Certainly subjective, but would say that Python is more ubiqutious, as its frequently taught as a beginner language, easier to program in (more expressive + easier to debug) and underlies far more core scientific libraries than Groovy.
to say it is a lingua franca is to say that it is the common language of bioinformatics, which to me is not only not to say its the best, it's not even an endorsement.
having said that i do think the comment is fairly arbitrary.
the languages of the world may be divided on, say, geography and ethnicity, but the languages of the computer are divided by purpose.
to me this confounds the simple interpretation of any one language as a lingua franca for not just bioinformatics, but for any sufficiently broad field in general...
in other words im still not sure about the original claim, i just also dont think that the term itself has a positive connotation per se
Dear Kenny, I also experienced the learning curve of Nextflow to be quite steep. However, I would say this post is a rather lacking attempt to place your company blog post, and this is reflected in the pointed responses it has received. Also, I would expect a more balanced approach to the pros and cons. I understand that some design decisions had to be made for your product, but I assume it would depict more competence to carefully weigh the merits of each WF system against each other and then point out the rationale for your decision.
Personally, I prefer writing a new workflow DSL + engine for each project. If I'm working a language with a particularly strong type system I try to embed them entirely in the type system.
I understand that this post is meant to promote a bioinformatics startup. I honestly do wish any startup in the field all the best (though most are likely to vanish in a short period of time); it is hard enough to make a profitable business model in this field. I am just wondering if that post is possibly rather detrimental being so one-sided. Also, I am not aware of us having a policy against such posts so far.
i agree. it is disappointing - after all, this is like, 2005-level clickbait at best. Where's the buzzfeed-inspired use of questions? Where's the claim that point 2 will AMAZE me? Do it better, Kenny. The world deserves a better kind of bait-post.
The general mood seems quite negative on that post.
I'm a pro Nextflow person, but what's said in the summarize is mostly true. Maybe trivial but might be useful for a beginner that has to choose between learning Snakemake and Nextflow.
Companies has to find ways to exists, so I guess this post is an acceptable way to do it.
And it is not as one-sided as it looks, there is even a pro-Nextflo paragraph: What Nextflow is good at
Everyone is entitled to their opinions, even strong ones. I think the key is not to present (strong) opinions as facts, and to leave some room for debate by presenting both sides.
Not sure what prompted this, but
Pretty arguable.
Certainly subjective, but would say that Python is more ubiqutious, as its frequently taught as a beginner language, easier to program in (more expressive + easier to debug) and underlies far more core scientific libraries than Groovy.
What's interesting, if you dig into the nextflow source, is that the choice of Groovy seems to be predicated on the authors stumbling upon an existing dataflow framework rather than choosing the best language for the job and writing one from scratch (http://www.gpars.org/webapp/quickstart/index.html). See eg: https://github.com/nextflow-io/nextflow/blob/master/modules/nextflow/src/main/groovy/nextflow/dag/DAG.groovy.
What you mean: Python is better than Groovy in certain aspects
What you're saying: Python is the goddamn best language in the field
Your response should not have been "certainly subjective" but "I should have added my context of comparison, let me fix it now".
to say it is a lingua franca is to say that it is the common language of bioinformatics, which to me is not only not to say its the best, it's not even an endorsement.
having said that i do think the comment is fairly arbitrary.
the languages of the world may be divided on, say, geography and ethnicity, but the languages of the computer are divided by purpose.
to me this confounds the simple interpretation of any one language as a lingua franca for not just bioinformatics, but for any sufficiently broad field in general...
in other words im still not sure about the original claim, i just also dont think that the term itself has a positive connotation per se
and what is the best text editor : vim or emacs ?
The best text editor is vim (neovim is also acceptable).
and what is the best structured format ? yaml, json or xml ?
i think we can safely rule out xml
and what is the best indentation style : tabulations or spaces ?
and what is the best way to farm for thumbs up? three flame posts? or responding to three flame posts?
Dear Kenny, I also experienced the learning curve of Nextflow to be quite steep. However, I would say this post is a rather lacking attempt to place your company blog post, and this is reflected in the pointed responses it has received. Also, I would expect a more balanced approach to the pros and cons. I understand that some design decisions had to be made for your product, but I assume it would depict more competence to carefully weigh the merits of each WF system against each other and then point out the rationale for your decision.