{"id":36693,"date":"2005-01-14T07:00:00","date_gmt":"2005-01-14T07:00:00","guid":{"rendered":"https:\/\/blogs.msdn.microsoft.com\/oldnewthing\/2005\/01\/14\/cleaner-more-elegant-and-harder-to-recognize\/"},"modified":"2005-01-14T07:00:00","modified_gmt":"2005-01-14T07:00:00","slug":"cleaner-more-elegant-and-harder-to-recognize","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/oldnewthing\/20050114-00\/?p=36693","title":{"rendered":"Cleaner, more elegant, and harder to recognize"},"content":{"rendered":"

\nIt appears that some people interpreted the title of\n\none of my rants from many months ago,\n“Cleaner, more elegant, and wrong”<\/a>,\nto be\n\na reference to exceptions in general<\/a>.\n(See bibliography reference [35]; observe that the citer\neven changed the title of my article for me!)\n<\/p>\n

\nThe title of the article was a reference to a specific code snippet\nthat I copied from a book, where the book’s author claimed that\nthe code he presented was “cleaner and more elegant”.\nI was pointing out that the code fragment was not only cleaner\nand more elegant, it was also wrong.\n<\/p>\n

\nYou can<\/strong> write correct exception-based programming.\n<\/p>\n

\nMind you, it’s hard.\n<\/p>\n

\nOn the other hand, just because something is hard doesn’t mean\nthat it shouldn’t be done.\n<\/p>\n

\nHere’s a breakdown:\n<\/p>\n\n\n\n
Really easy<\/th>\nHard<\/th>\nReally hard<\/th>\n<\/tr>\n
Writing bad error-code-based code
\nWriting bad exception-based code<\/td>\n		Writing good error-code-based code<\/td>\nWriting good exception-based code<\/td>\n<\/tr>\n<\/table>\n

\nIt’s easy to write bad code, regardless of the error model.\n<\/p>\n

\nIt’s hard to write good error-code-based code since you have to check\nevery error code and think about what you should do when an error occurs.\n<\/p>\n

\nIt’s really<\/strong> hard to write good exception-based\ncode since you have to check every single line of code (indeed,\nevery sub-expression) and think\nabout what exceptions it might raise and how your code will react to it.\n(In C++ it’s not quite so bad because C++ exceptions are raised only\nat specific points during execution.\nIn C#, exceptions can be raised at any time.)\n<\/p>\n

\nBut that’s okay. Like I said, just because something is hard\ndoesn’t mean it shouldn’t be done.\nIt’s hard to write a device driver, but people do it,\nand that’s a good thing.\n<\/p>\n

\nBut here’s another table:\n<\/p>\n\n\n\n
Really easy<\/th>\nHard<\/th>\nReally hard<\/th>\n<\/tr>\n
\n\n\n\n
\n Recognizing that error-code-based code is badly-written<\/td>\n<\/tr>\n
\n Recognizing the difference between bad error-code-based code and\n not-bad error-code-based code.<\/td>\n<\/tr>\n<\/table>\n<\/td>\n\n\n\n
\n Recognizing that error-code-base code is not badly-written<\/td>\n<\/tr>\n<\/table>\n<\/td>\n\n\n\n\n\n\n
\n Recognizing that exception-based code is badly-written<\/td>\n
\n Recognizing that exception-based code is not badly-written<\/td>\n<\/tr>\n
\n Recognizing the difference between bad exception-based code\n and not-bad exception-based code<\/td>\n<\/tr>\n<\/table>\n<\/td>\n<\/table>\n

\nHere’s some imaginary error-code-based code. See if you can\nclassify it as “bad” or “not-bad”:\n<\/p>\n

