Community policy on listing resources that are plagiarised and violate the CC licence? #194
With reference to the PR #192 that was merged today, does the Julia community have a policy on listing duplicate resources that involve plagiarism and also breaking the original CC licence by changing it to Apache?
A quick comparision of
Unfortunately, instead of contributing and collaborating, he (melvin) has chosen to break the license terms and indulge in blatant plagiarism - Since last week he has forked
On a relevant note, I must mention that last month Stefan had emailed me asking if they could use Julia.jl as the basis for forming a list of unregistered packages and I was thrilled that it would be considered good enough. Fwiw, the core-devs didnt need my permission but I appreciate the fact that they are polite enough to ask. That said, I have no issues with folks who actually compete and create another awesome-* meme .. heck, I do too by curating a scientific Python resource repo (pythonidae) which is vastly different from
@melvin0008, copying my (and other contributors) work by rearranging them on a single README page and changing the licence does not change the fact that you are disrespectful of other peoples efforts, have broken the CC license terms and by getting yourself listed on the Julialang website, the project has (unknowingly) provided legitimacy to someone who is has knowingly broken the Licence terms. That isnt how FOSS works, where breaking licences and/or plagiarism is taken seriously.
The same situation can arise with packages too, so I hope the Julia community responds pro-actively and also formulates some guidelines (or policy) while responding and dealing with Foss licencing and plagiarism issues in future.
Thanks for reading!
 The name
FWIW, I had suggested to Melvin to list it on the Julia website in JuliaLang/julia#9667
Have you written to Melvin with your concerns? If it is derived work, I am sure he will follow the license guidelines and attribution. Only when a person otherwise ignores all communication, I would label it plagiarism.
Approaching this positively and respectfully will gain us more contributors, and lead towards the larger improvement of Julia.jl. I have some other thoughts on improving Julia.jl - which I will file as an issue there.
It is perfectly fine for someone to take pieces of other work and combine them in new and interesting ways. That is a fundamental basis of open source. We work hard on avoiding fragments, and often people are happy to collaborate. In some cases, the forks are meaningful and even invited.
TBH, I didnt expect a defense of plagiarism and license violation from
On Fri, Jan 9, 2015 at 11:10 AM, Viral B. Shah email@example.com wrote:
See the commit linked below which proves he has copied an entire
Please do. I would love to get feedback on some packages whose
TBH, I'm not sure what you mean by "the forks are meaningful and even
Would you still call this verbatim copying a meaningful fork? Viral, I
-SVAKSHA ॥ http://about.me/svaksha ॥
I feel like I am missing something about the CC license. Doesn't it allow sharing and remixing the content so long as the license conditions are followed? In my understanding, this would be completely ok, so long as the terms of the license are followed. If I am missing something, I would love to be corrected. I have myself released content under CC, with the expectation that it will be copied, modified, and shared.
In this case, Melvin should attribute the original work, and require further derived work to follow the same license. If that is done, does it not resolve the issue at hand?
Also, there are far too many jibes against various types of people in your comment above - which are personal views and not helping the discussion here, and strictly not civil. I am requesting you edit the post above accordingly.
On Fri, Jan 9, 2015 at 12:51 PM, Viral B. Shah firstname.lastname@example.org wrote:
Not really - there is a huge difference in context, and especially
BY-NC-SA 4.0, or a later version of the BY-NC-SA license.
So to answer your question, he (or anyone else) can fork, distribute
Best case, I'll try to ignore intent and hope we can agree to disagree
As I mentioned in the original BR, competition is certainly welcome,
Tbh, I had no clue that you asked him to list himself on the website,
Hope that explains!
Update: Another diff, the CC (and its different forms) are an artistic license whilst the Apache (and its ilk) are software licenses, hence incompatible.
By no intentions have I tried to take credit for your work.
My main motivation for awesome-julia is to help people get all the resources to start learning about it. Your Repository was focused mainly on listing packages. I have also added your link under packages. I was writing a list of resources for Julia as there is no repository with a prefix of awesome. Also, I had started this with an inspiration to start contributing to Julia as I am studying the Julia code and will soon start contributing to it.
I have just copied the IDE part of your Julia.jl repository to list the available IDE's for developers. If you would like me to remove the IDE section please let me know. I had no intentions in changing CC to Apache. I had just selected Apache without even thinking about it.
