Talk:SharpDevelop

Official web site?

There are two different links to an "official web site". Is this correct? --RenniePet (talk) 22:40, 7 June 2008 (UTC)

Recent rewrite

Tried to establish notability and added some external refs. Is it fine now? -- Tom Jenkins ^(reply) 19:39, 6 September 2011 (UTC)

Look, if you don't tell me how to improve it there's nothing I can do. You don't expect New York Times to start talking about SharpDevelop do you? Or maybe an article in MSDN from chief competitor Microsoft? What else do you need? -- Tom Jenkins ^(reply) 08:11, 7 September 2011 (UTC)

Start by reading the relevant guidelines, such as WP:RS, WP:Notability. Take article disputes to the topic's discussion page. TEDickey (talk) 08:14, 7 September 2011 (UTC)

Alright, it appears I've shown its notability with adequate refs. Cleanup? What? Where? You mean I've added too many refs? And whats a "promotional edit" anyways? -- Tom Jenkins ^(reply) 10:56, 8 September 2011 (UTC)

Tom, don't worry about it - overall, it's coming along. The "Cleanup" tag is, in my opinion, not helpful. In future, seek help from other editors at WP:Requests for feedback or at /Assessment at whatever project the article falls under, or with Request for comment, or even by using WP:IRC. So here's my buncha advice. The guidelines neutral point of view, WP:PROMO, and WP:PUFFERY, which includes words to avoid, are useful. I don't know if you're related to any part of the project, but WP:COI speaks to that case. To answer your question about too many refs, yes, more than three in a row^[1][2][3] per claim are too many. Usually just one per claim (max) is sufficient; one per paragraph is common, when a source is really deep. Feel free to spread those refs around to support specific claims throughout the article. I think you've established notability with good sources, so it's time to get jiggy, er, specific with them. If several sources say basically the same thing, just keep one or two. The General notability guideline emphasizes widespread coverage in notable, reliable sources, so don't strip out sources which support "widespread". Per WP:RS, lose any blogs of people who are not a) notable by themselves, b) established cited experts, c) published authors, or d) principals in the project. It would be good if you could add authors, and dates to those inline citations. To use a citation in multiple places in an article, name it:

<ref name="SG">Guy, Some (2011). ''Some book.'' Somepub.</ref>, and later, <ref name="SG"/> <-- note the extra slash

Finally, MOS:LEAD describes what the first paragraph in an article (above the table of contents) should try to accomplish. I personally like a "clean" lead, which summarizes the article's main points, but leaves the inline citations for later, in the article body. --Lexein (talk) 11:49, 8 September 2011 (UTC)

Don't "pull" paginated, specific book references out of inline citations. They're the best sources you've got - they should be inline cited where needed in the article. Sorry if you got bad advice from someone. --Lexein (talk) 12:50, 8 September 2011 (UTC)

Thank you ver' much! I'll be workin' on it! -- Tom Jenkins ^(reply) 07:52, 9 September 2011 (UTC)

Done paginated links - brought back into article in place of online links

However, reading through the list - some of the claims aren't supported. "Download stats" from sourceforge isn't necessarily related to the number of users, for instance. Likewise the link for ohloh doesn't point to information that supports the statement. At least two of the three authors of the book are developers of SharpDevelop. A good place to start cleanup would be to remove the questionable links. TEDickey (talk) 00:35, 9 September 2011 (UTC)

Done download stats - reworded

Done ohloh link - removed many unimportant reviews/links

Done self-authored book - reworded

Ditto the links given for "generally favorable reviews". The first one appears to be talking about like/dislike from a couple of dozen people (not even close to the "10,000" claimed earlier) with Visual Studio the only clear choice, and the second looks pretty neutral, with as many minuses as pluses. TEDickey (talk) 00:51, 9 September 2011 (UTC)

Done "favorable reviews" - removed

The links given for AvalonEdit likewise don't support the comments (a) it is a separate project and (b) it can be used by other programs. TEDickey (talk) 00:59, 9 September 2011 (UTC)

Done (a) - removed

Done (b) - well the entire CodeProject article is about using the component in your own application. Can I can make it any clearer? (in the link) -- Tom Jenkins ^(reply) 09:46, 9 September 2011 (UTC)

