Talk:Fork (blockchain)

Fix lede

User:Ladislav Mecir I'm in the middle of fixing the order that you rightly pointed out is broken, but I see you've undone again as I'm fixing it! Did you see MOS:FIRST? regards Widefox; talk 12:49, 7 March 2018 (UTC)

Hi. The problem with your edit is that it destroyed readability of the first paragraph. I updated the text to satisfy MOS:FIRST, while maintaining the sense of the text. Ladislav Mecir (talk) 13:02, 7 March 2018 (UTC)

The history of the article shows that the lede was OK before [1] but was broken here [2] by your edit. Yes I should have explained on this talk that I'm in the middle of editing to include your concern (which I presumed other editors would fix anyhow), but I didn't expect you to repeat undo my edit when it can be fixed. As I'm contesting your original bad edit (in part) please give me an opportunity to complete my attempt #2. If you don't agree, please discuss here as it is about undoing your original contested edit. Widefox; talk 13:16, 7 March 2018 (UTC)

Well, after looking to fix this, we need more sources. I've rewritten the lede as best I can (not brilliant) to broaden the scope to include both the old version and the previous one. So that's including both 1. fork (software) and 2. blockchain "fork"/"split". That's just following the source. It should probably get rewritten with better sources. Is that better? Do you know some sources about the difference between blockchain "split" and "fork"? Widefox; talk 16:05, 7 March 2018 (UTC)

"The history of the article shows that the lede was OK before [3]" - actually, it wasn't, since the information given was misleading. Per article sources forks are of two kinds: hard forks and soft forks. Both types are blockchain rule changes, hard forks are not backwards compatible rule changes, while soft forks are backwards compatible rule changes. As explained in the article, in consequence of the change(s) the blockchain can either split (=diverge into two paths forward) or continue as one chain without splitting. This information is substantial. It is not a good idea to say that a rule change (a fork) is the same as a chain split. There are examples (already cited in the article) when a rule change (a fork) occurred, but the blockchain did not split. There are also examples when a rule change (a fork) occurred and the blockchain did split. We can either mess up and mix these notions in such a way that nobody knows which is which, or try to be as informative as possible and discern these notions. Ladislav Mecir (talk) 07:23, 8 March 2018 (UTC)

I see that your edits did exactly what I was afraid of - you deleted the split notion description from the article. That messes the things up completely. Ladislav Mecir (talk) 07:38, 8 March 2018 (UTC)

Please check the source [4]. a fork is what happens when a blockchain diverges into two potential paths forward -- either with regard to a network's transaction history or a new rule in deciding what makes a transaction valid.. That is correctly portrayed in the article, isn't it? (in fact, I too closely paraphrased it). That source clearly includes in the scope of fork both the blockchain and the rules A byproduct of distributed consensus, forks happen anytime two miners find a block at nearly the same time. The ambiguity is resolved when subsequent blocks are added to one, making it the longest chain, while the other block gets "orphaned" (or abandoned) by the network. But forks also can be willingly introduced to the network. This occurs when developers seek to change the rules the software uses to decide whether a transaction is valid or not. That's also correctly represented in the article, right?

We need more good sources to make a change from this, as the current wording just follows the source. Previously, statements did not follow the cited source which is totally unacceptable here per WP:OR / failed verification. I have no opinion about the correct definition of this as it is reflected in sources. I repeat, do you have some? Feel free to change or discuss if you do. Regards Widefox; talk 14:25, 8 March 2018 (UTC)

Per this CoinTelegraph article, forks are protocol changes. This definition is much more practical than the definition using the "diverges into two potential paths forward" one, since there always is an unlimited number of potential paths forward, depending on who creates the new block first. It is also compatible with the soft fork and hard fork definition, as opposed to the "potential paths forward" formulation, which is totally incompatible with the soft fork or hard fork definition essentially common to both the CoinDesk and CoinTelegraph articles. I shall try to find more sources defining the notions. Ladislav Mecir (talk) 16:57, 8 March 2018 (UTC)

Widefox, you are right that the current wording of the lead section violates the copyright of the cited source. Ladislav Mecir (talk) 09:59, 22 March 2018 (UTC)

Definition by Amy Castor

Amy Castor defines a fork as "fork is what happens when a blockchain diverges into two potential paths forward". In my opinion:

• While the Castor's definition of a fork is based on "blockchain potential paths forward", the particular fork types (hard fork and soft fork) are all defined differently by Castor—they are defined as "a software upgrade that introduces a new rule".
• The word "potential" means "likely", "unrealized", while the examples of a fork in the majority of sources are such changes that the "two paths forward" either do not exist and were not likely to exist (e.g. Segregated Witness), or are already realized (e.g. Bitcoin Cash). Ladislav Mecir (talk) 09:59, 22 March 2018 (UTC)