BTW I was of the thought that collecting resources, documenting and contributing to the open source community was something out of personal interest and betterment of the developers community. I did not know this would create a fuss and let you berate me on what sexism is. [How is this even remotely related to the issue at hand?]
I haven't read the links in question yet, but: could everyone please make an effort to stay on the main issue, which is the copyright question.
I don't think it is the goal of the community to promote copyright violations if such exists. At the same time, I would hope that we can give the benefit of the doubt and try to understand and educate on the requirements of each licensing scheme. This is a very important topic for the continued strength of open-source software, but it also takes time to understand the details. My own (very limited) understanding is that CC-BY-SA can be considered "like the GPL for documentation", and as such any direct copying means that the same license should be applied (as well as attribution provided).
I will do the following .
@svaksha Does changing the license and attributing you to the IDE part resolve the issue.
Please let me know ?
To add to what @ihnorton has said, while many of us have years of experience working with FOSS, others don't – and the ins and outs of etiquette and licensing aren't always self-evident. Mistakes will be made even with the best of intentions, but as long as people are willing to fix them and learn (and it looks like that's the case here), no harm done.
On Fri, Jan 9, 2015 at 3:03 PM, Melvin email@example.com wrote:
Usually, I always assume good intent, but your responses in this
AFAIK, one cannot compare a CC license with a software license where a
No, it does not (see above and below). But, before I delve into the
As for the CC legalese, IANAL and no two licenses (even among the CC)
Instead your response(s) in this thread and the commit message
Hope that explains my thoughts on the CR and CC license violations.
On Fri, Jan 9, 2015 at 2:57 PM, Isaiah firstname.lastname@example.org wrote:
Thank you for clarifying that Isaiah and thanks for alerting me to the
I'm glad you think licensing issues are an important discussion worth
IANAL, and yet in my response I've tried to explain the copyright and
It is disingenious to not notice that while he was easily able to
That said, ignoring the CC and CR violations is not the right way to
Thanks, SVAKSHA ॥ http://about.me/svaksha ॥
I have no disagreement with having one great collection, and I would prefer that it be curated.
Now, for my own understanding, If I make a copy of Julia.jl, then modify some parts, retain the original license, and credit the original author, mentioning my changes - isn't this something the CC license allows? Would I be considered engaging in plagiarism then?
@melvin0008 has changed his license. He needs to credit the original source, and mention the changes (perhaps in his README.md). Is there anything else? I am removing awesome-julia temporarily until the licensing is properly done. I will reinstate the work when it is compliant, and over time some of his stuff can be merged into Julia.jl too.
However, I still request (and this request of mine has been consistently ignored) that the discussion be kept to the technicalities of licensing, and no attacks be made.
@svaksha The difference between Julia.jl and awesome-julia is Julia.jl concentrates on packages from various fields and awesome-julia concentrates more on the learning perspective . Most of my listing have been done using julialang.org as the primary resource. Others have been done using google. There was only one part i.e the IDE which I had copied. I removed that part yesterday and have written a new IDE list, if you compare it to yours you will clearly see the differences.
So technically currently I haven't copied anything from anywhere.
@Viral As of yesterday, I changed the IDE part which was copied. I curated an entire new list as you can see. I curated this IDE list after searching for the respected languages cited at link .
I have kept the conversation technical but would like to note that the constant jibes of me being a sexist are inappropriate.
On Sat, Jan 10, 2015 at 4:38 AM, Viral B. Shah email@example.com wrote:
It does. In such situations, people automatically submit a PR (if they
The plagiarism (which violates the Indian CR, which violates the
Thank you Viral. A clear policy and a strong stand on CR and
Fwiw, I thought I had explained why inspite of the license change, the
Hope this clears the air.
Again, please be specific about which comment in my BR is not civil -
SVAKSHA ॥ http://about.me/svaksha ॥
On Sat, Jan 10, 2015 at 7:00 AM, Melvin firstname.lastname@example.org wrote:
Melvin, If you wish to play the denial game, I am willing to contest
I am going to ignore this as its not inappropriate to ask a question
@svaksha you can go ahead and look for copied commits .[I compiled a new IDE list yesterday] I can assure you everything has been found using google or information already existing on julialang.org. What content do you think is plagiarised? Please let me know. I will delete it.