The link given for the form designer doesn't mention F#, IronPython, IronRuby. The list after that link is unsourced. TEDickey (talk) 01:01, 9 September 2011 (UTC)

Done form designer links - added refs

Thanks for being more specific. Yes, I see what you mean in that I've done a hasty job and connected any ref to prove any point. I'll improve the refs or correct the statements. -- Tom Jenkins ^(reply) 07:52, 9 September 2011 (UTC)

I see that the download-stats has been interpreted as one download equating to one developer. The source doesn't state that there are no duplicates, and in practice something that goes through many releases will have a corresponding number of duplicates. (Since the assertion regarding the number of developers is not found in the source, it should be modified to reflect the source) TEDickey (talk) 22:04, 9 September 2011 (UTC)

A recent edit moves the chunk beginning "SharpDevelop contains an equivalent feature for almost every important Visual Studio feature" to make it more prominent. The source for that statement is of course the developer's website (noting that a statement like that should be sourced to an independent well-known reviewer). The second source for that statement (to this comment) doesn't relate to the statement, and can be removed. TEDickey (talk) 13:00, 2 October 2011 (UTC)

It was arguably more prominent in the lead paragraph. Since my motives have been presumed to be in bad faith, with the language "to make it more prominent", I hereby testify that I moved it to the features section with the motive of putting it with features, just as stated in the edit summary, and nothing else. Anyways, getting on with actually improving the article, I've sourced a rephrase of the claim from a different primary source page, and quoted the Google book source, rather than removing the source as suggested. Looking forward to your positive, constructive, content adding, source adding, contributions to this article, Tedickey. --Lexein (talk) 18:27, 2 October 2011 (UTC)

Making it the first sentence of the first paragraph in a section make it more visible than buried halfway through an overly-long paragraph. Citing WP:AGF in response to a standard editorial comment is not being cooperative TEDickey (talk) 18:33, 2 October 2011 (UTC)

If you imply that an editor acts in bad faith, expect a reminder of WP:AGF. If you don't know someone's motives, either ask, or just agf, and do not publicly tar someone.

Continuing to just go ahead and improve the article, I added RS primary source (book) for claim. --Lexein (talk) 19:37, 2 October 2011 (UTC)

Book references

Added book references inline and cleaned up minor/major points and reference links as requested by TEDickey and Lexein. Is there anything else that needs to be done? --- Tom Jenkins ^(reply) 09:49, 9 September 2011 (UTC)

The (renamed) Tool suite details section needs refs. Other than that it's looking good. I like the references list at the end, as you've done. If you like, all the refs can go there, for a very tidy wikitext while editing, but that's of low importance. I fixed up the prose/punc/ref order (inline refs go after punctuation). Hope you don't mind that I've recast the citations themselves in the way that {{cite book}} renders them (Last, First (date). "Title"/Title. Publisher. p.), italicized book titles, quoted article titles, capitalized titles, and added language descriptors. Frequently, ISBN numbers are also used in book citations, but it's no big deal.

Keep a lookout for magazine and journal articles about the project - they often appear in Google Books searches. And "online" sources by notable authors, people involved in the project, or newspaper or magazine-related blogs, can be reliable as primary or secondary sources, depending. A careful re-reading of WP:RS will reveal some usable gold. --Lexein (talk) 11:00, 9 September 2011 (UTC)

Lexein, you've been an angel; thank you very, very much! Well, for one, the book refs look much harder to read, its like "Gates, Bill", which is contrary to the usual "Bill Gates", so I don't like that very much (can I flip it?), but apart from those tiny things, its much better, and so sorry for the awful typo; I'm not usually the one to make such mistakes but still very sorry for that. On the whole, thanks a million! I'm happy this is coming up and extremely happy to be getting some assistance. Have a very nice day. -- Tom Jenkins ^(reply) 19:15, 9 September 2011 (UTC)

"Architecture" is an improvement. I guess I'm biased, but I hoped you might let the refs sit for a while, as you look around at other articles, including WP:Good articles, WP:Featured articles and other software articles, to see how their refs are written. The last name first style is fairly universal in academia and encyclopedias (which WP strives to be.) Two advantages come to mind. One, you can write "Smith humorously referred to the software as "not confusing at all".", and the reader can find the ref by name, number, or by clicking on the number. Another is when multiple cites from a source are needed; then, "Smith, p. 27." and "Smith, p. 82." are quite preferable to repetitions of the full cite, only differing in the page number. --Lexein (talk) 19:56, 9 September 2011 (UTC)

