This article is within the scope of WikiProject Computer science, a collaborative effort to improve the coverage of Computer science related articles on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.Computer scienceWikipedia:WikiProject Computer scienceTemplate:WikiProject Computer scienceComputer science
This article is within the scope of WikiProject Software, a collaborative effort to improve the coverage of software on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.SoftwareWikipedia:WikiProject SoftwareTemplate:WikiProject Softwaresoftware
This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing
Latest comment: 2 years ago6 comments3 people in discussion
Just saw the edit summary: "Where is the content written like an advertisement?" Well here is a meaningless marketing like statement for a start:
Translating DOOM from C to V and building it takes less than a second.
building it on what hardware? compared to what? what secondary sources benchmark this? I haven't read all this page yet, but statements like that do sound rather "advert like". Thank you for your attempts to improve the article though. Sirfurboy🏄 (talk) 17:27, 13 March 2023 (UTC)Reply
Agree – things like "it does this so fast!" or "it does this in this really unique and revolutionary way that deserves an expansive mention!" are largely subjective and do not belong in an encyclopedic article.
Thanks Sirfurboy for the input. I can see your perspective, but aren't the 2nd and 3rd statements offering proof of the statement above it? Each link, is placed to verify the validity of the statement. Thus it seems like we are caught in a circular catch-22. The section is C translation. Simply stating that, offers no context to the reader. The statements below are expanding on what that means in respect to features of the V language. Thus, "V can translate your entire C project and offer you the safety, simplicity, and compilation speed-up (via modules)." But now is that statement merely a claim or a fact? Therefore the next statement and link are proof of the claim, "Translating DOOM from C to V and building it takes less than a second." Furthermore, V is not selling anything by this. There is no product. That it can quickly translate a C program to V, is a proven statement, that is backed up by visual proof.
Don't get me wrong, I'm perfectly fine with us removing statements to not look like advertising or promotional content, or rewording statements with flowery sales-like language. But worried that removal of key explanatory statements, can lead to the next editor taking a position that statements are mere claims without merit or proof.
"What secondary sources benchmark this?" A primary source was used to show visual proof, along with secondary sources, in the context of capability to translate C projects to V. As is it can be interpreted that you are objecting to speed or type of hardware, the statement was removed. The statement of fact, on the capability of the language to do such a translation remains.
LVDP01, thank you for giving input as well. You mentioned a few things. "it does this so fast!" or "it does this in this really unique and revolutionary way that deserves an expansive mention!". I'm trying to narrow down actual specifics, so that we can come to a consensus as to what language is objectionable or acceptable. Can you please refer to specific lines in the draft that you might object to and give an explanation. Would greatly appreciate it.
Based on the comments you both have made so far, anything that could reasonably give an appearance of being written like an advertisement or promotional has been removed. Additional secondary sources have been added.
So for secondary sourcing, what I would like to see is someone who has objectively compared like for like compilation between, say, a C compiler and code compiled in V. I did look last night and found a 2019 article, which is a bit dated for this subject, which did give some figures, but they too appeared to be just quoting the documentation which says "V compiles between ≈100k and 1.2 million lines of code per second per CPU core (without hardware optimization)." Again, no comparison with other languages nor details about the CPU core. There are benchmarks out there. for instance,[1], but these are WP:PRIMARY and really not very clear to a reader. They are independent, so that is something, but isn't there a paper somewhere, or at least a trade press article, where someone has put V through its paces to see how it performs? If not, then maybe just say that optimising speed is a design goal, and leave it at that. Sirfurboy🏄 (talk) 15:17, 14 March 2023 (UTC)Reply
Thanks, as this airing out can lead to progress and demonstrate to everyone what is going, from what appears as a months long standstill. In the context to the original objection, concerning C2V speed of translating a C project, that was removed to accommodate the point made. My concern about what appears to be a new objection, is over fairness, and creating an additional barrier that is not based on the draft document nor is a standard other programming languages on Wikipedia are being held to.
Example, in checking the Zig Wikipedia (done to compare what might be acceptable), it has such statements as, "Zig is intended to improve code safety." Nim Wikipedia has, "Nim was created to be a language as fast as C". There was no secondary sources or references given nor were these statements challenged. I can not find any requirement nor examples on Wikipedia, that a programming language must provide a speed comparison between itself and other languages. I don't know where we would be going with V vs other language, as beyond the original context of objections that were supposedly based on advertisement-like statements.
"V compiles between ≈100k and 1.2 million lines of code per second per CPU core (without hardware optimization)." This statement does not exist in the draft document, but appears to be an attempt to challenge the validity of claims on an external website. That seems like it would be outside the scope of the present discussion and the issue of if statements being made are advertisements.
Related to this line of objections, are statements like, "it does this so fast!" or "it does this in this really unique and revolutionary way that deserves an expansive mention!" These also don't exist in the draft document, but appears to show a particular kind of reaction. I can't edit or objectively address, what does not actually exist in the draft. However, I did remove language from the draft, based on the sentiment of what may cause such an reaction.
"I would like to see is someone who has objectively compared like for like compilation between, say, a C compiler and code compiled in V". There are a few things going on here. The benchmark link to, has nothing to do with how fast V can compile itself, but rather is a speed comparison between languages, based on various tasks. Of which, there are many other such sites, Kostya Benchmarks. Additionally, the performance shown and scores of languages constantly change depending on week or month, compiler used (which get updated), algorithms used, quality of code submitted, varying number of submissions (some languages have several versus just one), and being optimized for the task.
The draft statements are showing that V can compile to C. If we search for additional information outside of the scope of what's presented in the draft, V can use various C compilers. V can compile to C, then use various C compilers to create executables. Speed would be indistinguishable, between V or C source code, relative to the algorithms used or adding of any additional features (like optional GC). "V's main backend compiles to human readable C", is a statement of fact about what the language does and provide. The references are for validating that statement and features of the language.
"If not, then maybe just say that optimizing speed is a design goal". I removed the original statement entirely, which was "DOOM from C to V and building it takes less than a second". Some form of it could be put back later, but if this is to ever move beyond a draft (after all these months), probably for the best that it's removed.
"V compiles between ≈100k and 1.2 million lines of code per second per CPU core (without hardware optimization)." "Again, no comparison with other languages nor details about the CPU core."
Actually, V's website does give details about CPU (refer to fast compilation section). "V compiles ≈110k (Clang backend) and ≈500k (x64 and tcc backends) lines of code per second. (Intel i5-7500, SM0256L SSD, no optimization)" Differences in CPU, OS, compiler/backend, version of V, and year of test may give different results. Didn't address this statement particularly, as not in the draft itself. However, adding the additional information and context, as mentioned here and checked on it.
For clarification, "V is written in V and compiles itself in under a second.", was addressed as mentioned in the draft. This is referred to by secondary sources (Marcos Oliveira and Navule Rao) and demonstrated visually on two different videos. An older video from March 2021 (Building V from source in 0.3 seconds) and a newer video from a demonstration for IBM (...presentation of V's features at IBM).