@svaksha: regarding the extent of copyright, it is my understanding that facts are not copyrightable. As an example, Julia includes tables of mathematical data which are not subject to copyright even though they were almost certainly typed in from a book (at least originally). A list of Julia IDEs is not a copyrightable thing (IMHO, IANAL, etc.), but your original descriptors and arrangement certainly are. There is some kind of gray area I don't understand in between "phone numbers" and "photocopy of the phonebook", but clearly the verbatim copying was on the wrong side and was unacceptable. As such, I support Viral's decision to remove the link.
(for the sake of completeness, the exception to the "arrangement" criterion would probably be alphabetical ordering, which is a fact of English)
Regarding collaboration, there are many examples I can think of where people do their own thing: Keno built Cxx.jl instead of collaborating within "my" Clang.jl (which was obviously the correct decision as Cxx.jl is superb). Doug Bates is building a new RCall even though there are 2-3 (plus?) others written already written at various levels of maturity. How many Julia autodiff libraries are there? And, don't get me started on testing frameworks (15?). I can't imagine open-source working in any other way than "you work on what you want to work on."
Regarding CC-BY-NC-SA (thanks for the correction), per the CC FAQ it does not seem accurate to me to say that modifications/derivatives may only be made on a fork of Julia.jl. I could be wrong, and would appreciate the clarification if so, but I believe the CC-ND variant would provide stronger protections if that is your goal. Also see the section on license revocability. To the extent that I can understand the license (as per above), "awesome-julia" would qualify for the automatic reinstatement by adding the attribution (BY) and license (SA) -- both steps of which has been done, although Melvin has gone further and seems to have deleted all verbatim text in the IDE section and replaced it with his own descriptive text (I have not done a line-by-line comparison elsewhere).
Regarding the other point you made earlier:
This sounds fine. I have some wording in mind and will discuss in a separate pull request.
On Sat, Jan 10, 2015 at 7:26 AM, Melvin email@example.com wrote:
I will file the BR here later. Meanwhile, why this PR when I
Core-devs, please step in and stop this. TBH, I'm finding this
SVAKSHA ॥ http://about.me/svaksha ॥
@svaksha I have stated many times. My motivation for creating awesome-julia has been to let new developers with the set of resources to learn and start development to Julia. I made a mistake of copying the IDE section of Julia.jl . Consider it to be an amateur mistake. But I have corrected my mistake and changed the entire IDE part of awesome-julia. Julia.jl focuses on libraries and resources needed for specific fields. Mine is a overview and a great start for developers who are excited about Julia. All of the content I have curated has been referred from Julialang.org and simple google searches. You have ignored my kind request to kindly point out what you feel is plagiarized as of today. My view is for both the lists to exist in parallel and help the developers to get the required information , they are looking for. For me that's the spirit of Opensource and would be in the interest of propelling Julia as one of the top dynamic languages.
Other than this, you have berated me and accused me of sexism which by all means is a stretch and an attempt towards defamation. This is clearly inappropriate.
Please be cordial and point out issues you still feel are plagiarized. I would be happy to solve this issue once and for all.
I think we agree on this:
IANAL, but here is my understanding of the copyright situation. My view represents a background of American and European copyright. If this is different in India (as was indicated it might be in early discussion), please indicate so, so we may understand the subtleties involved.
Now, for the legal parts of the license itself. This is my interpretation from studying the license text. Again, please let me know if (either side) interprets this differently.
As well as 3.b which states
(this is the Share Alike part) as well as (this is the attribution part) 3.a
Now, there is a question of whether the given attribution is sufficient. From my reading of the license it probably isn't, as the license requires four explicit things to be part of the attribution.
So the two core questions are (which, @svaksha I would ask you to answer)
Also, I don't think anybody here doubts the importance of copyright and licensing. We've had so many discussions where we couldn't use code because of it's license and we take the issue seriously. That said, licenses are legal documents and it is best to argue based on the verbiage of the license (I recognize you have done this).
I hope this post can serve as a basis of further discussion. I would encourage either side to pick out any specific bullets they disagree with.
@melvin0008 I understand your frustration, but in this matter I don't think it is relevant, as the license does not have a "good intention" clause. Also, I would encourage both parties to temporarily ignore the non-technical issues of the discussion (and not respond to them) until the legal question is resolved. If either side still wants to make a statement one that is done, by all means do (but keep it civilized), but let's resolve the legal issue first to avoid further resentment. If there's anything I can do to help out with this, please let me know.