Definition by Andrew Tarr

An article by Andrew Tarr defines forks as "changes of protocol". In my opinion:

• The definition differs substantially from the definition proposed by Castor. (The definition does not refer to potential divergences in the blockchain.)
• The definition agrees with the definition of fork types proposed by Castor. (Protocol changes are equivalent to software upgrades, since the protocol is implemented in software.)
• The definition agrees with the definition of fork types proposed by Tarr.
• The definition describes the majority of fork examples that can be found in the independent sources. (In particular, it describes both the Segregated Witness and Bitcoin Cash.) Ladislav Mecir (talk) 09:59, 22 March 2018 (UTC)

Definition by Andreas Antonopoulos

A book by Andreas Antonopoulos defines a fork as "Fork, also known as accidental fork, occurs when two or more blocks have the same block height, forking the block chain." In my opinion:

• The definition is not compatible with Castor's definition. (Only realized divergences are called forks, Castor's unrealized divergences are not accepted as forks by this definition.)
• The definition is not compatible with Tarr's definition. (Protocol changes not causing blockchain divergences are not considered forks by this definition.)
• The definition is not compatible with the majority of fork examples in independent sources. (In particular, it describes neither Segregated Witness nor Bitcoin Cash, since none of them were accidental, in fact.) Ladislav Mecir (talk) 09:59, 22 March 2018 (UTC)

Ladislav Mecir As I didn't say it violated the copyright, that's a straw man. It is close paraphrasing, which can be improved. As you can see, the current broad definition covers the variety. It's both fork of data and fork of code. Widefox; talk 11:27, 22 March 2018 (UTC)

"I didn't say it violated the copyright" - you did say that you paraphrased the text too close, i.e. that you actually did violate the copyright, and I do agree. Ladislav Mecir (talk) 13:46, 22 March 2018 (UTC)

"As you can see, the current broad definition covers the variety." - It provably does not, as illustrated on the examples. We should respect that different definitions exist. Ladislav Mecir (talk) 13:46, 22 March 2018 (UTC)

Yes I said I paraphrased it (I'm repeating myself). You are making the claim of copyright violation, not me, which I reject and suggest you read WP:PARAPHRASE Limited close paraphrasing is appropriate within reason...Limited close paraphrasing is also appropriate if there are only a limited number of ways to say the same thing. and to quickly fix this highlighting that it should ideally be rewritten, but I was unable to is OK IMHO. If that's not your opinion that's up to you, but do not make the factually incorrect assertion that I claimed it was copyright violation, which you should withdraw and decese from and this argumentative line. I do, however, agree that it could/should be replaced with something else, but my aim (repeating myself) was to quickly fix an article that failed verification. It may be optimal to quote the different definitions, which you seem to have done work collecting, but I leave that for others. Widefox; talk 16:30, 26 March 2018 (UTC)

References

Recent revert

This revert by YechezkelZilber reinstates an edit of a WP:STATUSQUO content. I propose to revert the text to WP:STATUSQUO again based on these points:

• The WP:STATUSQUO content is based on a citation of a WP:IRS.
• The non status quo content is based on a citation of Investopedia, which is not accepted as a reliable source here. In my opinion, it also misrepresents the content of the cited Investopedia article.
• The non status quo content is non-encyclopedic - its format does not adhere to WP:MOS.
• The non status quo content is non-encyclopedic, using such a non-encyclopedic term as "vise versa".
• The grammar of the non status quo content is incorrect, making the text rather meaningless.
• Even if the grammar of the non status quo content was corrected, the text would still remain factually incorrect. Ladislav Mecir (talk) 12:55, 26 May 2020 (UTC)

1) The former text was simply wrong. Unreadable and factually wrong text cannot be used under the "here is a reference" logic. Note that a lot of the Bitcoin entry uses a couple of those hard to read sources with bad language. This is encyclopedia. Not a reciting of a couple of badly written technical books.

2) Stop wikilawyering. You deleted by arguing "reference". I gave reference. Now you start lawyering "credible source". There are ten more sources showing the same content roughly. Am I on the hook to satisfy all your demands for every edit?

PS. I think your edits are in good faith. But I disagree. And please do not force your edits like you did in the last day. You can argue. I can see that your style is different. But there is no need to be aggressive. You can discuss before taking unilateral actions.... Respectfully Jazi Zilber (talk) 13:10, 26 May 2020 (UTC)

@YechezkelZilber: "There are ten more sources showing the same content roughly." Why don't you cite a source that can be accepted as reliable, then?

"But there is no need to be aggressive." Indeed. Especially reverting the WP:STATUSQUO to your preferred wording without discussion is not the suggested way how to proceed. Ladislav Mecir (talk) 21:59, 26 May 2020 (UTC)