The book reference for "Windows developer power tools" has more than a few lines discussing SharpDevelop. The total in the other half-dozen in the lede paragraph amounts to no more coverage than the paragraph immediately above this - across more than a thousand pages. Those are what's referred to as "incidental" mention, rather than "in depth". Google finds text in 7 of the 8 books cited; looks like a few pages in the power tools book. It finds nothing in the book about Microsoft SQL Server 2005 TEDickey (talk) 22:57, 9 September 2011 (UTC)

Broad coverage may be interpreted as deep in a few sources, or shallowly in many, many sources, or a combination of the two, as we have here. Windows Developer Power Tools references the software 45 times, and discusses it in depth. The rest either recommend it, or list it among the recommended. I'm satisfied with the WP:N notability of SharpDevelop in its field.

However, it is true that the plethora of mentions could be trimmed to a selection of the best quality, but I'm not going to press that issue: it's minor. We should WP:AGF about the Microsoft SQL Server 2005 citation, because WP:V#Access to sources does not mandate google verifiability. URLs in cites are for convenience of verifiability, but most books are not online. Tom - could you please

• quote the text used from pg 580 of ref #4?
• locate 2 or 3 magazine reviews or tutorials? Thanks. --Lexein (talk) 10:36, 2 October 2011 (UTC)

I'll see what I can do. Thanks for supporting this article, Lexien! Its really appreciated. WP needs more positive and helpful editors like you. -- Tom Jenkins ^(reply) 14:05, 3 October 2011 (UTC)

References

primary source used for notability claim

That statement regarding "almost every important feature" is apparently using a primary source, and is a notability claim. That needs some substantial third-party sources, which delineate the features which are and are not part of Visual Studio. Bear in mind that the express versions are much-reduced from Visual Studio, and referring to the two as the same is misleading. TEDickey (talk) 08:14, 5 October 2011 (UTC)

Even limiting to Visual Studio Express, the statement is from the developers and is used in a context that makes it a notability claim. By "substantial" sources, we mean that it's more than a line or so from sources that deal mostly with related programs. TEDickey (talk) 09:21, 5 October 2011 (UTC)

Yes, primary claims should literally be quotes directly attributed to the developers in the prose, as in how I left it a few edits back - primary sourcing should never be hidden from the reader of the prose. Should revert this to avoid WP:ADV. --Lexein (talk) 17:14, 5 October 2011 (UTC)

In essence, I'm just trying to explain that it contains every basic feature needed for development like VS. I used the word "important" to show that the basics are workable and you can use it like normal. Basics = project templates, code editing, GUI editing, intellisense, debugging, unit testing, etc. That is the reason I originally described all the major features in the lead, to justify/explain that statement. Someone put that statement into features, and then the statement becomes weak without explanation and you cry out for primary sources. I don't need primary sources to prove that a Pencil is a Writing Tool, because if I can explain how it is a writing tool, that statement justifies itself. So you tell me what you think should be done. But I ask that you look from a reader's POV, one who comes from a VS background and how you can explain to him/her that SD is a viable alternative easily. -- Tom Jenkins ^(reply) 10:02, 6 October 2011 (UTC)

I'm not trying to stress you out. Basically, Wikipedia's voice is that of independent secondary sources. We try to avoid being mouthpieces for primary sources. The quote was primary sourced, and attributed in the prose to the developers - this is proper distancing language. If secondary sources do not directly support the product's feature-for-feature matchup then neither can we - to do so would be WP:OR, or if only primary sources say so, and we don't quote them, WP:UNDUE. If secondary sources separately support claims of support of separate subsets of features, we must be careful to avoid WP:SYNTH, and not say that all features are replicated. We could at this time say some variation that the developers intend or plan to match the product's feature set with commercial offerings, unquoted, while citing the primary source. --Lexein (talk) 10:26, 6 October 2011 (UTC)