To clarify this point:
3.a.1.A specifically refers to the required attribution:
In other words, the licensor may ask the adaptor to remove the attribution which is otherwise legally required. This does not apply to the copyrighted content itself, however (and it wouldn't be open source if it did).
While it's polite to avoid reusing people's content if they request it, it's certainly not legally required by this license, as long as the attribution is in place.
Also, I don't see any terms that reference the possible motives of the adaptor, nor any licensing implications those motives would have. That line of discussion lies outside of the technical and legal concern.
Also, RE attribution I would point out 3.a.2:
Since the Julia.jl repository is itself a resource that includes the required information (the five bullets above), this suggests that simply linking to it is enough to satisfy the attribution requirement.
Have I complied to all the rules? Or do I need to change something?
You've attributed the Julia.jl repository in your readme, which from my reading of the license gives you the right to copy and paste as much of it as you want. Given that you have removed the copied content anyway, I would say that you've gone above and beyond what the license strictly requires and don't need to do anything more.
However, it would of course be best to wait for @svaksha's input – we want to make sure everyone is happy with the outcome here, insofar as we're able to.
(Though let's bear in mind that going beyond what the license requires is a favour to be asked, not a responsibility to be demanded.)
(Svaksha and I have been talking about this issue yesterday and I promised my input. Unfortunately writing it took a bit longer than anticipated, some information has been covered already.)
On Plagiarism and Licensing
Taking Svaksha's IDE list and copying it into Melvin's repository, without attribution, is Plagiarism. By omiting attribution, the list was passed off as Melvin's own work, instead of Svaksha's work. I'm glad to see attribution is fixed.
The license on Svaksha's work detailed one way in how to do this right, in section 3a of the license. The attribution section is explicit on the information you have to provide. I would argue the text 'Credits for curating the IDE List given to Svaksha' is insufficient based on that. It suggests to me, that Svaksha did a pull request into your repository (or mailed a diff or similar), instead of this being part of a larger work Svaksha compiled. It may even suggest she's actively involved in that section of your list, which she is not.
Another argument you could try to make, is that copying the list falls under Fair Use. Fair Use is a way of using the work, without following the license requirements or violating copyright. Note that fair use still requires a citation reference (attribution). Without this reference, it is still plagiarism. Fair Use does not cover including part of a work into a larger work, so Fair Use does not apply in this instance.
Circling back to Plagiarism, Plagiarism is a term originating from the Academic and Journalistic world. Note that Plagiarism has nothing to do with the author of the original work complaining or not. When you copy a text and don't attribute, irrespective of doing that intentional or accidental, the act is considered Plagiarism. Plagiarists are usually cast out by their peers. Back when I went to university, people commiting plagiarism were expelled, whereas people publishing scientific and journalistic works with plagiarism were expelled from their (scientific and journalistic) communities.
In most cases, edits to an existing work don't even create a new work. An obvious example is fixing a typo. Just because you changed a letter doesn't make it an original work. The same goes for rewording. For example, changing the text "Yoda, I talk like" to "I talk like Yoda" doesn't suddenly make it your original work. Taking a single chapter (like the IDE section) and adding a lot of other text, may make the resulting text an original work, but won't make that copied chapter an original work. Copying the IDE section and then rewording it puts awesome-julia clearly in the danger zone of a license and copyright violation. Wether it is actually a violation, is up to a court of law, but I certainly wouldn't want to be in Melvin's position.
You are right that combining work in new and interesting ways, is a pillar of open source. Usually, this involves software. However note that this ability usually comes with restrictions. For instance, if I link against a GPL library, I must make my software GPL. If I adapt someone's code, I must maintain the original copyright and license. These requirements are spelled out in the license document. The same holds for text.
Why Contacting JuliaLang
Svaksha spent a lot of work on creating her list. It involves finding projects, understanding what they are about, thinking of a one-line description and categorizing them. Unattributed copying of the list removes the recognition she'll get for this work and instead directs it to Melvin. The instance specifially addressed here, is the recognition and endorsement Melvin received from the JuliaLang project. Not only does JuliaLang praise Melvin for the list (including it in your list of learning materials), by including it, it also legitimizes it. For these reasons, Svaksha is right to bring it to the attention of JuliaLang.
