available in the current (i.e., different) R / Bioconductor.
The transition from DESeq to DESeq2 was very protracted, and �at the end� there were still new users starting with the very outdated DESeq, and still requests to support DESeq that were no longer relevant.
From: Bioc-devel <bioc-devel-bounces using r-project.org> on behalf of Matteo Tiberti <tiberti using cancer.dk>
Date: Tuesday, February 7, 2023 at 6:54 AM
To: Kevin Coombes <kevin.r.coombes using gmail.com>, Herv� Pag�s <hpages.on.github using gmail.com>
Cc: bioc-devel using r-project.org <bioc-devel using r-project.org>
Subject: Re: [Bioc-devel] name for new BioC package
thanks for your input
just to answer on your different points:
@Kevin Coombes<mailto:kevin.r.coombes using gmail.com> I agree it can be annoying - we don't intend to heavily change the API, we mostly want to add functionality and refactoring some of the internals (including basic information that the package uses), which means results of Moonlight2R will be significantly different from MoonlightR
@Laurent, I think we will eventually deprecate MoonlightR but we would prefer having both of them on Bioconductor for now, as e.g. it makes it easier to run comparisons which are reproducible without too much pain
@Herv� Pag�s<mailto:hpages.on.github using gmail.com> thanks a lot! I think we will keep the name with 2 for now
Danish Cancer Society
Telefon: +45 35 25 73 07
www.cancer.dk<https://www.cancer.dk/international/> | Vores privatlivspolitik<https://www.cancer.dk/om-os/privatlivspolitik/>
From: Kevin Coombes <kevin.r.coombes using gmail.com>
Sent: 04 February 2023 14:26
To: Herv� Pag�s <hpages.on.github using gmail.com>
Cc: Matteo Tiberti <tiberti using cancer.dk>; bioc-devel using r-project.org <bioc-devel using r-project.org>
Subject: Re: [Bioc-devel] name for new BioC package
For the record, as a user, I *hated* the move from MOFA to MOFA2. Not the new package name, but the fact that they also Schanged all the function names and argument names. Mostly, they switched from using periods to underscores. But this meant having to tediously hand-edit every script that used MOFA in order to continue using that script in newer versions of R (since they also discontinued supporting the MOFA package in newer versions). Also, some of the changes produced less useful graphical summaries, to the extent that I took the time to write my own code to reproduce the original versions.
So, I would suggest that you at least think about how much work you are creating for your established users before making the change. And make choices that minimize the burden you are imposing on them.
On Sat, Feb 4, 2023, 1:03 AM Herv� Pag�s <hpages.on.github using gmail.com<mailto:hpages.on.github using gmail.com>> wrote:
We had DESeq2 after DESeq, Rbowtie2 after Rbowtie, MOFA2 after MOFA,
etc.. so I don't see any problem, but thanks for asking!
On 03/02/2023 00:08, Matteo Tiberti wrote:
> dear maintainers,
> I am currently listed as maintainer of Bioconductor package MoonlightR, designed for the prediction of cancer driver genes, which implements the Moonlight workflow.
> We are currently working on a second version of our workflow, called Moonlight2, and would like to have it released on Bioconductor as well, in form of the Moonlight2R package. The new package uses similar principles as the current one, but will have significant changes and updates, both in terms of new functionality and revision of old functionalities. The Moonlight2R project/paper will also have in part a different corresponding authorship respect to the current one. MoonlightR and Moonlight2R currently reside in two separate GitHub repositories.
> Ideally we would like to have both packages on BioConductor for the moment, the old one (called MoonlightR) and the new one that we intend to submit before the April cut-off for 3.17 (called Moonlight2R), where the number signifies the version of the protocol rather than the software. However on the package submission list, I see that having package names that "imply a temporal relationship" respect to an existing package is discouraged. Given the circumstances, do you think it would be possible to use the Moonlight2R name for the package (i.e. would it be a reason for rejection or object of revision during submission) or is it fair to keep it as is?
> Many thanks
> Matteo Tiberti
> Danish Cancer Society Research Center
> Strandboulevarden 49
> DK-2100 Copenhagen
> Telephone: +45 35 25 73 07
> www.cancer.dk<http://www.cancer.dk><https://www.cancer.dk/international/<http://www.cancer.dk%3e%3chttps:/www.cancer.dk/international/>> | Vores privatlivspolitik<https://www.cancer.dk/om-os/privatlivspolitik/>
> [[alternative HTML version deleted]]
> Bioc-devel using r-project.org<mailto:Bioc-devel using r-project.org> mailing list
Bioconductor Core Team
hpages.on.github using gmail.com<mailto:hpages.on.github using gmail.com>
Bioc-devel using r-project.org<mailto:Bioc-devel using r-project.org> mailing list
Bioc-devel using r-project.org mailing list
[[alternative HTML version deleted]]
More information about the Bioc-devel