Sorry for sounding stressed out, and apologies for my lack of understanding in such a complicated situation. Thank you for the explanation, but this only toughens the point for me. I'm trying to show that its a viable alternative, nothing more. Since I come from a VS background, it took me a couple of years to think of trying such a tool out. And when I did, turns out it runs much faster and generally works fine for small-medium apps, the GUI is more responsive and it supports the default/custom .NET components. And the install requires 50 MB, not 5 GB on your hard drive. Now I'm trying to transfer such experience to the reader, struggling to get past the requirements of WP. Sorry for being such a dolt but I'm rather simple minded in my approach, usually being WP:BOLD and trying to get the flat facts on the paper, in as unbiased a tone as possible. -- Tom Jenkins ^(reply) 08:17, 7 October 2011 (UTC)

able to compile

Isn't the compiler itself part of the SDK (from Microsoft)? If that's the case, it's a pointless statement, except for promotional purposes. TEDickey (talk) 08:16, 5 October 2011 (UTC)

What?! How is anyone going to know about that? Whether its part of the MS SDK or Apple SDK doesn't mean anything to anyone. If it doesn't compile .NET apps, Sharpdevelop is entirely useless. It is extremely important to explain that it is actually workable, and functions like VS. Most devs come from a VS background so you're trying to explain that SD is able to work approximately the same. An important part of WP articles is that you break down the topic and explain each part separately, whether it being obvious or not. It seems like you've got some hatred towards Sharpdevelop that you want to make it as hard to understand as possible. Do you work for MS or something? -- Tom Jenkins ^(reply) 09:58, 6 October 2011 (UTC)

The point is whether they wrote their own compiler or provided (or adapted) configuration files. I'm pointing out places where the statements made don't appear to follow the sources. Going beyond the sources can be construed as promotional - that's essentially what marketing does, requiring readers to discount much of what's said. By following the sources more closely, and checking for reliable secondary sources (not download sites, which cut/paste from the developer's adverts), it's possible to get a reasonably neutral topic. Taking a note from another editor, you might read WP:AGF to get some additional context. TEDickey (talk) 10:54, 6 October 2011 (UTC)

You've got some good points, ted, and I'm no expert in WP policies, or sources for that matter. I'm trying to place what I believe to be fact - that SD is about as fine as VS (which I was willing to back up with a detailed list of features), although it may not support advanced features, the basics are workable. I don't know what else I can do at this point. Sorry for ABF. -- Tom Jenkins ^(reply) 08:20, 7 October 2011 (UTC)

Well, reviewing the available sources and pruning out the ones that only casually mention, or balancing out those that do/don't agree with your viewpoint are a general way to proceed. TEDickey (talk) 10:43, 7 October 2011 (UTC)

Full WPF Redesign

This statement is actually false. A good amount of Sharp Develop still uses Windows Forms inside of WPF wrappers. A quick rifle through the source code shows this. Sircmpwn (talk) 19:45, 27 December 2011 (UTC)

That was my impression (which is why I raised the issue of SDK, etc). The secondary sourcing is still very weak, and the topic's promoters don't seem inclined to dig this deeply. TEDickey (talk) 21:29, 27 December 2011 (UTC)

Re: "SharpDevelop 4.0 has been rewritten ..." Without defending it, I think this is SOP for MS to say to devs, "leverage your Forms by just wrapping them, for a whole new interface!" So it's not technically false, just very misleading, and worthy of deletion as unsourced. --Lexein (talk) 00:12, 28 December 2011 (UTC)

Unsourced section moved here for discussion, sourcing

Architecture

SharpDevelop uses its own parsers for C# and VB.NET code completion. The parsers were generated using a grammar description and a modified version of the Coco/R compiler generator of the University of Linz, which is included in the source code. For Boo, the parser from the Boo compiler is used, but resolving the type of expressions and type inference is done with custom code that supports lazy evaluation of the types.

The integrated SharpDevelop debugger uses its own debugging library that communicates with the Microsoft .NET runtime using COM interop. SharpDevelop 4.0 has been rewritten to have a full WPF interface and is based on .NET 4.0. It also uses Microsoft's new MSBuild API.

The last sentence has been flagged (above) as (at best) misleading. --Lexein (talk) 00:12, 28 December 2011 (UTC)