Svaksha has pointed out that Melvin choose to cooperate with aweomse-python and awesome-php lists, but choose not to do this with Julia.jl. She also pointed out her experience with Indian men was bad, therefor prefering an open discussion instead. This happens in open source communities and this happens in tech.
They're not views, but they're experiences. They validate her reluctance to handle this quitly. These experiences are also shared by a lot of people, making them not personal either. She even pointed this out in the piece where you responded to her. :( I don't see what is uncivil about it either, she calmly explained why she wanted a public discussion, included background information and even pointers on how to read up more on the subject. If anything, she went above and beyond the minimal requirements for an answer and kept it short too.
The commit message reads as if Melvin doesn't care much about the issue. Obviously the license problem is not on his priority list. But by extension, he doesn't respect Svaksha either, otherwise her finding this a big deal would have made the license problem a big issue by proxy. Instead, he refers to the whole thing as 'weird'.
It's non-trivial to explain this issue well. From your point of view, this is an isolated incident. Melvin is just someone who made some unfortunate mistakes. From our viewpoint (Svaksha and I) this is part of a pattern, a systemic problem that turns up again and again. Being ignored, being thought of as non-technical, being questioned on our technical or mathematical skills... On top of that, many people aren't even aware they're doing it, it's that deep ingrained in society.
I touched on this briefly in the 'Why Contacting JuliaLang' section, but I need to reiterate it here. Men taking credit for the work a woman did happens. A lot.
From our point of view, Melvin looks like someone who'll happily collaborate with men, but won't with women, instead copying their content. This may not even have been a conscious decision of him, but that doesn't mean it's not a sexist act.
Thanks @nahratzah for your comment. I will not touch on the sexism, because I do not feel like I'm in a position to comment. I recognize sexism exists in tech and that it is a significant problem.
I agree with everything you said. Specific comments/followups inline.
I think you'll find widespread agreement in this community. A significant number of Julia contributors are academics and we take these issues seriously.
Yes, as I mentioned I agree that the attribution was probably insufficient. Would you and @svaksha agree with the verbiage
Yes. Copyright law is tricky and people sometimes need to be reminded of it. I hope (and think) @melvin0008 has learned from this experience.
Yes, I understand this and it makes sense. However, please understand that this puts us in a tricky position. We are not a court of law and thus cannot make legal judgements. There are other avenues for that including filing a DMCA claim with GitHub. We hope to make sure that the community we have built up is as friendly and welcoming as possible, and in this capacity we can try to resolve the issue (ideally by talking to both parties and educating where misinformation exists thus trying to resolve the conflict amicably). I hope the received response shows that we care about these issues.
As was discussed early, it would make sense to have a general policy on plagarism and license violation in the officially curated lists, but we should probably discuss that in a separate issue.
Ultimately, I would like @svaksha to comment on whether she still considers any of the material in @melvin0008's repository as it stands to be plagiarized. Afterwards, I think a discussion on collaboration (for which an issue on Julia.jl would probably be a more appropriate venue) can happen if both sides are still interested. We do like to see people join forces, but of course we can't require anybody to collaborate.
@Keno I have definitely learnt a lot regarding Licenses and Plagiarism. I will never make such a mistake again.
@nahratzah, Thanks for the detailed response. Sorry about the slow response - been travelling with patchy internet access.
@Keno, Thank you for acknowledging the CC/CR violation. While I'm glad that the core-devs take this seriously enough, it would be nicer and clearer if there were a public policy outlining the do's and dont's - there have been some discussions in the past on the mailing list and usually the persons involved step back and stop. That isnt the case here, so to answer your question:
YES, please see the commits linked below.
The crux of the problem is that its not just one attribution, rather a number of commits (see below) that have been plagiarised over a short period into
I agree with @nahratzah who has already explained the copyright and CC license violations very nicely with the 'fair use' and the "yoda" example for plagiarism. Ofcourse, Melvin disagrees: "So the part of IDE list still being plagiarized is out of question" (#194 (comment)), which means he intends to continue copying stuff over if
Since the bug report was closed on Melvin's demand, it kills the discussion route and feels like silencing. Not at all what I expected. In this specific case, the reality is that Melvin didnt dare to break the license and plagiarise a man's work, so within the FOSS community, it sends the signal that its acceptable to steal a womans work. Feel free to disagree with this POV but that is my take-away from this experience which isnt very different from what other women have experienced.