\nBOOL ComputeChecksum(LPCTSTR pszFile, DWORD* pdwResult)\n{\n  HANDLE h = CreateFile(pszFile, GENERIC_READ, FILE_SHARE_READ,\n       NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);\n  HANDLE hfm = CreateFileMapping(h, NULL, PAGE_READ, 0, 0, NULL);\n  void *pv = MapViewOfFile(hfm, FILE_MAP_READ, 0, 0, 0);\n  DWORD dwHeaderSum;\n  CheckSumMappedFile(pvBase, GetFileSize(h, NULL),\n           &dwHeaderSum, pdwResult);\n  UnmapViewOfFile(pv);\n  CloseHandle(hfm);\n  CloseHandle(h);\n  return TRUE;\n}\n<\/pre>\n
\nThis code is obviously bad.  No error codes are checked.\nThis is the sort of code you might write when in a hurry,\nmeaning to come back to and improve later.\nAnd it’s easy to spot that this code needs to be improved big time\nbefore it’s ready for prime time.\n<\/p>\n
\nHere’s another version:\n<\/p>\n
\nBOOL ComputeChecksum(LPCTSTR pszFile, DWORD* pdwResult)\n{\n  BOOL fRc = FALSE;\n  HANDLE h = CreateFile(pszFile, GENERIC_READ, FILE_SHARE_READ,\n       NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);\n  if (h != INVALID_HANDLE_VALUE) {\n    HANDLE hfm = CreateFileMapping(h, NULL, PAGE_READ, 0, 0, NULL);\n    if (hfm) {\n      void *pv = MapViewOfFile(hfm, FILE_MAP_READ, 0, 0, 0);\n      if (pv) {\n        DWORD dwHeaderSum;\n        if (CheckSumMappedFile(pvBase, GetFileSize(h, NULL),\n                               &dwHeaderSum, pdwResult)) {\n          fRc = TRUE;\n        }\n        UnmapViewOfFile(pv);\n      }\n      CloseHandle(hfm);\n    }\n    CloseHandle(h);\n  }\n  return fRc;\n}\n<\/pre>\n
\nThis code is still wrong, but it clearly looks like\nit’s trying to be right.  It is what I call “not-bad”.\n<\/p>\n
\nNow here’s some exception-based code you might write in a hurry:\n<\/p>\n
\nNotifyIcon CreateNotifyIcon()\n{\n NotifyIcon icon = new NotifyIcon();\n icon.Text = \"Blah blah blah\";\n icon.Visible = true;\n icon.Icon = new Icon(GetType(), \"cool.ico\");\n return icon;\n}\n<\/pre>\n
\n(This is actual code from a real program in an article about\ntaskbar notification icons, with minor changes\nin a futile attempt to disguise the source.)\n<\/p>\n
\nHere’s what it might look like after you fix it to be correct\nin the face of exceptions:\n<\/p>\n
\nNotifyIcon CreateNotifyIcon()\n{\n NotifyIcon icon = new NotifyIcon();\n icon.Text = \"Blah blah blah\";\n icon.Icon = new Icon(GetType(), \"cool.ico\");\n icon.Visible = true;\n return icon;\n}\n<\/pre>\n
\nSubtle, isn’t it.\n<\/p>\n
\nIt’s easy to spot the difference between bad error-code-based code\nand not-bad error-code-based code:  The not-bad error-code-based code\nchecks error codes.  The bad error-code-based code never does.\nAdmittedly, it’s hard to tell whether the errors were handled\ncorrectly, but at least you can tell the difference between\nbad code and code that isn’t bad.  (It might not be good, but at least\nit isn’t bad.)\n<\/p>\n
\nOn the other hand, it is extraordinarily difficult to see the\ndifference between bad exception-based code and not-bad\nexception-based code.\n<\/p>\n
\nConsequently, when I write code that is exception-based,\nI do not have the luxury of writing bad code first and then\nmaking it not-bad later.  If I did that, I wouldn’t be able to find\nthe bad code again, since it looks almost identical to not-bad code.\n<\/p>\n
\nMy point isn’t that exceptions are bad.\nMy point is that exceptions are too hard and I’m not smart\nenough to handle them.  (And neither, it seems, are book authors,\neven when they are trying to teach you how to program with exceptions!)\n<\/p>\n
\n(Yes, there are programming models like RAII and transactions,\nbut rarely do you see sample code that uses either.)<\/p>\n","protected":false},"excerpt":{"rendered":"
It appears that some people interpreted the title of one of my rants from many months ago, “Cleaner, more elegant, and wrong”, to be a reference to exceptions in general. (See bibliography reference [35]; observe that the citer even changed the title of my article for me!) The title of the article was a reference […]<\/p>\n","protected":false},"author":1069,"featured_media":111744,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[26],"class_list":["post-36693","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-oldnewthing","tag-other"],"acf":[],"blog_post_summary":"
It appears that some people interpreted the title of one of my rants from many months ago, “Cleaner, more elegant, and wrong”, to be a reference to exceptions in general. (See bibliography reference [35]; observe that the citer even changed the title of my article for me!) The title of the article was a reference […]<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/36693","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/users\/1069"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/comments?post=36693"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/36693\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media\/111744"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media?parent=36693"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/categories?post=36693"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/tags?post=36693"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}