Before delving into the commits, I do want to address and describe why it takes a lot more work than "I got the link from the Julia website and via google". Julia being a new language is constantly in a state of flux and the easy part is listing packages listed in the official package list as they would have clear-cut descriptions of what the package does, the deps, tests, supported versions, etc... ala, say PyPI or CPAN or CRAN, where you can easily figure out what the package does and get back to work. BUT what of those packages (for lack of a better term) that are not on METADATA? And there are numerous such packages (erm...code repos?) that are not a part of the official list but I intentionally list these unregistered packages in
I dont use a large majority of the packages, but still obsess over figuring out what the package does when the repo has no README or description, and if the package name is a cryptic acronym (like 'HMM') it can be anything - a lib used in Math, Biology, Statistics or something else entirely. Assuming its the former, I have to start reading the code to figure out the classification of the parent category. If the HMM package was written with a focus on bioinformatics AND/OR another type of "science" research, it can be again sub-classified under say "Cancer Research", that is further sub-categorized under say, genetics, epidemiology, cell biology, etc.., respectively. If the HMM package has a Statistics/Math focus, it has to be listed as per the sub-cat on the top level page. Sometimes the information is available if the devel wrote tests (not really, it would then be in the official metadata list, though not necessarily), or via comments within the code, the function names, etc which I co-relate with a search engine result or a paper I read, or the work the dev does - Example, their profile may state they are a researcher working in a Cancer lab or a Statistics lab and co-relating that with similar 'category' packages in their repo page helps me guessimate what the package does (or not - I make mistakes too). Some developers may not agree with the category chosen and they do write asking me to relist their package under a different CAT than the one I chose (thanks), or some may ask me to remove the listing as its not meant for public consumption, or just write to say thanks for listing it. When the latter happens, I assume they are ok with the categorization :-)
Roughly, that is the most simplistic way to describe the effort that goes into the categorization and while the category tagging may seem like a trivial obsession over very little (the package description), therein lies the key difference. Boot-strapping a new language and getting it accepted into an existing production system is hard and depends on a large library ecosystem support that ensures that people find the library the developer spent hours writing (/testing/developing), use it and improve it and add more features and contribute upstream if they are so inclined. I'm happy that
List of commits that plagiarise
We've drafted a community standards document here:
It's not linked from the front page yet – that's the next step.
@svaksha: I didn't go through all of the links in your post, but I did go through a significant sampling of them – 15 or so. It seems that, aside from the commit that copies the list of IDEs, including descriptions, which @melvin0008 has acknowldged and removed, they do not share copyrightable text with your repository. Many of the links are the same, but links are facts, which cannot be copyrighted and don't fall under the scope of plagiarism either. Given that the copyrightable text about IDEs has been removed and @melvin0008 now links to Julia.jl (thereby qualifying to use the material under its creative commons license), do you agree that there is no longer any copyright violation occurring here?
@melvin0008, although lists of links (like tables of facts) are not copyrightable, I do think that it would be nice if you added at the top of the README, after the phrase "Inspired by awesome-php.", a sentence along the lines of "Many of these resources were originally collected and curated by @svashka and others – thank you for the hard work." Even if not legally necessary, as @svaksha has pointed out, curating such lists is a lot of work, so it seems appropriate to give as much credit as possible. You might also want to include a sentence about your motivation for making a different list – what perspective does your collection bring to the issue of curating Julia resources?
Links that have been curated from Julialang.org/community and Julialang.org/learning are as follows:
The repo maintainers reiterate that we do not condone and will not tolerate personal attacks in any forum. Please forgive past mistakes and please exercise restraint in all comments.
This issue was opened to address the following:
(1) a copyright violation and related link from JuliaLang/* resources. This violation has been acknowledged and the following corrective action was taken:
(2) codification of community standards with respect to copyright/attribution. To this end:
We feel that the specific issues raised above have been addressed as much as possible on this issue tracker. The comment area remains open, in case there are any further, actionable items related to the above issues.
Please be aware that this is not the appropriate forum for general copyright disputes, nor are any maintainers able or competent to adjudicate such issues.
We expect everyone to follow the community standards in any future comments and refrain from personally-directed comments.